Methods in Enzymology 472 - Single Molecule Tools: Fluorescence Based Approaches

METHODS IN ENZYMOLOGY Editors-in-Chief

JOHN N. ABELSON AND MELVIN I. SIMON Division of Biology California Institute of Technology Pasadena, California Founding Editors

SIDNEY P. COLOWICK AND NATHAN O. KAPLAN

Academic Press is an imprint of Elsevier 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 32 Jamestown Road, London NW1 7BY, UK First edition 2010 Copyright # 2010, Elsevier Inc. All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@ elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made For information on all Academic Press publications visit our website at elsevierdirect.com ISBN: 978-0-12-374954-3 ISSN: 0076-6879 Printed and bound in United States of America 10 11 12 10 9 8 7 6 5 4 3 2 1

CONTRIBUTORS

John Abelson Department of Biochemistry and Biophysics, University of California, San Francisco, California, USA Ichiro Amitani Department of Microbiology, and Department of Molecular and Cellular Biology, University of California, Davis, California, USA William M. Atkins Department of Medicinal Chemistry, University of Washington, Seattle, Washington, USA Ronald J. Baskin Department of Molecular and Cellular Biology, University of California, Davis, California, USA Jaime J. Benı´tez Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York, USA Keith P. Bjornson Pacific Biosciences, Menlo Park, California, USA Mario Blanco Department of Chemistry, Single Molecule Analysis Group, and Program in Cellular and Molecular Biology, University of Michigan, Ann Arbor, Michigan, USA Mario Brameshuber Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Adina R. Buxbaum Anatomy and Structural Biology and Gruss-Lipper Biophotonics Center, Albert Einstein College of Medicine, New York, USA Anirban Chakraborty Department of Chemistry and Chemical Biology, Waksman Institute, and Howard Hughes Medical Institute, Rutgers University, Piscataway, New Jersey, USA Bidhan P. Chaudhuri Pacific Biosciences, Menlo Park, California, USA

xiii

xiv

Contributors

Peng Chen Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York, USA Yan Chen School of Physics and Astronomy, University of Minnesota, Minneapolis, USA Ronald L. Cicero Pacific Biosciences, Menlo Park, California, USA Ashok A. Deniz Department of Molecular Biology, The Scripps Research Institute, La Jolla, California, USA Christopher C. Dombrowski Department of Microbiology, and Department of Molecular and Cellular Biology, University of California, Davis, California, USA Richard H. Ebright Department of Chemistry and Chemical Biology, Waksman Institute, and Howard Hughes Medical Institute, Rutgers University, Piscataway, New Jersey, USA Yon W. Ebright Department of Chemistry and Chemical Biology, Waksman Institute, and Howard Hughes Medical Institute, Rutgers University, Piscataway, New Jersey, USA Margaret M. Elvekrog Department of Chemistry, Columbia University, New York, USA Michael T. Englander Department of Chemistry, and Integrated Program in Cellular, Molecular, and Biomedical Sciences, Columbia University, New York, USA Teresa Fazio Department of Applied Physics and Applied Mathematics, Center for Electron Transport in Molecular Nanostructures, NanoMedicine Center for Mechanical Biology, Columbia University, New York, USA Jingyi Fei Department of Chemistry, Columbia University, New York, USA Allan Chris M. Ferreon Department of Molecular Biology, The Scripps Research Institute, La Jolla, California, USA Benjamin A. Flusberg Pacific Biosciences, Menlo Park, California, USA

xv

Contributors

Yann Gambin Department of Molecular Biology, The Scripps Research Institute, La Jolla, California, USA Lori S. Goldner Department of Physics, University of Massachusetts, Amherst, Massachusetts, USA Ruben L. Gonzalez Jr. Department of Chemistry, Columbia University, New York, USA Jason Gorman Department of Biological Sciences, Columbia University, New York, USA Jeremy J. Gray Pacific Biosciences, Menlo Park, California, USA Eric C. Greene The Howard Hughes Medical Institute, and Department of Biochemistry and Molecular Biophysics, Columbia University, New York, USA Max Greenfeld Department of Chemical Engineering, and Department of Biochemistry, Stanford University, Stanford, California, USA ¨rgen Groll Ju DWI e.V. and Institute of Technical and Macromolecular Chemistry, RWTH Aachen University, Aachen, Germany Christine Guthrie Department of Biochemistry and Biophysics, University of California, San Francisco, California, USA Haralambos Hadjivassiliou Department of Biochemistry and Biophysics, University of California, San Francisco, California, USA Christopher Hart Helicos BioSciences Massachusetts, USA

Corporation,

One

Kendall

Square,

Cambridge,

¨bler Thomas Haselgru Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Bettina Heise Department of Knowledge-based Mathematical Systems, Johannes Kepler University Linz, Linz, Austria Daniel Herschlag Department of Chemical Engineering, and Department of Biochemistry, Stanford University, Stanford, California, USA

xvi

Contributors

Clemens Hesch Biophysics Institute, Johannes Kepler University Linz, Linz, Austria David Holden Pacific Biosciences, Menlo Park, California, USA Ana M. Jofre Department of Physics and Optical Science, University of North Carolina, Charlotte, North Carolina, USA Jolene Johnson School of Physics and Astronomy, University of Minnesota, Minneapolis, USA Martin Kaltenbrunner Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Aaron M. Keller Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York, USA Samuel Kim Department of Chemistry, Stanford University, Stanford, California, USA Peter Koo Department of Physics, Yale University, New Haven, Connecticut, USA Jonas Korlach Pacific Biosciences, Menlo Park, California, USA Stephen C. Kowalczykowski Department of Microbiology, and Department of Molecular and Cellular Biology; Biophysics Graduate Group, University of California, Davis, California, USA Manuela Lehner Center for Biomedical Nanotechnology, Upper Austrian Research GmbH, Linz, Austria Doron Lipson Helicos BioSciences Massachusetts, USA

Corporation,

One

Kendall

Square,

Cambridge,

Bian Liu Department of Microbiology, and Department of Molecular and Cellular Biology; Biophysics Graduate Group, University of California, Davis, California, USA Patrick Macdonald School of Physics and Astronomy, University of Minnesota, Minneapolis, USA Daniel D. MacDougall Department of Chemistry, Columbia University, New York, USA

xvii

Contributors

Patrice M. Milos Helicos BioSciences Massachusetts, USA

Corporation,

One

Kendall

Square,

Cambridge,

Andrew D. Miranker Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut, USA Martin Moeller DWI e.V. and Institute of Technical and Macromolecular Chemistry, RWTH Aachen University, Aachen, Germany Crystal R. Moran Department of Molecular Biology, The Scripps Research Institute, La Jolla, California, USA Joachim D. Mueller School of Physics and Astronomy, University of Minnesota, Minneapolis, USA Abhinav Nath Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut, USA, and Department of Medicinal Chemistry, University of Washington, Seattle, Washington, USA Fatih Ozsolak Helicos BioSciences Massachusetts, USA

Corporation,

One

Kendall

Square,

Cambridge,

Christian Paar Center for Biomedical Nanotechnology, Upper Austrian Research GmbH, Linz, Austria Hye Yoon Park Anatomy and Structural Biology and Gruss-Lipper Biophotonics Center, Albert Einstein College of Medicine, New York, USA Wolfgang Paster Department of Molecular Immunology, Center for Physiology, Pathophysiology, and Immunology, Medical University of Vienna, Vienna, Austria Zdeneˇk Petra´sˇek Biotec, TU Dresden, Dresden, Germany Dileep K. Pulukkunat Department of Chemistry, Columbia University, New York, USA Arjun Raj Department of Bioengineering, University of Pennsylvania, Philadelphia, Pennsylvania, USA

xviii

Tal Raz Helicos BioSciences Massachusetts, USA

Contributors

Corporation,

One

Kendall

Square,

Cambridge,

Elizabeth Rhoades Department of Molecular Biophysics and Biochemistry, and Department of Physics, Yale University, New Haven, Connecticut, USA Jonas Ries Biotec, TU Dresden, Dresden, Germany Ravi Saxena Pacific Biosciences, Menlo Park, California, USA ¨tz Gerhard J. Schu Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Michaela Schwarzenbacher Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Petra Schwille Biotec, TU Dresden, Dresden, Germany Robert H. Singer Anatomy and Structural Biology and Gruss-Lipper Biophotonics Center, Albert Einstein College of Medicine, New York, USA Alois Sonnleitner Center for Biomedical Nanotechnology, Upper Austrian Research GmbH, Linz, Austria Kathleen Steinmann Helicos BioSciences Massachusetts, USA

Corporation,

One

Kendall

Square,

Cambridge,

Samuel H. Sternberg Department of Chemistry, Columbia University, New York, USA Hannes Stockinger Department of Molecular Immunology, Center for Physiology, Pathophysiology, and Immunology, Medical University of Vienna, Vienna, Austria Stefan Sunzenauer Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Jianyong Tang Janelia Farm Research Campus, Howard Hughes Medical Institute, Ashburn, Virginia, USA

xix

Contributors

John Thompson Helicos BioSciences Massachusetts, USA

Corporation,

One

Kendall

Square,

Cambridge,

Adam J. Trexler Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut, USA Stephen W. Turner Pacific Biosciences, Menlo Park, California, USA Sanjay Tyagi Public Health Research Institute, New Jersey Medical School-UMDNJ, Newark, New Jersey, USA Mari-Liis Visnapuu Department of Biochemistry and Molecular Biophysics, Columbia University, New York, USA Nils G. Walter Department of Chemistry, Single Molecule Analysis Group, University of Michigan, Ann Arbor, Michigan, USA Dongye Wang Department of Chemistry and Chemical Biology, Waksman Institute, and Howard Hughes Medical Institute, Rutgers University, Piscataway, New Jersey, USA Jiangning Wang Department of Chemistry, Columbia University, New York, USA Jeffrey Wegener Pacific Biosciences, Menlo Park, California, USA Julian Weghuber Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Shalom Wind Department of Applied Physics and Applied Mathematics, Center for Electron Transport in Molecular Nanostructures, NanoMedicine Center for Mechanical Biology, Columbia University, New York, USA Bin Wu School of Physics and Astronomy, University of Minnesota, Minneapolis, and Albert Einstein College of Medicine, Bronx, New York, USA Richard N. Zare Department of Chemistry, Stanford University, Stanford, California, USA

PREFACE

Ever since Feynman’s suggestion in the early 1960s that ‘‘there’s plenty of room at the bottom’’, single-molecule tools have seen an exponential rise in popularity (note that exponentially increasing rates are characteristic of explosions!). One can hardly go to a Biophysical Society meeting these days without being impressed by the literally thousands of posters and seminars that show data exploiting the unique capabilities of single-molecule probing techniques. Among their benefits are that they (i) can directly observe events at the molecular level; (ii) reveal rare and/or transient species and heterogeneities along a reaction pathway, which are often lost in ensemble averages; (iii) can directly access the low copy numbers (typically 1–1000) of any specific biopolymer in a single cell; (iv) afford counting and nanometer-accuracy localization of molecules in spatially distributed samples such as a cell; (v) enable the ultimate miniaturization and multiplexing of biological assays such as DNA sequencing; (vi) allow for the direct measurement of the mechanical forces affecting and enacted by biopolymers; and (vii) yield standard populationaveraged information from the statistics of many single-molecule observations. A half-century of single-molecule tool development has yielded technical advances that have demonstrated each of these advantages, and more are sure to emerge. Yet in any field enjoying increasing popularity, there inevitably comes a crossroads, which inspired MIE volumes 472 and 475. To advance beyond being used or studied only by a limited (and eventually vanishing) group of specialists, a set of tools or area of research needs to find more widespread appreciation. Many methods that are commonplace in labs today—such as gel electrophoresis, PCR, and sequencing—made that transition from specialist’s art to general practitioner’s basic tool by a combination of being very appealing and becoming easy to master. The two MIE volumes are aimed to facilitate this transition by, often for the first time, revealing for a broad selection of single-molecule tools those details that pioneering specialists rarely have the space to cover in their research publications. Compiling methods from an emerging field is a daunting task, since new tools are developed nearly daily. The resulting selection is, by necessity, incomplete, limited by both the availability of contributors and my gaps in knowledge. Yet through the vigorous response to my solicitation of articles, what was planned as one volume became two, somewhat loosely organized by theme. While editing each of these works, I became increasingly impressed by the consistently superb quality of the contributions, in terms of both style and substance. I am therefore very grateful to John Abelson for xxi

xxii

Preface

convincing me to take on the job as editor, and trusting me with it, to the phenomenal group of authors (some of which even made the deadline), and to the staff at Elsevier for allowing me to divide the contributions into two volumes and supporting me in numerous other ways. My hope is that the hard work by everyone involved bears fruit and helps spread the word and enthusiasm about the power of single-molecule tools. NILS G. WALTER

METHODS IN ENZYMOLOGY

VOLUME I. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME II. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME III. Preparation and Assay of Substrates Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME IV. Special Techniques for the Enzymologist Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME V. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VI. Preparation and Assay of Enzymes (Continued) Preparation and Assay of Substrates Special Techniques Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VII. Cumulative Subject Index Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VIII. Complex Carbohydrates Edited by ELIZABETH F. NEUFELD AND VICTOR GINSBURG VOLUME IX. Carbohydrate Metabolism Edited by WILLIS A. WOOD VOLUME X. Oxidation and Phosphorylation Edited by RONALD W. ESTABROOK AND MAYNARD E. PULLMAN VOLUME XI. Enzyme Structure Edited by C. H. W. HIRS VOLUME XII. Nucleic Acids (Parts A and B) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XIII. Citric Acid Cycle Edited by J. M. LOWENSTEIN VOLUME XIV. Lipids Edited by J. M. LOWENSTEIN VOLUME XV. Steroids and Terpenoids Edited by RAYMOND B. CLAYTON

xxiii

xxiv

Methods in Enzymology

VOLUME XVI. Fast Reactions Edited by KENNETH KUSTIN VOLUME XVII. Metabolism of Amino Acids and Amines (Parts A and B) Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME XVIII. Vitamins and Coenzymes (Parts A, B, and C) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME XIX. Proteolytic Enzymes Edited by GERTRUDE E. PERLMANN AND LASZLO LORAND VOLUME XX. Nucleic Acids and Protein Synthesis (Part C) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME XXI. Nucleic Acids (Part D) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XXII. Enzyme Purification and Related Techniques Edited by WILLIAM B. JAKOBY VOLUME XXIII. Photosynthesis (Part A) Edited by ANTHONY SAN PIETRO VOLUME XXIV. Photosynthesis and Nitrogen Fixation (Part B) Edited by ANTHONY SAN PIETRO VOLUME XXV. Enzyme Structure (Part B) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVI. Enzyme Structure (Part C) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVII. Enzyme Structure (Part D) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVIII. Complex Carbohydrates (Part B) Edited by VICTOR GINSBURG VOLUME XXIX. Nucleic Acids and Protein Synthesis (Part E) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XXX. Nucleic Acids and Protein Synthesis (Part F) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME XXXI. Biomembranes (Part A) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME XXXII. Biomembranes (Part B) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME XXXIII. Cumulative Subject Index Volumes I-XXX Edited by MARTHA G. DENNIS AND EDWARD A. DENNIS VOLUME XXXIV. Affinity Techniques (Enzyme Purification: Part B) Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK

Methods in Enzymology

VOLUME XXXV. Lipids (Part B) Edited by JOHN M. LOWENSTEIN VOLUME XXXVI. Hormone Action (Part A: Steroid Hormones) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XXXVII. Hormone Action (Part B: Peptide Hormones) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XXXVIII. Hormone Action (Part C: Cyclic Nucleotides) Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY VOLUME XXXIX. Hormone Action (Part D: Isolated Cells, Tissues, and Organ Systems) Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY VOLUME XL. Hormone Action (Part E: Nuclear Structure and Function) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XLI. Carbohydrate Metabolism (Part B) Edited by W. A. WOOD VOLUME XLII. Carbohydrate Metabolism (Part C) Edited by W. A. WOOD VOLUME XLIII. Antibiotics Edited by JOHN H. HASH VOLUME XLIV. Immobilized Enzymes Edited by KLAUS MOSBACH VOLUME XLV. Proteolytic Enzymes (Part B) Edited by LASZLO LORAND VOLUME XLVI. Affinity Labeling Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK VOLUME XLVII. Enzyme Structure (Part E) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XLVIII. Enzyme Structure (Part F) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XLIX. Enzyme Structure (Part G) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME L. Complex Carbohydrates (Part C) Edited by VICTOR GINSBURG VOLUME LI. Purine and Pyrimidine Nucleotide Metabolism Edited by PATRICIA A. HOFFEE AND MARY ELLEN JONES VOLUME LII. Biomembranes (Part C: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER

xxv

xxvi

Methods in Enzymology

VOLUME LIII. Biomembranes (Part D: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LIV. Biomembranes (Part E: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LV. Biomembranes (Part F: Bioenergetics) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVI. Biomembranes (Part G: Bioenergetics) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVII. Bioluminescence and Chemiluminescence Edited by MARLENE A. DELUCA VOLUME LVIII. Cell Culture Edited by WILLIAM B. JAKOBY AND IRA PASTAN VOLUME LIX. Nucleic Acids and Protein Synthesis (Part G) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME LX. Nucleic Acids and Protein Synthesis (Part H) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME 61. Enzyme Structure (Part H) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 62. Vitamins and Coenzymes (Part D) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 63. Enzyme Kinetics and Mechanism (Part A: Initial Rate and Inhibitor Methods) Edited by DANIEL L. PURICH VOLUME 64. Enzyme Kinetics and Mechanism (Part B: Isotopic Probes and Complex Enzyme Systems) Edited by DANIEL L. PURICH VOLUME 65. Nucleic Acids (Part I) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME 66. Vitamins and Coenzymes (Part E) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 67. Vitamins and Coenzymes (Part F) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 68. Recombinant DNA Edited by RAY WU VOLUME 69. Photosynthesis and Nitrogen Fixation (Part C) Edited by ANTHONY SAN PIETRO VOLUME 70. Immunochemical Techniques (Part A) Edited by HELEN VAN VUNAKIS AND JOHN J. LANGONE

Methods in Enzymology

xxvii

VOLUME 71. Lipids (Part C) Edited by JOHN M. LOWENSTEIN VOLUME 72. Lipids (Part D) Edited by JOHN M. LOWENSTEIN VOLUME 73. Immunochemical Techniques (Part B) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 74. Immunochemical Techniques (Part C) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 75. Cumulative Subject Index Volumes XXXI, XXXII, XXXIV–LX Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 76. Hemoglobins Edited by ERALDO ANTONINI, LUIGI ROSSI-BERNARDI, AND EMILIA CHIANCONE VOLUME 77. Detoxication and Drug Metabolism Edited by WILLIAM B. JAKOBY VOLUME 78. Interferons (Part A) Edited by SIDNEY PESTKA VOLUME 79. Interferons (Part B) Edited by SIDNEY PESTKA VOLUME 80. Proteolytic Enzymes (Part C) Edited by LASZLO LORAND VOLUME 81. Biomembranes (Part H: Visual Pigments and Purple Membranes, I) Edited by LESTER PACKER VOLUME 82. Structural and Contractile Proteins (Part A: Extracellular Matrix) Edited by LEON W. CUNNINGHAM AND DIXIE W. FREDERIKSEN VOLUME 83. Complex Carbohydrates (Part D) Edited by VICTOR GINSBURG VOLUME 84. Immunochemical Techniques (Part D: Selected Immunoassays) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 85. Structural and Contractile Proteins (Part B: The Contractile Apparatus and the Cytoskeleton) Edited by DIXIE W. FREDERIKSEN AND LEON W. CUNNINGHAM VOLUME 86. Prostaglandins and Arachidonate Metabolites Edited by WILLIAM E. M. LANDS AND WILLIAM L. SMITH VOLUME 87. Enzyme Kinetics and Mechanism (Part C: Intermediates, Stereo-chemistry, and Rate Studies) Edited by DANIEL L. PURICH VOLUME 88. Biomembranes (Part I: Visual Pigments and Purple Membranes, II) Edited by LESTER PACKER

xxviii

Methods in Enzymology

VOLUME 89. Carbohydrate Metabolism (Part D) Edited by WILLIS A. WOOD VOLUME 90. Carbohydrate Metabolism (Part E) Edited by WILLIS A. WOOD VOLUME 91. Enzyme Structure (Part I) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 92. Immunochemical Techniques (Part E: Monoclonal Antibodies and General Immunoassay Methods) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 93. Immunochemical Techniques (Part F: Conventional Antibodies, Fc Receptors, and Cytotoxicity) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 94. Polyamines Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME 95. Cumulative Subject Index Volumes 61–74, 76–80 Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 96. Biomembranes [Part J: Membrane Biogenesis: Assembly and Targeting (General Methods; Eukaryotes)] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 97. Biomembranes [Part K: Membrane Biogenesis: Assembly and Targeting (Prokaryotes, Mitochondria, and Chloroplasts)] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 98. Biomembranes (Part L: Membrane Biogenesis: Processing and Recycling) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 99. Hormone Action (Part F: Protein Kinases) Edited by JACKIE D. CORBIN AND JOEL G. HARDMAN VOLUME 100. Recombinant DNA (Part B) Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE VOLUME 101. Recombinant DNA (Part C) Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE VOLUME 102. Hormone Action (Part G: Calmodulin and Calcium-Binding Proteins) Edited by ANTHONY R. MEANS AND BERT W. O’MALLEY VOLUME 103. Hormone Action (Part H: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 104. Enzyme Purification and Related Techniques (Part C) Edited by WILLIAM B. JAKOBY

Methods in Enzymology

xxix

VOLUME 105. Oxygen Radicals in Biological Systems Edited by LESTER PACKER VOLUME 106. Posttranslational Modifications (Part A) Edited by FINN WOLD AND KIVIE MOLDAVE VOLUME 107. Posttranslational Modifications (Part B) Edited by FINN WOLD AND KIVIE MOLDAVE VOLUME 108. Immunochemical Techniques (Part G: Separation and Characterization of Lymphoid Cells) Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS VOLUME 109. Hormone Action (Part I: Peptide Hormones) Edited by LUTZ BIRNBAUMER AND BERT W. O’MALLEY VOLUME 110. Steroids and Isoprenoids (Part A) Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 111. Steroids and Isoprenoids (Part B) Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 112. Drug and Enzyme Targeting (Part A) Edited by KENNETH J. WIDDER AND RALPH GREEN VOLUME 113. Glutamate, Glutamine, Glutathione, and Related Compounds Edited by ALTON MEISTER VOLUME 114. Diffraction Methods for Biological Macromolecules (Part A) Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 115. Diffraction Methods for Biological Macromolecules (Part B) Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 116. Immunochemical Techniques (Part H: Effectors and Mediators of Lymphoid Cell Functions) Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS VOLUME 117. Enzyme Structure (Part J) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 118. Plant Molecular Biology Edited by ARTHUR WEISSBACH AND HERBERT WEISSBACH VOLUME 119. Interferons (Part C) Edited by SIDNEY PESTKA VOLUME 120. Cumulative Subject Index Volumes 81–94, 96–101 VOLUME 121. Immunochemical Techniques (Part I: Hybridoma Technology and Monoclonal Antibodies) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 122. Vitamins and Coenzymes (Part G) Edited by FRANK CHYTIL AND DONALD B. MCCORMICK

xxx

Methods in Enzymology

VOLUME 123. Vitamins and Coenzymes (Part H) Edited by FRANK CHYTIL AND DONALD B. MCCORMICK VOLUME 124. Hormone Action (Part J: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 125. Biomembranes (Part M: Transport in Bacteria, Mitochondria, and Chloroplasts: General Approaches and Transport Systems) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 126. Biomembranes (Part N: Transport in Bacteria, Mitochondria, and Chloroplasts: Protonmotive Force) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 127. Biomembranes (Part O: Protons and Water: Structure and Translocation) Edited by LESTER PACKER VOLUME 128. Plasma Lipoproteins (Part A: Preparation, Structure, and Molecular Biology) Edited by JERE P. SEGREST AND JOHN J. ALBERS VOLUME 129. Plasma Lipoproteins (Part B: Characterization, Cell Biology, and Metabolism) Edited by JOHN J. ALBERS AND JERE P. SEGREST VOLUME 130. Enzyme Structure (Part K) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 131. Enzyme Structure (Part L) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 132. Immunochemical Techniques (Part J: Phagocytosis and Cell-Mediated Cytotoxicity) Edited by GIOVANNI DI SABATO AND JOHANNES EVERSE VOLUME 133. Bioluminescence and Chemiluminescence (Part B) Edited by MARLENE DELUCA AND WILLIAM D. MCELROY VOLUME 134. Structural and Contractile Proteins (Part C: The Contractile Apparatus and the Cytoskeleton) Edited by RICHARD B. VALLEE VOLUME 135. Immobilized Enzymes and Cells (Part B) Edited by KLAUS MOSBACH VOLUME 136. Immobilized Enzymes and Cells (Part C) Edited by KLAUS MOSBACH VOLUME 137. Immobilized Enzymes and Cells (Part D) Edited by KLAUS MOSBACH VOLUME 138. Complex Carbohydrates (Part E) Edited by VICTOR GINSBURG

Methods in Enzymology

xxxi

VOLUME 139. Cellular Regulators (Part A: Calcium- and Calmodulin-Binding Proteins) Edited by ANTHONY R. MEANS AND P. MICHAEL CONN VOLUME 140. Cumulative Subject Index Volumes 102–119, 121–134 VOLUME 141. Cellular Regulators (Part B: Calcium and Lipids) Edited by P. MICHAEL CONN AND ANTHONY R. MEANS VOLUME 142. Metabolism of Aromatic Amino Acids and Amines Edited by SEYMOUR KAUFMAN VOLUME 143. Sulfur and Sulfur Amino Acids Edited by WILLIAM B. JAKOBY AND OWEN GRIFFITH VOLUME 144. Structural and Contractile Proteins (Part D: Extracellular Matrix) Edited by LEON W. CUNNINGHAM VOLUME 145. Structural and Contractile Proteins (Part E: Extracellular Matrix) Edited by LEON W. CUNNINGHAM VOLUME 146. Peptide Growth Factors (Part A) Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 147. Peptide Growth Factors (Part B) Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 148. Plant Cell Membranes Edited by LESTER PACKER AND ROLAND DOUCE VOLUME 149. Drug and Enzyme Targeting (Part B) Edited by RALPH GREEN AND KENNETH J. WIDDER VOLUME 150. Immunochemical Techniques (Part K: In Vitro Models of B and T Cell Functions and Lymphoid Cell Receptors) Edited by GIOVANNI DI SABATO VOLUME 151. Molecular Genetics of Mammalian Cells Edited by MICHAEL M. GOTTESMAN VOLUME 152. Guide to Molecular Cloning Techniques Edited by SHELBY L. BERGER AND ALAN R. KIMMEL VOLUME 153. Recombinant DNA (Part D) Edited by RAY WU AND LAWRENCE GROSSMAN VOLUME 154. Recombinant DNA (Part E) Edited by RAY WU AND LAWRENCE GROSSMAN VOLUME 155. Recombinant DNA (Part F) Edited by RAY WU VOLUME 156. Biomembranes (Part P: ATP-Driven Pumps and Related Transport: The Na, K-Pump) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER

xxxii

Methods in Enzymology

VOLUME 157. Biomembranes (Part Q: ATP-Driven Pumps and Related Transport: Calcium, Proton, and Potassium Pumps) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 158. Metalloproteins (Part A) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 159. Initiation and Termination of Cyclic Nucleotide Action Edited by JACKIE D. CORBIN AND ROGER A. JOHNSON VOLUME 160. Biomass (Part A: Cellulose and Hemicellulose) Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 161. Biomass (Part B: Lignin, Pectin, and Chitin) Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 162. Immunochemical Techniques (Part L: Chemotaxis and Inflammation) Edited by GIOVANNI DI SABATO VOLUME 163. Immunochemical Techniques (Part M: Chemotaxis and Inflammation) Edited by GIOVANNI DI SABATO VOLUME 164. Ribosomes Edited by HARRY F. NOLLER, JR., AND KIVIE MOLDAVE VOLUME 165. Microbial Toxins: Tools for Enzymology Edited by SIDNEY HARSHMAN VOLUME 166. Branched-Chain Amino Acids Edited by ROBERT HARRIS AND JOHN R. SOKATCH VOLUME 167. Cyanobacteria Edited by LESTER PACKER AND ALEXANDER N. GLAZER VOLUME 168. Hormone Action (Part K: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 169. Platelets: Receptors, Adhesion, Secretion (Part A) Edited by JACEK HAWIGER VOLUME 170. Nucleosomes Edited by PAUL M. WASSARMAN AND ROGER D. KORNBERG VOLUME 171. Biomembranes (Part R: Transport Theory: Cells and Model Membranes) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 172. Biomembranes (Part S: Transport: Membrane Isolation and Characterization) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER

Methods in Enzymology

xxxiii

VOLUME 173. Biomembranes [Part T: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 174. Biomembranes [Part U: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 175. Cumulative Subject Index Volumes 135–139, 141–167 VOLUME 176. Nuclear Magnetic Resonance (Part A: Spectral Techniques and Dynamics) Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES VOLUME 177. Nuclear Magnetic Resonance (Part B: Structure and Mechanism) Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES VOLUME 178. Antibodies, Antigens, and Molecular Mimicry Edited by JOHN J. LANGONE VOLUME 179. Complex Carbohydrates (Part F) Edited by VICTOR GINSBURG VOLUME 180. RNA Processing (Part A: General Methods) Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 181. RNA Processing (Part B: Specific Methods) Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 182. Guide to Protein Purification Edited by MURRAY P. DEUTSCHER VOLUME 183. Molecular Evolution: Computer Analysis of Protein and Nucleic Acid Sequences Edited by RUSSELL F. DOOLITTLE VOLUME 184. Avidin-Biotin Technology Edited by MEIR WILCHEK AND EDWARD A. BAYER VOLUME 185. Gene Expression Technology Edited by DAVID V. GOEDDEL VOLUME 186. Oxygen Radicals in Biological Systems (Part B: Oxygen Radicals and Antioxidants) Edited by LESTER PACKER AND ALEXANDER N. GLAZER VOLUME 187. Arachidonate Related Lipid Mediators Edited by ROBERT C. MURPHY AND FRANK A. FITZPATRICK VOLUME 188. Hydrocarbons and Methylotrophy Edited by MARY E. LIDSTROM VOLUME 189. Retinoids (Part A: Molecular and Metabolic Aspects) Edited by LESTER PACKER

xxxiv

Methods in Enzymology

VOLUME 190. Retinoids (Part B: Cell Differentiation and Clinical Applications) Edited by LESTER PACKER VOLUME 191. Biomembranes (Part V: Cellular and Subcellular Transport: Epithelial Cells) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 192. Biomembranes (Part W: Cellular and Subcellular Transport: Epithelial Cells) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 193. Mass Spectrometry Edited by JAMES A. MCCLOSKEY VOLUME 194. Guide to Yeast Genetics and Molecular Biology Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 195. Adenylyl Cyclase, G Proteins, and Guanylyl Cyclase Edited by ROGER A. JOHNSON AND JACKIE D. CORBIN VOLUME 196. Molecular Motors and the Cytoskeleton Edited by RICHARD B. VALLEE VOLUME 197. Phospholipases Edited by EDWARD A. DENNIS VOLUME 198. Peptide Growth Factors (Part C) Edited by DAVID BARNES, J. P. MATHER, AND GORDON H. SATO VOLUME 199. Cumulative Subject Index Volumes 168–174, 176–194 VOLUME 200. Protein Phosphorylation (Part A: Protein Kinases: Assays, Purification, Antibodies, Functional Analysis, Cloning, and Expression) Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON VOLUME 201. Protein Phosphorylation (Part B: Analysis of Protein Phosphorylation, Protein Kinase Inhibitors, and Protein Phosphatases) Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON VOLUME 202. Molecular Design and Modeling: Concepts and Applications (Part A: Proteins, Peptides, and Enzymes) Edited by JOHN J. LANGONE VOLUME 203. Molecular Design and Modeling: Concepts and Applications (Part B: Antibodies and Antigens, Nucleic Acids, Polysaccharides, and Drugs) Edited by JOHN J. LANGONE VOLUME 204. Bacterial Genetic Systems Edited by JEFFREY H. MILLER VOLUME 205. Metallobiochemistry (Part B: Metallothionein and Related Molecules) Edited by JAMES F. RIORDAN AND BERT L. VALLEE

Methods in Enzymology

xxxv

VOLUME 206. Cytochrome P450 Edited by MICHAEL R. WATERMAN AND ERIC F. JOHNSON VOLUME 207. Ion Channels Edited by BERNARDO RUDY AND LINDA E. IVERSON VOLUME 208. Protein–DNA Interactions Edited by ROBERT T. SAUER VOLUME 209. Phospholipid Biosynthesis Edited by EDWARD A. DENNIS AND DENNIS E. VANCE VOLUME 210. Numerical Computer Methods Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 211. DNA Structures (Part A: Synthesis and Physical Analysis of DNA) Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 212. DNA Structures (Part B: Chemical and Electrophoretic Analysis of DNA) Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 213. Carotenoids (Part A: Chemistry, Separation, Quantitation, and Antioxidation) Edited by LESTER PACKER VOLUME 214. Carotenoids (Part B: Metabolism, Genetics, and Biosynthesis) Edited by LESTER PACKER VOLUME 215. Platelets: Receptors, Adhesion, Secretion (Part B) Edited by JACEK J. HAWIGER VOLUME 216. Recombinant DNA (Part G) Edited by RAY WU VOLUME 217. Recombinant DNA (Part H) Edited by RAY WU VOLUME 218. Recombinant DNA (Part I) Edited by RAY WU VOLUME 219. Reconstitution of Intracellular Transport Edited by JAMES E. ROTHMAN VOLUME 220. Membrane Fusion Techniques (Part A) Edited by NEJAT DU¨ZGU¨NES, VOLUME 221. Membrane Fusion Techniques (Part B) Edited by NEJAT DU¨ZGU¨NES, VOLUME 222. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part A: Mammalian Blood Coagulation Factors and Inhibitors) Edited by LASZLO LORAND AND KENNETH G. MANN

xxxvi

Methods in Enzymology

VOLUME 223. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part B: Complement Activation, Fibrinolysis, and Nonmammalian Blood Coagulation Factors) Edited by LASZLO LORAND AND KENNETH G. MANN VOLUME 224. Molecular Evolution: Producing the Biochemical Data Edited by ELIZABETH ANNE ZIMMER, THOMAS J. WHITE, REBECCA L. CANN, AND ALLAN C. WILSON VOLUME 225. Guide to Techniques in Mouse Development Edited by PAUL M. WASSARMAN AND MELVIN L. DEPAMPHILIS VOLUME 226. Metallobiochemistry (Part C: Spectroscopic and Physical Methods for Probing Metal Ion Environments in Metalloenzymes and Metalloproteins) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 227. Metallobiochemistry (Part D: Physical and Spectroscopic Methods for Probing Metal Ion Environments in Metalloproteins) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 228. Aqueous Two-Phase Systems Edited by HARRY WALTER AND GO¨TE JOHANSSON VOLUME 229. Cumulative Subject Index Volumes 195–198, 200–227 VOLUME 230. Guide to Techniques in Glycobiology Edited by WILLIAM J. LENNARZ AND GERALD W. HART VOLUME 231. Hemoglobins (Part B: Biochemical and Analytical Methods) Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW VOLUME 232. Hemoglobins (Part C: Biophysical Methods) Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW VOLUME 233. Oxygen Radicals in Biological Systems (Part C) Edited by LESTER PACKER VOLUME 234. Oxygen Radicals in Biological Systems (Part D) Edited by LESTER PACKER VOLUME 235. Bacterial Pathogenesis (Part A: Identification and Regulation of Virulence Factors) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 236. Bacterial Pathogenesis (Part B: Integration of Pathogenic Bacteria with Host Cells) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 237. Heterotrimeric G Proteins Edited by RAVI IYENGAR VOLUME 238. Heterotrimeric G-Protein Effectors Edited by RAVI IYENGAR

Methods in Enzymology

xxxvii

VOLUME 239. Nuclear Magnetic Resonance (Part C) Edited by THOMAS L. JAMES AND NORMAN J. OPPENHEIMER VOLUME 240. Numerical Computer Methods (Part B) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 241. Retroviral Proteases Edited by LAWRENCE C. KUO AND JULES A. SHAFER VOLUME 242. Neoglycoconjugates (Part A) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 243. Inorganic Microbial Sulfur Metabolism Edited by HARRY D. PECK, JR., AND JEAN LEGALL VOLUME 244. Proteolytic Enzymes: Serine and Cysteine Peptidases Edited by ALAN J. BARRETT VOLUME 245. Extracellular Matrix Components Edited by E. RUOSLAHTI AND E. ENGVALL VOLUME 246. Biochemical Spectroscopy Edited by KENNETH SAUER VOLUME 247. Neoglycoconjugates (Part B: Biomedical Applications) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 248. Proteolytic Enzymes: Aspartic and Metallo Peptidases Edited by ALAN J. BARRETT VOLUME 249. Enzyme Kinetics and Mechanism (Part D: Developments in Enzyme Dynamics) Edited by DANIEL L. PURICH VOLUME 250. Lipid Modifications of Proteins Edited by PATRICK J. CASEY AND JANICE E. BUSS VOLUME 251. Biothiols (Part A: Monothiols and Dithiols, Protein Thiols, and Thiyl Radicals) Edited by LESTER PACKER VOLUME 252. Biothiols (Part B: Glutathione and Thioredoxin; Thiols in Signal Transduction and Gene Regulation) Edited by LESTER PACKER VOLUME 253. Adhesion of Microbial Pathogens Edited by RON J. DOYLE AND ITZHAK OFEK VOLUME 254. Oncogene Techniques Edited by PETER K. VOGT AND INDER M. VERMA VOLUME 255. Small GTPases and Their Regulators (Part A: Ras Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL

xxxviii

Methods in Enzymology

VOLUME 256. Small GTPases and Their Regulators (Part B: Rho Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 257. Small GTPases and Their Regulators (Part C: Proteins Involved in Transport) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 258. Redox-Active Amino Acids in Biology Edited by JUDITH P. KLINMAN VOLUME 259. Energetics of Biological Macromolecules Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS VOLUME 260. Mitochondrial Biogenesis and Genetics (Part A) Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN VOLUME 261. Nuclear Magnetic Resonance and Nucleic Acids Edited by THOMAS L. JAMES VOLUME 262. DNA Replication Edited by JUDITH L. CAMPBELL VOLUME 263. Plasma Lipoproteins (Part C: Quantitation) Edited by WILLIAM A. BRADLEY, SANDRA H. GIANTURCO, AND JERE P. SEGREST VOLUME 264. Mitochondrial Biogenesis and Genetics (Part B) Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN VOLUME 265. Cumulative Subject Index Volumes 228, 230–262 VOLUME 266. Computer Methods for Macromolecular Sequence Analysis Edited by RUSSELL F. DOOLITTLE VOLUME 267. Combinatorial Chemistry Edited by JOHN N. ABELSON VOLUME 268. Nitric Oxide (Part A: Sources and Detection of NO; NO Synthase) Edited by LESTER PACKER VOLUME 269. Nitric Oxide (Part B: Physiological and Pathological Processes) Edited by LESTER PACKER VOLUME 270. High Resolution Separation and Analysis of Biological Macromolecules (Part A: Fundamentals) Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK VOLUME 271. High Resolution Separation and Analysis of Biological Macromolecules (Part B: Applications) Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK VOLUME 272. Cytochrome P450 (Part B) Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN VOLUME 273. RNA Polymerase and Associated Factors (Part A) Edited by SANKAR ADHYA

Methods in Enzymology

xxxix

VOLUME 274. RNA Polymerase and Associated Factors (Part B) Edited by SANKAR ADHYA VOLUME 275. Viral Polymerases and Related Proteins Edited by LAWRENCE C. KUO, DAVID B. OLSEN, AND STEVEN S. CARROLL VOLUME 276. Macromolecular Crystallography (Part A) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 277. Macromolecular Crystallography (Part B) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 278. Fluorescence Spectroscopy Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 279. Vitamins and Coenzymes (Part I) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 280. Vitamins and Coenzymes (Part J) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 281. Vitamins and Coenzymes (Part K) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 282. Vitamins and Coenzymes (Part L) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 283. Cell Cycle Control Edited by WILLIAM G. DUNPHY VOLUME 284. Lipases (Part A: Biotechnology) Edited by BYRON RUBIN AND EDWARD A. DENNIS VOLUME 285. Cumulative Subject Index Volumes 263, 264, 266–284, 286–289 VOLUME 286. Lipases (Part B: Enzyme Characterization and Utilization) Edited by BYRON RUBIN AND EDWARD A. DENNIS VOLUME 287. Chemokines Edited by RICHARD HORUK VOLUME 288. Chemokine Receptors Edited by RICHARD HORUK VOLUME 289. Solid Phase Peptide Synthesis Edited by GREGG B. FIELDS VOLUME 290. Molecular Chaperones Edited by GEORGE H. LORIMER AND THOMAS BALDWIN VOLUME 291. Caged Compounds Edited by GERARD MARRIOTT VOLUME 292. ABC Transporters: Biochemical, Cellular, and Molecular Aspects Edited by SURESH V. AMBUDKAR AND MICHAEL M. GOTTESMAN

xl

Methods in Enzymology

VOLUME 293. Ion Channels (Part B) Edited by P. MICHAEL CONN VOLUME 294. Ion Channels (Part C) Edited by P. MICHAEL CONN VOLUME 295. Energetics of Biological Macromolecules (Part B) Edited by GARY K. ACKERS AND MICHAEL L. JOHNSON VOLUME 296. Neurotransmitter Transporters Edited by SUSAN G. AMARA VOLUME 297. Photosynthesis: Molecular Biology of Energy Capture Edited by LEE MCINTOSH VOLUME 298. Molecular Motors and the Cytoskeleton (Part B) Edited by RICHARD B. VALLEE VOLUME 299. Oxidants and Antioxidants (Part A) Edited by LESTER PACKER VOLUME 300. Oxidants and Antioxidants (Part B) Edited by LESTER PACKER VOLUME 301. Nitric Oxide: Biological and Antioxidant Activities (Part C) Edited by LESTER PACKER VOLUME 302. Green Fluorescent Protein Edited by P. MICHAEL CONN VOLUME 303. cDNA Preparation and Display Edited by SHERMAN M. WEISSMAN VOLUME 304. Chromatin Edited by PAUL M. WASSARMAN AND ALAN P. WOLFFE VOLUME 305. Bioluminescence and Chemiluminescence (Part C) Edited by THOMAS O. BALDWIN AND MIRIAM M. ZIEGLER VOLUME 306. Expression of Recombinant Genes in Eukaryotic Systems Edited by JOSEPH C. GLORIOSO AND MARTIN C. SCHMIDT VOLUME 307. Confocal Microscopy Edited by P. MICHAEL CONN VOLUME 308. Enzyme Kinetics and Mechanism (Part E: Energetics of Enzyme Catalysis) Edited by DANIEL L. PURICH AND VERN L. SCHRAMM VOLUME 309. Amyloid, Prions, and Other Protein Aggregates Edited by RONALD WETZEL VOLUME 310. Biofilms Edited by RON J. DOYLE

Methods in Enzymology

VOLUME 311. Sphingolipid Metabolism and Cell Signaling (Part A) Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN VOLUME 312. Sphingolipid Metabolism and Cell Signaling (Part B) Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN VOLUME 313. Antisense Technology (Part A: General Methods, Methods of Delivery, and RNA Studies) Edited by M. IAN PHILLIPS VOLUME 314. Antisense Technology (Part B: Applications) Edited by M. IAN PHILLIPS VOLUME 315. Vertebrate Phototransduction and the Visual Cycle (Part A) Edited by KRZYSZTOF PALCZEWSKI VOLUME 316. Vertebrate Phototransduction and the Visual Cycle (Part B) Edited by KRZYSZTOF PALCZEWSKI VOLUME 317. RNA–Ligand Interactions (Part A: Structural Biology Methods) Edited by DANIEL W. CELANDER AND JOHN N. ABELSON VOLUME 318. RNA–Ligand Interactions (Part B: Molecular Biology Methods) Edited by DANIEL W. CELANDER AND JOHN N. ABELSON VOLUME 319. Singlet Oxygen, UV-A, and Ozone Edited by LESTER PACKER AND HELMUT SIES VOLUME 320. Cumulative Subject Index Volumes 290–319 VOLUME 321. Numerical Computer Methods (Part C) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 322. Apoptosis Edited by JOHN C. REED VOLUME 323. Energetics of Biological Macromolecules (Part C) Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS VOLUME 324. Branched-Chain Amino Acids (Part B) Edited by ROBERT A. HARRIS AND JOHN R. SOKATCH VOLUME 325. Regulators and Effectors of Small GTPases (Part D: Rho Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 326. Applications of Chimeric Genes and Hybrid Proteins (Part A: Gene Expression and Protein Purification) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON VOLUME 327. Applications of Chimeric Genes and Hybrid Proteins (Part B: Cell Biology and Physiology) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON

xli

xlii

Methods in Enzymology

VOLUME 328. Applications of Chimeric Genes and Hybrid Proteins (Part C: Protein–Protein Interactions and Genomics) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON VOLUME 329. Regulators and Effectors of Small GTPases (Part E: GTPases Involved in Vesicular Traffic) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 330. Hyperthermophilic Enzymes (Part A) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 331. Hyperthermophilic Enzymes (Part B) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 332. Regulators and Effectors of Small GTPases (Part F: Ras Family I) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 333. Regulators and Effectors of Small GTPases (Part G: Ras Family II) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 334. Hyperthermophilic Enzymes (Part C) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 335. Flavonoids and Other Polyphenols Edited by LESTER PACKER VOLUME 336. Microbial Growth in Biofilms (Part A: Developmental and Molecular Biological Aspects) Edited by RON J. DOYLE VOLUME 337. Microbial Growth in Biofilms (Part B: Special Environments and Physicochemical Aspects) Edited by RON J. DOYLE VOLUME 338. Nuclear Magnetic Resonance of Biological Macromolecules (Part A) Edited by THOMAS L. JAMES, VOLKER DO¨TSCH, AND ULI SCHMITZ VOLUME 339. Nuclear Magnetic Resonance of Biological Macromolecules (Part B) Edited by THOMAS L. JAMES, VOLKER DO¨TSCH, AND ULI SCHMITZ VOLUME 340. Drug–Nucleic Acid Interactions Edited by JONATHAN B. CHAIRES AND MICHAEL J. WARING VOLUME 341. Ribonucleases (Part A) Edited by ALLEN W. NICHOLSON VOLUME 342. Ribonucleases (Part B) Edited by ALLEN W. NICHOLSON VOLUME 343. G Protein Pathways (Part A: Receptors) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT VOLUME 344. G Protein Pathways (Part B: G Proteins and Their Regulators) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT

Methods in Enzymology

xliii

VOLUME 345. G Protein Pathways (Part C: Effector Mechanisms) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT VOLUME 346. Gene Therapy Methods Edited by M. IAN PHILLIPS VOLUME 347. Protein Sensors and Reactive Oxygen Species (Part A: Selenoproteins and Thioredoxin) Edited by HELMUT SIES AND LESTER PACKER VOLUME 348. Protein Sensors and Reactive Oxygen Species (Part B: Thiol Enzymes and Proteins) Edited by HELMUT SIES AND LESTER PACKER VOLUME 349. Superoxide Dismutase Edited by LESTER PACKER VOLUME 350. Guide to Yeast Genetics and Molecular and Cell Biology (Part B) Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 351. Guide to Yeast Genetics and Molecular and Cell Biology (Part C) Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 352. Redox Cell Biology and Genetics (Part A) Edited by CHANDAN K. SEN AND LESTER PACKER VOLUME 353. Redox Cell Biology and Genetics (Part B) Edited by CHANDAN K. SEN AND LESTER PACKER VOLUME 354. Enzyme Kinetics and Mechanisms (Part F: Detection and Characterization of Enzyme Reaction Intermediates) Edited by DANIEL L. PURICH VOLUME 355. Cumulative Subject Index Volumes 321–354 VOLUME 356. Laser Capture Microscopy and Microdissection Edited by P. MICHAEL CONN VOLUME 357. Cytochrome P450, Part C Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN VOLUME 358. Bacterial Pathogenesis (Part C: Identification, Regulation, and Function of Virulence Factors) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 359. Nitric Oxide (Part D) Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 360. Biophotonics (Part A) Edited by GERARD MARRIOTT AND IAN PARKER VOLUME 361. Biophotonics (Part B) Edited by GERARD MARRIOTT AND IAN PARKER

xliv

Methods in Enzymology

VOLUME 362. Recognition of Carbohydrates in Biological Systems (Part A) Edited by YUAN C. LEE AND REIKO T. LEE VOLUME 363. Recognition of Carbohydrates in Biological Systems (Part B) Edited by YUAN C. LEE AND REIKO T. LEE VOLUME 364. Nuclear Receptors Edited by DAVID W. RUSSELL AND DAVID J. MANGELSDORF VOLUME 365. Differentiation of Embryonic Stem Cells Edited by PAUL M. WASSAUMAN AND GORDON M. KELLER VOLUME 366. Protein Phosphatases Edited by SUSANNE KLUMPP AND JOSEF KRIEGLSTEIN VOLUME 367. Liposomes (Part A) Edited by NEJAT DU¨ZGU¨NES, VOLUME 368. Macromolecular Crystallography (Part C) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 369. Combinational Chemistry (Part B) Edited by GUILLERMO A. MORALES AND BARRY A. BUNIN VOLUME 370. RNA Polymerases and Associated Factors (Part C) Edited by SANKAR L. ADHYA AND SUSAN GARGES VOLUME 371. RNA Polymerases and Associated Factors (Part D) Edited by SANKAR L. ADHYA AND SUSAN GARGES VOLUME 372. Liposomes (Part B) Edited by NEJAT DU¨ZGU¨NES, VOLUME 373. Liposomes (Part C) Edited by NEJAT DU¨ZGU¨NES, VOLUME 374. Macromolecular Crystallography (Part D) Edited by CHARLES W. CARTER, JR., AND ROBERT W. SWEET VOLUME 375. Chromatin and Chromatin Remodeling Enzymes (Part A) Edited by C. DAVID ALLIS AND CARL WU VOLUME 376. Chromatin and Chromatin Remodeling Enzymes (Part B) Edited by C. DAVID ALLIS AND CARL WU VOLUME 377. Chromatin and Chromatin Remodeling Enzymes (Part C) Edited by C. DAVID ALLIS AND CARL WU VOLUME 378. Quinones and Quinone Enzymes (Part A) Edited by HELMUT SIES AND LESTER PACKER VOLUME 379. Energetics of Biological Macromolecules (Part D) Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS VOLUME 380. Energetics of Biological Macromolecules (Part E) Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS

Methods in Enzymology

xlv

VOLUME 381. Oxygen Sensing Edited by CHANDAN K. SEN AND GREGG L. SEMENZA VOLUME 382. Quinones and Quinone Enzymes (Part B) Edited by HELMUT SIES AND LESTER PACKER VOLUME 383. Numerical Computer Methods (Part D) Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 384. Numerical Computer Methods (Part E) Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 385. Imaging in Biological Research (Part A) Edited by P. MICHAEL CONN VOLUME 386. Imaging in Biological Research (Part B) Edited by P. MICHAEL CONN VOLUME 387. Liposomes (Part D) Edited by NEJAT DU¨ZGU¨NES, VOLUME 388. Protein Engineering Edited by DAN E. ROBERTSON AND JOSEPH P. NOEL VOLUME 389. Regulators of G-Protein Signaling (Part A) Edited by DAVID P. SIDEROVSKI VOLUME 390. Regulators of G-Protein Signaling (Part B) Edited by DAVID P. SIDEROVSKI VOLUME 391. Liposomes (Part E) Edited by NEJAT DU¨ZGU¨NES, VOLUME 392. RNA Interference Edited by ENGELKE ROSSI VOLUME 393. Circadian Rhythms Edited by MICHAEL W. YOUNG VOLUME 394. Nuclear Magnetic Resonance of Biological Macromolecules (Part C) Edited by THOMAS L. JAMES VOLUME 395. Producing the Biochemical Data (Part B) Edited by ELIZABETH A. ZIMMER AND ERIC H. ROALSON VOLUME 396. Nitric Oxide (Part E) Edited by LESTER PACKER AND ENRIQUE CADENAS VOLUME 397. Environmental Microbiology Edited by JARED R. LEADBETTER VOLUME 398. Ubiquitin and Protein Degradation (Part A) Edited by RAYMOND J. DESHAIES VOLUME 399. Ubiquitin and Protein Degradation (Part B) Edited by RAYMOND J. DESHAIES

xlvi

Methods in Enzymology

VOLUME 400. Phase II Conjugation Enzymes and Transport Systems Edited by HELMUT SIES AND LESTER PACKER VOLUME 401. Glutathione Transferases and Gamma Glutamyl Transpeptidases Edited by HELMUT SIES AND LESTER PACKER VOLUME 402. Biological Mass Spectrometry Edited by A. L. BURLINGAME VOLUME 403. GTPases Regulating Membrane Targeting and Fusion Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 404. GTPases Regulating Membrane Dynamics Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 405. Mass Spectrometry: Modified Proteins and Glycoconjugates Edited by A. L. BURLINGAME VOLUME 406. Regulators and Effectors of Small GTPases: Rho Family Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 407. Regulators and Effectors of Small GTPases: Ras Family Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 408. DNA Repair (Part A) Edited by JUDITH L. CAMPBELL AND PAUL MODRICH VOLUME 409. DNA Repair (Part B) Edited by JUDITH L. CAMPBELL AND PAUL MODRICH VOLUME 410. DNA Microarrays (Part A: Array Platforms and Web-Bench Protocols) Edited by ALAN KIMMEL AND BRIAN OLIVER VOLUME 411. DNA Microarrays (Part B: Databases and Statistics) Edited by ALAN KIMMEL AND BRIAN OLIVER VOLUME 412. Amyloid, Prions, and Other Protein Aggregates (Part B) Edited by INDU KHETERPAL AND RONALD WETZEL VOLUME 413. Amyloid, Prions, and Other Protein Aggregates (Part C) Edited by INDU KHETERPAL AND RONALD WETZEL VOLUME 414. Measuring Biological Responses with Automated Microscopy Edited by JAMES INGLESE VOLUME 415. Glycobiology Edited by MINORU FUKUDA VOLUME 416. Glycomics Edited by MINORU FUKUDA VOLUME 417. Functional Glycomics Edited by MINORU FUKUDA

Methods in Enzymology

xlvii

VOLUME 418. Embryonic Stem Cells Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 419. Adult Stem Cells Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 420. Stem Cell Tools and Other Experimental Protocols Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 421. Advanced Bacterial Genetics: Use of Transposons and Phage for Genomic Engineering Edited by KELLY T. HUGHES VOLUME 422. Two-Component Signaling Systems, Part A Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 423. Two-Component Signaling Systems, Part B Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 424. RNA Editing Edited by JONATHA M. GOTT VOLUME 425. RNA Modification Edited by JONATHA M. GOTT VOLUME 426. Integrins Edited by DAVID CHERESH VOLUME 427. MicroRNA Methods Edited by JOHN J. ROSSI VOLUME 428. Osmosensing and Osmosignaling Edited by HELMUT SIES AND DIETER HAUSSINGER VOLUME 429. Translation Initiation: Extract Systems and Molecular Genetics Edited by JON LORSCH VOLUME 430. Translation Initiation: Reconstituted Systems and Biophysical Methods Edited by JON LORSCH VOLUME 431. Translation Initiation: Cell Biology, High-Throughput and Chemical-Based Approaches Edited by JON LORSCH VOLUME 432. Lipidomics and Bioactive Lipids: Mass-Spectrometry–Based Lipid Analysis Edited by H. ALEX BROWN VOLUME 433. Lipidomics and Bioactive Lipids: Specialized Analytical Methods and Lipids in Disease Edited by H. ALEX BROWN

xlviii

Methods in Enzymology

VOLUME 434. Lipidomics and Bioactive Lipids: Lipids and Cell Signaling Edited by H. ALEX BROWN VOLUME 435. Oxygen Biology and Hypoxia Edited by HELMUT SIES AND BERNHARD BRU¨NE VOLUME 436. Globins and Other Nitric Oxide-Reactive Protiens (Part A) Edited by ROBERT K. POOLE VOLUME 437. Globins and Other Nitric Oxide-Reactive Protiens (Part B) Edited by ROBERT K. POOLE VOLUME 438. Small GTPases in Disease (Part A) Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 439. Small GTPases in Disease (Part B) Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 440. Nitric Oxide, Part F Oxidative and Nitrosative Stress in Redox Regulation of Cell Signaling Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 441. Nitric Oxide, Part G Oxidative and Nitrosative Stress in Redox Regulation of Cell Signaling Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 442. Programmed Cell Death, General Principles for Studying Cell Death (Part A) Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN, AND MAURO PIACENTINI VOLUME 443. Angiogenesis: In Vitro Systems Edited by DAVID A. CHERESH VOLUME 444. Angiogenesis: In Vivo Systems (Part A) Edited by DAVID A. CHERESH VOLUME 445. Angiogenesis: In Vivo Systems (Part B) Edited by DAVID A. CHERESH VOLUME 446. Programmed Cell Death, The Biology and Therapeutic Implications of Cell Death (Part B) Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN, AND MAURO PIACENTINI VOLUME 447. RNA Turnover in Bacteria, Archaea and Organelles Edited by LYNNE E. MAQUAT AND CECILIA M. ARRAIANO VOLUME 448. RNA Turnover in Eukaryotes: Nucleases, Pathways and Analysis of mRNA Decay Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN

Methods in Enzymology

xlix

VOLUME 449. RNA Turnover in Eukaryotes: Analysis of Specialized and Quality Control RNA Decay Pathways Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN VOLUME 450. Fluorescence Spectroscopy Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 451. Autophagy: Lower Eukaryotes and Non-Mammalian Systems (Part A) Edited by DANIEL J. KLIONSKY VOLUME 452. Autophagy in Mammalian Systems (Part B) Edited by DANIEL J. KLIONSKY VOLUME 453. Autophagy in Disease and Clinical Applications (Part C) Edited by DANIEL J. KLIONSKY VOLUME 454. Computer Methods (Part A) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 455. Biothermodynamics (Part A) Edited by MICHAEL L. JOHNSON, JO M. HOLT, AND GARY K. ACKERS (RETIRED) VOLUME 456. Mitochondrial Function, Part A: Mitochondrial Electron Transport Complexes and Reactive Oxygen Species Edited by WILLIAM S. ALLISON AND IMMO E. SCHEFFLER VOLUME 457. Mitochondrial Function, Part B: Mitochondrial Protein Kinases, Protein Phosphatases and Mitochondrial Diseases Edited by WILLIAM S. ALLISON AND ANNE N. MURPHY VOLUME 458. Complex Enzymes in Microbial Natural Product Biosynthesis, Part A: Overview Articles and Peptides Edited by DAVID A. HOPWOOD VOLUME 459. Complex Enzymes in Microbial Natural Product Biosynthesis, Part B: Polyketides, Aminocoumarins and Carbohydrates Edited by DAVID A. HOPWOOD VOLUME 460. Chemokines, Part A Edited by TRACY M. HANDEL AND DAMON J. HAMEL VOLUME 461. Chemokines, Part B Edited by TRACY M. HANDEL AND DAMON J. HAMEL VOLUME 462. Non-Natural Amino Acids Edited by TOM W. MUIR AND JOHN N. ABELSON VOLUME 463. Guide to Protein Purification, 2nd Edition Edited by RICHARD R. BURGESS AND MURRAY P. DEUTSCHER VOLUME 464. Liposomes, Part F Edited by NEJAT DU¨ZGU¨NES,

l

Methods in Enzymology

VOLUME 465. Liposomes, Part G Edited by NEJAT DU¨ZGU¨NES, VOLUME 466. Biothermodynamics, Part B Edited by MICHAEL L. JOHNSON, GARY K. ACKERS, AND JO M. HOLT VOLUME 467. Computer Methods Part B Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 468. Biophysical, Chemical, and Functional Probes of RNA Structure, Interactions and Folding: Part A Edited by DANIEL HERSCHLAG VOLUME 469. Biophysical, Chemical, and Functional Probes of RNA Structure, Interactions and Folding: Part B Edited by DANIEL HERSCHLAG VOLUME 470. Guide to Yeast Genetics: Functional Genomics, Proteomics, and Other Systems Analysis, 2nd Edition Edited by GERALD FINK, JONATHAN WEISSMAN, AND CHRISTINE GUTHRIE VOLUME 471. Two-Component Signaling Systems, Part C Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 472. Single Molecule Tools, Part A: Fluorescence Based Approaches Edited by NILS G. WALTER

C H A P T E R

O N E

Star Polymer Surface Passivation for Single-Molecule Detection ¨rgen Groll and Martin Moeller Ju Contents 2 2 4

1. 2. 3. 4.

Introduction Surface Grafting of PEO and Protein Repellence The NCO–sP(EO-stat-PO) System Preparation of sP(EO-stat-PO)-Coated Substrates for Single-Molecule Experiments 5. Analysis of Protein Structure and Function on NCO–sP(EO-stat-PO) Surfaces Acknowledgments References

6 11 16 16

Abstract Poly(ethylene oxide) (PEO) is known as an excellent coating material to minimize nonspecific protein adsorption. For an examination of biomolecules attached to surfaces with sensitivities down to the single-molecule level, demands on the surface additionally comprise low-intrinsic fluorescence of the coating material and a possibility to immobilize biomolecules in their functional conformation. One strategy that combines the protein-resistant properties of PEO with chemical functionality is the use of star-shaped PEOs that allow for interpolymer crosslinking. Our system consists of six-arm PEO-based star polymers functionalized with reactive isocyanate groups at the ends of the polymer chain. The isocyante groups allow intermolecular cross-linking so that high grafting densities may be achieved, which render the surfaces extremely resistant to protein adsorption. Application by spin coating offers a simple procedure for the preparation of minimally interacting surfaces. The reactive end groups may be further biofunctionalized to recognize specific biomolecules such as streptavidin or His-tagged proteins in specific geometries or as single isolated molecules. These properties, together with the advantageous chemical properties of PEO, render the surfaces ideal for immobilizing proteins with detection limits down to the single molecule level. This chapter focuses on the preparation of substrates that DWI e.V. and Institute of Technical and Macromolecular Chemistry, RWTH Aachen University, Aachen, Germany Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72019-X

#

2010 Elsevier Inc. All rights reserved.

1

2

¨rgen Groll and Martin Moeller Ju

are suitable for single-molecule experiments. Besides a detailed description of surface preparation, two examples for the single-molecule detection of immobilized proteins, nucleosomes and RNase H, are presented that demonstrate the advantages of the star-polymer derived coatings over lineargrafted PEO.

1. Introduction Poly(ethylene oxide) (PEO) is a hydrophilic uncharged polymer that has been recognized as particularly efficient for achieving protein-resistant surfaces (Harris and Zalipsky, 1997). While most studies deal with the grafting of linear PEO chains to the material surface, changes in the molecular architecture of the polymer are beneficial for higher polymer segment density on the surface and a higher density of end-group-functionalities. In this chapter, we focus on star-shaped PEO molecules and especially on their use for the preparation of biofunctional surfaces that can be used for single-molecule experiments. Layers consisting of cross-linked, end-functionalized, six-armed, PEO-based star molecules were found to be as good as or better than linear-PEO-modified surfaces at reducing protein adsorption (Groll et al., 2004, 2005a). In contrast to the effort that is necessary to obtain surfaces with high grafting density by the surface grafting of linear PEO, such films can be prepared by simple spin or dip coating from aqueous solution, since chemical cross-linking ensures a high polymer segment density on the substrate (Gasteier et al., 2007). Since the reactive end-groups can be used for further functionalization, particular attention is paid to the ability to generate biofunctional films by a one-step preparation method through spin-casting. The experimental procedure for substrate preparation will be discussed in detail, and examples are given for single-molecule experiments demonstrating the negligible nonspecific interaction of proteins with the surface and the unperturbed function of proteins that are immobilized on such coatings.

2. Surface Grafting of PEO and Protein Repellence The general strategy to render surfaces inert toward proteins is to introduce a coating layer that prevents protein adsorption either thermodynamically, so that attractive surface interactions are overcompensated by repulsive interactions with the layer, or at least kinetically by creating a free energy barrier of sufficient height that cannot be overcome on relevant time scales (Halperin, 1999). An abundance of studies have shown that PEO-coated surfaces display exceptional protein resistance (Gasteier et al.,

Star Polymer Surface Passivation for Single-Molecule Detection

3

2007; Harder et al., 1998; Harris and Zalipsky, 1997; Kingshott et al., 2002; Malmsten et al., 1998; McPherson et al., 1998; Sofia et al., 1998; Unsworth et al., 2008). On substrates such as glass or silicon, protein-repellant coatings are frequently made from long, randomly coiling linear PEO chains terminally anchored to the surface. Only in the brush regime, the grafting density of linear chains is high enough that the attached polymer chains stretch out perpendicularly to the surface, thus avoiding unfavorable monomer–monomer interactions and maintaining optimal solvation. Consequently, the grafted chains provide adequate coverage and thickness to form a very effective steric barrier against protein adsorption (Szleifer and Carignano, 2000; Yang et al., 1999). Grafting density and chain length are thus the two essential experimental control parameters by which the degree of protein resistance is governed (Malmsten et al., 1998; McPherson et al., 1998; Unsworth et al., 2006). It has recently been shown that, for molecular weights between 600 and 2000 g/mol, a grafting density of 0.5 linear OH-terminal PEO chains/nm2 is the threshold for minimal protein adsorption (Unsworth et al., 2008). Star-shaped PEO molecules (star PEO) have a central core region from which the PEO arms extend. Due to this constraint, their density is higher than that of a linear chain, which offers the opportunity to produce PEO surfaces with higher grafting density (Douglas et al., 1990; Sofia et al., 1998). Moreover, the ends of the arms are preferentially located near the periphery due to the steric constraints in the interior of the star (Irvine et al., 1996). Therefore, the probability is increased for end-functionalized groups to bind to the surface. Star PEO systems thus appear an attractive choice to confer protein resistance to surfaces. Indeed, star PEO with 70 arms and a molecular weight of 5200 g/mol per arm have been reported to pack closely on the surface and to efficiently reduce protein adsorption, although the efficiency is reduced for small proteins such as cytochrome c (Sofia et al., 1998). In another study, star PEO with 24 arms and a molecular weight of 9700 g/mol per arm as well as star PEO with 72 arms and a molecular weight of 4500 g/mol per arm were surface grafted (Irvine et al., 1998). Atomic force microscopy and reflectivity measurements show that the hydrated star molecules are overlapping, and since the star segments are depleted near the substrate, the authors explain the residual adsorption by the diffusion of small proteins such as the 12,000 g/mol cytochrome c through the low PEO density seams between molecules and subsequent surface adsorption. As a result, surfaces obtained by grafting the smaller star PEO molecules prevent protein adsorption better since their packing on the surface is denser. By contrast, due to the spherical shape of the highmolecular-weight PEO stars, gaps remain between the stars that appear sufficiently large for small proteins to reach the surface and adsorb to it. From these protein-adsorption properties of surfaces obtained by the grafting of PEO stars with different sizes, two strategies become clear and these have to be followed in order to improve star PEO-derived surface

4

¨rgen Groll and Martin Moeller Ju

coatings. First, both the number and molecular weight of the arms have to be reduced. Smaller stars seem better for the modification of surfaces since their packing is denser on the surface. In addition, a more flexible core than the poly-divinylbenzene core used in the aforementioned studies would make the stars more flexible and, in combination with a smaller number of arms, less spherical. A second approach is to functionalize the stars with reactive groups at the ends of the arms that enable intermolecular crosslinking of the PEO stars. Covalently cross-linked stars on the surface will result in a more homogeneous lateral PEO density profile and higher surface coverage. Moreover, cross-linking would allow for a more variable control of the layer thickness and the preparation of layers that are thicker than monolayers, since the end-groups are not restricted to lateral reactions. A further advantage concerns the functionalization of stars with molecules that react with the end-groups. The spatial distribution of such molecules in the layer can be controlled on the nanoscale by attaching them to the stars before or after surface grafting (Maheshwari et al., 2000). In this way, either a random distribution or a nanoclustered ligand pattern can be achieved.

3. The NCO–sP(EO-stat-PO) System We followed both strategies mentioned in the last paragraph for the development of our six-arm, star-shaped molecules with terminal reactive functional groups. The backbone consists of a statistical copolymer of ethylene oxide and propylene oxide in a ratio of 4:1; each arm has a molecular mass of 2000 g/mol with low polydispersity and is attached to a sorbitol core (sP(EO-stat-PO)). The arms of the star molecules are initially terminated with OH and can be functionalized with various reagents to yield molecules with different reactivities. This chapter will concentrate on isocyanate (NCO–) terminated star molecules (NCO–sP(EO-stat-PO)) that are obtained through functionalization with isophorone diisocyanate (IPDI; Goetz et al., 2002). As IPDI is chemically attached to the star molecules without a catalyst, the primary, less reactive aliphatic isocyanate groups remain as functional groups at the ends of the arms, so that the NCO–sP (EO-stat-PO) molecules can be dissolved in water and coatings can be applied from aqueous solutions. When the NCO–sP(EO-stat-PO) material is dissolved in water, hydrolysis of the isocyanate groups lead to the formation of carbaminic acid which, at neutral pH, instantly decarboxylates to form amine groups. These amines react with unreacted isocyanate groups to form urea bridges between the NCO–sP(EO-stat-PO) molecules. Since the kinetics of amine addition to isocyanate is much faster than hydrolysis (Caraculacu and Coseri, 2001), urea bridge formation occurs preferentially until steric restrictions significantly lower the reaction probability.

5

Star Polymer Surface Passivation for Single-Molecule Detection

The aqueous NCO–sP(EO-stat-PO) solution can be used for coating surfaces either by simple dip-, spin-, or spray-coating. However, due to the ongoing hydrolysis and aminolysis of isocyanate groups, a time window of maximum 20 min after the addition of water to NCO–sP(EO-stat-PO) should not be exceeded. After coating, the system requires at least 12 h for completion of the cross-linking reaction within the layer. During this time, all isocyanate groups hydrolyze and then either react with other isocyanate groups to form urea bridges or remain as free amino groups that can be further functionalized (Groll et al., 2005a,b,c). Figure 1.1 shows the chemical reactions that occur during film formation and presents a model of the resulting surface coatings. One particular advantage of the NCO–sP(EO-stat-PO) system is the change in reactivity from isocyanate groups, which are reactive toward nucleophilic groups, such as alcohols, amines, and thiols, to amine groups during the layer preparation and curing of the coating. Addition of watersoluble compounds that bear nucleophilic groups to the aqueous NCO–sP (EO-stat-PO) solution before coating thus results in covalent attachment of these molecules to the coating. It is important to emphasize that, after A NCO-sP(EO-stat-PO) EO-stat-PO RO

4 OR

R=

O m

IPDI

O O

n

NH

NCO

OR Mn = 2 kDa/arm; m = 0.8; n = 0.2 Cross-linking:

=

R

NH2

R2

NCO + H2N R

R2

-

NCO + H2O

-

R

O

N-C-N H H

R

B

Figure 1.1 Surface coating with star polymers. (A) Schematic of the chemical composition of the NCO–sP(EO-stat-PO) system and the cross-linking reaction induced by water. (B) Schematic of a cross-linked sP(EO-stat-PO) surface coating where all isocyanate groups are hydrolyzed or aminolyzed resulting in urea bridges between the stars (yellow) and free amino groups (green). Figure partially reprinted with permission from Heyes et al. (2007), copyright Royal Society of Chemistry.

6

¨rgen Groll and Martin Moeller Ju

complete hydrolysis of the isocyanate groups, the compounds that are covalently bound to the layers are embedded in a coating that inhibits nonspecific interactions with proteins and cells, so that the immobilized molecules can interact specifically. This feature is achieved in a one-step layer preparation without the use of further chemical blocking-agents. The remainder of this chapter will focus on the preparation of substrates for single-molecule experiments according to this layer preparation method; a broader overview of different strategies for functionalization and application of the NCO–sP(EO-stat-PO) system for cell culture and biomaterials has recently been given elsewhere (Gasteier et al., 2007).

4. Preparation of sP(EO-stat-PO)-Coated Substrates for Single-Molecule Experiments Fluorescence is among the most sensitive analytical tools available and has been extensively applied to study biomolecules. In order to achieve single-molecule detection on surfaces, it is of utmost importance that the substrates have extremely low background fluorescence. Consider that a standard wide field fluorescent microscope has a point spread function approximately half the wavelength of the emitted light, which is
300 nm for visible fluorophores. The fluorescent label has a diameter of
1 nm. Therefore, in order to observe a single molecule, the number of emitted photons from the fluorophore covering approximately a 1  1 nm2 area must be significantly higher than the integrated background fluorescence over the entire 300  300 nm2 area. This stringent requirement necessitates that all preparation steps are carried out either in a clean room, or under a laminar flow hood with an air-filtration system. Glass slides, chemicals, solvents, and especially, the aminofunctional material used to functionalize the substrate and the PEO polymers, must be of the highest purity possible. The substrates are prepared according to the following steps:   

Ultrasonication in water for 2 min. Ultrasonication in isopropanol for 2 min. Cleaning and activation in a plasma oven or by UV/ozone treatment.

We have used different plasma machines, for example, the TePla 100-E system with 100 W at a process gas pressure of 0.5 mbar, with either oxygen or filtered air as flow gas, and process times between 10 and 15 min. For UV/ozone treatment, we have used a home-made device using a 40 W UV lamp (main emission 185 nm; UV-Technik Speziallampen GmbH) in an oxygen stream of 350 ml/min with a sample distance of 5 mm to the lamp for 10–15 min. Both processes result in clean substrates that were suitable for single-molecule experiments.

Star Polymer Surface Passivation for Single-Molecule Detection

7

To covalently bind NCO–sP(EO-stat-PO) to the substrate, the surface must be functionalized with isocyanate-reactive groups, such as alcohols, amines, or thiols, as anchor points. Since the NCO–sP(EO-stat-PO) molecules cross-link on the substrates, the system tolerates a much lower number of functional groups on the surface than other systems that use classical grafting techniques. Furthermore, the cross-linking also results in a certain stability of the film against desorption from the surface in aqueous solution even without covalent attachment. For contact times with aqueous solutions of less than 10 h, films are sufficiently stable and may simply be coated on cleaned glass substrates. If covalent attachment is necessary, amino functionalization may be achieved by a variety of methods, for example, electrostatic adhesion of polyamines such as poly-L-lysine, chemical vapor deposition of 4-amino [2,20 ]-paracyclophanes, or aminosilanization. We have used aminosilanization according to the following protocol:   

Cleaning and activation of the substrates as described earlier. Transfer of the substrates into a glove box under nitrogen atmosphere. Immersion into a solution of 3 ml N-[3-(trimethoxysilyl)propyl] ethylenediamine in 50 ml dry toluene for 2 h. Aminopropylsilane (APTES) may be used alternatively.  Repeated rinsing of the substrates with dry toluene.  Storage of the substrates in dry toluene until further use. Alternatively, the commercial reagent VectabondTM from Vector Labs may be used according to the detailed protocol provided by the supplier. While this method can be performed without the need of a glove box, it is important to use a new, unopened bottle of Vectabond each time, as it does not stay clean for long after it has been opened. Spin-coating of aqueous NCO–sP(EO-stat-PO) solutions provides precise control of the layer thickness via rotation speed and prepolymer concentration and results in homogeneous layers (Groll et al., 2005b). NCO–sP(EO-stat-PO) is preweighed in portions of typically 50 mg in nitrogen atmosphere and provided in airtight glass vials. For spin-coating, anhydrous THF is added to the vial to predissolve the prepolymer, so that upon the addition of water, a homogeneous mixture is formed immediately. Typically, THF is added to result in solutions with a concentration of 10–20 mg/ml NCO–sP(EO-stat-PO). Then, water is added until the aqueous solution is diluted to typically 1 mg/ml NCO–sP(EO-stat-PO). After gentle shaking to ensure homogeneous mixing, the solution is left to react for exactly 5 min. After this time, the solution is slightly opaque. In the meantime, the aminosilanized glass coverslip is placed on the spin coater. The solution is drawn up into a syringe, and a 0.02-mm membrane filter (Whatman Anotop 10) is attached to the syringe Luer connection. The solution is pressed through the filter directly onto the amino functional coverslip placed on the spin coater. The whole coverslip is homogeneously

8

¨rgen Groll and Martin Moeller Ju

covered with solution (
0.5 ml is needed per 24  24 mm coverslip). It is important not to touch the coverslip with the syringe as this may cause problems with the homogeneity of the spin-coating. The spin coater is turned on and the substrate rotated at 2500 rpm for 45 s. After that, the coverslip is placed into a glass petri dish that has been cleaned either by plasma- or UV/ozone treatment. The petri dish is covered, sealed with parafilm, and stored at ambient conditions over night for cross-linking of the surface. Figure 1.2 presents a flow diagram for the layer preparation

1

2

3

4

5

Figure 1.2 Flowchart of the coating procedure for generation of NCO–sP(EO-statPO)-coated slides. Glass microscopy slides are cleaned by ultrasonication, activated either by oxygen plasma or UV/ozone treatment and, if long-term stability of the films on the substrates in aqueous conditions is required, subsequently aminofunctionalized (1). These substrates are fixed on a spin coater (2), ideally in dust-free or (not of) dust-reduced conditions, and dropwise covered with an aqueous solution of NCO–sP (EO-stat-PO) 5 min after dissolution of the star-shaped prepolymers in a typical concentration of 1 mg/ml. The solution is dropped onto the substrate from a syringe and filtered through 0.02 mm syringe filters (3). When the substrate is completely covered by the solution, spin-casting is initiated (4) with 2500 rpm and left rotating for 45 s. Afterward, the samples are carefully taken off the spin-coater using a teflon tweezer (5) and stored for 24 h in clean glass petri dishes under ambient conditions (ideally dust free or dust reduced) for cross-linking of the coating.

Star Polymer Surface Passivation for Single-Molecule Detection

9

procedure. As hydrolysis and subsequent cross-linking between the starmolecules continuously proceed in solution, only a limited number of maximally five substrates can be produced from each solution. The remaining aqueous NCO–sP(EO-stat-PO) solution has to be discarded. It was shown using scanning confocal fluorescence microscopy with single-molecule detection sensitivity that the background fluorescence of cross-linked sP (EO-stat-PO) surfaces carefully prepared according to this protocol is low enough to detect the fluorescence of single molecules (Groll et al., 2004). The nonfouling (passivation) properties of sP(EO-stat-PO) coatings were compared to surfaces modified with linear methoxy-terminated PEO chains and BSA (Groll et al., 2004; Koopmans et al., 2008). In both studies, the sP (EO-stat-PO) films exhibited superior nonfouling properties (Fig. 1.3). For binding of proteins to such cross-linked sP(EO-stat-PO) films, ligands for specific immobilization have to be covalently linked to the polymer. One often-used coupling-system of proteins to surfaces relies on the complex formation between biotin and streptavidin. Streptavidin recognizes biotin with high specificity and affinity, binding with a Kd of
10 15 M. Streptavidin is tetravalent toward biotin and is able to subsequently bind a biotinylated protein of interest. Biotinylated surfaces may be prepared by the addition of biocytin, a biotin derivative that contains a free amine group, to the aqueous NCO–sP(EO-stat-PO) solution prior to spin coating. Experimentally, only one variation of the layer preparation protocol above is necessary. Instead of adding pure water to the NCO–sP(EOstat-PO) material dissolved in THF, an aqueous solution of biocytin is added. During the 5- min interval between mixing and spin-casting, the amine group of biocytin reacts with the isocyanate groups of the polymer. After spin-casting, the covalently linked biotin moieties are homogeneously distributed in the polymer film. The amount of biocytin that is dissolved in water determines the degree of layer functionality. Typically, 2 mg of biocytin are used per 50 mg NCO–sP(EO-stat-PO) prepolymer. In order to perform single-molecule measurements on such substrates, first streptavidin has to be bound to the biotin, followed by binding of the biotinylated and fluorescently labeled protein to the streptavidin. All these steps should be performed in aqueous solution, so that sequential flushing of solutions has to be enabled. This can be achieved by constructing substrate ‘‘sandwiches’’ following the procedure schematically shown in Fig. 1.4. One 32  24 mm coverlip and one 20  20 mm coverslip are required for each sandwich. At least one of the substrates has to be coated with a biotinylated cross-linked sP(EO-stat-PO) film, whereas the second one may be coated with a nonfunctionalized sP(EO-stat-PO) layer. The sandwiches are prepared by taking the large coverslip from the petri dish and placing two pieces of double-sided tape on the upside to leave a 2–3 mm channel in the center. Then, the small coverslip is placed on top of the tape. Care should be taken to ensure that the polymer-coated upside now faces

10

A

¨rgen Groll and Martin Moeller Ju

RNase H 1000 Background

Density of spots (mm−2)

100

Unspecific adsorption

10

1

0.1

0.01

sP(EO-stat-PO) B 100

Linear PEO5000

BSA

Nucleosomes sP(EO-stat-PO)

100

Linear PEO5000

100

75

75

75

50

50

50

25 0

5 mm

25 0

5 mm

25

BSA

5 mm

0

Figure 1.3 Quality of surface passivation. Nonspecific protein adsorption on crosslinked sP(EO-stat-PO) films is compared to that on grafted linear PEO5000 and physisorbed BSA using single molecules of Alexa Fluor 546 labeled RNase H (A, density of spots) and Cy3-ATTO647N labeled nucleosomes (B, fluorescence images and normalized fluorescence signal intensities). Significantly higher levels of nonspecific adsorption were observed on BSA than on PEO surfaces. sP(EO-stat-PO) surfaces showed negligible nonspecific adsorption. Figure partially reprinted with permission from Groll et al. (2004), copyright American Chemical Society and Koopmans et al. (2008), copyright Wiley-VCH Verlag GmbH& Co. KGaA.

down into the channel. The tape is cut along the edges of the coverslip and the sandwich is placed in a clean, sealed glass petri dish to transport to the microscope. For the experiments, the streptavidin solution (typically 20 mg/ml) and biotinylated/labeled protein (typically 200 pM) are flushed through the channel by adding a droplet of solution to one open side of the channel and waiting for 10 min each time before the removal of the droplet and addition of the next one.

11

Star Polymer Surface Passivation for Single-Molecule Detection

A

1 B

C

2

4

3

5

Figure 1.4 Schematic of substrate preparation for single-molecule experiments. For optimal results, preparation should be carried out under clean room conditions. The substrate-sandwiches are prepared by taking the large coverslip (32 mm  24 mm; (1)) from the petri dish, and placing two pieces of double-sided tape (2) on the upside to leave a 2–3 mm channel in the center (A). Then, the small coverslip (20  20 mm; (3)) is placed on top of the tape with the coated side facing down toward the big coverslip (B). The tape is cut along the edges of the coverslip, and the sample can be placed in a clean, sealed glass petri dish to transport to the microscope. The channel between the two coated coverslips (4) allows for repeated flushing with solutions by adding a droplet at the edge of the small glass slide (5) and waiting for 10 min.

5. Analysis of Protein Structure and Function on NCO–sP(EO-stat-PO) Surfaces To take advantage of the applications of surface-immobilized proteins, for example, as biosensors, it is necessary to achieve: (i) high preference of specific adsorption over nonspecific adsorption, and (ii) binding of the protein in its native, functional structure. Ideally, the protein should completely refold into its native form if it is temporarily exposed to denaturing conditions. Maintenance of the folded state of the protein and its ability to refold after denaturation can be addressed using single-molecule fluorescence (or Fo¨rster) resonance energy transfer (FRET). More specifically, the labeling of proteins with two spectrally different fluorophores enables experiments in which the fluorophore with the higher excitation energy, the donor, is selectively excited. The energy is then nonradiatively transferred via dipole–dipole coupling to the lower energy fluorophore, the acceptor (Fo¨rster, 1948; Stryer and Haugland, 1967). Since the efficiency of this process is proportional to

12

¨rgen Groll and Martin Moeller Ju

R 6, where R is the distance between the two chromophores, structural information of the protein may be inferred from the FRET efficiency. In order to study single-molecule FRET of surface-immobilized molecules, scanning confocal microscopy or total internal reflection microscopy may be used to limit background fluorescence. A schematic representation on how FRET is used to infer the structural details of immobilized biomolecules is presented in Fig. 1.5. The surface is coated with a protein-resistant material, such as linear PEO or crosslinked sP(EO-stat-PO), which contains a bioactive group such as biotin (represented as a red antenna). In order to compare the cross-linked sP (EO-stat-PO) with other surface preparation techniques, linear-PEOcoated surfaces were prepared by grafting linear PEO (MW ¼ 5000 Da) with amine-reactive NHS end-groups from aqueous solution, of which a small fraction (
1%) was also functionalized with biotin. Cross-linked sP (EO-stat-PO) surfaces were prepared as described earlier in a convenient, single-step layer preparation. Subsequent addition of streptavidin to these surfaces and then of biotinylated, FRET-labeled protein as described earlier allows specific immobilization of the protein. Maintenance of a functional protein structure can be followed through the FRET efficiency. Upon structural changes and unfolding of the peptide chain, the distance between the chromophores increases and the efficiency of energy transfer decreases so that an increased fluorescence of the donor dye can be detected. By this method, the binding specificity of biotinylated and FRET-labeled nucleosomes to three different surfaces and especially, the structural integrity of the surface-bound nucleosomes, were measured (Koopmans et al., 2008). On biotinylated BSA, binding specificity was only 2%, and only 28% of the immobilized nucleosomes retained structural integrity. On linear PEO5000, the values increased to 60% and 53%, respectively. Cross-linked sP(EO-stat-PO) films, by contrast, resulted in 90% specificity of binding, and 78% of the immobilized nucleosomes were bound in their intact structure. Spontaneous unwrapping of nucleosomal DNA, so called nucleosome breathing, was measured on the surfaceimmobilized nucleosomes and the dynamics of the process compared to that in solution. The lifetimes of both the closed and the open states are approximately fivefold faster on cross-linked sP(EO-stat-PO) films than the dynamics observed on the linear-PEO-coated surface (1.5 s closed state, 120 ms open state), and they perfectly agree with the breathing kinetics of nucleosomes in solution (Li et al., 2005). Thus, nucleosomes can be specifically immobilized on sP(EO-stat-PO) coatings while maintaining their structural integrity and their dynamic nature. In order to study the interaction of specifically bound proteins with the NCO–sP(EO-stat-PO) surface and ask whether immobilized proteins refold into their functional structure after temporary denaturation, biotinylated RNase H was labeled with Alexa Fluor 546 and Alexa Fluor 647 and

13

Star Polymer Surface Passivation for Single-Molecule Detection

Alexa546

FRET

Alexa647

h•n h•n

h•n

h•n

Unfolding Refolding

6 M GdmCl

Buffer

sP(EO-stat-PO)

Buffer

0.5 1.0 0.0 0.5 1.0 0.0 0.5 1.0 FRET efficiency FRET efficiency FRET efficiency

0.0

0.5 1.0 0.0 0.5 1.0 0.0 0.5 1.0 FRET efficiency FRET efficiency FRET efficiency

Linear PEO

0.0

Figure 1.5 Schematic illustration of the FRET technique used to monitor structural information of immobilized proteins. In the folded state, the dye molecules, which are placed at specific sites on the protein, are close together. Upon excitation of the donor dye (green), a high transfer efficiency of the energy to the acceptor dye (red) occurs. If the protein is unfolded, the distance between the dyes increases and consequently, the energy transfer efficiency decreases. The distance dependence of the FRET efficiency can be used to infer details of the protein structure. The pictures show example scanning confocal fluorescence microscopy images of RNase H proteins immobilized on NCO–sP(EO-stat-PO)-coated surfaces and the resulting single-molecule FRET histograms of RNase H immobilized on NCO–sP(EO-stat-PO) and linear-PEO5000coated surfaces under buffer conditions, 6 M guanidinium chloride (GDmCl) and then subsequently reimmersed in buffer. The green fraction centered at zero FRET efficiency represents molecules that have no acceptor molecule due to incomplete labeling or dye photobleaching. The red fraction at high (
0.9)-FRET efficiency are molecules in their folded state. The yellow fraction at low-to-intermediate FRET efficiency (
0.4) represents molecules in their unfolded state. Figure partially reprinted with permission from Groll et al. (2004), copyright American Chemical Society.

14

¨rgen Groll and Martin Moeller Ju

bound to the biotinylated NCO–sP(EO-stat-PO) layers via streptavidin. The FRET dye pair was chosen and placed at sites on the protein so that a high-FRET efficiency is observed as long as the protein maintains its folded state. If the protein loses its compact-folded structure, the average distance between the dyes increases and the FRET efficiency is reduced. RNase H was labeled with Alexa Fluor 546 and Alexa Fluor 647 (Molecular Probes, ˚ in buffer) at amino acids 3 and 135, respectively Eugene, OR, R0 ¼ 66 A (Kuzmenkina et al., 2006). As a result, when unfolding of the protein chain is induced, the distance between the dyes increases and the donor fluorescence becomes detectable. If the peptide chain is able to refold to its initial conformation, the energy is again transferred to the acceptor dye and the fluorescence of this chromophore dominates. Exemplary scanning confocal fluorescence microscopy images of single, specifically immobilized, FRET-labeled RNase H on cross-linked sP(EO-stat-PO) surfaces are shown in Fig. 1.5, together with histograms of the calculated FRET efficiency of many single molecules on both linear PEO and cross-linked sP (EO-stat-PO) surfaces. The histograms in Fig. 1.5 are colored, based on their FRET efficiency ranges, to indicate proteins that are folded (red) or unfolded (yellow) or contain no acceptor dye and thus, cannot be structurally interrogated (green). For RNase H immobilized on linear PEO surfaces under buffer conditions, it is evident that there is a large distribution of FRET efficiencies from molecule to molecule, indicating that the proteins adopt a wide range of structures (many of them in a low-to-intermediate (unfolded) FRET state), while on cross-linked sP (EO-stat-PO) surfaces, practically all molecules that contain both dyes are in a high-FRET state, indicating that the folded state of RNase H is maintained for most molecules (Amirgoulova et al., 2004; Groll et al. 2004). Upon exposure of immobilized RNase H on either linear or crosslinked sP(EO-stat-PO) surfaces to high concentrations of guanidinium chloride (GdmCl), a reduction in the FRET efficiency was observed, indicating that the molecules unfold. Upon exchanging the denaturant for normal buffer conditions once again, practically all the RNase H proteins immobilized on cross-linked sP(EO-stat-PO) surfaces were able to refold to their compact high-FRET state. In contrast, on the linear PEO surfaces, once the protein molecules completely unfolded, practically none were able to refold to their high-FRET folded state. Moreover, on the cross-linked sP (EO-stat-PO) surfaces, the unfolding–refolding of RNase H molecules was found to be completely reversible over at least 50 cycles of adding and removing GdmCl (Amirgoulova et al., 2004). In Fig. 1.6, the FRET state of the same single molecules before and after 50 cycles of adding and removing 6 M GdmCl was followed. The ability of the molecules to fold to the same high-FRET state after multiple denaturation–renaturation cycles is suggestive that the proteins are in or at least close to their native state. Further evidence stems from whether the protein is still able to function on the

15

Star Polymer Surface Passivation for Single-Molecule Detection

B

A

C DNA RNA

Dye quenched Q Quencher

D

E

Fluorescence (a.u.)

Folded on sP(EO-stat-PO) Refolded on sP(EO-stat-PO) Enzyme in solution 0

100 200 Time (s)

300

0

20

40

60

ka/s-1mM-1

Figure 1.6 Scanning confocal fluorescence microscopy image of RNase H molecules immobilized on cross-linked sP(EO-stat-PO) surfaces before (A) and after (B) 50 cycles of unfolding–refolding. The same molecules that were folded before the 50 cycles were shown to completely refold after, and are highlighted in both figures. There are always a fraction of molecules that have no acceptor dye but have a donor dye. These green molecules are ignored as we cannot infer structural information from them. In the right image, some of the molecules that were visible in the left frame are no longer visible in the right frame, due to photobleaching during the scanning. Panel (C) shows a schematic representation of the enzymatic assay of RNase H immobilized on cross-linked sP(EOstat-PO) surfaces. A fluorescently-quenched DNA–RNA hybrid dissociates upon RNA cleavage by the RNase H. (D) Increase in fluorescence of the dye following RNA cleavage with time upon exposure to RNase H immobilized on cross-linked sP(EO-stat-PO) both before and after an unfolding–refolding cycle. The refolding curve has been offset vertically for presentation purposes. (E) Calculated activity coefficients of the RNase H in solution, upon immobilization on cross-linked sP(EO-stat-PO), and following an unfolding– refolding cycle on the cross-linked sP(EO-stat-PO). Figure partially reprinted with permission from Heyes et al., 2007, copyright Royal Society of Chemistry.

16

¨rgen Groll and Martin Moeller Ju

surface. The function of RNase H is to cleave RNA–DNA hybrids (Kanaya, 1998). An enzymatic assay for RNase H was previously developed that uses a fluorescently labeled RNA–DNA construct (Hogrefe et al., 1990). The hybrid contains a fluorescent dye and a quencher, so that it does not fluoresce in its uncleaved state. Once exposed to RNase H, the RNA is cleaved and the DNA–RNA hybrid dissociates. This cleavage separates the quencher from the dye, and the dye fluoresces. By following the increase in fluorescence over time for the DNA–RNA hybrid exposed to RNase H specifically immobilized on sP(EO-stat-PO) surfaces, one is able to determine the activity coefficient of the enzyme. The results of this assay are shown in Fig. 1.6. Compared to the enzyme in solution, the activity coefficient of the enzyme is both unaffected by immobilization and returns to the same value after unfolding and refolding on the sP (EO-stat-PO) surface. More details of these FRET experiments with surface-bound RNase H are described elsewhere (Heyes et al., 2007).

ACKNOWLEDGMENTS We thank G. Ulrich Nienhaus, Colin D. Heyes, Wiepke J. A. Koopmans, and John van Noort for their cooperation. This work was financially supported by the VW-foundation (project self-assembled hydrogel layers) and the German Science Foundation (DFG, graduate school 1035 ‘‘Biointerface’’ and Project B1 in the TR-SFB 37).

REFERENCES Amirgoulova, E. V., Groll, J., Heyes, C. D., Ameringer, T., Ro¨cker, C., Mo¨ller, M., and Nienhaus, G. U. (2004). Biofunctionalized polymer surfaces exhibiting minimal interaction towards immobilized proteins. ChemPhysChem 5, 552–555. Caraculacu, A. A., and Coseri, S. (2001). Isocyanates in polyaddition processes. Structure and reaction mechanisms. Prog. Polym. Sci. 26, 799–851. Douglas, J. F., Roovers, J., and Freed, K. F. (1990). Characterization of branching architecture through "universal" ratios of polymer solution properties. Macromolecules 23, 4168–4180. Fo¨rster, T. (1948). Zwischenmolekulare Energiewanderung und Fluoreszenz. Ann. Physik. 2, 55–75. Gasteier, P., Reska, A., Schulte, P., Salber, J., Offenhaeusser, A., Moeller, M., and Groll, J. (2007). Surface grafting of PEO-based star-shaped molecules for bioanalytical and biomedical applications. Macromol. Biosci. 7, 1010–1023. Goetz, H., Beginn, U., Bartelink, C. F., Gruenbauer, H. J. M., and Mo¨ller, M. (2002). Preparation of isophorone diisocyanate terminated star polyethers. Macromol. Mater. Eng. 287, 223–230. Groll, J., Amirgoulova, E., Ameringer, T., Heyes, C. D., Ro¨cker, C., Nienhaus, G. U., and Mo¨ller, M. (2004). Biofunctionalized, ultrathin coatings of crosslinked star-shaped poly (ethylene oxide) allow reversible folding of immobilized proteins. J. Am. Chem. Soc. 126, 4234–4239.

Star Polymer Surface Passivation for Single-Molecule Detection

17

Groll, J., Ademovic, Z., Ameringer, T., Klee, D., and Mo¨ller, M. (2005a). Comparison of coatings from reactive star shaped PEG-stat-PPG prepolymers and grafted linear PEG for biological and medical applications. Biomacromolecules 6, 956–962. Groll, J., Ameringer, T., Spatz, J. P., and Mo¨ller, M. (2005b). Ultrathin coatings from isocyanate-terminated star PEG prepolymers: Layer formation and characterization. Langmuir 21, 1991–1999. Groll, J., Haubensak, W., Ameringer, T., and Moeller, M. (2005c). Ultrathin coatings from isocyanate terminated star PEG prepolymers: Patterning of proteins on the layers. Langmuir 21, 3076–3083. Halperin, A. (1999). Polymer brushes that resist adsorption of model proteins: Design parameters. Langmuir 15, 2525–2533. Harder, P., Grunze, M., Dahint, R., Whitesides, G. M., and Laibinis, P. E. (1998). Molecular conformation in oligo(ethylene glycol)-terminated self-assembled monolayers on gold and silver surfaces determines their ability to resist protein adsorption. J. Phys. Chem. B 102, 426–436. Harris, J. M., and Zalipsky, S. (1997). Poly(ethylene glycol): Chemistry and Biological Applications. American Chemical Society, Washington, DC. Heyes, C. D., Groll, J., Moeller, M., and Nienhaus, G. U. (2007). Synthesis, patterning and applications of star-shaped poly(ethylene glycol) biofunctionalized surfaces. Mol. Biosyst. 3, 419–430. Hogrefe, H. H., Hogrefe, R. I., Walder, R. Y., and Walder, J. A. (1990). Kinetic analysis of Escherichia coli RNase H using DNA-RNA-DNA/DNA substrates. J. Biol. Chem. 265, 5561–5566. Irvine, D. J., Mayes, A. M., and Griffith-Cima, L. (1996). Self-consistent field analysis of grafted star polymers. Macromolecules 29, 6037–6043. Irvine, D. J., Mayes, A. M., Satija, K. S., Barker, G. J., Sofia-Allgor, S. J., and Griffith, L. G. (1998). Comparison of tethered star and linear poly(ethylene oxide) for control of biomaterials surface properties. Biomed. Mater. Res. 40, 498–509. Kanaya, S. (1998). Enzymatic activity and protein stability of E. coli ribonuclease HI. In ‘‘Ribonucleases H,’’ (R. J. Crouch and J. J. Toulme, eds.), INSERM, Paris. Kingshott, P., Thissen, H., and Griesser, H. (2002). Effects of cloud-point grafting, chain length, and density of PEG layers on competitive adsorption of ocular proteins. Biomaterials 23, 2043–2056. Koopmans, W. J. A., Schmidt, T., and van Noort, J. (2008). Nucleosome immobilization strategies for single-pair FRET microscopy. ChemPhysChem 9, 2002–2009. Kuzmenkina, E. V., Heyes, C. D., and Nienhaus, G. U. (2006). Single-molecule FRET study of denaturant induced unfolding of RNase H. J. Mol. Biol. 327, 313–324. Li, G., Levitus, M., Bustamante, C., and Widom, J. (2005). Rapid spontaneous accessibility of nucleosomal DNA. Nat. Struct. Mol. Biol. 12, 46–53. Maheshwari, G., Brown, G., Lauffenburger, A. D., Wells, A., and Griffith, L. G. (2000). Cell adhesion and motility depend on nanoscale RGD clustering. J. Cell Sci. 113, 1677–1686. Malmsten, M., Emoto, K., and Alstine, J. M. V. (1998). Effect of chain density on inhibition of protein adsorption by poly(ethylene glycol) based coatings. J. Colloid Interface Sci. 202, 507–517. McPherson, T., Kidane, A., Szleifer, I., and Park, K. (1998). Prevention of protein adsorption by tethered poly(ethylene oxide) layers: Experiments and single-chain mean-field analysis. Langmuir 14, 176–186. Sofia, S. J., Premnath, V., and Merrill, E. W. (1998). Poly(ethylene oxide) grafted to silicon surfaces: Grafting density and protein adsorption. Macromolecules 31, 5059–5070. Stryer, L., and Haugland, R. P. (1967). Energy transfer: A spectroscopic ruler. Proc. Natl. Acad. Sci. USA 58, 719–726.

18

¨rgen Groll and Martin Moeller Ju

Szleifer, I., and Carignano, M. A. (2000). Tethered polymer layers: Phase transitions and reduction of protein adsorption. Macromol. Rapid. Commun. 21, 423–448. Unsworth, L. D., Tun, Z., Sheardown, H., and Brash, J. L. (2006). In situ neutron reflectometry investigation of gold-chemisorbed PEO layers of varying chain density: Relationship of layer structure to protein resistance. J. Colloid Interface Sci. 296, 520–526. Unsworth, L. D., Sheardown, H., and Brash, J. L. (2008). Protein-resistant poly(ethylene oxide)-grafted surfaces: Chain density-dependent multiple mechanisms of action. Langmuir 24, 1924–1929. Yang, Z., Galloway, J. A., and Yu, H. (1999). Protein interactions with poly(ethylene glycol) self-assembled monolayers on glass substrates: Diffusion and adsorption. Langmuir 15, 8405–8411.

C H A P T E R

T W O

Azide-Specific Labeling of Biomolecules by Staudinger– Bertozzi Ligation: Phosphine Derivatives of Fluorescent Probes Suitable for Single-Molecule Fluorescence Spectroscopy Anirban Chakraborty,*,†,1 Dongye Wang,*,†,1 Yon W. Ebright,*,† and Richard H. Ebright*,† Contents 20 21 21 22 23 25 25 26 26 27 27 28 28

1. Introduction 2. Materials and Methods 2.1. Materials 2.2. General methods 2.3. Synthesis of Alexa488-phosphine (Fig. 2.1) 2.4. Synthesis of Cy3B-phosphine (Fig. 2.2) ˚ 2.5. Synthesis of Alexa647-phosphine20 A (Fig. 2.3A) 24 A˚ 2.6. Synthesis of Alexa647-phosphine (Fig. 2.3B) 2.7. Azide-specific labeling 2.8. Quantitation of labeling efficiency 2.9. Quantitation of labeling specificity Acknowledgments References

Abstract We describe the synthesis of phosphine derivatives of three fluorescent probes that have a brightness and photostability suitable for single-molecule fluorescence spectroscopy and microscopy: Alexa488, Cy3B, and Alexa647. In addition, we describe procedures for use of these reagents in azide-specific, bioorthogonal * Department of Chemistry and Chemical Biology, Waksman Institute, Rutgers University, Piscataway, New Jersey, USA Howard Hughes Medical Institute, Rutgers University, Piscataway, New Jersey, USA 1 Contributed equally {

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72018-8

#

2010 Elsevier Inc. All rights reserved.

19

20

Anirban Chakraborty et al.

labeling through Staudinger–Bertozzi ligation, as well as procedures for the quantitation of labeling specificity and labeling efficiency. The reagents and procedures of this report enable chemoselective, site-selective labeling of azidecontaining biomolecules for single-molecule fluorescence spectroscopy and microscopy.

1. Introduction The Staudinger–Bertozzi ligation involves reaction between a first compound containing an azide moiety and a second compound containing a phosphine moiety with an adjacent methyl ester, and results in coupling of the compounds through an amide linkage (Kiick et al., 2002; Saxon and Bertozzi, 2000; reviewed in Kohn and Breinbauer, 2004; Sletten and Bertozzi, 2009). The reaction is bioorthogonal, since azides and phosphines are not present in natural biomolecules and since azides and phosphines do not react with moieties present in natural biomolecules. The reaction is biocompatible, since it proceeds in aqueous solution under mild conditions at moderate temperatures and moderate pH ranges. The reaction is efficient, and yields of
90% are achieved routinely. The bioorthogonality, biocompatibility, and high efficiency of the Staudinger–Bertozzi ligation render the reaction suitable for two applications: (i) biomolecule-specific labeling of engineered biomolecules containing randomly located azide moieties and (ii) biomolecule-specific, site-specific labeling of engineered biomolecules containing site-specifically incorporated azide moieties. In published work, the reaction has been used for labeling of engineered azide-containing biomolecules in vitro with single proteins, in vitro with mixtures of proteins, in vivo in living cells, and in vivo in living organisms (Kiick et al., 2002; Prescher et al., 2004; Saxon and Bertozzi, 2000). Multiple strategies have been reported for the incorporation of azides into biomolecules, providing potential targets for azide-specific, bioorthogonal labeling through use of the Staudinger–Bertozzi ligation. For example, azides have been randomly incorporated into carbohydrates and proteinlinked carbohydrates by supplying cells with azide-functionalized carbohydrate precursors (Chang et al., 2007; Dube et al., 2006; Hang et al., 2003; Hangauer and Bertozzi, 2008; Laughlin and Bertozzi, 2007; Laughlin et al., 2006; Prescher et al., 2004; Saxon and Bertozzi, 2000; Saxon et al., 2002; Vocadlo et al., 2003); azides have been randomly incorporated into proteins by supplying cells or organisms with azide-functionalized methionine (Kiick et al., 2002; Link and Tirrell, 2003; Link et al., 2004; Ngo et al., 2009); azides have been site-specifically incorporated into proteins in vitro by ligation with azide-functionalized farnesyl, lipoyl, or puromycin surrogates (Baruah et al., 2008; Gauchet et al., 2006; Humenik et al., 2007); and azides have been sitespecifically incorporated into proteins in vitro and in vivo by use of unnatural

Phosphine Derivatives of Fluorescent Probes

21

amino acid mutagenesis (Chin et al., 2002; Deiters et al., 2003; Krieg et al., 1986; Nguyen et al., 2009; Ohno et al., 2007; Tsao et al., 2005). Multiple phosphine derivatives suitable for azide-specific, bioorthogonal labeling through use of the Staudinger–Bertozzi ligation have been reported, including phosphine derivatives of the affinity probe biotin and phosphine derivatives of the fluorescent probes fluorescein, coumarin, tetraethylrhodamine, and Cy5.5 (Chang et al., 2007; Hangauer and Bertozzi, 2008; Lemieux et al., 2003; Saxon and Bertozzi, 2000; Wang et al., 2003). Single-molecule fluorescence spectroscopy requires fluorescent probes that have exceptionally high brightness and exceptionally high photostability (fluorescent probes of greater brightness and photostability than fluorescein, coumarin, and tetraethylrhodamine; reviewed in Ha 2001; Kapanidis and Weiss, 2002; Roy et al., 2008). Single-molecule fluorescence resonance energy transfer (FRET) experiments further require pairs of fluorescent probes capable of serving as an efficient donor/acceptor, wherein the fluorescence emission spectrum of the donor overlaps the fluorescence excitation spectrum of the acceptor. In FRET experiments, the lengths and flexibilities of the linkers between biomolecule and fluorescent probes can significantly affect results; therefore, maximum flexibility in experimental design requires sets of reagents that yield different lengths and flexibilities of linkers between biomolecules and fluorescent probes. Here we report the synthesis of phosphine derivatives of fluorescent probes that have a brightness and photostability suitable for single-molecule fluorescence spectroscopy (Alexa488, Cy3B, and Alexa647; Cooper et al., 2004; Leung et al., 2005; Panchuk-Voloshina et al., 1999), that have a spectral overlap suitable to serve as donor/acceptor pairs for FRET (Alexa488/Cy3B, Alexa488/Alexa647, and Cy3B/Alexa647), and that, in one case, yield alternatively either a moderate-length, flexible biomolecule-probe linker or a ˚ and 9 rotatable bonds longer, more flexible, biomolecule-probe linker (20 A ˚ vs. 24 A and 12 rotatable bonds; Figs. 2.1–2.3). In addition, we report procedures for the application of these reagents in azide-specific, bioorthogonal labeling through use of the Staudinger–Bertozzi ligation and procedures for the quantitation of labeling specificity and labeling efficiency. The reagents and procedures of this report enable chemoselective, site-selective labeling of azide-containing biomolecules for single-molecule spectroscopy.

2. Materials and Methods 2.1. Materials 1-Methyl-2-diphenylphosphinoterephthalate (MDPT) was synthesized as described in Kiick et al. (2002). Alexa Fluor 488 cadaverine, Alexa Fluor 647 cadaverine, Alexa Fluor 647 NHS ester, 1-ethyl-3-

22

Anirban Chakraborty et al.

H2N

SO3−

SO3−

O

I

NH2+

COOH

O

N H

H2N

MDPT EDAC/NHSS/DIPEA 3 h, 37 ⬚C H2N

SO3−

SO3−

O

NH2+

COOH

II O

H3COOC

N H

N H

O

PPh2

Figure 2.1 Synthesis of Alexa488-phosphine. Staudinger–Bertozzi ligation between ˚ linker between the compound II and an azide-containing biomolecule yields an  18 A biomolecule and the fluorophore (distance measured from first nitrogen atom of azide to fused ring system of fluorophore with fully extended conformation of linker).

(3-dimethylaminopropyl)-carbodiimide (EDAC), and N-hydroxysulfosuccinimide (NHSS) were purchased from Invitrogen (Carlsbad, CA). Cy3B-NHS was purchased from GE Healthcare (Piscataway, NJ). Monotrityl-ethelenediamine (acetic acid salt) was purchased from Novabiochem (Madison, WI). N-Trityl-1,2-ethanediamine (hydrobromide salt), N,N0 - trifluoroacetic acid (TFA), triethylamine (TEA), diisopropylethylamine (DIPEA), and N,N0 -dimethylformamide (DMF) were purchased from Sigma-Aldrich (Milwaukee, WI). Bio-Gel P30 was purchased from BioRad (Hercules, CA).

2.2. General methods Reversed-phase HPLC was performed on a Hitachi L7100 instrument using Supelco Discover Bio C18 column (25 cm  10 mm, 10 mm). All solutions for HPLC were deoxygenated by bubbling argon for 15 min. Flash chromatography was performed using silica gel (230–400 mesh, 60 A˚). MALDI-MS was performed on an Applied Biosystems MDS SCIEX 4800 instrument.

23

Phosphine Derivatives of Fluorescent Probes

O SO3−

N-O

III

O

O

N+

N O

TrNH(CH2)2NH2 1 h, 25 ⬚C TrHN

IV

H N

SO3− O

N+

N O

20% TFA 1 h, 25 ⬚C H2N

V

H N

SO3− O

N+

N O

O N H

VI H3COOC

PPh2

MDPT EDAC/NHSS/DIPEA 3 h, 37 ⬚C SO3−

H N O

N+

N O

Figure 2.2 Synthesis of Cy3B-phosphine. Staudinger–Bertozzi ligation between com˚ linker between the pound VI and an azide-containing biomolecule yields an  15 A biomolecule and the fluorophore.

2.3. Synthesis of Alexa488-phosphine (Fig. 2.1) 2.3.1. Alexa488-carboyl-pentylenediaminyl-phosphine (Alexa488-phosphine; II) EDAC (4.2 mg; 21 mmol) in 50 ml degassed water and NHSS (4.2 mg; 16 mmol) in 50 ml degassed water were mixed, and MDPT (5.9 mg; 15 mmol) in 50 ml DMF was added. A precipitate was observed. Degassed water (50 ml) was added, followed by DMF (200 ml), resulting in dissolution of the precipitate. Compound I (Alexa Fluor 488 cadaverine; 1.0 mg; 1.5 mmol; Fig. 2.1) in 50 ml DMF was added, followed by DIPEA (5.6 ml; 31 mmol), and the reaction mixture was incubated for 3 h at 37  C. Product II was purified by reversed-phase HPLC (solvent A: 0.1% TFA in water; solvent B: 100% acetonitrile; gradient: 30–100% B in 30 min at 2 ml/min) and lyophilized. MS (MALDI): calculated, m/z 964.9 (MHþ); found 964.9.

24

Anirban Chakraborty et al.

A

-

(CH2)3SO3N

O3S(H2C)3 N+ -O

3

S

B

-

-O

SO3-

VII

O

NO

O

(CH2)3SO3-

O3S(H2C)3 N+

N

S

SO3-

3

XI

HN

O

O TrNH(CH2)2NH2 TEA 1 h, 25 ⬚C -O

3

N+

3

N

S

-O

SO3-

VIII HN

S

SO3-

3

HN

33% TFA 30 min, 25 ⬚C

-O

N+

XII

(CH2)3SO3N

O3S(H2C)3 N+

3S

(CH2)3SO3N

3S(H2C)3

-O

O

NHTr

-

MDPT EDAC/NHSS/DIPEA 3 h, 37 ⬚C

(CH2)3SO3-

S(H2C)3

-O

H2N

O

O H3COOC

N H PPh2

SO3-

IX HN

O

NH2 MDPT EDAC/NHSS/DIPEA 3 h, 37 ⬚C -O

(CH2)3SO3N

3S(H2C)3

-O

N+

3S

SO3-

X HN O

O

NH

PPh2 COOCH3 ˚

˚

Figure 2.3 Synthesis of Alexa647-phosphine20 A (A) and Alexa488-phosphine24 A (B). Staudinger–Bertozzi ligation between compound X or XII and an azide-contain˚ linker between the biomolecule ing biomolecule yields, respectively, an  20 or  24 A and the fluorophore.

25

Phosphine Derivatives of Fluorescent Probes

2.4. Synthesis of Cy3B-phosphine (Fig. 2.2) 2.4.1. Cy3B-carboyl-ethylenediaminyl-trityl (IV) Mono-trityl-ethelenediamine (acetic acid salt; 23.5 mg; 65 mmol), compound III (Cy3B-NHS; 5.0 mg; 6.5 mmol), and TEA (60 ml; 430 mmol) were added, in turn, to 200 ml anhydrous DMF, and the reaction mixture was incubated for 1 h at room temperature. Product IV was purified by reversed-phase HPLC (solvent A: water; solvent B: 90% acetonitrile, 10% water; gradient 30–80% B in 30 min at 2 ml/min) and lyophilized. MS (MALDI): calculated, m/z 845.6 (MHþ); found, 845.6. 2.4.2. Cy3B-carboyl-ethylenediamine (V) TFA (50 ml; 0.65 mmol) was added to compound IV (4.2 mg; 5.0 mmol) in 200 ml chloroform, and the reaction mixture was incubated for 1 h at room temperature, and the solvent was evaporated. Product V was purified by reversed-phase HPLC (solvent A: 0.1% TFA in water; solvent B: 100% acetonitrile; gradient: 20–80% B in 30 min at 2 ml/min) and lyophilized. MS (MALDI): calculated, m/z 603.3 (MHþ); found, 603.3. 2.4.3. Cy3B-carboyl-ethylenediaminyl-phosphine (Cy3B-phosphine; VI) EDAC (12.5 mg; 65 mmol) in 50 ml DMF, NHSS (8.8 mg; 65 mmol) in 50 ml DMF, and MDPT (24 mg; 60 mmol) in 50 ml DMF were combined. Compound V (2.4 mg; 4.0 mmol) in 50 ml DMF was added, followed by DIPEA (23 ml; 130 mmol), and the reaction mixture was incubated for 3 h at 37  C. Product VI was purified by reversed-phase HPLC (solvent A: 0.1% TFA in water; solvent B: 100% acetonitrile; gradient: 30–100% B in 30 min at 2 ml/min) and lyophilized. MS (MALDI): calculated, m/z 948.5 (MHþ); found, 948.5. ˚

2.5. Synthesis of Alexa647-phosphine20 A (Fig. 2.3A) 2.5.1. Alexa647-pentanoyl-ethylenediaminyl-trityl (VIII) N-Trityl-1,2-ethanediamine (hydrobromide salt) (23 mg; 60 mmol) was added to compound VII (Alexa Fluor 647 NHS ester; 5.0 mg; 5.0 mmol) in 1 ml DMF. TEA (10.0 ml; 71 mmol) was added and the reaction mixture was incubated for 30 min at room temperature. The reaction mixture was dried under vacuum, redissolved in 0.5 ml ethanol and 20 ml ammonium hydroxide. Product VIII was isolated by flash chromatography and dried under vacuum. MS (MALDI): calculated, m/z 1143.4 (MHþ); found, 1143.4.

26

Anirban Chakraborty et al.

2.5.2. Alexa647-pentanoyl-ethylenediamine (IX) TFA (100 ml; 1.3 mmol) was added to compound VIII (5.0 mg; 4.4 mmol) in 200 ml chloroform, and the reaction mixture was incubated for 30 min at room temperature. The reaction mixture was dried under vacuum, and product IX was purified using flash chromatography. MS (MALDI): calculated, m/z 901 (MHþ); found, 901. 2.5.3. Alexa647-pentanoyl-ethylenediaminyl-phosphine ˚ (Alexa647-phosphine20 A; X) EDAC (21 mg; 110 mmol) in 250 ml degassed water, NHSS (21 mg; 78 mmol) in 250 ml degassed water, compound IX (5.0 mg; 5.5 mmol) in 200 ml DMF and 50 ml degassed water, and MDPT (30 mg; 75 mmol) in 250 ml DMF were combined. A precipitate was observed. DMF (700 ml) was added, resulting in dissolution of the precipitate. DIPEA (28 ml; 160 mmol) was added, and the mixture was incubated for 3 h at 37  C. Product X was purified by reversed-phase HPLC (solvent A: 0.1% TFA in water; solvent B: 100% acetonitrile; gradient: 30–100% B in 30 min at 2 ml/min) and dried under vacuum. MS (MALDI): calculated, m/z 1248.4 (MHþ); found, 1248.4.

2.6. Synthesis of Alexa647-phosphine24 A (Fig. 2.3B) ˚

2.6.1. Alexa647-pentanoyl-pentylenediaminyl-phosphine ˚ (Alexa647-phosphine24 A; XII) EDAC (21 mg; 110 mmol) in 250 ml degassed water, NHSS (21 mg; 78 mmol) in 250 ml degassed water, compound XI (Alexa Fluor 647 cadeverine; 5.0 mg; 5.5 mmol) in 200 ml DMF and 50 ml degassed water, and MDPT (30 mg; 75 mmol) in 250 ml DMF were combined. A precipitate was observed. DMF (700 ml) was added, resulting in dissolution of the precipitate. DIPEA (28 ml; 160 mmol) was added, and the mixture was incubated for 3 h at 37  C. The product was purified by reversed-phase HPLC (solvent A: 0.1% TFA in water; solvent B: 100% acetonitrile; gradient: 30–100% B in 30 min at 2 ml/min) and was dried under vacuum. MS (MALDI): calculated, m/z 1290.5 (MHþ); found, 1290.5.

2.7. Azide-specific labeling Reaction mixtures (3 ml) contained 20 mM P-azide (derivative of protein P containing a single azide moiety) and 200 mM probe phosphine ˚ (Alexa488-phosphine, ˚ Cy3B-phosphine, Alexa647-phosphine20A, or 24 A Alexa647-phosphine ) in 50 mM Tris–HCl, pH 7.9, 6 M guanidine– HCl, and 5% (v/v) glycerol. Reaction mixtures were incubated for 15 h at 37  C. Reaction mixtures were then applied to 10 ml columns of Bio-Gel

Phosphine Derivatives of Fluorescent Probes

27

P30 preequilibrated in 50 mM Tris–HCl, pH 7.9, 6 M guanidine–HCl, and 5% (v/v) glycerol; columns were washed with 3 ml of the same buffer; and products were eluted in an additional 3 ml of the same buffer.

2.8. Quantitation of labeling efficiency The concentration of the product of the labeling reaction and the efficiency of labeling reaction are determined from UV/Vis-absorbance measurements and are calculated as A280  2F;280 ðA max =2F; max Þ 2P;280   A max =2F; max labeling efficiency ¼ 100%  concentration of product

concentration of product ¼

where A280 is the measured absorbance at 280 nm, Amax is the measured absorbance at the long-wavelength absorbance maximum of fluorescent probe F (493, 559, and 652 nm for Alexa488, Cy3B, and Alexa647, respectively), 2P,280 is the molar extinction coefficient of protein P at 280 nm (calculated as in Gill and von Hippel, 1989), 2F,280 is the molar extinction coefficient of fluorescent probe F at 280 nm (8,030 M 1 cm 1, 10,400 M 1 cm 1, and 7,350 M 1 cm 1 for Alexa488, Cy3B, and Alexa647, respectively), and 2F,max is the extinction coefficient of fluorescent probe F at its long-wavelength absorbance maximum (73,000 M 1 cm 1 at 493 nm, 130,000 M 1 cm 1 at 559 nm, and 245,000 M 1 cm 1 at 652 nm for Alexa488, Cy3B, and Alexa647, respectively). Typical labeling efficiencies are
90%.

2.9. Quantitation of labeling specificity The specificity of labeling is determined from the efficiencies of labeling (see preceding section) of (i) the product of the labeling reaction with P-azide and (ii) the product of a parallel labeling reaction with P. The specificity of labeling is calculated as labeling specificity ¼ 100%  f1  ½ðlabeling efficiency with PÞ ðlabeling efficiency with P-azideÞg Alternatively, the specificity of labeling can be determined from the fluorescence intensities at the emission maximum of fluorescent probe F (516 nm upon excitation at 493 nm, 570 upon excitation at 559 nm, or 672 nm upon excitation at 652 nm for Alexa488, Cy3B, and Alexa647, respectively) of (i) the product of the labeling reaction with P-azide and

28

Anirban Chakraborty et al.

(ii) an equal concentration of the product of a parallel labeling reaction with P. In this case, the specificity of labeling is calculated as labeling specificity ¼ 100%  f1  ½ðfluorescence with PÞ ðfluorescence with P-azideÞg Typical labeling specificities are
90%.

ACKNOWLEDGMENTS We thank S. Weiss for suggesting identities of fluorescent probes suitable for single-molecule detection. This work was supported by NIH grants GM41376 and AI72766 and a Howard Hughes Investigatorship to R. H. E.

REFERENCES Baruah, H., Puthenveetil, S., Choi, Y. A., Shah, S., and Ting, A. Y. (2008). An engineered aryl azide ligase for site-specific mapping of protein-protein interactions through photocross-linking. Angew. Chem. Int. Ed. Engl. 47, 7018–7021. Chang, P. V., Prescher, J. A., Hangauer, M. J., and Bertozzi, C. R. (2007). Imaging cell surface glycans with bioorthogonal chemical reporters. J. Am. Chem. Soc. 129, 8400–8401. Chin, J. W., Santoro, S. W., Martin, A. B., King, D. S., Wang, L., and Schultz, P. G. (2002). Addition of p-azido-L-phenylalanine to the genetic code of Escherichia coli. J. Am. Chem. Soc. 124, 9026–9027. Cooper, M., Ebner, A., Briggs, M., Burrows, M., Gardner, N., Richardson, R., and West, R. (2004). Cy3B: Improving the performance of cyanine dyes. J. Fluoresc. 14, 145–150. Deiters, A., Cropp, T. A., Mukherji, M., Chin, J. W., Anderson, J. C., and Schultz, P. G. (2003). Adding amino acids with novel reactivity to the genetic code of Saccharomyces cerevisiae. J. Am. Chem. Soc. 125, 11782–11783. Dube, D. H., Prescher, J. A., Quang, C. N., and Bertozzi, C. R. (2006). Probing mucintype O-linked glycosylation in living animals. Proc. Natl. Acad. Sci. USA 103, 4819–4824. Gauchet, C., Labadie, G. R., and Poulter, C. D. (2006). Regio- and chemoselective covalent immobilization of proteins through unnatural amino acids. J. Am. Chem. Soc. 128, 9274–9275. Gill, S. C., and von Hippel, P. H. (1989). Calculation of protein extinction coefficients from amino acid sequence data. Anal. Biochem. 182, 319–326. Ha, T. (2001). Single-molecule fluorescence resonance energy transfer. Methods 25, 78–86. Hang, H. C., Yu, C., Kato, D. L., and Bertozzi, C. R. (2003). A metabolic labeling approach toward proteomic analysis of mucin-type O-linked glycosylation. Proc. Natl. Acad. Sci. USA 100, 14846–14851. Hangauer, M. J., and Bertozzi, C. R. (2008). A FRET-based fluorogenic phosphine for livecell imaging with the Staudinger ligation. Angew. Chem. Int. Ed. 47, 2394–2397. Humenik, M., Huang, Y., Wang, Y., and Sprinzl, M. (2007). C-terminal incorporation of bio-orthogonal azide groups into a protein and preparation of protein-oligodeoxynucleotide conjugates by Cu’-catalyzed cycloaddition. ChemBioChem 8, 1103–1106.

Phosphine Derivatives of Fluorescent Probes

29

Kapanidis, A. N., and Weiss, S. (2002). Fluorescent probes and bioconjugation chemistries for single-molecule fluorescence analysis of biomolecules. J. Chem. Phys. 117, 10953–10964. Kiick, K. L., Saxon, E., Tirrell, D. A., and Bertozzi, C. R. (2002). Incorporation of azides into recombinant proteins for chemoselective modification by the Staudinger ligation. Proc. Natl. Acad. Sci. USA 99, 19–24. Kohn, M., and Breinbauer, R. (2004). The Staudinger ligation–A gift to chemical biology. Angew. Chem. Int. Ed. 43, 3106–3116. Krieg, U. C., Walter, P., and Johnson, A. E. (1986). Photocrosslinking of the signal sequence of nascent preprolactin to the 54-kilodalton polypeptide of the signal recognition particle. Proc. Natl. Acad. Sci. USA 83, 8604–8608. Laughlin, S. T., and Bertozzi, C. R. (2007). Metabolic labeling of glycans with azido sugars and subsequent glycan-profiling and visualization via Staudinger ligation. Nat. Protoc. 2, 2930–2944. Laughlin, S. T., Agard, N. J., Baskin, J. M., Carrico, I. S., Chang, P. V., Ganguli, A. S., Hangauer, M. J., Lo, A., Prescher, J. A., and Bertozzi, C. R. (2006). Metabolic labeling of glycans with azido sugars for visualization and glycoproteomics. Methods Enzymol. 415, 230–250. Lemieux, G. A., De Graffenried, C. L., and Bertozzi, C. R. (2003). A fluorogenic dye activated by the Staudinger ligation. J. Am. Chem. Soc. 125, 4708–4709. Leung, W.-Y., Cheung, C.-Y., and Yue, S. (2005). Modified carbocyanine dyes and their conjugates. US Patent 6, 977, 305. Link, A. J., and Tirrell, D. A. (2003). Cell surface labeling of Escherichia coli via copper(I)catalyzed [3þ2] cycloaddition. J. Am. Chem. Soc. 125, 11164–11165. Link, A. J., Vink, M. K., and Tirrell, D. A. (2004). Presentation and detection of azide functionality in bacterial cell surface proteins. J. Am. Chem. Soc. 126, 10598–10602. Ngo, J. T., Champion, J. A., Mahdavi, A., Tanrikulu, I. C., Beatty, K. E., Connor, R. E., Yoo, T. H., Dieterich, D. C., Schuman, E. M., and Tirrell, D. A. (2009). Cell-selective metabolic labeling of proteins. Nat. Chem. Biol. 5, 715–717. Nguyen, D. P., Lusic, H., Neumann, H., Kapadnis, P. B., Deiters, A., and Chin, J. W. (2009). Genetic encoding and labeling of aliphatic azides and alkynes in recombinant proteins via a pyrrolysyl-tRNA synthetase/tRNA(CUA) pair and click chemistry. J. Am. Chem. Soc. 131, 8720–8721. Ohno, S., Matsui, M., Yokogawa, T., Nakamura, M., Hosoya, T., Hiramatsu, T., Suzuki, M., Hayashi, N., and Nishikawa, K. (2007). Site-selective post-translational modification of proteins using an unnatural amino acid, 3-azidotyrosine. J. Biochem. 141, 335–343. Panchuk-Voloshina, N., Haugland, R. P., Bishop-Stewart, J., Bhalgat, M. K., Millard, P. J., Mao, F., Leung, W.-Y., and Haugland, R. P. (1999). Alexa dyes, a series of new fluorescent dyes that yield exceptionally bright, photostable conjugates. J. Histochem. Cytochem. 47, 1179–1188. Prescher, J. A., Dube, D. H., and Bertozzi, C. R. (2004). Chemical remodelling of cell surfaces in living animals. Nature 430, 873–877. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Saxon, E., and Bertozzi, C. R. (2000). Cell surface engineering by a modified Staudinger reaction. Science 287, 2007–2010. Saxon, E., Luchansky, S. J., Hang, H. C., Yu, C., Lee, S. C., and Bertozzi, C. R. (2002). Investigating cellular metabolism of synthetic azidosugars with the Staudinger ligation. J. Am. Chem. Soc. 124, 14893–14902. Sletten, E. M., and Bertozzi, C. R. (2009). Bioorthogonal chemistry: Fishing for selectivity in a sea of functionality. Angew. Chem. Int. Ed. 48, 6974–6998.

30

Anirban Chakraborty et al.

Tsao, M. L., Tian, F., and Schultz, P. G. (2005). Selective Staudinger modification of proteins containing p-azidophenylalanine. ChemBioChem 6, 2147–2149. Vocadlo, D. J., Hang, H. C., Kim, E. J., Hanover, J. A., and Bertozzi, C. R. (2003). A chemical approach for identifying O-GlcNAc-modified proteins in cells. Proc. Natl. Acad. Sci. USA 100, 9116–9121. Wang, C. C., Seo, T. S., Li, Z., Ruparel, H., and Ju, J. (2003). Site-specific fluorescent labeling of DNA using Staudinger ligation. Bioconjug. Chem. 14, 697–701.

C H A P T E R

T H R E E

Preparation of Fluorescent Pre-mRNA Substrates for an smFRET Study of Pre-mRNA Splicing in Yeast John Abelson, Haralambos Hadjivassiliou, and Christine Guthrie Contents 1. Introduction 2. Identification of a Yeast Pre-mRNA with a Small Intron that is Spliced Efficiently In Vitro 3. Synthetic Fluorescent Ubc4 Pre-mRNA 4. Do the Dyes Affect the Efficiency of Splicing? 5. Mutant Pre-mRNAs 6. Tethering the Pre-mRNA to the Microscope Slide 7. Summary and Conclusion Acknowledgments References

32 32 33 37 37 38 39 40 40

Abstract The spliceosome is a complex small nuclear (sn)RNA–protein machine that removes introns from pre-mRNAs via two successive phosphoryl transfer reactions. For each splicing event, the spliceosome is assembled de novo on a premRNA substrate and a complex series of assembly steps leads to the active conformation. To comprehensively monitor pre-mRNA conformational dynamics during spliceosome assembly, we developed a strategy for single-molecule FRET (smFRET) that utilizes a small, efficiently spliced yeast pre-mRNA, Ubc4, in which donor and acceptor fluorophores are placed in the exons adjacent to the 50 and 30 splice sites. In this chapter, we describe the identification of Ubc4 pre-mRNA that is efficiently spliced in vitro and the methods we have developed for the chemical synthesis of fluorescent Ubc4 pre-mRNA for smFRET.

Department of Biochemistry and Biophysics, University of California, San Francisco, California, USA Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72017-6

#

2010 Elsevier Inc. All rights reserved.

31

32

John Abelson et al.

1. Introduction An in vitro splicing system in yeast was developed almost 25 years ago, enabling the study of spliceosome assembly (Lin et al., 1985). Much has been learned about the steps in spliceosome assembly since that time, but little is known about the kinetics of spliceosome assembly, largely because no progress has been made on developing a well-defined in vitro system. Experiments are still done in a crude extract where multistep reversible processes are obscured by asynchronous progression along the pathway. In addition, in vitro splicing in a yeast extract is generally inefficient, leading to only a fraction of molecules undergoing one or both steps of splicing. As abundantly demonstrated in this volume, single-molecule fluorescence resonance energy transfer (smFRET) is a technique that could overcome these difficulties. Accordingly, we have set out to establish an smFRET assay for spliceosome assembly in the yeast system. In principle, spliceosome assembly could be monitored by transfer between donor and acceptor probes in two proteins, in two RNAs, in a single RNA or in a single RNA and a single protein. For the development of this system, we have elected to monitor FRET between donor and acceptor probes in the pre-mRNA. For the first experiments, we have located the donor probe in exon 1 near the 50 splice site and the acceptor in exon 2 near the 30 splice site. In the premRNA, these probes may be (depending on the secondary structure of the pre-mRNA) too far away from each other to result in smFRET, but in the spliced mRNA product, they should be brought into proximity producing an smFRET signal. This experiment, then, is designed to monitor the relative movement of the 50 and 30 splice sites during spliceosome assembly. We have elected to follow the FRET changes that occur during assembly at the single-molecule level (Abelson et al., 2010). This chapter describes the identification and synthesis of a pre-mRNA substrate that is suitable for these experiments.

2. Identification of a Yeast Pre-mRNA with a Small Intron that is Spliced Efficiently In Vitro We first had to identify a yeast pre-mRNA with a small intron that is spliced efficiently in the yeast in vitro splicing system. A small intron is required in order to efficiently assemble the fluorescent substrate from synthetic RNA molecules. Most yeast in vitro splicing experiments have utilized the actin pre-mRNA. The actin intron is 308 nucleotides, too long for easy synthesis via synthetic RNA molecules.

Preparation of Fluorescent Pre-mRNA Substrates

33

To find a pre-mRNA in yeast with a small intron that is spliced well in vitro, we employed the yeast temperature-sensitive mutant, prp2-1. In this mutant, splicing is blocked before the first catalytic step when cells are shifted to the nonpermissive temperature. This results in the efficient accumulation of yeast pre-mRNAs. RNA was extracted from prp2-1 cells that had been grown at the permissive temperature (30  C) and shifted to the nonpermissive temperature (37  C) for 30 min. Bulk RNA was extracted from these cells and used as substrate in an in vitro splicing reaction. The RNA was extracted from these reaction mixtures and analyzed by RNA microarray analysis. The Guthrie lab has established a microarray assay that specifically monitors intron and mRNA features of all of the 250 or so yeast transcripts that contain introns (Pleiss et al., 2007). The presence of mRNA is monitored by hybridization to an immobilized DNA fragment whose sequence spans the spliced junction. For this experiment, cDNA was prepared from the splicing reactions described earlier. The control RNA was from reactions that did not contain ATP and thus would not contain transcripts spliced during the period of growth at the nonpermissive temperature. A tabulation of the results is shown in Table 3.1. According to this analysis, the pre-mRNA for the ribosomal protein RPS6A is best spliced in vitro. The actin pre-mRNA is 20th on the list. Several candidates with small introns emerged from this analysis. Ubc4, with an intron of 95 nucleotides, is second on the list. We ultimately chose the pre-mRNA for Ubc4 as the substrate for our FRET analysis. We next determined the minimum exon size for efficient splicing in vitro. A number of 32P labeled UBC 4 pre-mRNAs, differing in the size of their exons, were transcribed in vitro from DNA templates by T7 RNA polymerase and used as substrates for in vitro splicing (Fig. 3.1). The Ubc4 pre-mRNA (with exon lengths of 47 nucleotides) is the best of these substrates, but a precursor with exon lengths of only 20 nucleotides was about 70% as active as the former. In a trade-off between ease of synthesis and activity as a substrate, we chose to develop as an smFRET substrate the UBC 4 pre-mRNA consisting of two 20 nucleotide exons and the 95 nucleotide intron for a total size of 135 nucleotides.

3. Synthetic Fluorescent Ubc4 Pre-mRNA The pre-mRNA was synthesized by joining two oligonucleotides synthesized by Dharmacon: I (the 50 76 nucleotides of the pre-mRNA) and II (the 30 59 nucleotides; Fig. 3.2). For dye-labeled pre-mRNA, allyl-amine uridine was substituted for uridine at the appropriate positions. Oligonucleotide I was coupled to Cy3 and II to Cy5. For a 0.2 mM synthesis of oligonucleotides in this size

34

John Abelson et al.

Table 3.1 Relative splicing efficiencies of yeast pre-mRNAs in vitroa Rank

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 a

Systematic name

Gene name

ORF length

Intron length

YPL090C YBR082C YDR471W YKL081W YNL050C YLR061W YDR424C YBL040C YML056C YPL143W YAL030W YBR048W YPL081W YBL059C-A YBR181C YKL186C YHR101C YLR344W YER133W YFL039C

RSP6A UBC4 RPL27B TEF4 Unknown RPL22A DYN2 ERD2 IMD4 RPL33A SNC1 RPS11B RPS9A CMC2 RPS6B MTR2 BIG1 RPL26A GLC7 ACT1

1105 542 795 1565 904 755 455 757 1983 849 467 982 1095 415 1063 792 1095 831 1464 1436

394 95 384 326 91 389 96 97 408 525 113 511 501 85 352 154 87 447 525 308

Shown are the top 20 pre-mRNAs (out of about 250 genes in yeast that contain introns) in descending order of splicing efficiency in vitro as assayed in a microarray assay. RNA was extracted from prp2-1 grown at the permissive temperature (2-1 RNA) and from in vitro splicing reactions in which the pre-mRNA added to the reaction mixture was RNA isolated from prp2-1 grown at the permissive temperature and then shifted to the nonpermissive temperature for 30 min. Splicing reactions containing 240 mg/ml of the RNA extracted from cells grown at the nonpermissive temperature were incubated for 30 min at room temperature with (þATP RNA) and without ATP (ATP RNA). cDNA synthesized from the 2-1 control RNA was labeled with Cy3 and cDNA from þATP RNA and  ATP RNA with Cy5 as described (Pleiss et al., 2007). Splicing microarrays were hybridized with a mixture of 2-1 cDNA and þATP cDNA or 2-1 cDNA and ATP cDNA. After hybridization the microarrays were washed and analyzed for the ratio of ATP to 2-1. For this analysis we only considered hybridization to the set of oligonucleotides specific for the mRNAs of genes containing introns. The order of the genes in the table was determined by subtracting the rank orders of þATP/2-1 minus –ATP/2-1 for each gene (e.g., Ubc4 was in the 64th percentile in the (þ)ATP ratios and in the 0.8th percentile in the ()ATP ratios). Ubc4 pre-mRNA is second in this list and the canonical substrate for yeast in vitro pre-mRNA splicing, actin pre-mRNA, is 20th.

range, the yield is typically about 100 nmol. The quality of these oligonucleotides from Dharmacon is remarkable, given their length, but in order to obtain a homogeneous product, we first purify the deprotected oligonucleotides on a denaturing, 8 M urea, 6% polyacrylamide gel. The correct product has the slowest mobility and there is a streak of faster moving contaminants. The bands are visualized by UV shadowing or by staining

35

Preparation of Fluorescent Pre-mRNA Substrates

UBC4 30/30

30/20

30/10

20/20

20/10

20/30 10/10

10/20 10/30

0⬘ 15⬘ 30⬘

Figure 3.1 Exon length requirement for efficient splicing of Ubc4 pre-mRNA in vitro. DNA templates containing the T7 RNA polymerase promoter and the Ubc4 gene with different exon lengths were prepared by PCR. 32P-labeled T7 RNA polymerase RNA transcripts of the different Ubc4 constructs were incubated for 0, 15, and 30 min and assayed for splicing activity in a yeast extract.

20 nt Cy3

95 nt

−7

20 nt +4

I

II

76 nt

59 nt

Cy5

Figure 3.2 Design of Ubc4 pre-mRNA. Sequences of I and II: I ¼ 50 GAACUAAGUGAUCUAGAAAGGUAUGUCUAAAGUUAUGGCCACGUUUCAAAUGCGUGCUUUUUUUUUAAAACUUAUG; II ¼ 50 PCUCUUAUUUACUAACAAAAUCAACAUGCUAUUGAACUAGAGAUCCACCUACUUCAUGUT. To provide the proper substrate for RNA ligase the 50 end of II is phosphorylated and we make the last nucleotide in II a 20 -deoxy T in order to prevent circularization and in an attempt to give stability against exonucleolysis in cell extracts.

with toluidine blue and eluted from crushed gel slices overnight in 0.3 M sodium acetate pH 5.3, 1 mM EDTA and 0.1% SDS. Acrylamide is removed by phenol extraction, followed by chloroform extraction and ethanol precipitation. The precipitate is washed twice in 70% (v/v) ethanol, dried, and resuspended in water. The final yield after these steps is typically 25–30%. The activated succinimidyl esters of the Cy3 and Cy5 dyes are obtained from GE Healthcare. The content of a gel pack varies from lot to lot from 0.1 to 0.3 mg. The dye is resuspended in 20 ml of DMSO and mixed with 2–5 nmol of oligonucleotide in 0.1 M sodium bicarbonate buffer at pH 9.0. (Either fresh stocks of the bicarbonate buffer should be prepared or stocks should be frozen and the pH of the buffer checked before using.) The condensation is done at 60  C for 30 min. Four hundred microliters of

36

John Abelson et al.

3⬘ A TATTT CATA C A ACTACA

P

5⬘...GU ACC GUAUAAA GUAUGUUGAUGU

3⬘ A ACB

5⬘U A

Acceptor

UCC

0.3 M Na-acetate, pH 5.3, is added and the derivitized oligonucleotides are precipitated with two volumes of ethanol by incubation for 20 min at 80  C. The precipitate is washed at least three times with 70% (v/v) ethanol until all of the free dye has been removed. The pellets are dried and resuspended in water at a concentration of 100 mM. Condensation of the dye with the precursor is usually about 50% efficient. The dye-labeled oligonucleotide can be separated from the uncoupled oligonucleotide by electrophoresis on a denaturing, 8 M urea, 20% polyacrylamide gel, but this is a step of low yield and we have usually used dye-coupled RNAs without further purification. The dye-coupled oligonucleotides were joined together by T4 RNA ligase 1. A DNA splint, partially complementary to the 30 end of oligonucleotide I and the 50 end of oligonucleotide II, loops out the ends of the RNA so that they resemble a nicked tRNA anticodon loop, a good substrate for T4 RNA ligase 1 (Stark et al., 2006; Fig. 3.3). We have also used T4 RNA Ligase 2 for the ligation of oligonucleotides (Ho and Shuman, 2002). In this case, the splint is exactly complementary to the donor and acceptor ends, producing a gapped substrate. Although we have less experience using T4 RNA Ligase 2, the efficiency of ligation for both ligases is generally 30–50%. The ligated pre-mRNA is separated from unligated oligonucleotides by electrophoresis on a denaturing, 8 M urea, 6% polyacrylamide gel (Fig. 3.4). The recovery of the ligated oligonucleotides from the gel at this stage is sometimes poor. We believe that this is because the hydrophobicity of the dye-labeled oligos leads to a loss of materials in the phenol extraction step. We have had better yields from the gel by employing electroelution of excised bands, using a device manufactured by International Biotechnologies Inc. (New Haven, CT). Because of the cumulative losses in the multiple steps involved in this procedure, the overall efficiency of labeled pre-mRNA synthesis is low, but because Dharmacon supplies us with nanomoles of synthetic oligonucleotide and femtomoles are required for a splicing assay, the yield is adequate. Of course far less material is required for the single-molecule assays.

Donor

A GCAGAUCAUGUUUUUUAAGCCG...3⬘ T CGTCTAGTACAAAAAAT 5⬘

Splint

Figure 3.3 Design of a ligation splint for T4 RNA ligase 1. The black line separating the two parts of the splint and opposite the looped out ends of donor and acceptor is a 9 carbon linker. Good ligation efficiency is obtained when the ratio of acceptor to donor to splint is 1:1:1.

37

Preparation of Fluorescent Pre-mRNA Substrates

Cy5

Cy3

I–II I II

Figure 3.4 Ligation of oligonucleotides I and II to obtain Ubc4 pre-mRNA. Oligonucleotides I and II were ligated together with T4 RNA Ligase 1 employing the splint shown in Fig. 3.3. The ratio of I:II:splint was 1:1:1 (250 pmol of each). The reaction was incubated with 150 units T4 RNA ligase 1 (New England Biolabs) for 3 h at 37  C and the products were separated on a denaturing, 8 M urea, 6% polyacrylamide gel. The Typhoon Fluorimager scan of Cy5 is shown.

4. Do the Dyes Affect the Efficiency of Splicing? It is crucial to demonstrate that the dyes themselves do not inhibit the splicing reaction. We have prepared pre-mRNAs with dyes located at uridines in positions 7 in exon 1 and þ4 in exon 2, as well as in positions 12 and þ4, and 12 and þ10. All of these pre-mRNAs are spliced with the same efficiency and with an efficiency similar to that of 32P labeled premRNA not containing dyes (Fig. 3.5). The single-molecule experiments described in Abelson et al. (2010) have employed pre-mRNAs with dyelabeled uridines at position 7 and þ4.

5. Mutant Pre-mRNAs To validate the single-molecule FRET results, we have employed mutant pre-mRNAs that affect either the first or the second step of splicing. Fluorescent Ubc4 pre-mRNAs were synthesized with an A-to-C substitution in the branch point (b.p.) adenosine or a G-to-C substitution at the

38

John Abelson et al.

I7–II4

I12II4

− + + − + +

ATP

I12II10 − + +

Lariat-intron 2 Pre-mRNA

mRNA

I (76)

II (60)

−7

+4

Figure 3.5 The position of dyes in the pre-mRNA does not affect the efficiency of splicing. Synthetic Ubc4 pre-mRNA was prepared with dyes coupled to uridines at positions  7, þ 4,  12, þ4 and  12, þ 10. Splicing reactions were assayed by denaturing polyacrylamide gel electrophoresis of the RNAs and the gels were scanned with a Typhoon Fluorimager. The Cy5 scan is shown.

30 -splice site (30 ss). Figure 3.6 shows that the mutant pre-mRNAs behave as expected (Vijayraghavan et al., 1989). The 30 ss mutant is a substrate for the first step of splicing but is blocked in the second step and therefore accumulates the reaction intermediates. The b.p. mutant is blocked in the first step.

6. Tethering the Pre-mRNA to the Microscope Slide In our initial approach, we have tethered the Ubc4 pre-mRNA to the slide via a tether oligonucleotide that is complementary to exon 2 and contains a 50 biotin. The sequence of that oligonucleotide is 0

5 biotin-mAmAmCmAmUmGmAmAmGmUmAmGmGmUmGmGmA

39

Preparation of Fluorescent Pre-mRNA Substrates

+4

−7

UACUAAC

Cy3

C

UAG

C

Cy5

Cy5

wt.

3⬘ss

b.p. Lariat-ex II Pre-mRNA

mRNA

Figure 3.6 Pre-mRNA mutations inhibit splicing as expected. Ubc4 pre-mRNAs were synthesized with mutant changes in the branch point adenosine and the 30 splice site guanosine as shown. The fluorescent RNAs were assayed for splicing in vitro and the RNA products were separated by denaturing polyacrylamide gel electrophoresis. Shown is the Cy5 Typhoon Fluorimager scan of the gel.

This is a ribooligonucleotide in which all of the 20 -hydroxyls are methylated (m). This provides resistance to nuclease digestion. DNA tethers cannot be used in these experiments because in the splicing extract they will direct RNaseH cleavage of the substrate. Incorporation of 50 -biotin to the Ubc4 pre-mRNA is a preferable approach. We have shown that this alteration does not affect splicing (data not shown).

7. Summary and Conclusion We have described the identification of a yeast pre-mRNA splicing substrate, Ubc4, that contains a small intron facilitating its chemical synthesis. The synthetic pre-mRNA can be labeled with fluorescent dyes, providing a substrate for smFRET studies of spliceosomal assembly. The preliminary use of this substrate to study the dynamics of splicing in vitro has been described (Abelson et al., 2010; Blanco and Walter, 2010). Our laboratory is currently working on the development of fluorescent U6 and U2 small nuclear (sn) RNAs. These RNAs, synthesized from synthetic oligonucleotides as described earlier, can be incorporated into the (sn)RNA–protein complexes (snRNPs) in vitro by a method of reconstitution in which the (sn)RNA in the

40

John Abelson et al.

extract is inactivated by DNA-directed RNase H. In principle, spliceosome assembly can also be studied via the incorporation of dyes into spliceosomal proteins.

ACKNOWLEDGMENTS We acknowledge the participation of Dan Ryan, Jeffrey Pleiss, and Tommaso Villa in various stages of this project. The application of these substrates to the study of spliceosome assembly via smFRET has been a collaboration with Nils Walter’s laboratory at the University of Michigan. The work at UCSF was supported by an American Cancer Society Research Professor of Molecular Genetics award to C. G., by NIH grant GM021119 to C. G., and by a grant from the Agouron Institute to J. A.

REFERENCES Abelson, J., Blanco, M., Ditzler, M. A., Fuller, F., Aravamudhan, P., Wood, M., Villa, T., Ryan, D., Pleiss, J. A., Maeder, C., Guthrie, C., and Walter, N. G. (2010). Conformational dynamics of single pre-mRNA molecules during spliceosome assembly and splicing. Nat. Struct. Mol. Biol. 17, 1767–1775. Blanco, M., and Walter, N. G. (2010). Analysis of complex single molecule FRET time trajectories. Methods Enzymol. 472, 153–178. Ho, C. K., and Shuman, S. (2002). Bacteriophage T4 RNA ligase 2 (gp24.1) exemplifies a family of RNA ligases found in all phylogenetic domains. Proc. Natl. Acad. Sci. USA 99, 12709–12714. Lin, R. J., Newman, A. J., Cheng, S. C., and Abelson, J. (1985). Yeast mRNA splicing in vitro. J. Biol. Chem. 260, 14780–14792. Pleiss, J. A., Whitworth, G. B., Bergkessel, M., and Guthrie, C. (2007). Transcript specificit in yeast pre-mRNA splicing revealed by mutations in core spliceosomal components. PLoS Biol. 5, e90. Stark, M. R., Pleiss, J. A., Deras, M., Scaringe, S. A., and Rader, S. D. (2006). An RNA ligase-mediated method for the efficient creation of large, synthetic RNAs. RNA 12, 2014–2019. Vijayraghavan, U., Company, M., and Abelson, J. (1989). Isolation and characterization of pre-mRNA splicing mutants of Saccharomyces cerevisiae. Genes Dev. 3, 1206–1216.

C H A P T E R

F O U R

Nanovesicle Trapping for Studying Weak Protein Interactions by Single-Molecule FRET Jaime J. Benı´tez, Aaron M. Keller, and Peng Chen Contents 42 44 45 46 47 47 49

1. Introduction 2. Nanovesicle Trapping Approach 2.1. Lipid selection 2.2. Lipid nanovesicle preparation and protein trapping 3. smFRET Measurements of Weak Protein–Protein Interactions 3.1. Surface immobilization of nanovesicles 3.2. Control experiments 3.3. Application to weak interactions between intracellular copper transporters 4. Single-Molecule Kinetic Analysis of Three-State Protein–Protein Interactions 5. Further Developments 6. Concluding Remarks Acknowledgments References

51 54 57 58 59 59

Abstract Protein–protein interactions are fundamental biological processes. While strong protein interactions are amenable to many characterization techniques including crystallography, weak protein interactions are challenging to study because of their dynamic nature. Single-molecule fluorescence resonance energy transfer (smFRET) can monitor dynamic protein interactions in real time, but are generally limited to strong interacting pairs because of the low concentrations needed for single-molecule detection. Here, we describe a nanovesicle trapping approach to enable smFRET study of weak protein interactions at high effective concentrations. We describe the experimental procedures, summarize the application in studying the weak interactions between intracellular copper transporters, and detail the single-molecule kinetic analysis Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York, USA Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72016-4

#

2010 Elsevier Inc. All rights reserved.

41

42

Jaime J. Benı´tez et al.

of bimolecular interactions involving three states. Both the experimental approach and the theoretical analysis are generally applicable to studying many other biological processes at the single-molecule level.

1. Introduction Protein–protein interactions are essential for cellular functions including protein folding, cell signaling, and metal trafficking (Gragerov et al., 1992; Hall, 1992; Huffman and O’Halloran, 2001). The strength of protein–protein interactions can vary widely depending on the proteins involved. Strong protein interactions can have equilibrium dissociation constants (KD) of a few picomolar (10
12M), for example antigen–antibody interactions, for which tight binding is crucial (Nooren and Thornton, 2003). Weak protein interactions can have KD’s of a few micromolar to millimolar (10
6–10
3M), for example, interactions between metallochaperones and their target proteins, for which dynamic binding and unbinding are necessary to have many interaction turnovers (Banci and Rosato, 2003; Cobine et al., 2006; Huffman and O’Halloran, 2001; Kim et al., 2008; Lutsenko et al., 2007; Rosenzweig, 2001; Strausak et al., 2003). For understanding their fundamental properties, strong protein interactions are amenable to characterization by ensemble measurements, as stable interaction complexes can form even at dilute solution conditions. Stable protein complexes can further be crystallized for structural determination down to atomic resolution. In contrast, weak protein interactions are challenging to characterize in ensemble measurements for several reasons: (1) they are dynamic and stochastic, making synchronization of molecular actions often necessary; (2) the steady-state concentrations of interaction intermediates are often low, making detection difficult; and (3) the presence of multiple interaction intermediates can complicate ensemble-averaged measurements. To study these weak protein interactions, single-molecule measurements offer several advantages: (1) no synchronization of molecular reactions is necessary; (2) the molecular reactions, including the formation, interconversion, and dissolution of interaction intermediates, are followed in real time; and (3) only one molecular state, be it an intermediate, is observed at any time point, enabling the resolution of complex reaction kinetics. Single-molecule fluorescence resonance energy transfer (smFRET), with its inherent distance dependence in the nanometer scale, is particularly suited for probing dynamic protein–protein interactions, which is accompanied by changes in protein–protein distances. There are challenges to overcome, however, before smFRET can be applied to study weak protein interactions. The primary challenge is the concentration limit. Single-molecule fluorescence measurements are generally done at low

Nanovesicle Trapping for Weak Protein Interactions

43

concentrations (<10
9M) to spatially separate fluorophores so that there is less than one fluorophore (or one pair of fluorophores) on average in the detection volume (about 10
16–10
15 l) monitored in confocal microscopy or total internal reflection microscopy. This low concentration range limits single-molecule protein interaction studies to strong interacting pairs, whereas weak protein interactions need to be studied at much higher concentrations (>10
6M) to favor complex formation. To overcome this concentration limit, one needs to decrease the effective detection volume to  10
19–10
21 l, so that at concentrations of up to 10
6–10
4M there is no more than one fluorophore on average found in it (Laurence and Weiss, 2003). This can be done by reducing the laser excitation volume or by confining molecules in space. For reducing the excitation volume, Webb, Craighead, and coworkers have fabricated zero-mode waveguides made of metal-clad wells on top of a silica substrate (Levene et al., 2003). The diameter of these wells is much smaller than the wavelength of the excitation light, and therefore, light shining at the silica substrate cannot propagate through the wells. Thisblockage of light propagation reduces the light excitation to an evanescent electromagnetic field close to the silica substrate surface, leading to reduction of the laser excitation volume to 10
21 l. Using this approach, Webb, Craighead, and coworkers have studied the reactions of individual DNA polymerase molecules that have substrate binding affinity in the micromolar range. As these zero-mode waveguides are open reaction containers, a big advantage is easy exchange of solutions for changing reaction conditions. A disadvantage is the proximity of a metal surface to the fluorophore; the metal surface can influence the fluorophore’s fluorescence properties, such as its intensity and fluorescence lifetime. To follow individual molecules over time, the molecules also have to be immobilized on the silica surface at the bottom of the wells, which can introduce nonspecific surface interactions. For confining molecules spatially, trapping with nanometer-sized lipid vesicles is an effective approach (Fig. 4.1), which was initially used in single-molecule studies of enzyme reactions (Chiu et al., 1999), protein folding (Boukobza et al., 2001; Haran, 2003; Rhoades et al., 2003, 2004), and nucleic acid conformation dynamics (Lee et al., 2005; Okumus et al., 2004). Because of the confined volume, the effective concentration of a single-molecule inside a nanovesicle can be as high as tens of micromolar, while the overall concentration of the nanovesicles can be kept low to maintain the single-molecule detection condition. Using this nanovesicle trapping approach combined with smFRET measurements, Ha and coworkers have studied dynamic protein–nucleic acid interactions (Cisse et al., 2007), and we have studied weak protein–protein interactions at high effective concentrations (Benitez et al., 2008, 2009). In this chapter, we describe in detail how nanovesicle trapping, combined with smFRET

44

Jaime J. Benı´tez et al.

D + A

Lipid bilayer/PEG/BSA Quartz

Biotin Streptavidin

Figure 4.1 Schematics of nanovesicle trapping of two proteins labeled with a FRET donor–acceptor pair for smFRET studies.

measurements, can be used to characterize weak, dynamic protein interactions at the single-molecule level. We also detail the single-molecule kinetic analysis of bimolecular interactions that show three FRET states.

2. Nanovesicle Trapping Approach Nanovesicle trapping is an effective approach in reducing the effective detection volume to enable high concentration studies at the single-molecule level. This approach also offers several other advantages: (1) the lipid membrane enclosure mimics biological environments inside cells or organelles; (2) the membrane prevents nonspecific interactions between the protein and the glass surface because molecule immobilization is done via tethering the nanovesicle (Fig. 4.1); nevertheless, nonspecific interactions with the lipid membrane may occur; control experiments must be performed to check this possibility (see below); (3) the diameter of vesicles can be varied from a few hundred nanometers down to 50 nm, covering effective concentrations of up to 24 mM for a single-molecule inside (Fig. 4.2); and (4) for protein–protein interaction studies, interactions between molecules of the same type, if occurring, can be selectively discarded in the data analysis stage by examining only the nanovesicles that contain molecules of different types.

45

Nanovesicle Trapping for Weak Protein Interactions

Effective concentrations (mM)

100 10 1 0.1 0.01 250

500

750

1000

Vesicle diameter (nm)

Figure 4.2 Dependence of the effective concentration of a single molecule on the diameter of the nanovesicle. The solid symbols indicate a few commercial available membrane pore sizes for preparing nanovesicles.

In this section, we describe the experimental details of preparing nanovesicles to trap two different proteins for protein interaction studies. The procedures largely follow those of Haran and Ha (Boukobza et al., 2001; Okumus et al., 2004).

2.1. Lipid selection The lipids for forming the membrane bilayer of the nanovesicles contain two components: one major lipid (99%) that dominates the behavior of the membrane bilayer and the other minor lipid (1%) that contains a biotin group for surface immobilization. The chemical nature and the gel-to-liquid phase transition temperature (Tm) of the major lipid are important here. The lipid must not significantly interact with the proteins and perturb the protein interactions. A lipid with a net zero charge is preferred, as it is less likely to interact with soluble, largely hydrophilic proteins (Boukobza et al., 2001). Usually, the Tm of the major lipid should be much lower than the temperature for the single-molecule experiments, so the lipid bilayer stays in the fluidic liquid phase. A common major lipid used in single-molecule applications is Egg PC, extracted from egg yolk and  99% of which is L-aphosphatidylcholine. Its Tm is about
2  C (Silvius, 1983), so its bilayer is in the liquid phase at room temperature. Many other lipids with different Tm and charge properties are available and can be used for preparing vesicles (Silvius, 1983). For the biotinylated minor lipid, Biotinyl-cap PE (1,2dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(cap biotinyl)) is commonly used.

46

Jaime J. Benı´tez et al.

2.2. Lipid nanovesicle preparation and protein trapping The procedure consists of two major steps: (1) preparation of dry lipid film and hydration of the lipid film with buffer containing fluorescently labeled proteins to form vesicles and trap proteins inside and (2) extrusion of the formed vesicles through a polycarbonate membrane with well-defined pore diameters to make unilamellar vesicles of defined size. Using 100-nm pore-size membranes for extrusion, the multilamellar vesicles were estimated to be less than 2% (Hope et al., 1985). 2.2.1. Lipid film preparation and hydration 1. Prepare lipid stock solutions in chloroform at 100 mg ml
1 and store them at
20  C in a desiccator. The biotinylated lipid stock solution is prepared at 1 mg ml
1. 2. Transfer aliquots of lipid solutions into a clean glass test tube, forming a solution of  99% major lipid and 1% biotinylated minor lipid. Use enough amounts for a final total lipid concentration of 5 mg ml
1 upon hydration. Dry under a nitrogen flow until a thin lipid film is formed on the wall of the test tube. The lipid film can be further put under vacuum for 1–2 h to remove residual chloroform. 3. Hydrate the lipid film with the solution containing a mixture of the two labeled proteins under study. The concentration of each of the fluorescently labeled proteins should be close to the targeted effective concentration when the protein is trapped inside the nanovesicle. For example, for trapping a pair of protein molecules in 100 nm diameter nanovesicles, the concentration of each protein should be 3 mM, whereas for trapping in 200 nm diameter nanovesicles, use 400 nM concentration. The efficiency of cotrapping a pair of molecules can be increased if the hydration and trapping are done under conditions where the protein pair has maximum binding affinity for each other. 4. Briefly vortex the hydrated solution to detach the lipid film from the wall of the test tube. Incubate the solution for 10 min to 1 h at a temperature of at least 10  C above the Tm of the major lipid. The hydrated lipids will spontaneously form large multilamellar vesicles. Further freeze–thaw cycles (5–10 times) of the vesicle solution using liquid nitrogen and warm water can induce cracks in the membrane, which can improve entrapment of small molecules. Here, care must be taken that the freeze–thaw cycles do not denature the proteins; circular dichroism spectroscopy can be used to check the folding state of the protein. 2.2.2. Preparation of unilamellar nanovesicles via extrusion Unilamellar nanovesicles are formed by extrusion of the vesicle solution through a polycarbonate membrane with nanometer-sized pores (Hope et al., 1985; Johnson et al., 2002; MacDonald et al., 1991). Extrusion should

Nanovesicle Trapping for Weak Protein Interactions

47

be performed at a temperature of at least 10  C above the Tm of the major lipid. The Avanti mini extruder (Avanti Polar Lipids, Inc.) is handy for this purpose. Polycarbonate membranes of different pore diameters are available, ranging from 50 nm to 1 mm. The nanovesicle diameter obtained after extrusion follows a Gaussian distribution, the width of which is dependent on the number of passes through the extruder; the more passes, the narrower the distribution (Hope et al., 1985; MacDonald et al., 1991). The diameter distribution can be checked using dynamic light scattering measurements. We normally perform tens of passes, significantly more than what is suggested by Avanti. The number of molecules trapped within the nanovesicles follows a Poisson distribution, the average occupation number depending on the protein/lipid ratio in the hydration step (Boukobza et al., 2001). The exact occupancy of each nanovesicle can be determined by single-molecule fluorescence imaging (see Section 3.2).

3. smFRET Measurements of Weak Protein–Protein Interactions The smFRET experiments consist of (1) immobilization of nanovesicles in a flow cell and (2) real-time imaging using total internal reflection fluorescence microscopy. The microscope is equipped with two-color detection for imaging the fluorescence of the FRET donor and acceptor simultaneously.

3.1. Surface immobilization of nanovesicles To follow the protein–protein interactions inside each nanovesicle over time, the nanovesicles need to be immobilized on a surface. Most commonly, a biotin–avidin linkage is used. Biotinylated lipids in the nanovesicle membrane are used to bind avidins (e.g., streptavidin or neutravidin), which in turn are bound to a biotin-modified surface. We have used three different schemes to modify the surface with biotins, all of which yield similar results: (1) coating the surface with a lipid bilayer containing biotinylated lipids, (2) coating with biotinylated bovine serum albumin (BSA), and (3) coating with partially biotinylated polyethylene glycol (PEG). 3.1.1. Lipid bilayer coating This surface modification scheme takes advantage of the spontaneous fusion of lipid vesicles onto clean quartz surfaces to form a lipid bilayer (Boxer, 2000; Brian and McConnell, 1984), over which the nanovesicles can be attached. The lipids used for this bilayer can be the same as those used for the nanovesicles, for example, 99% Egg PC and 1% Biotinyl-cap PE.

48

Jaime J. Benı´tez et al.

1. The procedure for lipid preparation is the same as described earlier. The lipid film is prepared first and then hydrated with buffer in the absence of proteins. 2. The hydrated solution is sonicated for 30 min to 1 h until clarity. The sonication here breaks large multilamellar vesicles to form small unilamellar vesicles, which can spontaneously fuse to clean quartz surfaces. The distribution of vesicle sizes is not important here. 3. Incubate the quartz substrate with the solution containing small unilamellar vesicles at a total lipid concentration of 1–5 mg ml
1 for 1 h. Wash out excess lipids from the quartz surface with buffer. 4. Vesicle preparation, surface coating, and washing should all be performed at a temperature of at least 10  C above the Tm of the major lipid. One problem with using Egg PC for the supported bilayer is that at room temperature the bilayer exists in the liquid phase, so the attached nanovesicles are mobile. The nanovesicle mobility can be reduced by increasing the percentage of biotinylated lipid so that each nanovesicle is anchored to the supported bilayer by multiple biotin–avidin linkages (Okumus et al., 2004). Nevertheless, many nanovesicles still remain mobile as we observed in our experiments. To alleviate this mobility problem, we have used another lipid, DPPC (1,2-dipalmitoyl-sn-glycero-3-phosphocholine), which has a Tm of 41  C. Because of its high Tm, DPPC exists in the gel phase at room temperature, resulting in a mostly immobile lipid support. 3.1.2. BSA coating BSA can bind to quartz surfaces strongly via nonspecific interactions (Rasnik et al., 2005), and therefore, biotinylated BSA can be used to coat the quartz surface to immobilize nanovesicles: 1. Prepare 1 mg ml
1 biotinylated BSA solution and incubate on the quartz substrate for 30 min to 1 h. 2. Wash out excess biotin–BSA with buffer. The BSA coating is easy to perform and can prevent rupture and fusion of nanovesicles to the quartz surface. (In case some bare patches on the glass surface exist because of incomplete coating with BSA, vesicle fusion to the glass surface can form patches of lipid bilayer to fill them up.) 3.1.3. PEG coating Covalent functionalization of a quartz surface with partially biotinylated PEG is another scheme for immobilizing nanovesicles. The quartz surface is first covalently functionalized with amine groups, which are then covalently linked to PEG via succinimidyl ester chemistry.

Nanovesicle Trapping for Weak Protein Interactions

49



For amine modification: 1. Prepare a fresh solution of 1.5–2% amino silane reagent (Vectabond, Vector Laboratories) in acetone. (Other types of amino silane reagents work too, for example, 3-aminopropyltriethoxysilane.) 2. Drop 200 ml of the amino silane solution onto the quartz slide. 3. Incubate for 5 min and then wash extensively with ultra-filtered, deionized water for 1 min. Dry the slides with nitrogen and store under a dry environment.  For PEG modification: 1. Prepare a solution of 98–99% m-PEG-SPA-5000 and 1–2% biotinPEG-NHS-3400 (Nektar Therapeutics, JenKem Technology Inc., or SunBio, USA) in 100 mM NaHCO3, pH 8.2. 2. Drop 200 ml of the PEG solution onto an amine-functionalized slide and sandwich it with another slide. Place parafilm spacers in between to prevent squeezing out the solution. Incubate for 4 h in the dark. 3. Wash slides thoroughly with nanopure water and dry with nitrogen for usage.

3.2. Control experiments 3.2.1. Lipid–protein interactions To check whether the fluorescently labeled proteins have nonspecific interactions with the lipid membrane, one can coat the quartz surface with a lipid bilayer and flow in solutions containing high concentrations (e.g., 100 nM) of labeled proteins. After washing the flow cell with fresh buffer and imaging the single-molecule fluorescence, the number of molecules that are immobilized on the lipid bilayer by nonspecific interactions can be counted. Comparing the number of nonspecifically bound molecules to the number of molecules detected using specific biotin–avidin immobilization of nanovesicles provides an estimate of the extent of nonspecific interactions between the protein and the lipid membrane (Benitez et al., 2008, 2009; Okumus et al., 2004). 3.2.2. Occupancy of nanovesicles The nanovesicle trapping procedure will result in a distribution of occupancy of individual nanovesicles. The nanovesicle occupation is important to verify when using smFRET to study weakly interacting pairs. Under normal smFRET measurements, only the FRET donor is continuously excited and exhibits fluorescence. The FRET acceptor emits ideally only when it is close to the donor-labeled protein (e.g., upon protein–protein interaction) and is excited via energy transfer. The possible presence of multiple acceptor-labeled proteins within a nanovesicle can adversely affect quantitative determination of protein–protein interaction kinetics. Control

50

Jaime J. Benı´tez et al.

experiments are necessary to determine the distribution of occupancy of nanovesicles under the trapping conditions. To do so, one can use two different lasers to excite the FRET donor and acceptor separately. For example, for the Cy3–Cy5 FRET pair, the Cy3 fluorescence can be directly imaged by excitation with a 532-nm laser and Cy5 fluorescence with a 637-nm laser. With the nanovesicles already loaded with fluorescent proteins and immobilized on the surface, the control experiments follow: 1. Directly excite the donor dye and record a movie of fluorescence intensity. 2. Switch to the second laser to excite the acceptor dye in the same area and record a fluorescence movie. 3. Analyze both movies to obtain fluorescence trajectories and positions of individual molecules. Use the number of photobleaching steps in the fluorescence intensity trajectory to determine the number of donor (or acceptor) molecules in the nanovesicle. 4. Check the position colocalization if the donor and acceptor dyes belong to the same nanovesicle. In our experience, incidental vesicle colocalization due to limited spatial resolution is minimal when the surface density of protein containing vesicles is smaller than 0.2 mm
2. For weakly interacting protein pairs, such as Hah1 and MBD4 (see Section 3.3), the cotrapping efficiency is low. Among 340 nanovesicles containing either Hah1–Cy5 or MBD4–Cy3, only 21 of them contain a Hah1–Cy5 and a MBD4–Cy3. The number of acceptor molecules can also be checked during normal smFRET measurements. One can first use the donor-exciting laser for smFRET while recording a fluorescence movie. In the later part of the movie, the acceptor-exciting laser is turned on to excite the acceptor dye until the acceptor photobleaches. The photobleaching events in the acceptor intensity will indicate the number of acceptor molecules in the nanovesicle. In this way, one is sure to examine only single pairs of protein molecules. 3.2.3. FRET differentiation of acceptor-blinked/bleached states from the dissociated state of protein interactions Organic fluorescent dyes show blinking behavior, that is, the fluorescence intensity sometimes switches off temporarily. Although fluorescence blinking can be suppressed significantly by using an oxygen scavenging system and triplet quenchers (e.g., Trolox; Rasnik et al., 2006), occasional blinking of the FRET acceptor is problematic, as it would result in an apparently low FRET efficiency (EFRET ¼ IA/(IA þ ID), where IA and ID are the acceptor and donor fluorescence intensities), which could be mistaken as that of the dissociated state of protein–protein interactions. Fortunately, using nanovesicle trapping and Cy3–Cy5 as the FRET pair,

Nanovesicle Trapping for Weak Protein Interactions

51

the Cy5-blinked state has clearly lower EFRET than that of the dissociated state from control experiments (Benitez et al., 2008, 2009). As far as the apparent EFRET is concerned, the acceptor-blinked state is effectively the same as that in the absence of the acceptor and that of the acceptor photobleached state. Therefore, the apparent EFRET from nanovesicles that merely contain a donor molecule serves as a control for signal from the acceptor blinked state (Fig. 4.3A). The determined apparent EFRET with one Cy3 only is 0.04  0.05, which is the same as Cy5-blinked/bleached state of a Cy3–Cy5 pair (Fig. 4.3C). The dissociated state can be mimicked by a nanovesicle containing a free donor and a free acceptor (Fig. 4.3B), as the free dyes do not interact with each other. Here, the existence of both a donor and an acceptor must be confirmed by separate laser excitations (Fig. 4.3B). Under 532-nm excitation, the apparent EFRET is 0.15  0.14 (Fig. 4.3C); the larger value here compared with that of Cy5-blinked state is likely due to the residual direct excitation of Cy5 fluorescence by the 532-nm laser and some energy transfer of Cy3–Cy5 due to their confined coexistence inside the nanovesicle.

3.3. Application to weak interactions between intracellular copper transporters We applied the nanovesicle trapping approach to enable smFRET studies of the weak, dynamic interactions between the human intracellular copper chaperone Hah1 and the fourth metal-binding domain (MBD4) of the copper transporting ATPase Wilson disease protein (WDP) (Benitez et al., 2008, 2009). The interactions between Hah1 and WDP mediate the copper transfer from Hah1 to the MBDs of WDP, an essential process for safe trafficking of copper ions in human cells (Banci and Rosato, 2003; Cobine et al., 2006; Huffman and O’Halloran, 2001; Kim et al., 2008; Lutsenko et al., 2007; Rosenzweig, 2001; Strausak et al., 2003). Because of the low affinity of the Hah1–WDP interaction (KD  10
6 M), their interaction dynamics have been challenging to quantify in ensemble measurements. Nanovesicle trapping offers an ideal platform to examine their interactions at the single-molecule level using smFRET. We labeled Hah1 with the acceptor dye Cy5 and MBD4 with the donor dye Cy3, using maleimide chemistry at specific cysteine residues, and cotrapped them in 100-nm diameter nanovesicles. smFRET trajectories reveal their dynamic interactions (Fig. 4.4A). These trajectories show three different EFRET states: E0 (0.2) is the dissociated state, and E1 ( 0.5) and E2 ( 0.9) are two different interaction complexes. Transitions between E0 and E1 and between E0 and E2 correspond to the binding/ unbinding processes for forming complexes 1 and 2. The transitions between E1 and E2 correspond to the interconversions between the two

52

Jaime J. Benı´tez et al.

Intensity (a.u.)

A 30 k

Cy3 Cy5

15 k 0 0

50

100

150

Time (s) B

637 nm

Intensity (a.u.)

532 nm 20 k

Cy3 Cy5

10 k 0 0

50

100

150

Time (s)

Occurrence

C 60

20

30

10

0

0.0

0.3

0 0.6

EFRET

Figure 4.3 smFRET control experiments for acceptor blinked/bleached states and the dissociated state. (A) Two-color fluorescence intensity trajectories of a nanovesicle containing a single Cy3 molecule using 532-nm laser excitation. The Cy3 molecule photobleaches at the  62th second. (B) Two-color fluorescence intensity trajectories of a nanovesicle containing a single Cy3 and a single Cy5. The 532-nm laser is on throughout; the 637-nm laser was turned on at the  75th second. The Cy3 photobleaches at the 25th second; the Cy5 molecule photobleaches at  125th second. The first 25 seconds mimics the dissociated state of a Cy3–Cy5 pair. (C) Histograms of the apparent EFRET (¼IA/(IA þ ID); ID and IA are the fluorescence intensities of the donor and acceptor, respectively) for nanovesicles containing a single Cy3 (line patterned columns) and for nanovesicles containing a free Cy3 and Cy5 molecule (clear columns).

complexes. Figure 4.4B gives the interaction scheme between Hah1 and MBD4. The kinetic constants of all interaction processes can be extracted by analyzing the distributions of dwell times in each FRET state (Fig. 4.4C– H, Section 4). The direct observation of the interconversion dynamics between the two interaction complexes is particularly exciting here, as it enables determination of both the forward and the reverse interconversion rate constants (see Section 4)—ensemble characterization can often only

53

Nanovesicle Trapping for Weak Protein Interactions

A Counts

B

Cy5-Hah1

Cy3-MBD4

20 k

1

(HAH1-MBD4) E1 k–1 HAH1 + MBD4 k k3 –3 E0 k2 k1

10 k

EFRET

0k 1.0

E2

0.5

E1 E0

0.0 0

10

C

20 Time (s)

20

(k1 + k2)[P] = 0.8 ± 0.1 s–1

0

40

15

2

t0 → 1

6

F

10

0

(k1 + k2)[P] = 1.0 ± 0.1 s–1

2

t1 → 0

6

10

k–1 + k3 = 1.4 ± 0.1 s–1

20

10

6

t0 → 2

10

0

2

6 t2 → 0

10

k–2 + k–3 = 1.7 ± 0.2 s–1

20 10

10 2

0

H 30

20

0

k–2 + k–3 = 2.1 ± 0.2 s–1

20

G

30

(HAH1-MBD4) E2

E k–1 + k3 = 1.2 ± 0.1 s–1

30

10

2

40

D 30

No of events

30

k–2

2

t1 → 2

6

10

0

2

6 t2 → 1

10

Figure 4.4 smFRET measurements of weak protein interaction dynamics in a nanovesicle. (A) Two-color fluorescence intensity (upper) and corresponding apparent EFRET (lower) trajectories of a Cy5–Hah1 and a Cy3–MBD4 trapped in a 100-nm nanovesicle. (B) Interaction scheme between Hah1 and MBD4. (C–H) Distributions of the six types of dwell times from the EFRET trajectories of Hah1–MBD4 interactions. Solid lines are exponential fits; insets give the exponential decay constants and their relations to the protein interaction rate constants in (B). [P] is the effective concentration ( 3 mM) of a single molecule in a 100-nm vesicle. The individual rate constants are k1 ¼ (1.6  0.2)  105 M
1 s
1, k
1 ¼ 0.88  0.04 s
1, k2 ¼ (1.4  0.2)  105 M
1 s
1, k
2 ¼ 1.3  0.1 s
1, k3 ¼ 0.42  0.04 s
1, and k
3 ¼ 0.7  0.1 s
1. Data in (A, C–H) adapted with permission from Benitez et al. (2008, 2009). Copyright 2008 American Chemical Society. (See Color Insert).

determine the sum of the forward and reverse rates for intermediate interconversion dynamics, as the interconversion dynamics are generally nonsynchronizable.

54

Jaime J. Benı´tez et al.

4. Single-Molecule Kinetic Analysis of Three-State Protein–Protein Interactions The interaction scheme between Hah1 and MBD4 can be generalized to that in Fig. 4.5A. An idealized EFRET trajectory showing three FRET states is given in Fig. 4.5B with different types of dwell times denoted. In this section, we derive the probability density functions of the dwell times involved in this three-state interactions, using single-molecule kinetic analysis (Benitez et al., 2008, 2009; Xie, 2001; Xu et al., 2009). We first consider the binding processes that occur during the dwell time t0 in the E0 state. Based on the interaction scheme shown in Fig. 4.5A, the processes occurring during t0 are summarized in Scheme 4.1. The ensemble rate equations for these kinetic processes are 0

d½A d½A  0
¼
¼ ðk1 þ k2 Þ½A½A  dt dt

E1 B

A k1 k–1 A + A⬘ E0

(4.1a)

k2 k–2

k3

k–3 C E2

B

EFRET

1.0 t2→ 0 0.5

t0→ 2

t1→ 2

E2

t2→ 1 t1→ 0 t 0→ 1

E1 E0

0.0 0

100

50

150

Time (s)

Figure 4.5 Generic kinetic scheme of protein interactions and corresponding EFRET trajectories. (A) Generalized kinetic scheme of a single-interacting pair with three FRET states: one dissociated state, A þ A0 , with a FRET value of E0; and two interaction complexes, B and C, with FRET values of E1 and E2, respectively. (B) Idealized three-state EFRET trajectories of an interacting pair; all six types of dwell times are denoted.

55

Nanovesicle Trapping for Weak Protein Interactions

k1

B

A + A⬘ k2

C

Scheme 4.1 Kinetic processes occurring during the dwell time t0 at the E0 state.

d½B 0 ¼ k1 ½A½A  dt

(4.1b)

d½C 0 (4.1c) ¼ k2 ½A½A  dt For the single-molecule reactions occurring in a nanovesicle, we have to consider the molecules in terms of their probabilities at time t, P(t). These rate equations then become


dP 0 ðtÞ dPA ðtÞ ¼ ðk1 þ k2 ÞPA ðtÞPv;A0 ;A ðtÞ ¼
A dt dt dPB ðtÞ ¼ k1 PA ðtÞPv;A0 ;A ðtÞ dt

(4.2a) (4.2b)

dPC ðtÞ (4.2c) ¼ k2 PA ðtÞPv;A0 ;A ðtÞ dt Here PA(t) is the probability of finding A at time t; PA0 (t), PB(t), and PC(t) are defined similarly; and PA(t) þ PB(t) þ PC(t) ¼ 1. Pv;A0 ;A ðtÞ is the conditional probability at time t of finding A0 within the same infinitesimal volume v where A is located, provided that A is found. Pv;A0 ;A ðtÞ is then Pv;A0 ;A ðtÞ ¼

PA0 ;A ðtÞ

(4.3) V Here PA0 ,A(t) is the conditional probability at time t of finding A0 within the entire space of the nanovesicle, provided that A is found; and V is the volume of the nanovesicle. Because whenever A is present, A0 is found, PA0 ,A(t) ¼ 1. Therefore, Pv;A0 ;A ðtÞ ¼ 1=V , which is the effective concentration (ceff) of one molecule inside the nanovesicle. We then have


dP 0 ðtÞ dPA ðtÞ ¼ ðk1 þ k2 Þceff PA ðtÞ ¼
A dt dt

(4.4a)

56

Jaime J. Benı´tez et al.

dPB ðtÞ ¼ k1 ceff PA ðtÞ dt

(4.4b)

dPC ðtÞ (4.4c) ¼ k2 ceff PA ðtÞ dt The initial conditions for solving these equations are PA(0) ¼ PA0 (0) ¼ 1, PB(0) ¼ 0, and PC(0) ¼ 0, with t ¼ 0 being the onset of each binding reaction. We can then evaluate the probability density f0(t) of the dwell time t0. The probability of finding a particular t0 is f0(t)Dt; and f0(t)Dt is equal to the sum of two probabilities: (1) the probability of molecule A and A0 to form B between t ¼ t and t þ Dt, which is DPB(t) ¼ k1ceffPA(t)Dt; and (2) the probability of molecule A and A0 to form C between t ¼ t and t þ Dt, which is DPC(t) ¼ k2ceffPA(t)Dt. In the limit of infinitesimal Dt, dðPB ðtÞ þ PC ðtÞÞ (4.5) ¼ ðk1 þ k2 Þceff PA ðtÞ dt Using the initial conditions to solve Eqs. (4.4a)–(4.4c) for PA(t), we get f0 ðtÞ ¼

R1

f0 ðtÞ ¼ ðk1 þ k2 Þceff exp½
ðk1 þ k2 Þceff t

(4.6a)

Clearly, 0 f0 ðtÞdt ¼ 1, as expected. The dwell time t0 can be further separated into two types: one, t0!1, that ends with a transition to the E1 state and the other, t0!2, that ends with a transition to the E2 state. We can also evaluate the corresponding probability densities f0!1(t) and f0!2(t) of the dwell times t0!1 and t0!2. The probability of finding a particular t0!1 is f0!1(t)Dt; and f0!1(t)Dt is equal to the probability for A and A0 to form B between t ¼ t and t þ Dt, which is DPB(t) ¼ k1ceffPA(t)Dt. The probability of finding a particular t0!2 is f0!2(t)Dt; and f0!2(t)Dt is equal to the probability for A and A0 to form C between t ¼ t and t þ Dt, which is DPC(t) ¼ k2ceffPA(t)Dt. In the limit of infinitesimal Dt: f0!1 ðtÞ ¼

dPB ðtÞ ¼ k1 ceff exp½
ðk1 þ k2 Þceff t dt

(4.6b)

f0!2 ðtÞ ¼

dPC ðtÞ ¼ k2 ceff exp½
ðk1 þ k2 Þceff t dt

(4.6c)

Expectedly, f0!1(t) þ f0!2(t) ¼ f0(t). Note the exponential decay constants of f0!1(t) and f0!2(t) are the same as that of f0(t), all equal to (k1þk2) ceff, the sum of the two parallel kinetic processes in Scheme 4.1. The ratio between the total occurrence N0!1 of dwell time t0!1 and the total

Nanovesicle Trapping for Weak Protein Interactions

57

occurrence N0!2 of dwell time t0!2 in the smFRET trajectories also carries important information: R1 f0!1 ðtÞdt k1 N0!1 ¼ R01 (4.6d) ¼ N0!2 k2 0 f0!2 ðtÞdt Similarly, we can derive the probability density function of the dwell time t1 on the E1 state, which can be separated into two types: t1!0 and t1!2, and that of the dwell time t2 on the E2 state, which can be separated into t2!0 and t2!1. The results are f1 ðtÞ ¼ ðk
1 þ k3 Þexp½
ðk
1 þ k3 Þt

(4.7a)

f1!0 ðtÞ ¼ k
1 exp½
ðk
1 þ k3 Þt

(4.7b)

f1!2 ðtÞ ¼ k3 exp½
ðk
1 þ k3 Þt

(4.7c)

N1!0 k
1 ¼ N1!2 k3

(4.7d)

f2 ðtÞ ¼ ðk
2 þ k
3 Þexp½
ðk
2 þ k
3 Þt

(4.8a)

f2!0 ðtÞ ¼ k
2 exp½
ðk
2 þ k
3 Þt

(4.8b)

f2!1 ðtÞ ¼ k
3 exp½
ðk
2 þ k
3 Þt

(4.8c)

N2!0 k
2 ¼ N2!1 k
3

(4.8d)

Equations (4.6a)–(4.6d), (4.7a)–(4.7d), and (4.8a)–(4.8d) can be used to fit the corresponding experimental results to obtain the rate constants. The caption of Fig. 4.4C gives the determined rate constants for each of the kinetic steps in the Hah1–MBD4 interaction, from which the KD’s of the interaction complexes can be calculated. In ensemble-averaged measurements, if the two interaction complexes cannot be differentiated but are detectable, the measured effective dissociation constant (KD,eff) is related to the KD’s of the two complexes as 1/KD,eff ¼ 1/KD1 þ 1/KD2.

5. Further Developments A limitation of using Egg PC for forming nanovesicles is the enclosed environment that prevents facile exchange of solution. Being able to change the solution condition and introduce additional chemical reagents is highly

58

Jaime J. Benı´tez et al.

desired, however. Ha and coworkers have developed two strategies to make the nanovesicles porous to allow exchange of solution into the nanovesicles (Cisse et al., 2007): (1) using a lipid with a higher Tm and performing experiments at its Tm, which induces defects in the lipid membrane and (2) incorporating into the bilayer membrane the bacterial toxin a-hemolysin that forms pores. The first strategy is based on the fact that lipid bilayer membranes form packing defects at Tm, making the membrane permeable to small molecules (Chakrabarti and Deamer, 1992; Monnard, 2003). Ha and coworkers used the lipid DMPC (1,2-dimyristoyl-sn-glycero-3-phosphocholine), which has a Tm of 23  C. They showed that at 23  C, the nanovesicles made of DMPC lipid membranes are permeable to molecules as large as ATP, but not to macromolecules such as proteins and DNA. The second strategy uses the natural pore-forming ability of the membrane protein a-hemolysin, a heptameric transmembrane channel from Staphylococcus aureus. The monomers of a-hemolysin self-assemble into the heptameric channel structure in a lipid bilayer, forming a stable pore of 1.4–2.4 nm diameter and allowing exchange of most solution components (Song et al., 1996). The lipid membrane of the nanovesicles also provides a natural platform for studying protein interactions that involve membrane-bound or membrane-anchored proteins. To do so, one can incorporate or anchor one protein to the lipid membrane of the nanovesicle and trap the other protein inside. smFRET measurements can then be employed to monitor their interactions at high effective concentrations. The confined volume of the nanovesicles can also be exploited to probe the crowding effects on protein interactions by cotrapping a larger number of different types of unlabeled macromolecules inside, for example polysaccharides. This crowding effect arguably mimics the intracellular environment, offering an opportunity to study biomacromolecule dynamics in a controlled and confined environment in vitro.

6. Concluding Remarks Nanovesicle trapping is a convenient approach to enabling singlemolecule studies at high effective concentrations. This approach also offers easy surface immobilization and minimization of nonspecific interactions with glass surfaces. Coupled with smFRET measurements, dynamic events of protein interactions with weak affinity can be monitored in real time at the single-molecule level. Single-molecule kinetic analysis allows extraction of quantitative kinetics of the protein interactions, some of which are challenging to quantify with ensemble techniques. The lipid membrane also mimics the cellular environment, as well as provides a natural platform

Nanovesicle Trapping for Weak Protein Interactions

59

for studying membrane-bound or membrane-anchored proteins. The confined volume can further be exploited to study crowding effects on macromolecule dynamics at the single-molecule level. With porous vesicles allowing solution exchange, many biological processes can be studied at high effective concentrations in situ. We expect that more biological studies using the nanovesicle trapping approach will emerge.

ACKNOWLEDGMENTS This research is supported by the National Science Foundation (CHE0645392), National Institute of Health (GM082939), the Wilson Disease Association, a Camille and Henry Dreyfus New Faculty Award, an Alfred P. Sloan Fellowship, and Cornell University. J. J. B. and A. M. K. are supported by Molecular Biophysics Traineeships from the National Institute of Health. We thank Profs. D. L. Huffman and A. R. Rosenzweig for their collaboration.

REFERENCES Banci, L., and Rosato, A. (2003). Structural genomics of proteins involved in copper homeostasis. Acc. Chem. Res. 36, 215–221. Benitez, J. J., Keller, A. M., Ochieng, P., Yatsunyk, L. A., Huffman, D. L., Rosenzweig, A. C., and Chen, P. (2008). Probing real-time transient metallochaperone-target protein interactions at the single-molecule level with nanovesicle trapping. J. Am. Chem. Soc. 130, 2446–2447. Benitez, J. J., Keller, A. M., Ochieng, P., Yatsunyk, L. A., Huffman, D. L., Rosenzweig, A. C., and Chen, P. (2009). Correction/addition: Probing real-time transient metallochaperone-target protein interactions at the single-molecule level with nanovesicle trapping. J. Am. Chem. Soc. 131, 871. Boukobza, E., Sonnenfeld, A., and Haran, G. (2001). Immobilization in surface-tethered lipid vesiclesasanewtoolforsinglebiomoleculespectroscopy.J.Phys.Chem. B. 105,12165–12170. Boxer, S. G. (2000). Molecular transport and organization in supported lipid membranes. Curr. Opin. Cell Biol. 4, 704–709. Brian, A. A., and McConnell, H. M. (1984). Allogeneic stimulation of cytotoxic T cells by supported planar mambranes. Proc. Natl. Acad. Sci. USA 81, 6159–6163. Chakrabarti, A. C., and Deamer, D. W. (1992). Permeability of lipid bilayers to amino acids and phosphate. Biochim. Biophys. Acta 1111, 171–177. Chiu, D. T., Wilson, C. F., Karlsson, A., Danielsson, A., Lundqvist, A., Stro¨mberg, A., Ryttse´n, F., Davidson, M., Nordholm, S., Orwar, O., and Zare, R. N. (1999). Manipulating the biochemical nanoenvironment around single molecules contained within vesicles. Chem. Phys. 247, 133–139. Cisse, I., Okumus, B., Joo, C., and Ha, T. (2007). Fueling protein–DNA interactions inside porous nanocontainers. Proc. Natl. Acad. Sci. USA 104, 12646–12650. Cobine, P. A., Pierrel, F., and Winge, D. R. (2006). Copper trafficking to the mitochondrion and assembly of copper metalloenzymes. Biochim. Biophys. Acta 1763, 759–772. Gragerov, A., Nudler, E., Komissarova, N., Gaitanaris, G. A., Gottesman, M. E., and Nikiforov, V. (1992). Cooperation of GroEL/GroES and DnaK/DnaJ heat shock proteins in preventing protein misfolding in Escherichia coli. Proc. Natl. Acad. Sci. USA 89, 10341–10344.

60

Jaime J. Benı´tez et al.

Hall, A. (1992). Ras-related GTPases and the cytoskeleton. Mol. Biol. Cell 3, 475–479. Haran, G. (2003). Single-molecule fluorescence spectroscopy of biomolecular folding. J. Phys. Condens. Matter 15, R1291–R1317. Hope, M. J., Bally, M. B., Webb, G., and Cullis, P. R. (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. Huffman, D. L., and O’Halloran, T. V. (2001). Function, structure, and mechanism of intracellular copper trafficking proteins. Annu. Rev. Biochem. 70, 677–701. Johnson, J. M., Ha, T., Chu, S., and Boxer, S. G. (2002). Early steps of supported bilayer formation probed by single vesicle fluorescence assays. Biophys. J. 83, 3371–3379. Kim, B.-E., Nevitt, T., and Thiele, D. J. (2008). Mechanisms for copper acqusition, distribution and regulation. Nat. Chem. Biol. 4, 176–185. Laurence, T. A., and Weiss, S. (2003). How to detect weak pairs. Science 299, 667–668. Lee, J. Y., Okumus, B., Kim, D. S., and Ha, T. (2005). Extreme conformational diversity in human telomeric DNA. Proc. Natl. Acad. Sci. USA 102, 18938–18943. Levene, M. J., Korlach, J., Turner, S. W., Foquet, M., Craighead, H. G., and Webb, W. W. (2003). Zero-mode waveguide for single-molecule analysis at high concentration. Science 299, 682–686. Lutsenko, S., LeShane, E. S., and Shinde, U. (2007). Biochemical basis of regulation of human copper-transporting ATPase. Arch. Biochem. Biophys. 463, 134–148. MacDonald, R. C., MacDonald, R. I., Menco, B. E., Takeshita, K., Subbarao, N. K., and Hu, L. R. (1991). Small-volume extrusion apparatus for preparation of large, unilamellar vesicles. Biochim. Biophys. Acta 1061, 297–303. Monnard, P.-A. (2003). Liposome-entrapped polymerases as models for microscale/nanoscale bioreactors. J. Membr. Biol. 191, 87–97. Nooren, I. M. A., and Thornton, J. M. (2003). Diversity of protein–protein interactions. EMBO J. 22, 3486–3492. Okumus, B., Wilson, T. J., Lilley, D. M. J., and Ha, T. (2004). Vesicle encapsulation studies reveal that single molecule ribozyme heterogeneities are intrinsic. Biophys. J. 87, 2798–2806. Rasnik, I., McKinney, S. A., and Ha, T. (2005). Surfaces and orientations: Much to FRET about? Acc. Chem. Res. 38, 542–548. Rasnik, I., McKinney, S. A., and Ha, T. (2006). Nonblinking and long-lasting single molecule fluorescence imaging. Nat. Methods 3, 891–893. Rhoades, E., Gussakovsky, E., and Haran, G. (2003). Watching proteins fold one molecule at a time. Proc. Natl. Acad. Sci. USA 100, 3197–3202. Rhoades, E., Cohen, M., Schuler, B., and Haran, G. (2004). Two-state folding observed in individual protein molecules. J. Am. Chem. Soc. 126, 14686–14687. Rosenzweig, A. C. (2001). Copper delivery by metallochaperone proteins. Acc. Chem. Res. 34, 119–128. Silvius, J. R. (1983). Thermotropic phase transitions of pure lipids in model membranes and their modifications by membrane proteins. In ‘‘Lipid–Protein Interactions,’’ (P. C. Jost and O. H. Griffish, eds.), pp. 239–281. Wiley, New York. Song, L., Hobaugh, M. R., Shustak, C., Cheley, S., Bayley, H., and Gouaux, J. E. (1996). Structure of staphylococcal-hemolysin a heptameric transmembrane pore. Science 274, 1859–1865. Strausak, D., Howies, M. K., Firth, S. D., Schlicksupp, A., Pipkorn, R., Multhaup, G., and Mercer, J. F. B. (2003). Kinetic analysis of the interaction of the copper chaperone Atox1 with the metal binding sites of the menkes protein. J. Biol. Chem. 278, 20821–20827. Xie, X. S. (2001). Single-molecule approach to enzymology. Single Mol. 2, 229–236. Xu, W., Kong, J. S., and Chen, P. (2009). Single-molecule kinetic theory of heterogeneous and enzyme catalysis. J. Phys. Chem. C 113, 2393–2404.

C H A P T E R

F I V E

Droplet Confinement and Fluorescence Measurement of Single Molecules Lori S. Goldner,* Ana M. Jofre,† and Jianyong Tang‡ Contents 1. Introduction 2. Methods for Droplet Generation 2.1. Emulsification 2.2. Injection 2.3. Microfluidics 3. Methods for Droplet Manipulation 3.1. Optical manipulation 3.2. Lab-on-chip methods for droplet manipulation 4. Droplet Coalescence and Mixing 5. Experimental Considerations for Single Fluorophore Detection 5.1. Protocol for aligning the apparatus 5.2. Protocol for preparation of emulsion samples 5.3. Protocol for droplet injection 6. Single-Molecule Measurements in Droplets 7. Future Prospects Acknowledgments References

62 65 65 66 68 69 69 71 73 73 76 77 77 79 82 83 84

Abstract We describe a method for molecular confinement and single-fluorophore sensitive measurement in aqueous nanodroplets in oil. The sequestration of individual molecules in droplets has become a useful tool in genomics and molecular evolution. Similarly, the use of single fluorophores, or pairs of fluorophores, to study biomolecular interactions and structural dynamics is now common. Most often these single-fluorophore sensitive measurements are performed on molecules that are surface attached. Confinement via surface * Department of Physics, University of Massachusetts, Amherst, Massachusetts, USA Department of Physics and Optical Science, University of North Carolina, Charlotte, North Carolina, USA Janelia Farm Research Campus, Howard Hughes Medical Institute, Ashburn, Virginia, USA

{ {

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72015-2

#

2010 Elsevier Inc. All rights reserved.

61

62

Lori S. Goldner et al.

attachment permits molecules to be located and studied for a prolonged period of time. For molecules that denature on surfaces, for interactions that are transient or out-of-equilibrium, or to observe the dynamic equilibrium of freely diffusing reagents, surface attachment may not be an option. In these cases, droplet confinement presents an alternative method for molecular confinement. Here, we describe this method as used in single-fluorophore sensitive measurement and discuss its advantages, limitations, and future prospects.

1. Introduction The advent of techniques and instrumentation for measuring the fluorescence from single fluorophores has opened many new doors in scientific inquiry. Among these is the ability to directly visualize the structural transformations and reaction dynamics of individual biomolecules and biomolecular complexes. Using intrinsic fluorescence, or labeling biomolecules with an appropriate dye molecule or molecules, polarization or spectroscopic techniques can be used to elucidate molecular motion, structure, structural transformations, and binding. In contrast to ensemble measurements, single-molecule sensitive measurements can be compared directly with microscopic models. Single-molecule sensitive measurements of structural transformations, reaction dynamics, and binding all require that a molecule under study be measured for some amount of time. This requires a method for locating or distinguishing individual molecules, and for confining them in a detection region. Depending on the measurement scheme, various methods have been devised for distinguishing or localizing biomolecules, including surface binding (Ha et al., 1999; Noji et al., 1997; Wennmalm et al., 1997), surface adsorption (Bopp et al., 1997; Jia et al., 1999; Talaga et al., 2000; Wazawa et al., 2000), or confinement in a porous material (Dickson et al., 1997; Lu et al., 1998). Confinement to surface tethered liposomes has also been used and offers the advantage of a less perturbative and more reproducible environment for the confined molecule (Boukobza et al., 2001; Okumus et al., 2004; Yoon et al., 2006). Most recently, femtoliter and subfemtoliter droplets in oil have been used to confine single fluorescent molecules for measurement (Reiner et al., 2006; Tang et al., 2008). In addition to offering a simple and convenient alternative to other methods for confining or localizing molecules, droplets offer the distinct advantage that they can be made to coalesce on contact, without loss of hydrophilic contents, providing a convenient means for fast mixing. For droplets in the absence of surfactant coalescence is intrinsically fast, and the timing of mixing will be limited by diffusion in the droplets. Diffusional mixing can occur in less than 1 ms for subfemtoliter

Droplet Confinement of Single Molecules

63

droplets. Combined with single-molecule sensitive fluorescence measurement, this opens new possibilities for the study of reaction dynamics or any out-of-equilibrium process that requires fast mixing. Coalescence also offers the possibility of assembling and studying short-lived molecular complexes as they undergo rapid assembly and dissociation, since the various components can be introduced in individual droplets and are all confined to a single slightly larger droplet after coalescence. The monodispersity, manipulability, repeatability, small size, and fast mixing afforded by aqueous nanodroplets in oil offer many opportunities for nanochemistry and observation of chemical reactions on a molecule-by-molecule basis. More generally, droplets offer convenient compartmentalization and localization useful in many applications in analytical chemistry and specifically enzymology. Unlike nature’s own cellular compartments, droplets can be engineered to be stable under conditions not commensurate with living systems. The use of emulsions makes massively parallel high-throughput measurement of bioreactions possible; a single microliter of sample can be aliquoted into 106–109 separate bioreactors. Confinement to a small reaction volume leads to higher reaction rates than in the bulk, both because reactant concentrations can be made arbitrarily high as droplets become arbitrarily small, and because diffusion-limited reaction rates increase as volume decreases. The utility of single aqueous droplets in oil for the study of single molecule reaction kinetics has long been recognized. Almost 50 years ago Rotman (1961) demonstrated that the activity of single b-D-galactosidase molecules confined to aqueous droplets in silicone oil could be measured by monitoring the increase in fluorescence of a fluorogenic substrate over time. While Rotman did not detect individual turnovers, his technique represents perhaps the first time that single enzyme kinetics could be measured using fluorescence. More recently, chymotrypsin kinetics have been studied using a similar technique (Lee and Brody, 2005). In both these works, droplets were used to confine individual molecules and indirectly study their kinetics using fluorogenic substrates. Here, we describe how the fluorescence from single droplet-confined molecules can be directly observed. Many applications have been found in recent years for droplet compartmentalization of biological reagents. Often these involve the use of single DNA templates in applications of directed evolution or gene sequencing. As many of the methodologies might be useful in detection of single biochemical interactions, we provide a brief outline here. More complete reviews can be found elsewhere in the literature (Griffiths and Tawfik, 2006; Kelly et al., 2007; Leamon et al., 2006; Song et al., 2006; Taly et al., 2007). A general method for expression and evolution of enzymes using droplet compartmentalization in bulk emulsions to link genotype and phenotype was first elucidated by Tawfik and Griffiths (1998). In this work and in subsequent works (Bernath et al., 2005; Cohen et al., 2004; Doi and

64

Lori S. Goldner et al.

Yanagawa, 1999; Ghadessy et al., 2001; Griffiths and Tawfik, 2003), much of which is reviewed in Griffiths and Tawfik (2006), a single gene is confined to individual droplets in an emulsion. Ghadessy et al. (2001) demonstrated the directed evolution of a polymerase in a droplet using a feedback loop whereby successful polymerases amplify their own single encoding gene inside a droplet. More recently, Courtois et al. (2008) demonstrated protein expression and detection on-chip from single DNA templates confined to droplets using a microfluidic technique introduced by Dittrich et al. (2005). In contrast to bulk emulsions, the use of microfluidic droplet generators make it possible to individually prepare, address, and analyze specific droplets on a single device. Compartmentalization using droplet fluidics, therefore, provides a particularly convenient method to link genotype and phenotype for in vitro molecular evolution. While PCR on a single gene confined to a droplet was demonstrated by Ghadessy et al. (2001) in the context of directed evolution of DNA polymerases, it was introduced specifically for single-molecule sensitive amplification by Nakano et al. (2003). Musyanovych et al. (2005) refined the technique to use a miniemulsion formulation that resulted in a more monodisperse distribution of droplets; they were able to demonstrate PCR in 200 nm emulsions where one-third of the droplets contained both a single DNA template and single DNA Taq polymerase. Emulsion PCR has been further refined for commercial deployment and used for gene sequencing by 454 Life Sciences (Margulies et al., 2005). Similar to an earlier protocol using magnetic beads (Dressman et al., 2003), the 454 technique uses droplet confinement to amplify single genomic fragments 107 times and to attach the copies to a single bead confined in the same droplet. In both the magnetic case, and the 454 protocol, beads are later purified for analysis. For more uniform droplet generation and therefore more uniform amplification, microfluidic devices have recently been adapted to generate monodisperse droplets for PCR in a similar application (Kumaresan et al., 2008). The invention of real-time quantification of PCR (Heid et al., 1996) and the use of microfluidic devices now means that PCR can be achieved and analyzed on-chip with nanoliter (Beer et al., 2007) or picoliter (Kiss et al., 2008) droplets providing isolation necessary to amplify individual nucleic acid templates. A recent review of microfluidic DNA amplification can be found in Zhang and Ozdemir (2009). More recently, droplet microfluidic devices have been adapted quite generally for droplet-confined chemistry. Microfluidics have the advantage of creating very monodisperse droplets, with rates of droplet formation of up to 3000/s now typical. Furthermore, by confining reactants to a droplet, dispersion that occurs by virtue of a parabolic flow field in continuous flow devices is eliminated. Many of the applications of droplet fluidic devices are discussed in pertinent reviews (Griffiths and Tawfik, 2006; Kelly et al., 2007; Leamon et al., 2006; Song et al., 2006; Taly et al., 2007).

Droplet Confinement of Single Molecules

65

Finally, it is worth noting that single cells have similarly been confined to droplets. Huebner et al. (2007) showed protein expression could be detected inside single intact cells confined to droplets in a microfluidic device. The directed evolution work of Ghadessy et al. (2001) also used single bacterial cells to provide single gene templates, but in this technique, intact cells are lysed with the first PCR cycle.

2. Methods for Droplet Generation 2.1. Emulsification Commonly called nano- or miniemulsions, metastable suspensions of water-in-oil droplets less than a micron in size can be generated by a variety of methods (Mason et al., 2006; Solans et al., 2005). Many of these methods involve extreme shear generated by ultrasound or high-pressure homogenizers. Droplets as small as a few tens of nanometers in diameter (zeptoliter volumes) can be generated this way, but the droplet size and size distribution are difficult to control, and the agitation process is damaging to many biomolecules. For example, large DNA molecules are known to break under ultrasonication (Musyanovych et al., 2005). A protocol for the formation of emulsions using ultrasonication is given in Section 5 and a similar protocol is described in Reiner et al. (2006). Triton X-100 (Sigma) is used as a surfactant, and Fluorinert FC-40, FC-70, or FC-77 (3M) as the continuous phase. As discussed later, we use perfluorinated continuous phases for their low refractive index, which facilitates optical trapping of aqueous droplets. The surfactant facilitates formation and serves to make the droplet smaller; it does not, in this case, stabilize the droplets against coalescence. The droplets formed were generally smaller than 1 mm in size, and in some cases as small as 50 nm, as measured by dynamic light scattering. Perfluorinated compounds sold by 3M under the Fluorinert trade name have low solubility in water, 10 ppm or less, and only slightly higher water solubility. Evaporation of droplets is therefore not a serious problem. We have not investigated the consequences of small amounts of Fluorinert dissolved in the aqueous phase, but we generally expect it to be inert. The viscosity of the various Fluorinert fluids vary widely; FC-70 is nearly 20 times more viscous than FC-77 and requires a longer time in the ultrasonic cleaner to form an emulsion (several minutes) than does FC-77 (several seconds). A schematic of the measurement for molecules confined to droplets in emulsion is shown in Fig. 5.1, and details are given in Section 5. Droplets are optically trapped out of the emulsion by translating the sample stage and positioning a droplet in the trap, which is colocalized with the fluorescence excitation volume of a confocal microscope. Fluorescence from a molecule

66

Lori S. Goldner et al.

Emulsify

Trap and probe Fluorescent molecule

Detect Fluorescence excitation laser Photon counting detector

Fluorescence emission Aqueous nanodroplet in perfluorinated continuous phase

1064 nm trap laser

Figure 5.1 Overall scheme for single-molecule measurement in preformed emulsion droplets.

or molecules inside the droplet is collected through the microscope and detected at a high quantum-yield photon-counting detector. For droplets that are stable against coalescence, Tawfik and Griffiths (1998) used 4.5% (v/v) Span 80 (mineral oil) in their continuous phase, followed by 0.5% (v/v) Tween 80. They formed their emulsion by stirring, with a resulting mean droplet size of 2.6 mm. Ghadessy et al. (2001) used a similar protocol but they changed the surfactant concentrations and added a third surfactant, Triton X-100, with a reported average droplet diameter of 15 mm and improved stability to changes in temperature. Droplet size depends dramatically on the details of stirring. Musyanovych et al. (2005) explored different continuous phases and used ultrasonication to create a stable miniemulsion with resulting droplet average diameters from 100 to 1200 nm.

2.2. Injection Droplets formed in bulk emulsion have considerable limitations. First and foremost, there will be a distribution of sizes. Second, the experimental procedure requires the user to ‘‘hunt and trap’’ individual droplets, which can be time-consuming. A solution to these limitations is to use the piezoelectric droplet injector developed and described by Tang et al. (2009) and shown in Fig. 5.2. The injector (Fig. 5.2) is built around a piezoelectric tube (EBL Products Inc., type EBL 2, 3.175 mm OD
0.508 mm wall
25.4 mm length) with single inner and outer nickel electrodes. The tube is fitted with Macor endcaps MC1 and MC2 using either cyanoacrylate glue or epoxy. MC1 holds a sharpened microcapillary tube (also called micropipette) in place with a set screw, and MC2 serves as a structural base for the injector. MC2 is mounted onto an aluminum (Al) holder that is fixed on a three-dimensional

67

Droplet Confinement of Single Molecules

s

MC2

Al holder

s PT

T

p

MC1

B

Figure 5.2 Schematic representation of the piezoelectric driven inertial injector showing the set screws (s) that hold the glass micropipette (p) in the front macor endcap (MC1) and the back macor endcap (MC2) in the Al holder. PT is the piezoelectric tube (Tang et al., 2009).

translation stage with micrometer adjusters. The sharpened end of the micropipette protrudes from the front of the injector and the microcapillary end extends out the back of MC2. The protocol for loading using the droplet injector is given in Section 5. After the injector has been loaded with sample, and the sample driven with a backing pressure to form a meniscus at the microcapillary tip, a sawtooth waveform is used to drive the piezoelectric tube and eject droplets on demand into the continuous phase. The waveform is generated by an HP function generator (model 33120A, 8 ms rise time typical with fall time <100 ns) and amplified (Krohn-Hite amplifier, model 7500, 200 V peak, 75 W, 1 MHz bandwidth) before being used to drive the piezoelectric. Each cycle of the sawtooth waveform causes a slow extension (over 8 ms) and quick retraction (less than 50 ms) of the glass micropipette tip. A droplet is injected into the continuous phase with each retraction. The pipette tip is silanized to make it hydrophobic to prevent sticking of injected droplets to the glass, which is necessary to ensure detachment of the droplet from the glass (see below). The retraction force is limited by the bandwidth of the amplifier (1 MHz) and the electrical–mechanical resonances of the device. Sharpened microcapillary tubes are prepared from borosilicate glass tube (OD 1.0 mm, ID 0.25 mm, A-M Systems, Inc. or OD 1.0 mm and ID 0.50 mm, Sutter Instrument) pulled by a micropipette puller (P-2000, Sutter Instrument Company). The internal diameters of the pulled micropipette tips are nominally around 0.5 mm. 2.2.1. Protocol for pulling micropipettes (1) The Sutter P-2000 should be allowed to warm up for 20 min. Micropipettes should be cleaned in UV/Ozone cleaner ( Jelight model 42) for about 30 min. (2) Wearing gloves, put micropipette onto pulling stage of micropipette puller.

68

Lori S. Goldner et al.

(3) Initial P-2000 parameters of heat ¼ 350, fil ¼ 4, vel ¼ 30, delay ¼ 200, corresponding to program 11 on the factory settings of the P-2000, are often a good choice for pulling. (4) Inspect the pulled tips using a brightfield microscope with 100
dry, long-working-distance lens (NA ¼ 0.85). The OD of the tip should be around approximately 2 mm or less. Change parameters or alignment on P-2000 as necessary to achieve symmetric, correctly sized tips. (5) Mount each freshly pulled micropipette on a commercial microinjector such as the Eppendorf Femtojet, and apply zero pressure, but be prepared to apply 4000 hPa backing pressure. On the Femtojet, this can be achieved by setting the backing pressure to 4000 hPa but putting the device in the ‘‘load pipette’’ mode. (6) Dip the tip into SigmaCote (Sigma-Aldrich Co.) to render the surface hydrophobic. After 5 s in SigmaCote, apply 4000 hPa backing pressure. On the Femtojet set as in (5), hitting the menu button is sufficient. Small air bubbles should be noticed near the tip in the SigmaCote solution. The intent here is to coat the entire tip but avoid clogging the opening. (7) With backing pressure on, take the tip out of SigmaCote solution, then hit the menu button on the pump again to zero the backing pressure. (8) Inspect the tip again using 100
microscope objective to confirm that it is neither broken nor clogged. (9) If the tip is good, carefully put it in a grooved metallic or other heatresistant holder. (10) Put all coated tips into oven, and bake them at 90  C for 1 h. (11) Tips may be stored in a low-humidity dust-free environment for up to 8 weeks. (12) When ready for use, load tips into droplet injector shown in Fig. 5.2 so that 10 mm of pipette at the sharpened end protrude from the holder. Tighten the set screw to hold the micropipette in place, but be careful not to break the glass. Tips produced in this fashion create droplets with diameters from 700 nm to several micrometers on demand. More information on the injector and its use can be found in Tang et al. (2009).

2.3. Microfluidics Microfluidics can now be used to generate droplets with volume in a range between 1 fL and 1 nL (Anna et al., 2003; Garstecki et al., 2006; He et al., 2005; Link et al., 2004; Song et al., 2003; Thorsen et al., 2001). T-shaped channels (Garstecki et al., 2006; Link et al., 2004; Song et al., 2003; Thorsen et al., 2001) or flow focusing techniques (Anna et al., 2003) provide good

Droplet Confinement of Single Molecules

69

control of droplet size and size distribution for larger emulsion droplets confined in a planar geometry. While droplet formation on demand is not yet common in microfluidic devices, it has recently been demonstrated for larger droplets using a piezo-actuated polydimethylsiloxane device (Xu and Attinger, 2008). There is in general no fundamental reason why smaller droplets cannot be produced in a microfluidic device, but there is the added difficulty of working with either smaller channels and/or droplets that are more difficult to detect in an optical microscope. Other techniques worth noting that may yet prove useful in the confinement and detection of single molecules include: (1) droplets (femtoliter and larger) produced by a commercially available microinjector in a microchannel device (Lorenz et al., 2006), (2) Microcapillary-based devices that produce similarly sized monodisperse droplets continuously in a true three-dimensional geometry (Umbanhowar et al., 2000; Utada et al., 2005), and (3) monodisperse emulsions with picoliter droplets formed using microchannel plates (Sugiura et al., 2001).

3. Methods for Droplet Manipulation One of the most compelling applications for nanodroplets is their utility as microreactors: they fuse easily on contact, their contents are isolated from external contamination, and their small volume allows for fast mixing and minimal reactant waste. Nanodroplet manipulation is imperative for inducing such controlled chemical reactions.

3.1. Optical manipulation The ability to use laser light to trap and manipulate small dielectric particles was first demonstrated by Arthur Ashkin’s group at Bell Labs in 1970 (Ashkin, 1996; Ashkin et al., 1986). Since then, optical trapping has found a significant place among the tools available for biophysical studies in a multitude of applications (Arai et al., 1999; Neuman and Block, 2004; Svoboda and Block, 1994). There are two forces exerted on a particle in a light field: the scattering force and the gradient force. The scattering force is caused by the particle’s absorption and reradiation of light, pushing the particle in the direction of the light beam propagation. The gradient force, also known as the dipole force, is due to the interaction between the light’s electric field and a particle’s induced dipole. The dipole force is proportional to the polarizability of the trapped particle, which is a measure of the particle’s induced dipole. Thus,

70

Lori S. Goldner et al.

F / arI

ð5:1Þ

where I is the intensity of the laser light field and a is the particle’s polarizability, which for a sphere is given by (Neuman and Block, 2004) a ¼ n2m r 3

n2p  n2m n2p þ 2n2m

ð5:2Þ

where r is the sphere’s radius, nm is the index of refraction of the medium in which the particle is suspended, and np is the index of refraction of the particle. If the particle’s refractive index is greater than that of the medium, then the particle’s induced dipole will be parallel to the external electric field. A dipole aligned parallel to an external electric field is attracted to the region of highest field intensity, and the driving force is proportional to the gradient of the field intensity. On the other hand, if the particle’s refractive index is lower than that of the medium, then the force is repulsive and the particle is driven away from the field. 3.1.1. Optical tweezers Optical tweezers (Ashkin et al., 1986) are created by bringing a laser to a tight focus using a high numerical aperture objective lens; particles are trapped in the region of highest intensity at the focal spot of the laser. The condition for trapping with optical tweezers requires that the particle’s refractive index be higher than that of the medium in which it is suspended. This condition is easily met for the case for colloidal particles, such as glass microspheres suspended in an aqueous solution, where optical tweezers find many applications (Svoboda and Block, 1994). However, this is not typically the case for water-in-oil emulsions, as hydrocarbon oils generally have an index of refraction higher than water, which means that light will exert a repulsive force on droplets suspended in hydrocarbon oil so that trapping at the focus of the laser beam is impossible. However, the repulsive force can be used to manipulate droplets, as shown by Katsura et al. (2001) and later by Hase et al. (2007) where water droplets were pushed toward one another by means of repulsive optical tweezers. This enabled controlled coalescence of droplets for the purpose of observing contained chemical reactions. Optical tweezers can be used to trap individual droplets within an emulsion if the continuous phase has lower refractive index than the droplet phase. This is generally the case for perfluorocarbons such as the Fluorinerts sold by 3M. Reiner et al. (2006) first demonstrated the utility of perfluorocarbon continuous phase by optically trapping droplets suspended in Fluorinert FC-77, and an example from their work is shown in Fig. 5.3. Other perfluorocarbons in 3M’s FC series, such as FC-40, FC-43, and FC-70,

Droplet Confinement of Single Molecules

71

Figure 5.3 Sequence of video images showing the fusion of two aqueous droplets, initially held in independent optical tweezers. The upper droplet is translated by the mobile trap to the location of the droplet held by the fixed trap, at which point the two droplets fuse into one. The fixed trap is then turned off and the single droplet is translated upward by the mobile trap. The mobile trap (upper) is slightly defocused from the fixed trap (lower). The solid bar in the first picture is 1 mm in length. From Reiner et al. (2006).

offer various viscosities and similarly low refractive indices. The high viscosity of FC-70, for example, was found to be useful when aqueous droplets were formed by means of inertial injection (Tang et al., 2008). 3.1.2. Other trap configurations To optically trap aqueous droplets in conventional water-in-oil emulsions, a different trap configuration must be used to create an enclosed low-intensity region for the droplet. This can be accomplished by means of rapidly scanning the focused laser beam in a tight circle around the trapped droplet (Sasaki et al., 1992; Yao et al., 1996a,b). Alternatively, aqueous droplets have been trapped using optical vortex beams (Curtis et al., 2002; Gahagan and Swartzlander, 1996, 1999; Prentice et al., 2004), whose intensity profile is characterized by a dark central region and a helical phase distribution. Optical vortices are also known as Laguerre-Gaussian modes or TEM01* modes, and can be created and actively controlled by means of computer generated holography (Curtis et al., 2002). The beam profile is such that each droplet is surrounded by a repulsive cage of light. This is convenient for isolating a single droplet from interactions with any other droplet. However, such a repulsive barrier surrounding the droplet prevents it from coalescing with another trapped droplet, as the two droplets will have a large repulsive barrier to overcome. This problem has been circumvented using spatial light modulators to dynamically shift the position of the dark core as the two traps are brought into contact (Lorenz et al., 2007).

3.2. Lab-on-chip methods for droplet manipulation While we have so far found optical manipulation to be expedient for manipulation of droplets used for single-molecule confinement and fluorescence measurement, electrical and acoustic droplet manipulation are becoming increasingly important in lab-on-a-chip applications (Luo et al., 2009). We therefore expect that one or more of these techniques

72

Lori S. Goldner et al.

may find usefulness in single-molecule sensitive fluorescence measurement and we note them here for completeness. The application of an electrostatic field actuates droplets by means of two effects: an electric field exerts a dielectrophoretic force on the droplets; and an electric field changes the contact angle of droplets on a surface ( Jones, 2002). Electrowetting is useful for applications where reagents need not necessarily be isolated from the surface and has been used primarily to transport aqueous droplets in air (Chiou et al., 2008; Cho et al., 2003; Chugh and Kaler, 2008; Zeng and Korsmeyer, 2004). Another lab-onchip droplet actuation technique that does not involve electric fields is the application of a surface acoustic wave on the substrate, which transfers its momentum to an isolated droplet (Guttenberg et al., 2005; Luo et al., 2009). While not directly applicable to droplets in a continuous oil phase, it nonetheless provides an interesting alternative for on-chip droplet manipulation. 3.2.1. Dielectrophoresis The dielectrophoretic force is based on the same principle as the dipole force described above; uncharged polarizable particles, such as aqueous droplets, experience a force along the gradient of an inhomogeneous electric field. The use of dielectrophoresis to manipulate droplets in oil is well established. An applied electric field can be used to sort the droplets, guiding selected droplets into a specific region (Ahn et al., 2006b). Programmable microchip devices containing a series of electrodes have been used to controllably transport and fuse droplets (Schwartz et al., 2004). Dielectrophoretic transport is also possible for free-floating droplets, as demonstrated by Velev et al. (2003) where the droplets were suspended on a layer of high-density fluorinated oil. More recently, Park et al. (2008) demonstrated that dielectrophoretic forces can be actively controlled by optical means. 3.2.2. Electrowetting A droplet can be rolled along a surface by actively modulating its contact angle with the surface; electrowetting is one convenient method by which to actively control the contact angle (Luo et al., 2009). The application of an electric field decreases the interfacial energy between the liquid droplet and solid substrate, which causes the droplet to decrease its contact angle and spread out onto the substrate. If the field is applied to one side of the droplet, the droplet will deform asymmetrically, and movement will be induced. Electrowetting-based actuation in microfluidic instrumentation was demonstrated by Pollack et al. (2002) using a grid-like arrangement of electrodes on a flat open surface. The featureless surface allows for flexibility in configuring a path of motion for the droplet. Electrowetting gradients can also be applied on a surface by electrochemical means (Yamada and Tada, 2005) and controlled by optical means (Chiou et al., 2008), which obviates the need for the electrode grid.

Droplet Confinement of Single Molecules

73

4. Droplet Coalescence and Mixing For many applications where droplets are used strictly for compartmentalization, droplet coalescence is considered a nuisance and surfactants are chosen to stabilize droplets against coalescence. However, droplet coalescence can also be utilized for fast mixing and accurate control of timing in the study of out-of-equilibrium complexes or molecular assembly. Fast mixing is crucial to the control and study of chemical reactions. In simple microfluidic devices, where flow is laminar, mixing is generally diffusion limited and dispersion of the reagents in the parabolic flow field is problematic. Turbulent flows solve these problems and provide very fast mixing but require very high pressure and flow rates and geometries that create additional complications (Shastry et al., 1998). The use of droplets to confine the reagents in a flow field has many advantages, including the elimination of dispersion. Song et al. (2003) used a simple geometric mixer to mix the contents of large ‘‘plug’’ droplets (45 mm deep by 28 mm wide by variable length) in about 2 ms. Tang et al. (2009) showed that the use of small droplets makes similar mixing times possible via diffusion. Ahn et al. (2006a) demonstrated how pairs of droplets formed at separate T junctions could be synchronized by size-dependent flow and mixed via electrocoalescence. Priest et al. (2006) also demonstrated controlled electrocoalescence using electrodes integrated into a microfluidic device. For small droplets such as those produced by ultrasonication or through injection as described above, diffusion-limited mixing after coalescence is intrinsically fast. For droplets formed without a stable emulsion formulation, coalescence occurs on contact, as in Fig. 5.3. As discussed in Section 3.1, for droplets in a low-index continuous phase, optical tweezers can be used to bring droplets into contact. Reiner et al. (2006) first demonstrated droplet coalescence in this way, and Tang et al. (2009) used this scheme to demonstrate droplet mixing using a diffusion-limited reaction of the calcium-sensitive dye Fluo-3 with calcium chloride (Eberhard and Erne, 1989). Initially, the Fluo-3 dye is contained in a small droplet with a radius 0.25 mm, and CaCl2 in a larger droplet with a radius 1.5 mm. Fluo-3 fluorescence increases rapidly after coalescence, with most of the rise occurring in less than 1 ms. Details can be found in Tang et al. (2009).

5. Experimental Considerations for Single Fluorophore Detection The experimental setup for optical manipulation and fluorescence detection is described in Fig. 5.4. The apparatus is built around a Zeiss microscope body (Zeiss, Axiovert S100TV). IR light for the optical trap is

74

Lori S. Goldner et al.

Lamp PC

SM

Injector

HWP Trapping laser

Condensor SH

L PC

OBJ

SM L SM

L

S Excitation laser

DF M

DM

IR-BF TL

L

RM

CCD

CL

F

F

FL

APD

M

Figure 5.4 Apparatus for single-fluorophore sensitive measurement in droplets. M, mirrors. SM, steering mirrors imaged onto the back focal plane by lenses L. DM is an IR mirror that passes visible light. DF is a dichroic mirror in the microscope slider. RM is a flip mirror or prism internal to the microscope. PC are polarizing beamsplitters. HWP is a half-wave plate. NF is a notch filter, F are one or more band-pass or short-pass filters, FL is a focusing lens and CL is a collimating lens. SH is the sample holder; OBJ is a 100
oil immersion objective lens with NA 1.30. IR-BF is an IR blocking filter and TL is the microscope tube lens. S is a shutter. APD denotes an avalanche photodiode detector.

provided by a Ytterbium fiber laser (IPG Photonics, YLD-5, 5W, l ¼ 1064 nm). Continuous wave excitation is preferred over pulsed excitation to minimize two-photon absorption from the trapping laser. Fluorescence excitation light of up to a few milliwatts is provided by an Argonion laser at 488 nm and a diode pumped Nd:YAG at 532 nm (Crystal Laser Inc.). Collimated light from the IR trapping and excitation lasers is combined using a dielectric IR mirror (DM) that passes visible light, and directed through the objective lens, OBJ (Olympus UPlanFl, 100X, numerical aperture NA ¼ 1.30) using steering mirrors (SMs) that are imaged onto the back focal plane of OBJ. The IR and excitation lasers are aligned using the procedure below so that a trapped droplet sits in the confocal detection volume. Fluorescence from the droplet is imaged onto

Droplet Confinement of Single Molecules

75

the APD, which here serves as its own confocal pinhole. For detectors with larger active areas, an external pinhole is used. The various SMs and collimation optics on the IR and excitation beams (Fig. 5.4) permit the IR and excitation light to be aligned with respect to the microscope field-of-view and each other. A dual-band dichroic mirror, DF (Chroma Technology, z488-1064 rpc for 488 nm excitation, or Omega Optical, XF2017 for 532 nm excitation), is used in the microscope slider to reflect IR and excitation lasers to OBJ and to pass fluorescence to the detector. We have used various detectors in this application, all of which have high quantum efficiency necessary for single-molecule sensitive measurement. These include avalanche photo diodes SPCM-14 or SPCM-15 (PerkinElmer Optoelectronics), which are the best choice for red dyes; Micro Photon Devices, Bolzano, Italy, model PDM 50ct, which have a timing accuracy of <50 ps, and Hamamatsu 7422-40 photon-counting modules that have a spectral response similar to the PDMs and timing accuracy only slightly better than the PerkinElmer SPCMs. For measurements of fluorescence lifetime or time-resolved fluorescence anisotropy of droplet-confined molecules (Tang et al., 2009), a pulsed excitation laser (we use a frequency-doubled Ti:Sapphire laser) replaces the CW excitation laser and the faster PDM detector is used with a TCSPC board (SPC630 or SPC830, Becker & Hickl) for data collection. Filters (F, Fig. 5.4) are used to remove the remaining excitation and IR light; these typically include one or more IR filters (FGS900, Thorlabs), a notch filter (488 nm or 532 nm, Kaiser Optical Systems Inc.), and band-pass filters (e.g., Omega Optical, 3RD millennium, 500–550 nm or 550–600 nm for different excitation lasers and dyes). More recently we have been using Semrock filters in this application. For the dual-trap setup, the 1064 nm laser beam is split into two beams by a polarizing beamsplitter cube (PC) to generate two independent optical traps, one with a picomotor-driven SM. The two IR beams are recombined by another polarizing beamsplitter cube (PC). A half-wave plate in front of the first polarizing beamsplitter is used to control the relative laser intensity of these two traps. The distance between two traps can be adjusted in the range of tens of microns. A removable mirror (RM, Fig. 5.4) is used to direct transmitted light from the lamp onto a CCD video camera (SONY, XC-ST50 or Cohu 4990 series RS-170). These inexpensive cameras are used to record a brightfield image and movie of the field of view. Images are collected by a frame grabber card (National Instruments, PCI1405). For fluorescence detection, RM is removed and fluorescence photons detected at the APD or PMT are reported as TTL pulses that are counted by a data acquisition board (National Instruments, PCI6602). All data and imaging acquisition processes are controlled by Labview software (National Instruments).

76

Lori S. Goldner et al.

To minimize photobleaching by avoiding IR absorption of single fluorescent molecules from the excited state (Dijk et al., 2004), the IR trapping laser and visible excitation laser can be alternated using acousto-optical modulators (AOMs). Additionally, a delay of several microseconds between excitation laser off and trapping lasers on gives the triplet state time to relax and minimizes photobleaching with IR photons from the triplet. A typical modulation rate of 50 kHz is much greater than trap oscillation frequencies (on the order of 100 Hz), meaning that the optically trapped droplet will not be affected by the modulation (Brau et al., 2006). A modulation rate of 50 kHz is also considerably faster than the typical bin time for fluorescence photon acquisition. If an AOM is used to alternate excitation and trapping light, the photon detector should be gated to avoid aliasing of the binned photon signal and to lower the background fluorescence.

5.1. Protocol for aligning the apparatus (1) With lasers blocked, load a sample of dye solution (typically 1–10 mM) on the microscope stage. Turn on brightfield illumination and focus on the top surface of the coverslip. Block the eyepieces and open the camera port, and cover the sample stage to prevent damage from IR light coming through the objective. (2) With very low-input power, unblock lasers one at a time and check that the IR and visible lasers are correctly focused on the coverslip. IR laser power or filters may need to be adjusted so that the optical trap or traps are visible and unsaturated on the CCD. Align and focus using the SMs and lenses in Fig. 5.4 as necessary. Note that if the apparatus is assembled properly, with lenses imaging the SMs into the back focal plane and with the excitation light filling the back focal plane of the objective, then the SM should change the position of the spot (angle of the beam at the back focal plane), and the other mirror can be used to correct the position of the beam. Position one trap spot and the excitation spot at the same position near the center of the field of view. (3) Using microscope fine focus, adjust focus 20–30 mm above the coverslip surface. Turn off or block brightfield lamp and IR laser. (4) Switch the detection path to fluorescence port and align the detector on the dye signal, adjusting excitation light down in intensity to keep the signal at the detector well below detector saturation. (5) Block the lasers, switch back to the camera port, and replace the dye sample with a sample of micron diameter fluorescent beads in water, or emulsified droplets of high-concentration (10–1 mM) dye solution in FC-70. Find the coverslip top surface as in step (1). Check the focus in step (2) but note that any change to the excitation beam from this point on requires repeating steps (1)–(4). If the spots look good, repeat step (3).

Droplet Confinement of Single Molecules

77

(6) With IR blocking filter (IR-BF, Fig. 5.4) in place, turn the IR power up to about 100 mW. Position the stage and adjust the IR power to trap a bead or droplet in the trap to be centered on the confocal detection volume. (7) Turn off the brightfield illumination and switch detection path to fluorescence port. Optimize the photon counts at the detector by adjustment of the SMs of IR laser and the lenses right after the SMs. The trap center should now be located at the confocal detection volume and the apparatus should now be ready for single-molecule sensitive detection. Note that alignment will shift slightly with droplets that vary in size or composition, but overall we find this method satisfactory. Please note that as with any IR laser, extreme caution, including OD 6 or better IR laser goggles and appropriate interlocks, should be used at all times when dealing with trap light.

5.2. Protocol for preparation of emulsion samples (1) Clean FC-70, FC-40, or FC-77 with water agitation to remove hydrophilic impurities and wait overnight to settle the oil phase (lower phase) and aqueous phase (upper phase). (2) Filter the Fluorinert liquid with a 0.2-mm filter. (3) Add sample solution to Fluorinert with 1:10 volume ratio in a microcentrifuge tube. Typically we use 100 ml of aqueous sample solution containing the species to be studied at or above a concentration of 50 nM, depending on final droplet size, with 1 ml of the FC-77 in a 2-ml microcentrifuge tube. Triton X-100 at 0.1% (v/v) is added either to the Fluorinert or sample solution, although common practice is to put the surfactant in the continuous phase. (4) Hold the tube with sample solution and Fluorinert mixture in a bath in an ultrasonic cleaner (Branson 5510). For FC-77 it takes several seconds and for (more viscous) FC-70 it takes several minutes to form an emulsion. For solutions containing sulforhodamine B (SRB) or short pieces of DNA or RNA, the water bath is at room temperature; an ice bath is used for proteins. (5) Load the emulsified solution into a sealed glass chamber with coverslip on the bottom for use in the inverted microscope.

5.3. Protocol for droplet injection (1) Prepare droplet injector and pulled micropipettes as in Section 2.2. Load a pipette into the injector. (2) Using a gel-loading pipette tip, load approximately 4 ml of sample into the back of the pipette. Sample should contain fluorescently labeled

78

(3) (4) (5) (6)

(7) (8)

(9)

(10)

Lori S. Goldner et al.

species at a concentration of approximately 3 nM, or whatever is needed to result in less than or equal to one fluorophore per droplet. Using cyanoacrylate superglue (Loctite), glue Tygon Microbore PVC tubing (ID 1.0 mm, OD 1.8 mm) to the back end of the pipette and apply accelerator (Loctite 7113). Let dry for at least 5 s. Mount the piezo assembly onto the Al holder or other modified commercial micromanipulator. We use a 3D roller bearing translation stage with a magnetic kinematic mount for easy positioning of the Al holder. Attach the far end of the PVC tubing to a microinjector pump (FemtoJet 5247). Attach the electrical leads to the piezo amplifier (Krohn-Hite 7500). Load continuous phase into a well on the microscope coverslip (e.g., Grace Bio-labs Fastwell). A low vapor-pressure perfluorocarbon like FC-70 minimizes evaporation loss and makes it easier to manipulate injected droplets. A layer of polyethylene glycol (PEG) can be placed on the top of the FC-70 to further stabilize the system. Mount the coverslip containing the continuous phase on the microscope. Lower the tip into the continuous phase at an angle sufficient to permit illumination through a long working distance condenser, taking care not to damage the tip. With brightfield transmission illumination, lower the tip to 20–30 mm above the coverslip bottom and focus on the tip with the high NA objective lens. Using the injector pump, apply pressure (300–1500 hPa) to drive the sample plug to the tip of the pipette. Under the microscope, air droplets bubbling rapidly out of the tip will be evident, followed by sample droplets. Adjust the backing pressure so that the sample forms a meniscus near the end of the micropipette tip. To inject a nanodroplet into the FC-70, the piezoelectric tube is driven with a single cycle of a sawtooth waveform as discussed above.

Several oxygen scavenging/triplet quenching recipes are now commonly used in single molecule detection (Aitken et al., 2008; Rasnik et al., 2006). The perfluorocarbons that we use for their low refractive index have oxygen solubility orders of magnitude higher than water, so the environmental oil phase can supply enormous amount of oxygen which are beyond the capacity of chemical oxygen scavengers. Degassing and purging the continuous phase with an inert gas like nitrogen might provide an answer, and experiments are underway to determine the best practice for removing oxygen and triplet quenching in this droplet system. Currently our protocol calls for degassed perfluorinated phase and trolox at 1–2 mM (Rasnik et al., 2006) as a triplet quencher. Note that one common oxygen scavenging system, 0.1 mg/ml glucose oxidase, 0.02 mg/ml catalase, 3% (w/v) glucose can change the pH of the droplet interior (Englander et al., 1987).

79

Droplet Confinement of Single Molecules

6. Single-Molecule Measurements in Droplets Figure 5.5 demonstrates stepwise photobleaching of individual dye molecules in droplets containing one, two, and three fluorophores, respectively (Reiner et al., 2006; Tang et al., 2008). The data in this figure come from three separate droplets trapped out of an emulsion as shown schematically in Fig. 5.1 and described in detail in Section 5. The first few seconds show dark counts from the detector (about 100 counts/s) and the initial jump in intensity occurs when a shutter on the excitation laser (532 nm) is opened, producing a combination of fluorescence emission and laser background counts. The counts irreversibly decrease in discrete steps until only dark counts plus background from the excitation light is measured. The discrete drops in emission are indicative of photobleaching of individual dye molecules, and the number of dye molecules in the droplet can be determined by the number of photobleaching steps. The various signal sizes

80

B 400

60

300

40

200

20

100

Counts per 50 ms

A

0

0 0

2

4

6

8

0

10

2

4 6 Time (s)

8

10

C 150 120 90 60 30 0 0

5

10

15

20

25

Figure 5.5 Three examples of single dye molecule detection in trapped emulsion droplets. (A)–(C) illustrate the trapping and detection of 1, 2, and 3 sulforhodamine B (SRB) molecules, respectively. Photobleaching events are indicated by arrows. The measurements are taken with different excitation strengths. For (A) and (C) the laser power sent into the back aperture of the microscope objective was 600 mW and for (B) the power used was 2.5 mW. The different laser powers resulted in different step sizes for photobleaching events. For these measurements a solution of 5 nM SRB was used for which a 1-mm droplet is calculated to contain an average of 1.6 dye molecules. From Reiner et al. (2006).

80

Lori S. Goldner et al.

are the result of different excitation power levels in the different droplets (see figure caption). Results showing single fluorophore detection in droplets were also obtained for Red Fluorescent Protein, dye-labeled DNA (Reiner et al., 2006), dye-labeled RNA, and green fluorescent protein (Tang et al., 2008). By adding a second detector and another dichroic beamsplitter and associated filters in the detection pathway of the apparatus shown in Fig. 5.4, single molecular-pair fluorescence resonance energy transfer (spFRET) can be measured for molecules confined to a droplet. Figure 5.6 demonstrates FRET from a single RNA molecule confined to an emulsion droplet, prepared with ultrasonication as described above. The RNA (Dharmacon) consisted of the 16mer 50 -Cy3-G CUC ACU GGU CAC UCG-30 hybridized with its complement 50 -Cy5-C GAG UGA CCA GUG AGC-30 to a final concentration of 2.5 nM for both strands in 20 mM Tris–HCl, 50 mM NaCl and 5 mM MgCl2. This sample also contained glucose oxidase at 0.1 mg/ ml (Roche Applied Science, Indianapolis, IN), catalase at 0.02 mg/ml (Roche Applied Science, Indianapolis, IN) and glucose at 0.1 mg/ml for oxygen scavenging, although see the warning above. Figure 5.6 (left) shows clear and quantitative FRET data. At about 0.5 s, the excitation light shutter is opened and signal is detected in both the donor (green) and acceptor (red) channels. At about 2.7 s, the acceptor dye photobleaches; the donor dye intensity then increases, since the energy transfer is now zero. At just over 6 s, the donor dye photobleaches and we can measure the background signal. Inspection of the signal between 3 and 6 s additionally gives us the cross talk from the donor channel into the acceptor channel. Knowledge of the background and cross talk permits us to extract a quantitative FRET measurement (Yim et al., 2005) from the donor and acceptor data between 0.5 and 2.7 s, shown as a histogram in Fig. 5.6 (right).

Probability of occurance

Counts per 100 ms

400 300 200 100 0

0

2

4

6

Time (s)

8

10

0.25 0.20 0.15 0.10 0.05 0.00 0.0

0.2

0.4

0.6

0.8

1.0

FRET efficiency

Figure 5.6 (Left) An example of FRET data from single RNA molecule in a dye droplet, showing stepwise photobleaching of the acceptor at about 2.7 s and the donor at about 6 s. (Right) Histogram of the FRET values calculated every 100 ms for the data when both dyes are active, between 0.5 and 2.7 s.

81

Droplet Confinement of Single Molecules

A

B 1000

8000

800

6000

600

4000

400

2000

200

0

0

0.4

D 0.4

0.3

0.3

0.2

0.2

Photon counts

10,000

Anisotropy

C

0.1

0.1 0

2

4

6

8

10 0 Time (ns)

2

4

6

8

10

Figure 5.7 Examples of time-resolved fluorescence anisotropy of EGFP at 3 mM (A, C) in free solution and (B, D) in a droplet, corresponding to about 1000 molecules in a 1 mm droplet. The anisotropy (green) was computed from the parallel (red) and perpendicular (blue) intensities every 3 ps. Fits using a single exponential decay (black line) yield rotational diffusion times of (C) 13.8  0.2 ns in free solution and (D) 12.6  1.0 ns in a droplet. For clarity, the anisotropy averaged every 150 ps (magenta) shows agreement of the fit with the data. From Tang et al. (2008).

For larger molecules and proteins, ultrasonication is often disruptive. Figure 5.7 demonstrates another type of measurement on molecules confined to injected droplets formed using the device described in Section 2.2 and the protocol above (Tang et al., 2008). Here the measurement is time-resolved fluorescence anisotropy, used to study the rotational dynamics of the confined molecules. This is not a single-molecule measurement: approximately 1000 enhanced green fluorescence protein (EGFP; Cormack et al., 1996) molecules are confined to the droplet. The measurement also requires that a second detector be added to the setup shown in Fig. 5.4; however, instead of a dichroic element, a polarizing beamsplitter is used to separate the fluorescence into components parallel to and perpendicular to the excitation polarization. We refer to the two detected intensities as Ijj(t) and I?(t). The excitation laser is a pulsed frequency-doubled Ti: Sapphire laser operating at 461 nm and the detectors are Micro Photon Devices PDM 50ct. Arrival time histograms are determined using timecorrelated single photon-counting electronics with 3 ps A/D resolution

82

Lori S. Goldner et al.

(Becker & Hickl, Berlin, Germany, model SPC-830). To quantify the rotational dynamics we determine the rotational diffusion time using time-resolved fluorescence anisotropy (Cross and Fleming, 1984; Heikal et al., 2000; Hu and Lu, 2003; Schaffer et al., 1999; Striker et al., 1999; Uskova et al., 2000; Volkmer et al., 2000). We computed the anisotropy, R(t) ¼ D(t)/S(t), from the sum, S(t) ¼ Ijj(t) þ 2gI?(t), and difference, D(t) ¼ Ijj(t)  gI?(t), of the fluorescence components. The factor g accounts for detector imbalance and is determined with Alexa 488. Since the rotational diffusion time is much greater than the APD response time, the anisotropy was calculated without deconvolving the instrument response and modeled by a single exponential decay, RðtÞ  aet=y , where y is the rotational diffusion time and a is the fundamental anisotropy of the molecule. For more information on the details of this measurement, see Tang et al. (2008). Figure 5.7 shows typical measurements of time-resolved fluorescence anisotropy for EGFP confined to a droplet and EGFP free in solution. In both cases, the EGFP concentration was 3 mM and Triton X-100 was present at 0.1% (v/v). The anisotropy decay and rotational diffusion times are similar for freely diffusing and droplet-confined EGFP. Fitting the anisotropy using a single exponential decay (black line) yielded rotational diffusion times of 13.8  0.2 ns in a typical free solution sample (Fig. 5.7C) and 12.6  1.0 ns in a typical droplet-confined sample (Fig. 5.7D). In free solution, the mean rotational diffusion time was determined to be 13.8  0.1 ns from 11 datasets at 3 mM, in reasonable agreement with previous reports of EGFP and wild-type GFP rotational diffusion times (Striker et al., 1999; Swaminathan et al., 1997; Uskova et al., 2000). In droplets, the mean rotational diffusion time was determined to be 12.6  1.0 ns from 41 datasets at 3 mM. Combined with fluorescence images of individual droplets that show no indication of EGFP sticking at the boundary, the authors therefore concluded (Tang et al., 2008) that EGFP does not congregate at the water–oil interface in the presence of surfactant and that the rotational motion of EGFP inside the nanodroplets is consistent with Brownian rotation of EGFP free in solution.

7. Future Prospects The techniques outlined in this work involve the formation of bulk emulsions from which individual droplets are chosen, or the injection on demand of individual droplets that are then positioned and mixed with the use of optical tweezers. While easy and fast to set up, these techniques have limitations that might be overcome with the use of microfluidic and other mechanical approaches for droplet generation and manipulation. Microfluidics offers an expanded choice of continuous phase, and the ability to manipulate

Droplet Confinement of Single Molecules

83

and mix droplets without the need for an optical tweezer. Work still remains to be done to adapt droplet fluidic devices for single-molecule measurements. In particular, isolating and immobilizing individual droplets in a microfluidic device is still challenging, as is introducing droplets on demand with a volume below 1 fL. While there are no fundamental limitations to prevent advances in this direction, a suitable device is not yet available. Several possibilities exist for positioning droplets in a confocal detection volume without an optical trap and many have been discussed in the text. One we have not yet mentioned is the physical droplet trap introduced by Huebner et al. (2009); using a scaled-down version of their device, it might even be possible to study in parallel the contents of droplets in large arrays using total internal reflection microscopy at a surface. Another possibility for positioning droplets in a confocal detection volume is to track instead of trapping or otherwise immobilizing them. A system similar to that introduced for particle tracking (Desai et al., 2008) may be convenient where microfabrication facilities are limited and useful where droplets need to be studied away from any surface. The sequestration of individual molecules in droplets has already demonstrated its usefulness in genomics and molecular evolution. Likewise, single-molecule sensitive fluorescence measurements have become invaluable for understanding molecular dynamics and interactions in cell biology, biochemistry, and biophysics. We have reported on progress and methods used thus far to study single biomolecules, or pairs of biomolecules, confined to aqueous droplets. Confinement in this fashion obviates the need for perturbative surface tethers, eliminates the effect of surface inhomogeneities and simplifies sample preparation. The use of confining environments both facilitates measurements of single molecules and molecular complexes, and enhances their relevance to biological systems, where interactions generally take place within the confinement of crowded cells and organelles. The methods described here should be generally applicable in the study of protein folding and remodeling under confinement or crowding. Similar to liposome confinement, they will facilitate the study of molecules and molecular complexes, such as many proteins and ribosomes, that are inactive or whose activity is perturbed on a surface. Finally, the ease with which droplets can be manipulated and mixed also opens another possibility in single-molecule measurements—the ability to study transient complexes and out-of-equilibrium dynamics that require fast mixing and good control of timing.

ACKNOWLEDGMENTS The authors thank Geoffrey Lowman for his work on the data shown in Fig. 5.6. Funding for this work was provided by the NIST physics laboratory, the NRC postdoctoral fellow program, and NSF grant MCB-0920139.

84

Lori S. Goldner et al.

REFERENCES Ahn, K., Agresti, J., Chong, H., Marquez, M., and Weitz, D. A. (2006a). Electrocoalescence of drops synchronized by size-dependent flow in microfluidic channels. Appl. Phys. Lett. 88. Ahn, K., Kerbage, C., Hunt, T. P., Westervelt, R. M., Link, D. R., and Weitz, D. A. (2006b). Dielectrophoretic manipulation of drops for high-speed microfluidic sorting devices. Appl. Phys. Lett. 88, 3. Aitken, C. E., Marshall, R. A., and Pulglisi, J. D. (2008). An oxygen scavenging system for improvement of dye stability in single-molecule fluorescence experiments. Biophys. J. 94, 1826–1835. Anna, S. L., Bontoux, N., and Stone, H. A. (2003). Formation of dispersions using "flow focusing" in microchannels. Appl. Phys. Lett. 82, 364–366. Arai, Y., Yasuda, R., Akashi, K., Harada, Y., Miyata, H., Kinosita, K., and Itoh, H. (1999). Tying a molecular knot with optical tweezers. Nature 399, 446–448. Ashkin, A. (1996). Optical trapping and manipulation of neutral particles using lasers. Proc. Natl Acad. Sci. USA 94, 4853. Ashkin, A., Dziedzic, J. M., Bjorkholm, J. E., and Chu, S. (1986). Observation of a singlebeam gradient force optical trap for dielectric particles. Opt. Lett. 11, 288–290. Beer, N. R., Hindson, B. J., Wheeler, E. K., Hall, S. B., Rose, K. A., Kennedy, I. M., and Colston, B. W. (2007). On-chip, real-time, single-copy polymerase chain reaction in picoliter droplets. Anal. Chem. 79, 8471–8475. Bernath, K., Magdassi, S., and Tawfik, D. S. (2005). Directed evolution of protein inhibitors of DNA-nucleases by in vitro compartmentalization (IVC) and nano-droplet delivery. J. Mol. Biol. 345, 1015–1026. Bopp, M. A., Jia, Y. W., Li, L. Q., Cogdell, R. J., and Hochstrasser, R. M. (1997). Fluorescence and photobleaching dynamics of single light-harvesting complexes. Proc. Natl. Acad. Sci. USA 94, 10630–10635. Boukobza, E., Sonnenfeld, A., and Haran, G. (2001). Immobilization in surface-tethered lipid vesicles as a new tool for single biomolecule spectroscopy. J. Phys. Chem. B 105, 12165–12170. Brau, R. R., Tarsa, P. B., Ferrer, J. M., Lee, P., and Lang, M. J. (2006). Interlaced optical force-fluorescence measurements for single molecule biophysics. Biophys. J. 91, 1069–1077. Chiou, P. Y., Park, S.-Y., and Wu, M. C. (2008). Continuous optoelectrowetting for picoliter droplet manipulation. Appl. Phys. Lett. 93, 221110. Cho, S. K., Moon, H., and Kim, H. M. (2003). Creating, transporting, cutting, and merging liquid droplets by electrowetting-based actuation for digital microfluidic circuits. J. Microelectromech. Syst. 12, 70. Chugh, D., and Kaler, K. (2008). Leveraging liquid dielectrophoresis for microfluidic applications. Biomed. Mater. 3, 034009. Cohen, H. M., Tawfik, D. S., and Griffiths, A. D. (2004). Altering the sequence specificity of HaeIII methyltransferase by directed evolution using in vitro compartmentalization. Protein Eng. Des. Sel. 17, 3–11. Cormack, B. P., Valdivia, R. H., and Falkow, S. (1996). FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173, 33–38. Courtois, F., Olguin, L. F., Whyte, G., Bratton, D., Huck, W. T. S., Abell, C., and Hollfelder, F. (2008). An integrated device for monitoring time-dependent in vitro expression from single genes in picolitre droplets. ChemBioChem 9, 439–446. Cross, A. J., and Fleming, G. R. (1984). Analysis of time-resolved fluorescence anisotropy decays. Biophys. J. 46, 45–56. Curtis, J. E., Koss, B. A., and Grier, D. G. (2002). Dynamic holographic optical tweezers. Opt. Commun. 207, 169.

Droplet Confinement of Single Molecules

85

Desai, K. V., Bishop, T. G., Vicci, L., O’Brien, E. T., Taylor, R. M., and Superfine, R. (2008). Agnostic particle tracking for three-dimensional motion of cellular granules and membrane-tethered bead dynamics. Biophys. J. 94, 2374–2384. Dickson, R. M., Cubitt, A. B., Tsien, R. Y., and Moerner, W. E. (1997). On/off blinking and switching behaviour of single molecules of green fluorescent protein. Nature 388, 355–358. Dittrich, P. S., Jahnz, M., and Schwille, P. (2005). A new embedded process for compartmentalized cell-free protein expression and on-line detection in microfluidic devices. ChemBioChem 6, 811. Doi, N., and Yanagawa, H. (1999). STABLE: Protein-DNA fusion system for screening of combinatorial protein libraries in vitro. FEBS Lett. 457, 227–230. Dressman, D., Yan, H., Traverso, G., Kinzler, K. W., and Vogelstein, B. (2003). Transforming single DNA molecules into fluorescent magnetic particles for detection and enumeration of genetic variations. Proc. Natl. Acad. Sci. USA 100, 8817–8822. Eberhard, M., and Erne, P. (1989). Kinetics of calcium-binding to fluo-3 determined by stopped-flow fluorescence. Biochem. Biophys. Res. Commun. 163, 309–314. Englander, S. W., Calhoun, D. B., and Englander, J. J. (1987). Biochemistry without oxygen. Anal. Biochem. 161, 300–306. Gahagan, K. T., and Swartzlander, G. A. (1996). Optical vortex trapping of particles. Opt. Lett. 21, 827. Gahagan, K. T., and Swartzlander, G. A. (1999). Simultaneous trapping of low-index and high-index microparticles observed with an optical-vortex trap. J. Opt. Soc. Am. B Opt. Phys. 16, 533–537. Garstecki, P., Fuerstman, M. J., Stone, H. A., and Whitesides, G. M. (2006). Formation of droplets and bubbles in a microfluidic T-junction—Scaling and mechanism of break-up. Lab Chip 6, 437–446. Ghadessy, F. J., Ong, J. L., and Holliger, P. (2001). Directed evolution of polymerase function by compartmentalized self-replication. Proc. Natl. Acad. Sci. USA 98, 4552–4557. Griffiths, A. D., and Tawfik, D. S. (2003). Directed evolution of an extremely fast phosphotriesterase by in vitro compartmentalization. EMBO J. 22, 24–35. Griffiths, A. D., and Tawfik, D. S. (2006). Miniaturising the laboratory in emulsion droplets. Trends Biotechnol. 24, 395–402. Guttenberg, Z., Muller, H., Habermuller, H., Geisbauer, A., Pipper, J., Felbel, J., Kielpinski, M., Scriba, J., and Wixforth, A. (2005). Planar chip device for PCR and hybridization with surface acoustic wave pump. Lab Chip 5, 308. Ha, T., Zhuang, X. W., Kim, H. D., Orr, J. W., Williamson, J. R., and Chu, S. (1999). Ligand-induced conformational changes observed in single RNA molecules. Proc. Natl. Acad. Sci. USA 96, 9077–9082. Hase, M., Yamada, A., Hamada, T., Baigl, D., and Yoshikawa, K. (2007). Manipulation of cell-sized phospholipid-coated microdroplets and their use as biochemical microreactors. Langmuir 23, 348–352. He, M. Y., Kuo, J. S., and Chiu, D. T. (2005). Electro-generation of single femtoliter- and picoliter-volume aqueous droplets in microfluidic systems. Appl. Phys. Lett. 87. Heid, C. A., Stevens, J., Livak, K. J., and Williams, P. M. (1996). Real time quantitative PCR. Genome Res. 6, 986–994. Heikal, A. A., Hess, S. T., Baird, G. S., Tsien, R. Y., and Webb, W. W. (2000). Molecular spectroscopy and dynamics of intrinsically fluorescent proteins: Coral red (dsRed) and yellow (Citrine). Proc. Natl. Acad. Sci. USA 97, 11996–12001. Hu, D., and Lu, H. P. (2003). Single-molecule nanosecond anisotropy dynamics of tethered protein motions. J. Phys. Chem. B 107, 618. Huebner, A., Srisa-Art, M., Holt, D., Abell, C., Hollfelder, F., Demello, A. J., and Edel, J. B. (2007). Quantitative detection of protein expression in single cells using droplet microfluidics. Chem. Commun. 1218–1220.

86

Lori S. Goldner et al.

Huebner, A., Bratton, D., Whyte, G., Yang, M., deMello, A. J., Abell, C., and Hollfelder, F. (2009). Static microdroplet arrays: A microfluidic device for droplet trapping, incubation and release for enzymatic and cell-based assays. Lab Chip 9. Jia, Y. W., Talaga, D. S., Lau, W. L., Lu, H. S. M., DeGrado, W. F., and Hochstrasser, R. M. (1999). Folding dynamics of single GCN4 peptides by fluorescence resonant energy transfer confocal microscopy. Chem. Phys. 247, 69–83. Jones, T. B. (2002). On the relationship of dielectrophoresis and electrowetting. Langmuir 18, 4437. Katsura, S., Yamaguchi, A., Inami, H., Matsuura, S., Hirano, K., and Mizuno, A. (2001). Indirect micromanipulation of single molecules in water-in-oil emulsion. Electrophoresis 22, 289–293. Kelly, B. T., Baret, J. C., Taly, V., and Griffiths, A. D. (2007). Miniaturizing chemistry and biology in microdroplets. Chem. Commun. 1773–1788. Kiss, M. M., Ortoleva-Donnelly, L., Beer, N. R., Warner, J., Bailey, C. G., Colston, B. W., Rothberg, J. M., Link, D. R., and Leamon, J. H. (2008). High-throughput quantitative polymerase chain reaction in picoliter droplets. Anal. Chem. 80, 8975–8981. Kumaresan, P., Yang, C. J., Cronier, S. A., Blazei, R. G., and Mathies, R. A. (2008). Highthroughput single copy DNA amplification and cell analysis in engineered nanoliter droplets. Anal. Chem. 80, 3522–3529. Leamon, J. H., Link, D. R., Egholm, M., and Rothberg, J. M. (2006). Overview: Methods and applications for droplet compartmentalization of biology. Nat. Methods 3, 541–543. Lee, A. I., and Brody, J. P. (2005). Single-molecule enzymology of chymotrypsin using water-in-oil emulsion. Biophys. J. 88, 4303–4311. Link, D. R., Anna, S. L., Weitz, D. A., and Stone, H. A. (2004). Geometrically mediated breakup of drops in microfluidic devices. Phys. Rev. Lett. 92. Lorenz, R. M., Edgar, J. S., Jeffries, G. D. M., and Chiu, D. T. (2006). Microfluidic and optical systems for the on-demand generation and manipulation of single femtolitervolume aqueous droplets. Anal. Chem. 78, 6433–6439. Lorenz, R. M., Edgar, J. S., Jeffries, G. D. M., Zhao, Y. Q., McGloin, D., and Chiu, D. T. (2007). Vortex-trap-induced fusion of femtoliter-volume aqueous droplets. Anal. Chem. 79, 224–228. Lu, H. P., Xun, L. Y., and Xie, X. S. (1998). Single-molecule enzymatic dynamics. Science 282, 1877–1882. Luo, J. K., Fu, Y. Q., Li, Y., Du, X. Y., Flewitt, A. J., Walton, A. J., and Milne, W. I. (2009). Moving-part-free microfluidic systems for lab-on-a-chip. J. Micromech. Microeng. 19, 054001. Margulies, M., Egholm, M., Altman, W. E., Attiya, S., Bader, J. S., Bemben, L. A., Berka, J., Braverman, M. S., Chen, Y. J., Chen, Z. T., Dewell, S. B., Du, L., et al. (2005). Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380. Mason, T. G., Wilking, J. N., Meleson, K., Chang, C. B., and Graves, S. M. (2006). Nanoemulsions: Formation, structure, and physical properties. J. Phys. Condens. Matter 18, R635–R666. Musyanovych, A., Mailander, V., and Landfester, K. (2005). Miniemulsion droplets as single molecule nanoreactors for polymerase chain reaction. Biomacromolecules 6, 1824–1828. Nakano, M., Komatsu, J., Matsuura, S., Takashima, K., Katsura, S., and Mizuno, A. (2003). Single-molecule PCR using water-in-oil emulsion. J. Biotechnol. 102, 117–124. Neuman, K. C., and Block, S. M. (2004). Optical trapping. Rev. Sci. Instrum. 75, 2787–2809. Noji, H., Yasuda, R., Yoshida, M., and Kinosita, K. (1997). Direct observation of the rotation of F-1-ATPase. Nature 386, 299–302. Okumus, B., Wilson, T. J., Lilley, D. M. J., and Ha, T. (2004). Vesicle encapsulation studies reveal that single molecule ribozyme heterogeneities are intrinsic. Biophys. J. 87, 2798–2806.

Droplet Confinement of Single Molecules

87

Park, S.-Y., Pan, C., Wu, T.-H., Kloss, C., Kalim, S., Callahan, C. E., Teitell, M., and Chiou, E. P. Y. (2008). Floating electrode optoelectronic tweezers: Light-driven dielectrophoretic droplet manipulation in electrically insulating oil medium. Appl. Phys. Lett. 92, 151101. Pollack, M. G., Shenderov, A. D., and Fair, R. B. (2002). Electrowetting-based actuation of droplets for integrated microfluidics. Lab Chip 2, 96. Prentice, P. A., MacDonald, M. P., Frank, T. G., Cuschieri, A., Spalding, G. C., Sibbett, W., Campbell, P. A., and Dholakia, K. (2004). Manipulation and filtration of low index particles with holographic Laguerre-Gaussian optical trap arrays. Opt. Express 12, 593. Priest, C., Herminghaus, S., and Seemann, R. (2006). Controlled electrocoalescence in microfluidics: Targeting a single lamella. Appl. Phys. Lett. 89. Rasnik, I., McKinney, S. A., and Ha, T. (2006). Nonblinking and longlasting singlemolecule fluorescence imaging. Nat. Methods 3, 891–893. Reiner, J. E., Crawford, A. M., Kishore, R. B., Goldner, L. S., Helmerson, K., and Gilson, M. K. (2006). Optically trapped aqueous droplets for single molecule studies. Appl. Phys. Lett. 89, 013904-1–013904-3. Rotman, B. (1961). Measurement of activity of single molecules of beta-D-galactosidase. Proc. Natl. Acad. Sci. USA 47, 1981–1991. Sasaki, K., Koshioka, M., Misawa, H., Kitamura, N., and Masuhara, H. (1992). Optical trapping of a metal-particle and a water droplet by a scanning laser-beam. Appl. Phys. Lett. 60, 807–809. Schaffer, J., Volkmer, A., Eggeling, C., Subramaniam, V., Striker, G., and Seidel, C. A. M. (1999). Identification of single molecules in aqueous solution by time- resolved fluorescence anisotropy. J. Phys. Chem. 103, 331–336. Schwartz, J. A., Vykoukal, J. V., and Gascoyne, P. R. C. (2004). Droplet-based chemistry on a programmable micro-chip. Lab Chip 4, 11. Shastry, M. C. R., Luck, S. D., and Roder, H. (1998). A continuous-flow capillary mixing method to monitor reactions on the microsecond time scale. Biophys. J. 74, 2714–2721. Solans, C., Izquierdo, P., Nolla, J., Azemar, N., and Garcia-Celma, M. J. (2005). Nanoemulsions. Curr. Opin. Colloid Interface Sci. 10, 102–110. Song, H., Tice, J. D., and Ismagilov, R. F. (2003). A microfluidic system for controlling reaction networks in time. Angew. Chem. Int. Ed. 42, 768–772. Song, H., Chen, D. L., and Ismagilov, R. F. (2006). Reactions in droplets in microflulidic channels. Angew. Chem. Int. Ed. 45, 7336–7356. Striker, G., Subramaniam, V., Seidel, C. A. M., and Volkmer, A. (1999). Photochromicity and fluorescence lifetimes of green fluorescent protein. J. Phys. Chem. B 103, 8612–8617. Sugiura, S., Nakajima, M., Iwamoto, S., and Seki, M. (2001). Interfacial tension driven monodispersed droplet formation from microfabricated channel array. Langmuir 17, 5562–5566. Svoboda, K., and Block, S. M. (1994). Biological applications of optical forces. Annu. Rev. Biophys. Biomol. Struct. 23, 247–285. Swaminathan, R., Hoang, C. P., and Verkman, A. S. (1997). Photobleaching recovery and anisotropy decay of green fluorescent protein GFP-S65T in solution and cells: Cytoplasmic viscosity probed by green fluorescent protein translational and rotational diffusion. Biophys. J. 72, 1900. Talaga, D. S., Lau, W. L., Roder, H., Tang, J. Y., Jia, Y. W., DeGrado, W. F., and Hochstrasser, R. M. (2000). Dynamics and folding of single two-stranded coiled-coil peptides studied by fluorescent energy transfer confocal microscopy. Proc. Natl. Acad. Sci. USA 97, 13021–13026. Taly, V., Kelly, B. T., and Griffiths, A. D. (2007). Droplets as microreactors for high-throughput biology. ChemBioChem 8, 263–272.

88

Lori S. Goldner et al.

Tang, J., Jofre, A. M., Lowman, G. M., Kishore, R. B., Reiner, J. E., Helmerson, K., Goldner, L. S., and Greene, M. E. (2008). Green fluorescent protein in inertially injected aqueous nanodroplets. Langmuir 24, 4975–4978. Tang, J., Jofre, A. M., Kishore, R., Reiner, J. E., Greene, M. E., Lowman, G. M., Denker, J. S., Willis, C., Helmerson, K., and Goldner, L. S. (2009). Generation and mixing of subfemtoliter volume aqueous droplets on demand. Anal. Chem. 81, 8041–8047. Tawfik, D. S., and Griffiths, A. D. (1998). Man-made cell-like compartments for molecular evolution. Nat. Biotechnol. 16, 652–656. Thorsen, T., Roberts, R. W., Arnold, F. H., and Quake, S. R. (2001). Dynamic pattern formation in a vesicle-generating microfluidic device. Phys. Rev. Lett. 86, 4163–4166. Umbanhowar, P. B., Prasad, V., and Weitz, D. A. (2000). Monodisperse emulsion generation via drop break off in a coflowing stream. Langmuir 16, 347–351. Uskova, M. A., Borst, J. W., Hink, M. A., van Hoek, A., Schots, A., Klyachko, N. L., and Visser, A. (2000). Fluorescence dynamics of green fluorescent protein in AOT reversed micelles. Biophys. Chem. 87, 73–84. Utada, A. S., Lorenceau, E., Link, D. R., Kaplan, P. D., Stone, H. A., and Weitz, D. A. (2005). Monodisperse double emulsions generated from a microcapillary device. Science 308, 537–541. van Dijk, M. A., Kapitein, L. C., van Mameren, J., Schmidt, C. F., and Peterman, E. J. G. (2004). Combining optical trapping and single-molecule fluorescence spectroscopy: Enhanced photobleaching of fluorophores. J. Phys. Chem. B 108, 6479–6484. Velev, O. D., Prevo, B. G., and Bhatt, K. H. (2003). On-chip manipulation of free droplets. Nature 426, 515–516. Volkmer, A., Subramaniam, V., Birch, D. J. S., and Jovin, T. M. (2000). One- and twophoton excited fluorescence lifetimes and anisotropy decays of green fluorescent proteins. Biophys. J. 78, 1589. Wazawa, T., Ishii, Y., Funatsu, T., and Yanagida, T. (2000). Spectral fluctuation of a single fluorophore conjugated to a protein molecule. Biophys. J. 78, 1561–1569. Wennmalm, S., Edman, L., and Rigler, R. (1997). Conformational fluctuations in single DNA molecules. Proc. Natl. Acad. Sci. USA 94, 10641–10646. Xu, J., and Attinger, D. (2008). Drop on demand in a microfluidic chip. J. Micromech. Microeng. 18. Yamada, R., and Tada, H. (2005). Manipulation of droplets by dynamically controlled wetting gradients. Langmuir 21, 4254. Yao, H., Ikeda, H., Inoue, Y., and Kitamura, N. (1996a). Optical control of fusion of microparticles in solution and simultaneous spectrophotometric measurements. Anal. Chem. 68, 4304. Yao, H., Inoue, Y., Ikeda, H., Nakatani, K., Kim, H.-B., and Kitamura, N. (1996b). Micrometer size effect on dye association in single laser-trapped water droplets. J. Phys. Chem. 100, 1494. Yim, P. B., Zhang, Z., DeJong, E. S., Carroll, J. M., Marino, J. P., and Goldner, L. S. (2005). Single RNA kissing complexes studied by fluorescence resonance energy transfer. Proc. SPIE 5699, 450–460. Yoon, T. Y., Okumus, B., Zhang, F., Shin, Y. K., and Ha, T. (2006). Multiple intermediates in SNARE-induced membrane fusion. Proc. Natl. Acad. Sci. USA 103, 19731–19736. Zeng, J., and Korsmeyer, T. (2004). Principles of droplet electrohydrodynamics for lab-ona-chip. Lab Chip 4, 265. Zhang, Y. H., and Ozdemir, P. (2009). Microfluidic DNA amplification—A review. Anal. Chim. Acta 638, 115–125.

C H A P T E R

S I X

Single-Molecule Fluorescence Spectroscopy Using Phospholipid Bilayer Nanodiscs Abhinav Nath,*,† Adam J. Trexler,* Peter Koo,‡ Andrew D. Miranker,* William M. Atkins,† and Elizabeth Rhoades*,‡ Contents 90 91 92

1. Introduction 2. Nanodiscs and HDL Particles 2.1. POPC–MSP1D1 discs 3. Single-Molecule Techniques and Applications to Membrane Proteins 4. Cytochrome P450 3A4 and Its Allosteric Behavior 4.1. Incorporation of CYP3A4 in Nanodiscs 4.2. Surface attachment of CYP3A4–Nanodiscs 5. Image Filtering by Singular-Value Decomposition 5.1. SVD-based image filtering pseudocode 6. Islet Amyloid Polypeptide Binding to Nanodiscs 6.1. FCS measurement of rIAPP binding Nanodiscs 7. a-Synuclein Conformations on Nanodiscs 7.1. smFRET measurement of aS bound to Nanodiscs 8. Summary Acknowledgments References

95 96 99 100 102 104 106 107 109 110 112 112 112

Abstract Nanodiscs are a new class of model membranes that are being used to solubilize and study a range of integral membrane proteins and membraneassociated proteins. Unlike other model membranes, the Nanodisc bilayer is bounded by a scaffold protein coat that confers enhanced stability and a narrow particle size distribution. The bilayer diameter can be precisely controlled by

* Department of Molecular Biophysics and Biochemistry, Yale University, New Haven, Connecticut, USA Department of Medicinal Chemistry, University of Washington, Seattle, Washington, USA Department of Physics, Yale University, New Haven, Connecticut, USA

{ {

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72014-0

#

2010 Elsevier Inc. All rights reserved.

89

90

Abhinav Nath et al.

changing the diameter of the protein coat. All these properties make Nanodiscs excellent model membranes for single-molecule fluorescence applications. In this chapter, we describe our work using Nanodiscs to apply total internal reflection fluorescence microscopy (TIRFM), fluorescence correlation spectro¨rster resonance energy transfer (FRET) to study the integral scopy (FCS), and Fo membrane protein cytochrome P450 3A4 and the peripheral membrane-binding proteins islet amyloid polypeptide (IAPP) and a-synuclein, respectively. The monodisperse size distribution of Nanodiscs enhances control over the oligomeric state of the membrane protein of interest, and facilitates accurate solution-based measurements as well. Nanodiscs also comprise an excellent system to stably immobilize integral membrane proteins in a bilayer without covalent modification, enabling a range of surface-based experiments where accurate localization of the protein of interest is required.

Abbreviations ANF CYP DOPG FCS FRET HDL IAPP LCAT MSP NR PCA PD POPC POPS SVD TAMRA-SE TIRFM

a-naphthoflavone cytochrome P450 dioleoylphosphatidylglycerol fluorescence correlation spectroscopy Fo¨rster resonance energy transfer high-density lipoprotein islet amyloid polypeptide lecithin:cholesterol acyltransferase membrane scaffold protein Nile Red principal components analysis Parkinson’s disease palmitoyloleoylphosphatidylcholine palmitoyloleoylphosphatidylserine singular-value decomposition tetramethylrhodamine-succinimidylester total internal reflection fluorescence microscopy

1. Introduction Phospholipid bilayer Nanodiscs (Bayburt and Sligar, 2009; Bayburt et al., 2002; Nath et al., 2007a; Ritchie et al., 2009) are an emerging model membrane system for the study of membrane-associated proteins.

Single-Molecule Spectroscopy with Nanodiscs

91

Nanodiscs consist of a phospholipid bilayer surrounded by a protein coat formed of membrane scaffold protein (MSP) and are derived from nascent (discoidal) high-density lipoprotein (HDL) particles. Nanodiscs are more stable and monodisperse than conventional model membranes such as liposomes, bicelles, and micelles, and are thus a very appealing model system for a range of biochemical and biophysical experiments with integral and peripheral membrane proteins. Given the importance of membrane proteins in so many biological and pharmacological questions, there has been an understandable interest in novel Nanodisc technology and a number of exciting developments in membrane protein biochemistry over the past few years (Alami et al., 2007; Boldog et al., 2006; Morrissey et al., 2008). Concurrent with the growing use of Nanodiscs, there has been a rise in the application of single-molecule fluorescence techniques to a range of biological problems, including movement of motor proteins (Park et al., 2007; Peterman et al., 2004), ribosome dynamics (Blanchard et al., 2004), and enzyme catalysis (Henzler-Wildman et al., 2007; Lu et al., 1998), that have provided fundamentally new mechanistic insights and a new appreciation for the role of stochasticity and nonlinear dynamics in a range of biological processes. Several groups have recently reported the application of single-molecule fluorescence to integral membrane protein incorporated in Nanodiscs (Nath et al., 2008b) or HDL particles (Kuszak et al., 2009; Whorton et al., 2007). In this chapter, we present detailed protocols from our published work, as well as new methods and results using Nanodiscs to study peripheral membrane-binding proteins, in the hope that this will prove useful to other investigators of membrane proteins.

2. Nanodiscs and HDL Particles Nanodisc technology was developed in the early part of this decade by the group of Stephen G. Sligar at the University of Illinois, UrbanaChampaign, and builds on the considerable body of knowledge contributed by numerous groups about the structure and characterization of HDL particle biology. While the biology of HDL particle formation, maturation, and maintenance has been reviewed elsewhere (Atkinson and Small, 1986; Ohashi et al., 2005), we briefly describe parts of the process relevant to Nanodisc preparation. HDL is involved in reverse cholesterol transport from various tissues to the liver. Apolipoprotein A-I (ApoA-I) can bind serum phospholipids to form a discoidal particle with two protein monomers wrapped around a lipid bilayer
10 nm in diameter (Segrest et al., 1999); as many as five varieties of this species have been proposed, each with slightly

92

Abhinav Nath et al.

different lipid:protein ratios or protein packing around the bilayer. Subsequent loading of cholesterol and cholesteryl esters into the discoidal HDL particles by lecithin:cholesterol acyltransferase (LCAT) transforms them into mature, spheroidal HDL particles. HDL maturation and cholesterol uptake have recently been studied by single-particle fluorescence (Sanchez et al., 2007). Nanodisc formation is achieved by combining detergent-solubilized lipids and MSP in appropriate molar ratios, essentially determined by the optimal fluid-phase surface area occupied by the lipids in a bilayer and the diameter of the desired Nanodisc (Ritchie et al., 2009). The length of the protein coat defines the diameter of the Nanodisc particle, resulting in a relatively narrow size distribution that makes them especially well-suited to single-molecule experiments. Additionally, the Sligar lab has assembled a library of MSP constructs of varying lengths, all of which are ultimately derived from human ApoA-I, that can be used to modulate Nanodisc size with high fidelity (Denisov et al., 2004). In recent work, Hoeprich and coworkers (Baker et al., 2009) have developed apolipoprotein E4-based scaffold proteins that can be used to solubilize membrane proteins in particles up to 20 nm in diameter, further extending the versatility of Nanodisc-type model membranes. Table 6.1 contains example lipid ratios for common combinations of MSP constructs and lipids. Nanodisc self-assembly is triggered upon detergent removal from the mixed micelles, by dialysis or using hydrophobic adsorbents such as Bio-Beads SM-2 (Bio-Rad, Hercules, CA) or Amberlite XAD-2 (Sigma Aldrich, St Louis, MO; Fig. 6.1). Section 2.1 describes the fabrication of Nanodiscs without an integrated target protein.

2.1. POPC–MSP1D1 discs 1. An appropriate volume of lipid dissolved in chloroform (e.g., 200 ml of 20 mg/ml POPC) in a glass test tube is dried to a thin film under a gentle stream of nitrogen, and then dried overnight in a vacuum dessicator or for 2 h in a lyophilizer. To incorporate functionalized lipids, an appropriate mixture is used instead (e.g., 190 ml of 20 mg/ml POPC þ 8 ml of 20 mg/ml biotinyl-cap-DPPE for a 5% biotinylated bilayer). 2. The lipid film is dissolved in 200 ml of a buffer of 20 mM Tris–HCl, pH 7.4, 100 mM NaCl, 40 mM sodium cholate. Gentle vortexing and sonication may be necessary to completely solubilize the lipid; the resulting solution should look completely clear. 3. An appropriate volume of MSP (e.g., 560 ml of 150 ml MSP1D1) in 20 mM Tris–HCl, pH 7.4, 100 mM NaCl (Buffer A) is added, and the mixture is incubated with gentle shaking close to the transition temperature of the

93

Single-Molecule Spectroscopy with Nanodiscs

Table 6.1 Lipid–protein ratios for Nanodisc reconstitution

Diameter (nm)

Lipid composition

Protein (relative amounta)

Lipid 1 (relative amounta)

Lipid 2 (relative amounta)

9.7

100% POPC

MSP1D1 (1)

POPC (65)

–

9.7 9.7

100% DPPC MSP1D1 (1) 40%:60% MSP1D1 (1) POPC:DOPG

DPPC (90) POPC (26)

– DOPG (39)

12.9 12.9

100% POPC 40%:60% POPC:POPS

MSP1E3D1 (1) POPC (130) – MSP1E3D1 (1) POPC (52) POPS (78)

a

The value in parentheses is the desired molar excess of the relevant species with respect to scaffold protein concentration. For example, 40%:60% POPC:DOPG Nanodiscs made with MSP1D1 require a reconstitution mixture consisting of 1:26:39 MSP1D1:POPC:DOPG, plus an appropriate concentration of detergent. More details can be found at the Sligar group website, http://sligarlab.life.uiuc.edu or in Ritchie et al. (2009).

~5

Scaffold protein

Lipid

nm

m

Detergent Mixed micelles

0n

~1

Nanodiscs

Figure 6.1 Schematic of Nanodisc self-assembly. Scaffold proteins, lipid, and detergent are combined in appropriate ratios to form mixed micelles. Integral membrane proteins can also be included at this stage, as long as they can be solubilized in a suitable detergent. Detergent removal from the mixed micelles triggers self-assembly of the Nanodisc particle. Approximate dimensions for Nanodiscs created with MSP1 or MSP1D1 are shown. (The model, kindly provided by Dr. S. C. Harvey of Georgia Tech, is of a discoidal HDL particle with ApoA-I[D1-33.] (Segrest et al., 1999). The images are not to scale.)

lipid (4  C for POPC) for 30–90 min. Note that the molar ratio MSP:lipid: detergent is approximately 1:65:100, and the final concentration of detergent is slightly above its critical micelle concentration (6 mM for cholic acid); the latter consideration facilitates both mixed micelle formation and easy removal of detergent either by dialysis or by adsorbent. 4. Bio-Beads are washed with 2 volumes of methanol and 4 volumes of Buffer A. Approximately 0.3 g of these wet Bio-Beads are added to the

94

Abhinav Nath et al.

mixture from step 3. A saturated suspension of Bio-Beads has a density of
1 g/ml, so an appropriate volume can be transferred using a disposable pipet if this is more convenient. 5. The mixture with Bio-Beads is incubated close to the transition temperature of the lipid (4  C for POPC) for 4–12 h, and the Bio-Beads are then separated using a small gravity column or careful pipeting. 6. The size and monodispersity of the Nanodisc prep is verified using sizeexclusion chromatography on a Superdex 200 10/300 (GE Healthcare, Piscataway, NJ). The dimensions, bilayer properties, and biophysical behavior of Nanodiscs have been extensively characterized (Denisov et al., 2004; Grinkova et al., 2004) and generally correspond to what is known of HDL structure. By size-exclusion chromatography and native PAGE, the hydrodynamic diameter is close to 10 nm for particles made with ApoA-I or the scaffold protein constructs MSP1 (ApoA-I[D1-43]) or MSP1D1 (ApoA-I[D1-54], also referred to as MSP1T2 in some papers). The constructs MSP1E3 and MSP1E3D1, in which three additional 22-mer helical segments have been inserted into the sequences of MSP1 and MSP1D1, respectively, produce larger particles
12.5 nm in diameter. Small-angle X-ray scattering, atomic force microscopy, and transmission electron microscopy all indicate disc-shaped particles with similar diameters and a thickness of
5 nm, depending on the lipid composition (Denisov et al., 2004). Finally, differential scanning calorimetry measurements indicate that the influence of the scaffold protein on the bilayer is such that the lipids display a broad phase transition, less like liposomes and more like biological membranes (Denisov et al., 2005). Nanodiscs have been used to study integral membrane proteins such as G-protein coupled receptors (Leitz et al., 2006), cytochrome P450s (CYPs) (Davydov et al., 2007; Denisov et al., 2007; Kijac et al., 2007; Nath et al., 2007b), and bacterial chemoreceptors (Boldog et al., 2006, 2007), as well as peripheral membrane-associated proteins such as coagulation factors (Morrissey et al., 2008; Shaw et al., 2007). The constrained, monodisperse size of the Nanodisc bilayer has greatly facilitated close study of the stoichiometry of membrane protein complexes and their functional importance (Bayburt et al., 2006; Boldog et al., 2006, 2007). In one study of bacterial transmembrane chemoreceptors, the authors were able to modulate the oligomerization state of the target Tar protein by varying the stoichiometry of Tar per Nanodisc (Boldog et al., 2006). They found that the minimal structural unit of Tar is a homodimer, and while homodimers were properly modified in response to ligand binding, further downstream signaling required a higher order oligomer: a trimer of dimers. This level of control over oligomerization in a membrane is difficult to achieve with traditional model membranes.

Single-Molecule Spectroscopy with Nanodiscs

95

3. Single-Molecule Techniques and Applications to Membrane Proteins As this volume of Methods in Enzymology is devoted to single-molecule techniques, we provide only a very brief overview of the methods we use; the other chapters in this volume or standard textbooks (Gell et al., 2006; Lakowicz, 2006) provide a much more detailed discussion. We have used Nanodiscs in experiments with fluorescence correlation spectroscopy (FCS), Fo¨rster resonance energy transfer (FRET), and total internal reflection fluorescence microscopy (TIRFM). In FCS (Magde et al., 1972), the fluctuations in fluorescence from a dilute solution of particles diffusing through a small (
1 fl) focal volume are autocorrelated. Any process that changes fluorescence causes a change in the calculated autocorrelation function, and the rate of this process determines the timescale at which autocorrelation decays. The most common contribution to autocorrelation decay is simply the diffusion of particles into and out of the confocal volume: measurement of the diffusion time through a given volume enables an estimation of the diffusion time of a particle, and hence (by the Stokes–Einstein equation) its hydrodynamic radius. Although FCS is not purely a single-molecule technique, it can be used on the singlemolecule level and it is frequently discussed in the same category as strictly single-molecule fluorescence methods. FRET is the nonradiative transfer of energy that occurs via a transition dipole interaction between two fluorophores, a donor and an acceptor (Roy et al., 2008). The efficiency of transfer (ETeff) is proportional to the inverse sixth power of the distance: for a given pair of fluorophores, a low ETeff corresponds to fluorophores that are further apart than those with a high ETeff. For our purposes, FRET is used to report on protein conformation and thus both the donor and acceptor fluorophore will be placed on a single protein. As with FCS, fluorescence emission is measured from the molecules as they diffuse through the focal volume. The sample is diluted so that only a single-labeled protein is present in the observation volume and the relative emission collected from the donor and acceptor fluorophores is used to calculate ETeff for each molecule. In TIRFM (Axelrod, 1981), a laser beam is totally internally reflected at a glass surface (using either a prism or a high numerical-aperture oil objective). This generates an evanescent electromagnetic field that decays exponentially from the surface of the slide, and thus specifically excites the fluorophores in the sample within 100–200 nm of the surface. If particles of interest are attached to the glass slide at sufficiently low surface concentrations that they are separated by at least the diffraction limit, then fluorescence from single particles can be individually observed. The three

96

Abhinav Nath et al.

techniques mentioned earlier are not mutually exclusive and can be used in combination: for example, FCS–FRET (Torres and Levitus, 2007) can provide the rates of conformation fluctuations of molecules in solution, while TIRFM–FRET (Kozuka et al., 2006) reports on the conformational states of surface-attached molecules. Single-molecule fluorescence has been used extensively to study membrane-binding peptides and integral membrane proteins, interacting with a range of model membranes and mimetics including micelles, unilamellar vesicles, and supported bilayers, or even in living cells (Brunger et al., 2009; Garcia-Saez and Schwille, 2007). To cite just a few recent examples from the literature, single-particle tracking has been used to study the mechanism of protein recruitment to supported bilayers (Knight and Falke, 2007) and the formation of signaling complexes in living cells (Murakoshi et al., 2004). FCS can measure the difference in diffusion rates between a free protein and a liposome-bound state, enabling measurements of membrane affinity with high sensitivity (Rhoades et al., 2006). Singlemolecule FRET has been used to study conformational states of intrinsically disordered proteins bound to detergent micelles (Ferreon et al., 2009) as well as liposome-bound transmembrane motors (Diez et al., 2004). Compared to other model membranes, Nanodiscs are unsuitable for investigations of membrane leakage or diffusion within a bilayer. They are, however, advantageous for solution-based techniques that would benefit from a monodisperse membrane size distribution, such as FCSbased binding measurements. Because the Nanodisc bilayer is relatively small and its size can be accurately controlled, Nanodiscs are also an excellent system for the characterization of oligomers and complexes of membrane proteins in situ. Finally, their usefulness is most evident in the unique ability to stably immobilize membrane proteins in a discrete spatial location without any covalent modification or encapsulation of the protein (Rhoades et al., 2003). This enables long-term observation of single-membrane proteins in something very close to a native lipid environment. In the following sections, we describe our results and protocols using Nanodiscs to study three different protein systems by TIRFM, FCS, and FRET.

4. Cytochrome P450 3A4 and Its Allosteric Behavior Cytochrome P450 3A4 (CYP3A4) is the major drug-metabolizing enzyme in humans, responsible for the clearance of
50% of pharmaceutical compounds (Thummel and Wilkinson, 1998). It is a member of the CYP superfamily of heme-thiolate monooxygenases, capable of catalyzing

Single-Molecule Spectroscopy with Nanodiscs

97

reactions including hydroxylation, heteroatom oxidative dealkylation, and epoxidation on a broad range of different substrates (Davydov and Halpert, 2008; Guengerich, 2001). CYP3A4 is an integral membrane protein with an N-terminal helical anchor and other hydrophobic membrane-associated regions, and was one of the first proteins successfully incorporated in Nanodiscs (Baas et al., 2004). CYP3A4 incorporated into Nanodiscs is stable, monomeric, and soluble, and this complex has enabled a range of experiments on CYP3A4 biophysics that were previously impossible or impractical, or suffered from complications caused by the tendency of CYP3A4 to aggregate or oligomerize. Experiments with CYP3A4–Nanodiscs include equilibrium ligand binding (Baas et al., 2004; Nath et al., 2007b), stopped-flow kinetic measurements (Davydov et al., 2005), pressure-perturbation spectroscopy (Davydov et al., 2007), and turnover kinetics measurements with Nanodiscs containing a complex of CYP3A4 and cytochrome P450:NADPH oxidoreductase, an accessory enzyme required for normal catalysis (Denisov et al., 2007). It has been suggested that interactions between two or more CYPs in a membrane can affect catalytic activity and ligand binding (Fernando et al., 2007; Subramanian et al., 2009). CYP3A4–Nanodiscs provide a minimal system to study CYP3A4 behavior in a native-like membrane, but without interference from other membrane proteins. Nanodiscs have enabled valuable new insights into CYP3A4 mechanisms. One particularly challenging aspect of CYP3A4 enzymology is that it displays atypical allosteric kinetics with a large fraction of its substrates. CYP3A4 metabolizes thousands of substrates, ranging in size from
150 Da (e.g., acetaminophen) to
1200 Da (cyclosporin), and with other physicochemical properties just as diverse. Because it needs to accommodate such a broad group of substrates, CYP3A4 is capable of binding one, two, or three substrate molecules simultaneously. If the different complexes have altered Vmax values (or different affinities for additional substrate molecules), the functional result is a deviation from standard Michaelis–Menten kinetics. Although not classical multisubunit allosterism in the sense of Monod– Wyman–Changeux (Monod et al., 1965) or Koshland–Nemethy–Filmer (Koshland et al., 1966) models, CYP3A4 does display quite a high degree of cooperativity both with a single chemical species acting homotropically and with heterotropic cases involving distinct substrates and effectors (Atkins, 2004). This is clinically relevant: in part because of CYP3A4 allosterism, many combinations of drugs can alter each other’s metabolism by inhibition or activation, in ways that are difficult to predict solely from in vitro data for each drug alone. These drug–drug interactions can lead to unfavorable pharmacokinetics, where one or more of a patient’s drugs are outside their therapeutic window: if clearance by CYP3A4 or other drug-metabolizing enzymes is activated, then the circulating drug concentration might be too

98

Abhinav Nath et al.

low to be efficacious; if clearance is inhibited, then the drug could build up to toxic levels. This has generated significant pharmacological interest in preventing adverse drug–drug interactions, and in harnessing favorable drug– drug interactions by coadministering adjuvants that extend the duration of the therapeutic window (Zeldin and Petruschke, 2004). Many steroids and small polycyclic aromatic molecules act as CYP3A4 effectors: at low concentrations, they boost the metabolism of other substrates, and their own turnover often displays sigmoidal kinetics, implying that they boost their own metabolism. There has been a growing consensus in the past few years that these effectors bind to a peripheral site (outside the active site) and thereby modulate CYP3A4 structure and/or dynamics, or protein–protein interactions within the membrane, so as to increase the affinity for substrate binding at the active site. There is strong evidence from X-ray crystallography (Williams et al., 2004) that progesterone can bind outside the active site. As reviewed in Davydov and Halpert (2008) and Denisov et al. (2009), a large body of biophysical and enzymological characterization (Denisov et al., 2007; Isin and Guengerich, 2007; Lampe and Atkins, 2006; Roberts and Atkins, 2007; Tsalkova et al., 2007) focuses on the mechanism of effectors such as testosterone, progesterone, and a-naphthoflavone (ANF). Our studies of allosteric mechanisms were greatly facilitated by the discovery that the membrane dye Nile Red (NR) was not only a highly active substrate of CYP3A4 but also an allosteric reporter (Lampe et al., 2008; Nath et al., 2008a). Using a combination of equilibrium binding experiments monitored by fluorescence and absorbance spectroscopy, and catalytic assays, we found that CYP3A4 can bind one or two NR molecules, with higher affinity for the active site than for the peripheral effector site (in contrast to archetypal effectors such as ANF). NR fluorescence is greatly enhanced upon binding CYP3A4, and the doubly bound species displays much brighter and red-shifted fluorescence than the 1:1 complex. Finally, because NR preferentially binds the active site and ANF preferentially binds the peripheral effector site, CYP3A4 can easily form a trimeric complex with both NR and ANF bound. This heterotrimeric species is spectrally distinct from both the singly and doubly occupied NR–CYP3A4 complexes (Fig. 6.2). NR fluorescence thus provides a rich set of probes to investigate different aspects of homotropic and heterotropic cooperativity displayed by CYP3A4. We applied single-molecule TIRFM to the CYP3A4–NR–ANF system as described in a recent paper (Nath et al., 2008b), and were able to measure residence times of NR in the CYP3A4–Nanodisc complex based on the marked increase in NR fluorescence intensity when bound to either the CYP3A4 active site or the Nanodisc bilayer. Here we describe in detail the methods used in these experiments. CYP3A4 incorporated into Nanodiscs containing biotinylated lipids as described in Section 4.1. Surface attachment of CYP3A4–Nanodiscs using biotin–avidin chemistry is described in Section 4.2.

99

Single-Molecule Spectroscopy with Nanodiscs

M

~0.3 mM

m ~10

~1

N

mM

O

N O

O O

NR

ANF

Figure 6.2 Our model of Nile Red (oval) and a-naphthoflavone (rectangle) binding to CYP3A4 (gray prism) incorporated into a model membrane such as Nanodiscs (white cylinder). In this model, based on data from previous work (Nath et al., 2007a, b, 2008a,b), NR binds with high affinity to the active site, and with lower affinity to the peripheral effector site. In contrast, ANF binds with high affinity to the effector site, and can form a heterocomplex with NR. All three species marked with an asterisk are fluorescent and can be spectrally resolved.

4.1. Incorporation of CYP3A4 in Nanodiscs 1. MSP1D1 is treated with AcTEV protease (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions to remove the N-terminal hexahistidine tag. Aliquots are stored at 80  C. The resulting protein is denoted MSP1D1(). 2. A suitable amount of lipid (e.g., 190 ml of 20 mg/ml POPC þ 8 ml of 20 mg/ml biotinyl-cap-DPPE for a 5% biotinylated bilayer) is dried to a thin film under a N2 stream and stored overnight in a vacuum dessicator. 3. CYP3A4 is expressed as previously described (Nath et al., 2007b). CYP3A4 (5 mM in 2 ml of 100 mM potassium phosphate, pH 7.4, with 20% glycerol) is solubilized with 20 ml of a 10% (v/v) solution of Emulgen-913 (Kao Chemicals, Osaka, Japan) for a final detergent concentration of 0.1% (v/v). The mixture is incubated with gentle shaking at room temperature for 1 h. 4. CYP3A4 is exchanged into Buffer A (from Section 2.1) with 0.1% Emulgen-913, using Amicon centrifugal filter columns (30 kDa cutoff; Millipore, Billerica, MA) until the final glycerol concentration is less than 3%. (Higher levels of glycerol interfere with the disc formation process.) At the final centrifugation step, the sample is concentrated to a final volume of
1 ml.

100

Abhinav Nath et al.

5. Lipid films are resuspended in 400 ml of Buffer A with 100 mM cholate by gentle vortexing and sonication. 6. MSP1D1() (580 ml of 150 mM solution) is added to the lipid-detergent solution. 7. The solubilized CYP3A4 sample from step 4 is added. The mixture now contains CYP3A4:MSP1D1():lipid in approximate 0.1:1:63 molar ratio with
20 mM cholate. This corresponds to an approximate fourfold excess of empty discs without CYP3A4, thereby disfavoring incorporating multiple CYP3A4 molecules into a single Nanodisc. 8. The mixture is incubated with gentle shaking at 4  C (Tm of lipid) for 1 h. 9. Wet Bio-Beads (2 g) are added to the mixture and incubated with gentle shaking at 4  C for at least 4 h. 10. Bio-Beads are removed using a gravity flow column, and the sample is further dialyzed three times against 1 l changes of Buffer A for 2 h each to remove residual detergent. 11. Nickel-affinity chromatography (Ni-NTA Superflow resin, Qiagen, Valencia, CA) is used to separate CYP3A4–Nanodiscs from empty Nanodiscs. CYP3A4–Nanodiscs are eluted with Buffer A with 300 mM imidazole, pH 7.4. 12. The CYP3A4–Nanodiscs are dialyzed overnight against Buffer A. Monodispersity of CYP3A4–Nanodiscs is verified using size-exclusion chromatography on a Superdex 200 10/300 (GE Healthcare). If the sample is not homogenous, fractions corresponding to the size of the CYP3A4–Nanodisc complex should be collected and pooled. For long-term storage, add glycerol to 10% and store at 80  C. 13. A successful preparation of monomeric CYP3A4–Nanodiscs will show approximately equal absorbance at 280 and 417 nm. (There can be some variability due to minor changes in the resting spin state of the CYP3A4 heme, which affects absorbance at 417 nm.) The addition of saturating bromocriptine will induce a complete conversion to the low-spin state with an absorbance peak at 388 nm, as shown in Fig. 6.3A.

4.2. Surface attachment of CYP3A4–Nanodiscs 1. Glass coverslips (22 mm  22 mm, #1 thickness, Fisher Scientific, Pittsburgh, PA) are cleaned by sonication for 30 min in 10% (v/v) 7 cleaning solution (MP Biomedical, Solon, OH) followed by another 30 min of sonication in Milli-Q water (i.e., deionized and 0.22 mmfiltered in a Milli-Q Integral system, Millipore). As an alternative, coverslips can be cleaned by 10 min incubation in Piranha solution, made by gradual addition of hydrogen peroxide (30%) to concentrated H2SO4 to a

101

Single-Molecule Spectroscopy with Nanodiscs

Absorbance

A

B

~200 nm

0.1

0

CCD

350

400 450 500 550 Wavelength (nm)

600

C

12 h wash = 16:0 biotinyl–cap–PE

10 min wash = Avidin

= CYP3A4-Nanodisc (with 5% biotinylated lipid)

Figure 6.3 Setup of a TIRFM experiment using CYP3A4–Nanodiscs. (A) Bromocriptine (8 mM) induced an almost complete shift in the heme absorbance signal of CYP3A4–Nanodiscs from low spin (solid line) to high spin (high spin). After verifying appropriate size and homogeneity by size-exclusion chromatography, this experiment serves as a convenient check that a preparation of CYP3A4–Nanodiscs is properly functional. (B) Schematic of a through-objective TIRFM experiment. Excitation laser light (green) is guided through one side of an oil objective so that it is incident at a glass slide at the critical angle. This generates an evanescent field that decays, selectively exciting fluorophores within a few hundred nanometers of the slide surface. Emitted light (red) from these particles is passed through a dichroic mirror to a CCD camera. (C) Schematic of the surface attachment procedure using biotin–avidin cross-linking described in Section 4.2.

final ratio of 1:3. Piranha-cleaned slides must be thoroughly rinsed with Milli-Q water before use. Clean coverslips are dried overnight at 60  C. 2. Flowcells are assembled by stacking two cleaned coverslips separated by two parallel thin strips of Parafilm (Pechiney Plastic Packaging, Chicago, IL) and heating for 15 s at
200  C, just until the Parafilm melts and secures the two coverslips together. 3. Flowcells are rinsed with two volumes of 50% NH4OH and 15 volumes of Milli-Q water, and then incubated overnight at 4  C with a 1 mg/ml aqueous solution of avidin (Sigma Aldrich).

102

Abhinav Nath et al.

4. Flowcells are rinsed with 10 volumes of Buffer A (from Section 2.1), and then Nanodiscs or CYP3A4–Nanodiscs containing 5% biotinylated lipids are added at a concentration of 30 nM. 5. After 10 min, flowcells are rinsed with 2 more volumes of Buffer A to remove any unbound particles, and then NR (
50 nM in Buffer A) is added (Fig. 6.3C). Loaded flowcells are immediately transferred to the microscope for TIRFM visualization. Our TIRFM instrumentation (Fig. 6.3B) is based on an Olympus IX-71 inverted microscope (Olympus America, Center Valley, PA). A 561 nm, 50 mW diode-pumped solid-state continuous-wave laser (Newport Corp, Irvine, CA) is passed through a neutral density filter (OD was varied from 0.04 to 2.0), then through a 5:1 beam expander. The beam is then focused into the back aperture of a 60/1.45 NA oil objective (Olympus) and the entry of the beam into the back aperture is adjusted to achieve total internal reflection, creating a
75-mm-diameter evanescent field. Emitted fluorescence is passed through a 585-nm long-pass filter (Chroma Technology Corp., Rockingham, VT) and then to an iXon back-thinned EMCCD camera (Andor USA, South Windsor, CT) for detection. Images are captured using Andor Solis software. EM gain is varied from 30 to 280 dB, and exposure time is varied from 6 to 100 ms in an effort to maximize sensitivity while still capturing rapid binding events. Trajectories are collected for 1000 frames each (i.e., 6–100 s in length).

5. Image Filtering by Singular-Value Decomposition Trajectories with short exposures (6–30 ms) showed high levels of background noise (Fig. 6.4A). One popular approach to improving sensitivity is to average neighboring frames (Ulbrich and Isacoff, 2007), but in our case, this runs the risk of discarding rapid binding events and thereby biasing the distribution of residence times. The residence time of NR in membranes has been reported to be
15 ms (Gao et al., 2006), and so rapid events are expected to be especially relevant when characterizing binding to the Nanodisc bilayer. Therefore, we chose to filter images with exposure times <50 ms, using singular-value decomposition (SVD), a popular technique in image processing (Hendler and Shrager, 1994; Protter and Elad, 2009). SVD is related to principal components analysis (PCA) and similarly seeks to identify common trends in a set of data. When considering a set of TIRFM images, all of the pixels corresponding to the diffraction-limited image of a single particle will, on average, display higher covariance (i.e., changes in intensity correlated with each other) than random noise, which is uncorrelated. Conceptually, SVD can be thought of as the determination of

103

Single-Molecule Spectroscopy with Nanodiscs

A

B

C Fluorescence (a.u.)

900 800 700 600 500 400 16

20

24

28

32

36

Time (s)

Figure 6.4 Nile Red binding CYP3A4–Nanodiscs as observed by TIRFM. (A) Three unprocessed single frames from TIRF trajectories of NR binding CYP3A4–Nanodiscs. (B) The same three frames after noise reduction performed by singular-value decomposition as described in Section 5.1. (C) Comparison of unprocessed (gray) and filtered (black) time traces from the particle indicated by the white circle in (A), showing how the SVD process preserves information on bright periods.

the eigenvectors and eigenvalues of the covariance matrix of a dataset. Each eigenvector signifies an observed trend in the data, and the corresponding eigenvalue represents the magnitude or significance of that trend. In the context of SVD, the eigenvalues of the covariance matrix are called ‘‘singular values.’’ Higher singular values generally reflect meaningful trends in the data, while lower singular values correspond to random, uncorrelated noise. If the SVD is inverted while neglecting the lower singular values, the reconstructed dataset will lack much of the noise present in the original data. The key parameter in SVD-based filtering is the number of singular values to retain in the reconstructed image: too many, and the reconstructed image will still contain noise; too few, and significant ‘‘meaningful’’ information will be lost. Section 5.1 is a pseudocode description of our SVD

104

Abhinav Nath et al.

filtering procedure. Sample source code in Python for SVD filtering is available at http://www.yale.edu/rhoadeslab. Figure 6.4B shows the reduction in noise achieved by SVD, while Fig. 6.4C that SVD-filtered intensity traces from single particles retain information on rapid binding events.

5.1. SVD-based image filtering pseudocode 1. Convert each frame of a trajectory into a one-dimensional array, so that a 256  256 pixel image, for example, becomes a 65,536-value vector, with the value at each position corresponding to the brightness of a particular pixel. 2. Assemble the resulting arrays from all the frames of a trajectory into a matrix D, so that for example, a 1000-frame trajectory of 256  256 pixel images becomes a 65,536  1000 matrix. (Note: matrix operations of this size can be very memory-intensive, so to improve performance, it may be necessary to spatially divide the original into, for example, four 1000frame 128  128 pixel trajectories and process each one separately.) 3. Perform SVD on the matrix D. Many efficient implementations exist for various programming environments; we used the linalg.svd() function of Scientific Python ( Jones et al., 2001). SVD produces three matrices U, S, and V such that USVT ¼ D. The matrix U (65,536  1000) contains all the eigenvectors of the covariance matrix (i.e., common trends in the images) and V (1000  1000) shows how they change over the course of the trajectory. S (1000  1000) contains the singular values along its diagonal (and zeroes everywhere else), reflecting the relative contribution of each trend to the information in the original trajectory. 4. Choose a number n greater than the estimated number of particles visible in the trajectory, and discard all the columns of U and S, and all the rows of VT, whose index is greater than n. Some trial and error in setting the value of n is necessary for each trajectory: if n is too large, then the filtered trajectory will still contain much of the noise present in the original data. However, if n is too small, the filtered trajectory will lack significant transitions from the original data (i.e., the data will be oversmoothed), and meaningful information will be lost. 5. Reconstruct the filtered trajectory by multiplying the truncated versions of U, S, and VT to generate R (65,536  1000). 6. Convert each column of R to an image with the same dimensions as the originals (256  256). 7. Assemble the images into a trajectory of the same length as the original (1000 frames). For further analysis, we averaged intensities from a 3  3 pixel area around well-separated diffraction limited spots and manually measured the extent of bright periods, which represent bound NR. The resulting

105

Single-Molecule Spectroscopy with Nanodiscs

distribution of residence times was fit to exponential decays to determine the off-rate of NR under different conditions. As described in our earlier paper (Nath et al., 2008b), we measured a residence time for NR in Nanodiscs of
30 s 1, in good agreement with the measurements of residence times in lipid bilayers made by others (Gao et al., 2006). Using CYP3A4–Nanodiscs, we found a new, slower (
1.5 s 1) phase which we ascribed to NR dissociation from the CYP3A4 active site. Miconazole, a competitive inhibitor, blocked NR binding to CYP3A4 and eliminated the slow phase. The presence of 5 mM ANF, which should predominantly bind the peripheral effector site while leaving the active site accessible, further slowed the putative off-rate (
0.3 s 1). This would seem to suggest that CYP3A4 effectors may act by increasing the residence time of substrates in the enzyme active site. ANF did not affect off-rates from plain Nanodiscs, suggesting that the observed effect is mediated by CYP3A4 and not the bilayer. A common consideration with single-molecule data of this type is whether they contain artifacts due to photophysical effects. Intermittent photoblinking associated with the triplet state should occur on a submillisecond timescale for oxazine dyes like NR in aerobic solutions (Vogelsang et al., 2009), so they are unlikely to contribute to the observed bright and dark periods. Photobleaching of bound NR can also contribute to the observed decay in bright periods. To test the effect of this phenomenon, we compared data collected with different illumination powers (Fig. 6.5) and found relatively minor differences.

Frequency (normalized)

1.2 1 0.8

OD 0.5 OD 0.04

0.6 0.4 0.2 0 0

500

1000

1500

2000 2500 Time (ms)

3000

3500

4000

Figure 6.5 Dwell-time histograms measured using neutral density filters of OD 0.04 (gray bars) and OD 0.5 (black) bars in the excitation beam path. The higher laser power corresponded to an apparent decrease in the residence time of NR in the CYP3A4– Nanodisc active site of
25%, which we attribute to photobleaching of the fluorophore. Although this is a relatively minor change relative to the effect of ANF (a fourfold increase in residence time), we used the lower laser power for our measurements to minimize the contribution of photobleaching.

106

Abhinav Nath et al.

6. Islet Amyloid Polypeptide Binding to Nanodiscs Islet amyloid polypeptide (IAPP; also called amylin) is an intrinsically disordered amyloidogenic peptide strongly implicated in the pathogenesis of type II diabetes (Hebda and Miranker, 2009). IAPP is a peptide hormone cosecreted with insulin by pancreatic b-cells in healthy individuals, but readily forms amyloid fibers in vitro. Preamyloid oligomeric intermediates may be significant effectors of b-cell death, and thereby contribute to the progression of type II diabetes. Understanding what factors mediate pathogenic misfolding of IAPP is therefore of deep interest for the design of therapeutics. IAPP binds anionic membranes with high affinity. Membrane binding catalyzes the formation of predominantly a-helical oligomers and also accelerates amyloid (cross-b) fibril formation (Knight and Miranker, 2004; Knight et al., 2006). In turn, membrane-bound oligomers appear to contribute to toxicity by inducing membrane leakage. Studies of this complicated thermodynamic landscape, with membrane partitioning linked to the cooperative formation of two different types of oligomer, are fortunately simplified by the fact that the rat IAPP homolog, rIAPP, also forms membrane-bound oligomers but does not form amyloid fibrils (Knight et al., 2006). Several groups (Ling et al., 2009; Mishra et al., 2009; Nanga et al., 2008) are attempting to characterize structural features of different states in this pathway, using NMR and infrared spectroscopy. However, our current thermodynamic understanding of IAPP–membrane interactions relies primarily on the ensemble spectroscopic techniques of circular dichroism and thioflavin T fluorescence. FCS-based binding measurements with Nanodiscs may provide a more sensitive and detailed way to dissect the thermodynamics of IAPP–membrane interactions. For the case of a single freely diffusing species observed by FCS, the autocorrelation decay is fit with the following equation:

 1 t 1 t 1=2 GðtÞ ¼ 1þ 2 1þ N tD s tD here N is the average number of molecules in the confocal volume, tD is the diffusion time of the species, and s is a measure of the prolateness of the observation volume (the ratio of Z-axial radius to XY-radius). For a system with two components possessing distinct diffusion times, such as a mixture of free and lipid-bound labeled protein, the equivalent equation is

Single-Molecule Spectroscopy with Nanodiscs

107

"

 1 t 1 t 1=2 GðtÞ ¼ 1þ 2 A 1þ N tf s tf
1
 # t t 1=2 1þ 2 þQð1  AÞ 1 þ tb s tb here A is the fraction of free protein, Q is the average ‘brightness’ of the lipid-bound species relative to free protein, tf is the diffusion time of the free protein, and tb is the diffusion time of the lipid-bound species. It should be noted that this equation assumes that the free and bound states of the protein have equal brightness. The diffusion time of the bound species will correspond to that of plain Nanodiscs, which can be determined with high precision by virtue of their monodisperse size distribution. The FCS-based binding approach enables sensitive direct measurements of free and bound concentrations, while requiring orders of magnitude less material than ensemble experiments. This facilitates precise measurements of high-affinity binding events, which can be difficult when binding partners are present at high concentrations. As a proof of principle, we describe FCS measurements of rIAPP binding to Nanodiscs containing 60% DOPG:40% POPC.

6.1. FCS measurement of rIAPP binding Nanodiscs 1. Tetramethylrhodamine-succinimidylester (TAMRA-SE) is used to label rIAPP at the e-amino position of Lys1, the sole lysine residue in the rIAPP sequence. TAMRA-SE (0.2 mg in 100 ml DMSO) is added to rIAPP (1 mg in 1 ml Buffer A from Section 2.1) and the mixture is incubated with gentle shaking in the dark for 4 h at room temperature. Excess free dye is separated from labeled rIAPP using a Sephadex G-25 desalting column (GE Healthcare). Efficient labeling is verified by reverse-phase HPLC on a C-18 column (Grace Vydac, Deerfield, IL). 2. Nanodiscs are prepared as in Section 2.1, except that the lipid mixture is 100 ml of 20 mg/ml and 100 ml of 25 mg/ml DOPG. 3. FCS measurements are performed on an Olympus IX-71 inverted microscope with 60/1.2 N.A. water objective. The output of a 561 nm, 50 mW diode-pumped solid-state continuous-wave laser (Newport Corp) is adjusted with neutral density filters to a measured power of
5 mW just prior to entry to the microscope. Emitted fluorescence is collected through a 585 nm long-pass filter (Chroma Technology Corp.) and a 50 mm optical fiber (Oz Optics, Ottawa, Canada) to an avalanche photodiode (PerkinElmer, Waltham, MA) coupled to a Flex03-LQ-12 correlator (Correlator.com, Bridgewater, NJ) (Fig. 6.6A).

108

Abhinav Nath et al.

B Normalized autocorrelation, G(t)

A

APD Pinhole

Fraction bound

C

1.2 1 0.8 0.6 0.4 0.2 0 0.001 0.01 0.1 1 10 Time t (ms)

100 1000

0.75

0.5

0.25

0

0

25 50 Nanodiscs (nM)

75

Figure 6.6 IAPP binding Nanodiscs as observed by FCS. (A) Schematic of an FCS experiment. Excitation light (green) is focused into a sample, and emitted light (red) from a
1 fl observation volume is directed through an optical fiber (that serves as a pinhole) to an avalanche photodiode (APD) and autocorrelated. Any change in the intensity of fluorescence manifests as a decay in autocorrelation that provides information on the rate of the underlying process. In our case, the relevant processes are the diffusion of labeled rIAPP in free and Nanodisc-bound states (not shown to scale). (B) Autocorrelation traces of free rIAPP (red) and Nanodiscs with trace NR (black), showing the rightward shift expected with larger hydrodynamic size and longer diffusion time. In cyan is a mixture of labeled rIAPP and Nanodiscs under conditions with about half of the peptide bound, showing the intermediate two-component autocorrelation decay. (C) Binding isotherm obtained as Nanodiscs containing 60% DOPG:40% POPC were titrated into 10 nM labeled rIAPP. Performing this experiment at a range of different peptide concentrations will allow further study of the complex binding cooperativity observed in rIAPP–membrane interactions (Knight et al., 2006).

Single-Molecule Spectroscopy with Nanodiscs

109

4. To measure the Nanodisc diffusion time, 10 autocorrelation traces of 10 nM NR in the presence of 10 mM Nanodiscs are collected, averaged, and fit with the single-component FCS equation to determine tb. Under these conditions, the only contribution to observed fluorescence is from bound NR, which therefore provides an accurate measurement of Nanodisc tD. Alternatively, this calibration can be performed on Nanodiscs containing 1% rhodamine-DOPE. 5. 30 autocorrelation traces of 10 nM TAMRA–rIAPP are collected, averaged, and fit to the single-component FCS to determine tf. 6. 30 autocorrelation traces of 10 nM TAMRA–rIAPP are collected at a range of Nanodisc concentrations (Fig. 6.6B). Each set is averaged and fit to the two-component FCS equation with N and A (free fraction) as the only variables. A is plotted against Nanodisc concentration to generate a binding curve (Fig. 6.6C). Typical standard errors of A from this fitting approach are on the order of 0.5–1% of the free fraction (i.e., ranging from about 0.0015 to 0.006 over the course of the titration). Given that the uncertainty in tf and tb values is also about 1%, the overall uncertainty in the value of the bound fraction is about 2%. The resulting binding isotherm has an apparent KD of
50 nM, corresponding to a membrane partition coefficient of 2  105. This is relatively close to the equivalent partition coefficient for 100% DOPG membranes of 5  104, measured using circular dichroism spectroscopy (Knight et al., 2006). It must be strongly emphasized that IAPP–membrane binding is a complex thermodynamic process, with separate parameters governing membrane partition, nucleation, and growth of membranebound oligomers. A meaningful understanding of the system requires global analysis of data collected over a range of different protein and lipid concentrations. The improved sensitivity afforded by single-molecule techniques enables experiments over a concentration range much broader than the
10-mM regime necessary for the ensemble techniques we previously used (Knight et al., 2006), and so should lead to a more complete and accurate understanding of IAPP–membrane interactions.

7. a-Synuclein Conformations on Nanodiscs a-Synuclein (aS) is the primary protein constituent of cytoplasmic Lewy bodies and Lewy neurites that are pathologically linked to Parkinson’s disease (PD) (Goedert, 2001; Ueda et al., 1993). Although aS is strongly implicated in disease progression (Cookson, 2005), its precise role in PD is unclear. The native function of aS is also poorly understood, although evidence suggests that it may play a role both in maintaining neuronal plasticity and in the regulation

110

Abhinav Nath et al.

of synaptic vesicle recycling (George et al., 1995; Lotharius and Brundin, 2002) and thus that binding to membranes may be important to its function. aS is intrinsically disordered in solution (Weinreb et al., 1996) but undergoes a conformational change to an a-helical structure upon association with negatively charged membranes (Davidson et al., 1998; Jo et al., 2000). Two contrasting models have been proposed for the conformation of membranebound aS, either an extended, continuous helix (Bussell and Eliezer, 2003; Georgieva et al., 2008; Jao et al., 2004, 2008), or two antiparallel, noninteracting helices (Borbat et al., 2006; Chandra et al., 2003; Drescher et al., 2008; Ulmer et al., 2005), with an unstructured loop region in between. We recently used single-molecule FRET to study aS membrane-bound conformations, using lipid vesicles and SDS micelles (Trexler and Rhoades, 2009). We observed that aS displays either an extended helix on 50- and 100-nm lipid vesicles, and a hairpin conformation on the much more highly curved detergent micelles, in good agreement with the findings of other recent single molecule (Ferreon et al., 2009) and EPR (Georgieva et al., 2008) studies. Here, the use of Nanodiscs as a model membrane allows us to decouple the effects of membrane size and curvature: unlike vesicles and micelles, it is possible to modulate Nanodisc size by appropriate choice of MSP while retaining a flat bilayer (Denisov et al., 2004). The double cysteine mutant, aS T33C/T72C, labeled with a donor and acceptor fluorophore, can differentiate between bent and extended helical conformations. Here, we use this aS mutant to report on the helicalconformation of aS bound to Nanodiscs made with the extended scaffold protein MSP1E3, which have a bilayer
13 nm in diameter. aS T33C/ T72C was expressed, purified, and labeled essentially as previously described (Rhoades et al., 2006; Trexler and Rhoades, 2009).

7.1. smFRET measurement of aS bound to Nanodiscs 1. Nanodiscs are prepared as described in Section 2.1, except with the following mixture of scaffold protein and lipid: MSP1E3 (265 ml of 155 mM) and POPS (200 ml of 20 mg/ml). 2. Single-molecule FRET measurements are performed on an Olympus IX-71 inverted microscope with 60/1.2 N.A. water objective. The output of a 488 nm, 50 mW diode-pumped solid-state continuous-wave laser (Newport Corp) is adjusted with neutral density filters to a measured power of
10 mW just prior to entry to the microscope. Fluorescence is split using a 585 nm dichroic mirror (Chroma Technology Corp.) and fiber-coupled (100 mm fibers, Oz Optics) to avalanche photodiodes (PerkinElmer). 3. Samples contain 90 pM double-labeled aS T33C/T72C, the low concentration ensuring that each observed burst represents a single particle passing through the observed volume. The Nanodisc concentration is

111

Single-Molecule Spectroscopy with Nanodiscs

600 nM, an excess that enhances the probability of having at most one protein bound per disc. 4. Data are collected in 1ms time bins and processed using MATLAB scripts as described in detail in our previous work (Trexler and Rhoades, 2009). For each burst of photons associated with a Nanodisc-bound protein, the energy transfer efficiency (ETeff) value is calculated using the following formula: ETeff ¼

Ia  bId Ia þ gId

here Ia and Id are intensity in the acceptor and donor channel, respectively, corrected for background counts. b accounts for bleed-through of donor photons to the acceptor channel, while g accounts for differences in quantum yield and detection efficiency of the donor and acceptor fluorophores and was experimentally determined for our instrument as 1.2. ETeff values were then compiled into a histogram for analysis. When bound to the Nanodisc bilayer, aS T33C/T72C showed a peak ETeff of 0.53 (Fig. 6.7A), close to that measured using 100-nm POPS vesicles (0.57), and considerably lower than the corresponding value for SDS micelles (0.72) (Trexler and Rhoades, 2009). This observation suggests that Nanodiscs induce an extended helix conformation in aS (Fig. 6.7B)

A

B

Probability

0.15

0.1

0.05

0

0

0.25

0.5

0.75

1

ETeff

Figure 6.7 Conformation of aS bound to Nanodiscs as observed by FRET. (A) FRET histogram collected for double-labeled aS T33C/T72C bound to MSP1E3D1–POPS Nanodiscs shows a single species with a peak ETeff close to that observed for an extended helix on large unilamellar vesicles. (B) Model for aS (in yellow) extended helix conformation when bound to a Nanodisc, scaled appropriately to account for the larger diameter of particles made with MSP1E3D1 scaffold protein. Red and green spheres denote sites of labeling: a hairpin conformation, such as that observed with SDS micelles, would bring the dyes closer together and increase ETeff.

112

Abhinav Nath et al.

similar to that induced by much larger vesicles, but unlike the hairpin conformation observed with detergent micelles
4 nm in diameter. Membrane curvature may be an important determinant of aS conformation, and thus may be relevant to both native function and PD-associated pathology.

8. Summary In this chapter, we discussed how Nanodiscs may be used in popular single-molecule fluorescence techniques. Nanodiscs are a powerful complement to conventional model membranes, especially well-suited to single-molecule approaches. Solution-based single-molecule experiments may be facilitated by the relatively precise control over particle size afforded by different MSP constructs. Nanodiscs may prove even more useful in surface-based experiments; they provide a system to stably immobilize singlemembrane proteins in a native-like bilayer without covalent modification.

ACKNOWLEDGMENTS We very gratefully acknowledge Prof. S. G. Sligar and members of his research group for helpful advice and generous gifts of Nanodisc scaffold proteins. Funding was provided by NIH grants GM-32165 to W. M. A. and E. R., and GM-084391 to E. R. and A. D. M., and by the Ellison Medical Foundation to E. R.

REFERENCES Alami, M., Dalal, K., Lelj-Garolla, B., Sligar, S., and Duong, F. (2007). Nanodiscs unravel the interaction between the SecYEG channel and its cytosolic partner SecA. EMBO J. 26, 1995–2004. Atkins, W. (2004). Implications of the allosteric kinetics of cytochrome P450s. Drug Discov. Today 9, 478–484. Atkinson, D., and Small, D. (1986). Recombinant lipoproteins—Implications for structure and assembly of native lipoproteins. Annu. Rev. Biophys. Biol. 15, 403–456. Axelrod, D. (1981). Cell-substrate contacts illuminated by total internal reflection fluorescence. J. Cell Biol. 89, 141–145. Baas, B., Denisov, I., and Sligar, S. (2004). Homotropic cooperativity of monomeric cytochrome P450 3A4 in a nanoscale native bilayer environment. Arch. Biochem. Biophys. 430, 218–228. Baker, S., Hopkins, R., Blanchette, C., Walsworth, V., Sumbad, R., Fischer, N., Kuhn, E., Coleman, M., Chromy, B., Letant, S., Hoeprich, P., Adams, M., et al. (2009). Hydrogen production by a hyperthermophilic membrane-bound hydrogenase in water-soluble nanolipoprotein particles. J. Am. Chem. Soc. 131, 7508–7509. Bayburt, T., and Sligar, S. (2009). Membrane protein assembly into Nanodiscs. FEBS Lett. 10.1016/j.febslet.2009.10.024.

Single-Molecule Spectroscopy with Nanodiscs

113

Bayburt, T. H., Grinkova, Y. V., and Sligar, S. G. (2002). Self-assembly of discoidal phospholipid bilayer nanoparticles with membrane scaffold proteins. Nano Lett. 2, 853–856. Bayburt, T., Grinkova, Y., and Sligar, S. (2006). Assembly of single bacteriorhodopsin trimers in bilayer nanodiscs. Arch. Biochem. Biophys. 450, 215–222. Blanchard, S., Kim, H., Gonzalez, R., Puglisi, J., and Chu, S. (2004). tRNA dynamics on the ribosome during translation. Proc. Natl. Acad. Sci. USA 101, 12893–12898. Boldog, T., Grimme, S., Li, M., Sligar, S., and Hazelbauer, G. (2006). Nanodiscs separate chemoreceptor oligomeric states and reveal their signaling properties. Proc. Natl. Acad. Sci. USA 103, 11509–11514. Boldog, T., Li, M., and Hazelbauer, G. (2007). Using Nanodiscs to create water-soluble transmembrane chemoreceptors inserted in lipid bilayers. Methods Enzymol. 423, 317–335. Borbat, P., Ramlall, T., Freed, J., and Eliezer, D. (2006). Inter-helix distances in lysophospholipid micelle-bound alpha-synuclein from pulsed ESR measurements. J. Am. Chem. Soc. 128, 10004–10005. Brunger, A., Weninger, K., Bowen, M., and Chu, S. (2009). Single-molecule studies of the neuronal SNARE fusion machinery. Annu. Rev. Biochem. 78, 903–928. Bussell, R., and Eliezer, D. (2003). A structural and functional role for 11-mer repeats in alpha-synuclein and other exchangeable lipid binding proteins. J. Mol. Biol. 329, 763–778. Chandra, S., Chen, X., Rizo, J., Jahn, R., and Sudhof, T. (2003). A broken alpha-helix in folded alpha-synuclein. J. Biol. Chem. 278, 15313–15318. Cookson, M. (2005). The biochemistry of Parkinson’s disease. Annu. Rev. Biochem. 74, 29–52. Davidson, W., Jonas, A., Clayton, D., and George, J. (1998). Stabilization of alpha-synuclein secondary structure upon binding to synthetic membranes. J. Biol. Chem. 273, 9443–9449. Davydov, D., and Halpert, J. (2008). Allosteric P450 mechanisms: Multiple binding sites, multiple conformers or both? Expert Opin. Drug Met. 4, 1523–1535. Davydov, D., Fernando, H., Baas, B., Sligar, S., and Halpert, J. (2005). Kinetics of dithionite-dependent reduction of cytochrome P450 3A4: Heterogeneity of the enzyme caused by its oligomerization. Biochemistry 44, 13902–13913. Davydov, D., Baas, B., Sligar, S., and Halpert, J. (2007). Allosteric mechanisms in cytochrome P450 3A4 studied by high-pressure spectroscopy: Pivotal role of substrateinduced changes in the accessibility and degree of hydration of the heme pocket. Biochemistry 46, 7852–7864. Denisov, I., Grinkova, Y., Lazarides, A., and Sligar, S. (2004). Directed self-assembly of monodisperse phospholipid bilayer nanodiscs with controlled size. J. Am. Chem. Soc. 126, 3477–3487. Denisov, I., McLean, M., Shaw, A., Grinkova, Y., and Sligar, S. (2005). Thermotropic phase transition in soluble nanoscale lipid bilayers. J. Phys. Chem. B 109, 15580–15588. Denisov, I., Baas, B., Grinkova, Y., and Sligar, S. (2007). Cooperativity in cytochrome P450 3A4: Linkages in substrate binding, spin state, uncoupling, and product formation. J. Biol. Chem. 282, 7066–7076. Denisov, I., Frank, D., and Sligar, S. (2009). Cooperative properties of cytochromes P450. Pharmacol. Therapeut. 124, 151–167. Diez, M., Zimmermann, B., Borsch, M., Konig, M., Schweinberger, E., Steigmiller, S., Reuter, R., Felekyan, S., Kudryavtsev, V., Seidel, C., and Graber, P. (2004). Protonpowered subunit rotation in single membrane-bound F0F1-ATP synthase. Nat. Struct. Mol. Biol. 11, 135–141. Drescher, M., Veldhuis, G., van Rooijen, B., Milikisyants, S., Subramaniam, V., and Huber, M. (2008). Antiparallel arrangement of the helices of vesicle-bound alphasynuclein. J. Am. Chem. Soc. 130, 7796–7797.

114

Abhinav Nath et al.

Fernando, H., Davydov, D., Chin, C., and Halpert, J. (2007). Role of subunit interactions in P450 oligomers in the loss of homotropic cooperativity in the cytochrome P450 3A4 mutant L211F/D214E/F304W. Arch. Biochem. Biophys. 460, 129–140. Ferreon, A., Gambin, Y., Lemke, E., and Deniz, A. (2009). Interplay of alpha-synuclein binding and conformational switching probed by single-molecule fluorescence. Proc. Natl. Acad. Sci. USA 106, 5645–5650. Gao, F., Mei, E., Lim, M., and Hochstrasser, R. (2006). Probing lipid vesicles by bimolecular association and dissociation trajectories of single molecules. J. Am. Chem. Soc. 128, 4814–4822. Garcia-Saez, A., and Schwille, P. (2007). Single molecule techniques for the study of membrane proteins. Appl. Microbiol. Biotechnol. 76, 257–266. Gell, C., Brockwell, D., and Smith, A. (2006). Handbook of Single Molecule Fluorescence Spectroscopy. Oxford University Press, Oxford, NY. George, J., Jin, H., Woods, W., and Clayton, D. (1995). Characterization of a novel protein regulated during the critical period for song learning in the zebra finch. Neuron 15, 361–372. Georgieva, E., Ramlall, T., Borbat, P., Freed, J., and Eliezer, D. (2008). Membrane-bound alpha-synuclein forms an extended helix: Long-distance pulsed ESR measurements using vesicles, bicelles, and rodlike micelles. J. Am. Chem. Soc. 130, 12856–12857. Goedert, M. (2001). Alpha-synuclein and neurodegenerative diseases. Nat. Rev. Neurosci. 2, 492–501. Grinkova, Y., Denisov, I., Bayburt, T., and Sligar, S. (2004). Structure and composition of self-assembled discoidal phospholipid bilayer nanoparticles formed with different membrane scaffold proteins. Biophys. J. 86, 252A. Guengerich, F. (2001). Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem. Res. Toxicol. 14, 611–650. Hebda, J., and Miranker, A. (2009). The interplay of catalysis and toxicity by amyloid intermediates on lipid bilayers: Insights from type II diabetes. Annu. Rev. Biophys. 38, 125–152. Hendler, R., and Shrager, R. (1994). Deconvolutions based on singular-value decomposition and the pseudoinverse—A guide for beginners. J. Biochem. Biophys. Methods 28, 1–33. Henzler-Wildman, K., Thai, V., Lei, M., Ott, M., Wolf-Watz, M., Fenn, T., Pozharski, E., Wilson, M., Petsko, G., Karplus, M., Hubner, C., and Kern, D. (2007). Intrinsic motions along an enzymatic reaction trajectory. Nature 450, 838–844. Isin, E., and Guengerich, F. (2007). Multiple sequential steps involved in the binding of inhibitors to cytochrome p450 3A4. J. Biol. Chem. 282, 6863–6874. Jao, C., Der-Sarkissian, A., Chen, J., and Langen, R. (2004). Structure of membrane-bound alpha-synuclein studied by site-directed spin labeling. Proc. Natl. Acad. Sci. USA 101, 8331–8336. Jao, C., Hegde, B., Chen, J., Haworth, I., and Langen, R. (2008). Structure of membranebound alpha-synuclein from site-directed spin labeling and computational refinement. Proc. Natl. Acad. Sci. USA 105, 19666–19671. Jo, E., McLaurin, J., Yip, C., St George-Hyslop, P., and Fraser, P. (2000). Alpha-synuclein membrane interactions and lipid specificity. J. Biol. Chem. 275, 34328–34334. Jones, E., Oliphant, T., and Peterson, P. (2001). SciPy: Open Source Scientific Tools for Python. http://scipy.org. Kijac, A., Li, Y., Sligar, S., and Rienstra, C. (2007). Magic-angle spinning solid-state NMR spectroscopy of nanodisc-embedded human CYP3A4. Biochemistry 46, 13696–13703. Knight, J., and Falke, J. (2007). Single-molecule fluorescence studies of a PH domain: New insights into the membrane docking reaction. Biophys. J. 96, 566–582. Knight, J., and Miranker, A. (2004). Phospholipid catalysis of diabetic amyloid assembly. J. Mol. Biol. 341, 1175–1187.

Single-Molecule Spectroscopy with Nanodiscs

115

Knight, J., Hebda, J., and Miranker, A. (2006). Conserved and cooperative assembly of membrane-bound alpha-helical states of islet amyloid polypeptide. Biochemistry 45, 9496–9508. Koshland, D., Nemethy, G., and Filmer, D. (1966). Comparison of experimental binding data and theoretical models in proteins containing subunits. Biochemistry 5, 365–385. Kozuka, J., Yokota, H., Arai, Y., Ishii, Y., and Yanagida, T. (2006). Dynamic polymorphism of single actin molecules in the actin filament. Nat. Chem. Biol. 2, 83–86. Kuszak, A., Pitchiaya, S., Anand, J., Mosberg, H., Walter, N., and Sunahara, R. (2009). Purification and functional reconstitution of monomeric mu-opioid receptors: Allosteric modulation of agonist binding by Gi2. J. Biol. Chem. 284, 26732–26741. Lakowicz, J. R. (2006). Principles of fluorescence spectroscopy. Springer, New York; Berlin. Lampe, J., and Atkins, W. (2006). Time-resolved fluorescence studies of heterotropic ligand binding to cytochrome P450 3A4. Biochemistry 45, 12204–12215. Lampe, J., Fernandez, C., Nath, A., and Atkins, W. (2008). Nile Red is a fluorescent allosteric substrate of cytochrome P450 3A4. Biochemistry 47, 509–516. Leitz, A., Bayburt, T., Barnakov, A., Springer, B., and Sligar, S. (2006). Functional reconstitution of Beta2-adrenergic receptors utilizing self-assembling Nanodisc technology. Biotechniques 40, 601–612. Ling, Y., Strasfeld, D., Shim, S., Raleigh, D., and Zanni, M. (2009). Two-dimensional infrared spectroscopy provides evidence of an intermediate in the membrane-catalyzed assembly of diabetic amyloid. J. Phys. Chem. B 113, 2498–2505. Lotharius, J., and Brundin, P. (2002). Impaired dopamine storage resulting from alphasynuclein mutations may contribute to the pathogenesis of Parkinson’s disease. Hum. Mol. Genet. 11, 2395–2407. Lu, H., Xun, L., and Xie, X. (1998). Single-molecule enzymatic dynamics. Science 282, 1877–1882. Magde, D., Webb, W., and Elson, E. (1972). Thermodynamic fluctuations in a reacting system—Measurement by fluorescence correlation spectroscopy. Phys. Rev. Lett. 29, 705–708. Mishra, R., Sellin, D., Radovan, D., Gohlke, A., and Winter, R. (2009). Inhibiting islet amyloid polypeptide fibril formation by the red wine compound resveratrol. ChemBioChem 10, 445–449. Monod, J., Wyman, J., and Changeux, J. (1965). On nature of allosteric transitions— A plausible model. J. Mol. Biol. 12, 88–118. Morrissey, J., Pureza, V., Davis-Harrison, R., Sligar, S., Ohkubo, Y., and Tajkhorshid, E. (2008). Blood clotting reactions on nanoscale phospholipid bilayers. Thromb. Res. 122(Suppl 1), S23–S26. Murakoshi, H., Iino, R., Kobayashi, T., Fujiwara, T., Ohshima, C., Yoshimura, A., and Kusumi, A. (2004). Single-molecule imaging analysis of Ras activation in living cells. Proc. Natl. Acad. Sci. USA 101, 7317–7322. Nanga, R., Brender, J., Xu, J., Veglia, G., and Ramamoorthy, A. (2008). Structures of rat and human islet amyloid polypeptide IAPP(1-19) in micelles by NMR spectroscopy. Biochemistry 47, 12689–12697. Nath, A., Atkins, W., and Sligar, S. (2007a). Applications of phospholipid bilayer nanodiscs in the study of membranes and membrane proteins. Biochemistry 46, 2059–2069. Nath, A., Grinkova, Y., Sligar, S., and Atkins, W. (2007b). Ligand binding to cytochrome P450 3A4 in phospholipid bilayer nanodiscs: The effect of model membranes. J. Biol. Chem. 282, 28309–28320. Nath, A., Ferna´ndez, C., Lampe, J., and Atkins, W. (2008a). Spectral resolution of a second binding site for Nile Red on cytochrome P4503A4. Arch. Biochem. Biophys. 474, 198–204.

116

Abhinav Nath et al.

Nath, A., Koo, P., Rhoades, E., and Atkins, W. (2008b). Allosteric effects on substrate dissociation from cytochrome P450 3A4 in Nanodiscs observed by ensemble and singlemolecule fluorescence spectroscopy. J. Am. Chem. Soc. 130, 15746–15747. Ohashi, R., Mu, H., Wang, X., Yao, Q., and Chen, C. (2005). Reverse cholesterol transport and cholesterol efflux in atherosclerosis. QJM—Int. J. Med. 98, 845–856. Park, H., Toprak, E., and Selvin, P. (2007). Single-molecule fluorescence to study molecular motors. Q. Rev. Biophys. 40, 87–111. Peterman, E., Sosa, H., and Moerner, W. (2004). Single-molecule fluorescence spectroscopy and microscopy of biomolecular motors. Annu. Rev. Phys. Chem. 55, 79–96. Protter, M., and Elad, M. (2009). Image sequence denoising via sparse and redundant representations. IEEE Trans. Image Proc. 18, 27–35. Rhoades, E., Gussakovsky, E., and Haran, G. (2003). Watching proteins fold one molecule at a time. Proc. Natl. Acad. Sci. USA 100, 3197–3202. Rhoades, E., Ramlall, T., Webb, W., and Eliezer, D. (2006). Quantification of alphasynuclein binding to lipid vesicles using fluorescence correlation spectroscopy. Biophys. J. 90, 4692–4700. Ritchie, T., Grinkova, Y., Bayburt, T., Denisov, I., Zolnerciks, J., Atkins, W., and Sligar, S. (2009). Chapter 11—Reconstitution of membrane proteins in phospholipid bilayer nanodiscs. Methods Enzymol. 464, 211–231. Roberts, A., and Atkins, W. (2007). Energetics of heterotropic cooperativity between alphanaphthoflavone and testosterone binding to CYP3A4. Arch. Biochem. Biophys. 463, 89–101. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Sanchez, S., Tricerri, M., and Gratton, E. (2007). Interaction of high density lipoprotein particles with membranes containing cholesterol. J. Lipid Res. 48, 1689–1700. Segrest, J., Jones, M., Klon, A., Sheldahl, C., Hellinger, M., De Loof, H., and Harvey, S. (1999). A detailed molecular belt model for apolipoprotein A-I in discoidal high density lipoprotein. J. Biol. Chem. 274, 31755–31758. Shaw, A., Pureza, V., Sligar, S., and Morrissey, J. (2007). The local phospholipid environment modulates the activation of blood clotting. J. Biol. Chem. 282, 6556–6563. Subramanian, M., Low, M., Locuson, C., and Tracy, T. (2009). CYP2D6-CYP2C9 protein-protein interactions and isoform-selective effects on substrate binding and catalysis. Drug Metab. Dispos. 10.1124/dmd.109.026500 (in press). Thummel, K., and Wilkinson, G. (1998). In vitro and in vivo drug interactions involving human CYP3A. Annu. Rev. Pharmacol. 38, 389–430. Torres, T., and Levitus, M. (2007). Measuring conformational dynamics: A new FCS-FRET approach. J. Phys. Chem. B 111, 7392–7400. Trexler, A., and Rhoades, E. (2009). Alpha-synuclein binds large unilamellar vesicles as an extended helix. Biochemistry 48, 2304–2306. Tsalkova, T., Davydova, N., Halpert, J., and Davydov, D. (2007). Mechanism of interactions of alpha-naphthoflavone with cytochrome p450 3A4 explored with an engineered enzyme bearing a fluorescent probe. Biochemistry 46, 106–119. Ueda, K., Fukushima, H., Masliah, E., Xia, Y., Iwai, A., Yoshimoto, M., Otero, D., Kondo, J., Ihara, Y., and Saitoh, T. (1993). Molecular-cloning of cDNA-encoding an unrecognized component of amyloid in Alzheimer-disease. Proc. Natl. Acad. Sci. USA 90, 11282–11286. Ulbrich, M., and Isacoff, E. (2007). Subunit counting in membrane-bound proteins. Nat. Methods 4, 319–321. Ulmer, T., Bax, A., Cole, N., and Nussbaum, R. (2005). Structure and dynamics of micellebound human alpha-synuclein. J. Biol. Chem. 280, 9595–9603.

Single-Molecule Spectroscopy with Nanodiscs

117

Vogelsang, J., Cordes, T., Forthmann, C., Steinhauer, C., and Tinnefeld, P. (2009). Controlling the fluorescence of ordinary oxazine dyes for single-molecule switching and superresolution microscopy. Proc. Natl. Acad. Sci. USA 106, 8107–8112. Weinreb, P., Zhen, W., Poon, A., Conway, K., and Lansbury, P. (1996). NACP, a protein implicated in Alzheimer’s disease and learning, is natively unfolded. Biochemistry 35, 13709–13715. Whorton, M., Bokoch, M., Rasmussen, S., Huang, B., Zare, R., Kobilka, B., and Sunahara, R. (2007). A monomeric G protein-coupled receptor isolated in a highdensity lipoprotein particle efficiently activates its G protein. Proc. Natl. Acad. Sci. USA 104, 7682–7687. Williams, P., Cosme, J., Vinkovic, D., Ward, A., Angove, H., Day, P., Vonrhein, C., Tickle, I., and Jhoti, H. (2004). Crystal structures of human cytochrome P450 3A4 bound to metyrapone and progesterone. Science 305, 683–686. Zeldin, R., and Petruschke, R. (2004). Pharmacological and therapeutic properties of ritonavir-boosted protease inhibitor therapy in HIV-infected patients. J. Antimicrob. Chemoth. 53, 4–9.

C H A P T E R

S E V E N

Single-Molecule Spectroscopy Using Microfluidic Platforms Samuel Kim1 and Richard N. Zare Contents 120 121 121 122 122 123 124 125 125 127 131 131

1. Introduction 2. Microchip Fabrication 2.1. Design drawing and photomask printing 2.2. Molding master fabrication 2.3. Fabrication of PDMS chip 3. Instrumentation for Fluorescence Detection 4. Detergent-Assisted Microchannel Electrophoresis 4.1. Preparation of separation buffer and sample solutions 4.2. Microchip electrophoresis with electrokinetic injection 5. Fluorescence Correlation Spectroscopy Acknowledgments References

Abstract Microfluidics serves as a convenient platform for single-molecule experiments by providing manipulation of small amounts of liquids and micron-sized particles. An adapted version of capillary electrophoresis (CE) on a microchip can be utilized to separate chemical species with high resolution based on their ionic mobilities (i.e., charges and sizes), but identification of separated species is not trivial, especially for complex mixtures of sticky biomolecules. We describe here how to use a surfactant mixture system for CE on a poly(dimethylsiloxane) (PDMS) microchip, capture separated peaks within a 50-pl chamber using microvalves, analyze the fluorescence signals with correlation spectroscopy to extract molecular diffusion characteristics, and to identify the biomolecular clusters in a model immunocomplex system.

Department of Chemistry, Stanford University, Stanford, California, USA Current address: Polymer Research Institute and National Core Research Center for Systems Bio-Dynamics, Pohang University of Science and Technology, Pohang, Kyungbuk, South Korea

1

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72013-9

#

2010 Elsevier Inc. All rights reserved.

119

120

Samuel Kim and Richard N. Zare

1. Introduction Microfluidics refers to the study and control of the fluidic behavior of a liquid within structures of micrometer dimensions. Because of its ability to handle extremely small amounts of samples and microscale objects such as cells, the use of the technique in various types of single-molecule experiments has rapidly increased. For instance, microfluidic channels were used to prepare compartmentalized neuron cell cultures for single-particle tracking of nerve growth factor molecules (Cui et al., 2007). Single-cell expression assay of b-galactosidase in live bacteria cells was achieved by confining fluorescent molecules produced by enzymatic activity within 100-pl chambers (Cai et al., 2006). A glass microfluidic cell was an essential component of a new type of single-molecule trapping system where the Brownian motion of an individual fluorescent molecule is cancelled electrokinetically (Cohen and Moerner, 2006). Measurement of protein folding kinetics based on single-molecule FRET efficiency was performed using a microfabricated rapid mixer (Lipman et al., 2003). Electrophoretic separation of biomolecules by their charges and sizes can be integrated into a microfluidic device by implementing structures for injection plug formation and capillary-like microchannels (Wu et al., 2004). Although microchip electrophoresis is a highly efficient separation technique, the conventional detection methods, which are based on optical or electrical signals only, do not provide a means to ‘‘identify’’ separated chemical species unless known standard samples are used for spiking. In the case of heterogeneous biological systems in dynamic equilibrium, where molecules bind weakly with each other, it is almost impossible to prepare a pure sample for identification purposes. We overcome this limitation by employing fluorescence fluctuation spectroscopy to extract molecular parameters of separated species. We regard the ability to characterize individual molecular behavior as one of the key advantages of single-molecule spectroscopy as opposed to the study of ensemble averages. More specifically, we combine a microfluidic separation based on a twocomponent detergent mixture containing charged micelles with correlation analysis of laser-induced fluorescence signals as a detection modality. In this chapter, optimized protocols for the application of this technique to a model immunocomplex system and the experimental results are described. We suggest that this approach is a general one and can be used in many other situations to advantage. It is necessary to mention that a similar but distinct approach using fluorescence correlation spectroscopy (FCS) for biomolecular analysis in continuous flow capillary electrophoresis (CE) system has been developed (Fogarty and Van Orden, 2009; Van Orden and Keller, 1998). Also, the use of photon counting histograms to study a protein charge ladder caused by different numbers of incorporated fluorescent probes has been reported previously (Kim et al., 2007).

121

Single-Molecule Spectroscopy on a Microchip

2. Microchip Fabrication A microfluidic device for simple CE with a capture valve can be constructed using a multilayer soft lithography technique (Anderson et al., 2000; Unger et al., 2000). The fabrication procedures are divided into three steps: design drawing and mask printing, making a photoresist master on a silicon wafer, and making poly(dimethylsiloxane) (PDMS) channel structures using a master molding technique. In the following paragraphs, an optimized protocol for the fabrication of a CE chip is described.

2.1. Design drawing and photomask printing Each layer of the microfluidic device is drawn electronically using vector graphics software. Macromedia FreeHandÒ 10 was used for the work presented, but more advanced versions of computer programs such as Adobe IllustratorÒ or Audodesk AutoCADÒ can be used as well. One microchip is designed to fit into a 24 mm
60 mm glass coverslip, as shown in Fig. 7.1. The standard 4-in. wafer can accommodate three such CE chips. At the final stage of design drawing, the objects that will be made with ‘‘negative’’ photoresists in the photolithography step should be converted into their ‘‘negative’’ images. In Fig. 7.1, the valve layer (shown in gray), which will be made with the SU-8 negative photoresist, requires such conversion.

V3 V2

b

a

V4 P1

V1 A

10 mm B

C 35 mm 20 mm 10 mm

200 mm

Figure 7.1 Drawing of a CE microchip with a capture valve: (A) the ‘‘double-T’’ region for injection plug formation; (B) region for capturing separated chemical species, where the distance from the double-T region to the capture region is 25 mm; and (C) the side-view of the capture region. There is a 20-mm thick PDMS membrane between the channel layer (black) and the valve layer (gray), which can be deformed pneumatically. V1–V4 stand for access ports for the channel layer where high-voltage electrodes are inserted. P1 is the access port for the valve layer where pressure for valve actuation is applied.

122

Samuel Kim and Richard N. Zare

The finished design is printed on a transparency film, using a highresolution photoplotter (32,512 DPI; Fineline Imaging, Colorado Springs, CO), which can achieve about 10-mm resolution. In case a photomask with a higher resolution is necessary, a chrome mask can be prepared at a higher cost either from an academic nanofabrication facility or from a commercial vendor equipped with a mask writer. It is capable of providing a resolution of features of about 1 mm if created with laser writing.

2.2. Molding master fabrication A polymeric layer of well-defined thickness and shape can be created on a silicon wafer, using photolithography. The standard photolithography procedure consists of three steps: spin-coating, exposure, and developing. In spin-coating, the viscosity of the photoresist and the spin speed are adjusted to control the thickness of the structure. In exposure, the photoresist layer is exposed to ultraviolet (UV) light that is patterned by the photomask. Photoresists can be categorized into two groups according to their responses to this UV irradiation. A positive photoresist becomes soluble upon exposure, whereas a negative photoresist becomes insoluble upon exposure. In developing, the soluble portion of the photoresist is dissolved by a solvent called ‘‘developer,’’ leaving the desired microstructure on the wafer surface. The protocol described here is optimized for creating a rounded channel structure, using SPR 220-7 positive photoresist (Shipley). 1. SPR 220-7 is spin-coated (3500 rpm, 40 s) on a 4-in. silicon wafer and prebaked at 90  C for 200 s to form a 7-mm thick layer. 2. The photoresist layer is exposed to UV light (365 nm) for 15 s, using the exposure instrument protocol (Karl Suss, Waterbury Center, VT). No postexposure bake is necessary. 3. The exposed layer is developed for 270 s with LDD26W developer (Rohm and Haas, Marlborough, MA). 4. The wafer is postbaked at 120  C for 10 min to reflow and transform the photoresist structure into a round shape. This step is critical to complete valve closure (Melin and Quake, 2007). 5. The photoresist master is inspected on a microscope to check its quality. Finally, the master mold is coated with perfluoro-1,1,2,2-tetrahydrooctyltrichlorosilane (United Chemical Technologies, Bristol, PA) by exposing it to vapor in a vacuum desiccator to prevent adhesion of PDMS onto the master surface. Specifically, one drop of silane liquid is placed in a separate container inside the desiccator together with the silicon wafer.

2.3. Fabrication of PDMS chip We use the following protocol to create multilayer channel structures in a PDMS device:

Single-Molecule Spectroscopy on a Microchip

123

1. PDMS prepolymers (RTV 615A and 615B; GE Silicones, Waterford, NY) are mixed in a 5:1 mass ratio, poured onto the master for valve layer, degassed under vacuum, and cured at 80  C for 1–2 h. 2. The PDMS slab is cut out with a scalpel, and holes for fluid access are punched with a syringe needle ground flat at the tip. Finished pieces are scotch-taped on both sides to prevent contamination. 3. PDMS prepolymers, mixed in a 20:1 mass ratio, is spin-coated on the master for making the channel layer, with the initial spin speed of 300 rpm for 9 s and the final spin speed of 3000 rpm for 45 s. The PDMS-coated wafer is placed on a flat surface at room temperature for 5 min to let the coating flow and flatten further. Then, the PDMS coating is cured at 80  C for 20 min. 4. The 5:1 PDMS slab prepared in Step 2 is placed on the 20:1 PDMScoated wafer and aligned by inspecting through a stereoscope. This alignment procedure can be repeated many times to achieve perfect spatial match because both pieces are fully cured. After alignment, gentle pressure is applied to seal completely the edges of the bonded pieces. 5. A 10:1 PDMS mix is poured around the aligned slabs, degassed, and cured at 80  C for 1 h. 6. Chips are cut out with a scalpel, and access holes are punched in the same way as described in Step 2. 7. Glass coverslips (No. 1 1/2; VWR International, West Chester, PA) are spin-coated with a diluted 5:1 PDMS mix at an initial spin speed of 1000 rpm for 9 s and a final spin speed of 2000 rpm for 30 s. The diluted PDMS mix is prepared by adding 20 g of cyclohexane (179191; SigmaAldrich) to 10 g of the 5:1 PDMS mix. The coated coverslips are cured at 80  C for 20 min. 8. The chips prepared in Step 6 are placed on the spin-coated coverslips and sealed with gentle pressure. The chips are placed in an 80  C oven overnight for final bonding. 9. The chips are inspected on a microscope for defects and tested for valve functions (complete valve closure at a designated pressure value). These chips can be stored for an extended period of time under a dry condition, at least for 3 months from our experience. When exposed to buffers and/or detergents, the surface property of PDMS usually changes and degrades over time. The quality of a microchip CE deteriorates after continuous usage of the chip for 8–10 h.

3. Instrumentation for Fluorescence Detection The laser-induced fluorescence detection is achieved by using the following optical setup. A 532 nm laser beam (Compass 215M; Coherent Inc., Santa Clara, CA) or a 638 nm laser beam (RCL-025-638; Crystalaser,

124

Samuel Kim and Richard N. Zare

Reno, NV), guided by a single-mode optical fiber (OZ Optics, Canada) and collimated with a 5
objective lens (Zeiss), is sent to an inverted microscope (TE300; Nikon, Melville, NY) as an excitation source for epi-illumination of the sample. The laser beam is reflected by a dichroic mirror (540DRLP or 400535-635TBDR; Omega Optical, Brattleboro, VT) and focused by a high numerical aperture (NA) objective (Plan Apo, 60
, NA 1.20; Nikon). The resulting fluorescence photons are collected by the same objective lens, passed through the same dichroic mirror, a 50-mm pinhole, and a band-pass filter (595AF60, Omega Optical; HQ675/50m, Chroma Technology), and detected by an avalanche photodiode (SPCM AQR15; EG&G, Canada) (Fig. 7.2). The fluorescence signals are recorded with a counter/timer data acquisition card (PCI-6602; National Instruments, Austin, TX) and displayed as a signal–time graph on a computer screen, using a LabViewÒ program. This real-time display of fluorescence intensity can be used for fine tuning the alignment of the optical components by adjusting them to maximize the fluorescence signal. Autocorrelation functions of the fluorescence signals are calculated by transferring the signals to a hardware digital correlator (Flex99R480; Correlator.com, Bridgewater, NJ). The obtained correlation curves are fitted with IGOR Pro software (WaveMetrics, Inc., Portland, OR).

4. Detergent-Assisted Microchannel Electrophoresis Capillary electrophoresis is a powerful separation technique utilizing the differences in ionic mobilities through a capillary under the action of an applied electric field ( Jorgenson and Lukacs, 1983; St. Claire, 1996). Microvalve

V3 V2 V1

V4

25 mm

Objective Optical fiber Dichroic mirror

Pinhole

Laser 532/638 nm

APD

Band-pass filter

Lenses

Figure 7.2 Optical setup for laser-induced fluorescence detection: APD, avalanche photodiode; Vi, the voltage applied to electrode i. Reproduced from Kim et al. (2007) by permission of The Royal Society of Chemistry.

Single-Molecule Spectroscopy on a Microchip

125

A similar structure having separation resolution can be easily integrated into a microfluidic device, enabling electrophoresis experiment on a PDMS microchip. The native PDMS surface, however, is not suitable for electrophoretic separation of biomolecules because (1) the electroosmotic flow is not stable owing to uneven surface charge density and (2) the analytes can be easily adsorbed onto the hydrophobic PDMS surface, contributing to significant loss of analytes and the tailing of peaks during electrophoresis. One remedy to this problem is to passivate the PDMS surface with detergent molecules. We developed a dynamic coating method using a mixture of a nonionic detergent (dodecyl maltoside, DDM) and an anionic detergent (sodium dodecyl sulfate, SDS) (Huang et al., 2007). This procedure allows not only nearly complete rejection of adsorption of organic dye and protein molecules but also control of both electroosmotic flow and separation efficiency by varying the relative concentration of SDS, creating an environment similar to micellar electrokinetic chromatography (MEKC), where the presence of micelles promotes further separation of analytes (Terabe et al., 1984). This surfactant system has been successfully applied to the separation of simple organic dyes, protein charge ladders, phycobiliproteins, and immunocomplexes. Detailed procedures for performing microchip CE experiments to separate the immunocomplexes between dye-labeled bovine serum albumin (BSA) and monoclonal antibody (mAb) are described below.

4.1. Preparation of separation buffer and sample solutions The separation buffer contains 20 mM HEPES, pH 7.5, 0.1% (w/v) n-dodecylb-D-maltoside (Anatrace, Inc., Maumee, OH), and 0.01% (w/v) SDS (SigmaAldrich). For simple dye mixtures, 100 mM stock solutions of Alexa Fluor 647 succinimidyl ester (A20006; Invitrogen) and Cy5 succinimidyl ester (GE Healthcare, Piscataway, NJ) are prepared in the separation buffer and diluted to obtain desired concentrations. For immunocomplexes, 0.09 mg/ml BSA tetramethylrhodamine conjugate (A23016; Invitrogen) is prepared using the separation buffer; varying amounts (1.4, 0.20, and 0.028 mg/ml) of monoclonal anti-BSA antibody (B2901; Sigma) are added and incubated at least for 1 h. The immunocomplex solutions are then diluted to achieve concentrations in a nanomolar range after electrophoresis and capture, which enables FCS measurement with a good signal-to-noise ratio and resolvability. The optimal composition of DDM and SDS in the separation buffer is sample-dependent. Therefore, a calibration experiment at an early stage will be necessary for different types of analytes.

4.2. Microchip electrophoresis with electrokinetic injection 1. One microliter of 1% (w/v) SDS solution is added to the bottom of each of the access ports, V1–V3, using a gel-loading pipet tip. The microchannels should be spontaneously filled with the solution via capillary

126

Samuel Kim and Richard N. Zare

action. Then, the SDS solution is removed, which will not affect the capillary action because the volume necessary to fill the entire microchannel is negligibly small (22 nl), and 200 ml of the separation buffer is added to each of the access ports, V1–V3. After it is visually confirmed with a microscope that the liquid front reaches the access port V4, 200 ml of the separation buffer is added to V4. Whenever solution is added to the access ports, the formation of air bubbles at the bottom of the reservoirs should be carefully avoided to maintain electrical connection. 2. Voltages are applied (V1–V3 ¼ 1 kV, V4 ¼ 0 V) and the through-thechannel current is measured. The current will gradually decrease and stabilize at a value between 0.5 and 1 mA as 1% (w/v) SDS solution is replaced electrokinetically by the separation buffer. The electrical connection for all the three input ports (V1–V3) should be tested individually by monitoring electrical current; for example, the voltage setting V1 ¼ 1 kV, V2–V4 ¼ 0 V can be used to check the V1 port. If an input port is found to have ‘‘zero’’ current, the air bubble at the bottom of that port should be removed by suction with a gel-loading pipet tip. 3. The objective lens is moved to the detection point, which is 25 mm away from the double-T region. The excitation laser is focused at the half height of the channel by monitoring the shape of the reflected beam. At this point, the fluorescence intensity obtained from the avalanche photodiode should be close to the background level, which is usually below 3000 photons/s (i.e., 3 kHz as in Fig. 7.3) in our optical setup. 4. The sample solution to be separated is added to V1. Voltages for electrokinetic injection (V1 ¼ 1 kV, V2 ¼ 700 V, V3 ¼ 0 V, V4 ¼ 700 V) are applied for 17 s. Then, voltages for electrophoretic separation (V1 ¼ 700 V, V2 ¼ 1 kV, V3 ¼ 700 V, V4 ¼ 0 V) are applied and B

A A647

400

Fluorescence (kHz)

Fluorescence (kHz)

Cy5 300 200 100 0 10

20

30 40 Time (s)

50

400 300 200 100 0 10

20

30 40 Time (s)

50

Figure 7.3 Separation and capture of a dye mixture: (A) electropherogram of the mixture of Cy5 and Alexa Fluor 647 (A647); (B) fluorescence signal after capturing the A647 peak. Once captured, the intensity drops because of diffusive mixing. Fluctuations caused by molecular diffusion can be observed. Reproduced from Kim et al. (2007) by permission of The Royal Society of Chemistry.

Single-Molecule Spectroscopy on a Microchip

127

the fluorescence signal acquisition is initiated simultaneously. Synchronized operation of the high-voltage power supply and the photodetector is critical to obtaining reproducible results. This process should be repeated for each sample at least three times to ensure that no systematic error in migration time is observed. It should be noted that the separation result from the first run with a new sample added to the chip is usually not reliable because of disturbed injection profiles. 5. The migration times of separated peaks, which are determined in the previous step, can be used to capture a peak of interest. The same voltage settings are used for injection and separation except that the separation voltages are applied only for a specified period of time (i.e., the migration time of the peak to be captured). Immediately after turning off separation voltages, pressure is applied to the P1 port to actuate the capture valves. The duration for applying separation voltages needs to be optimized depending on how fast the microvalves respond. When this scheme is used, a timing accuracy of 100 ms or better is sufficient for successful capture.

5. Fluorescence Correlation Spectroscopy The separated peaks can be captured with microvalves and analyzed further by measuring molecular parameters based on fluorescence fluctuation signals. Here, we demonstrate the use of FCS for obtaining the diffusion coefficients of the immunocomplex species to determine the identities of the peaks. In FCS experiments, fluctuations of the fluorescence intensity from the molecules that are passing through the focus of a laser beam are analyzed by calculating the autocorrelation function of the signal. The autocorrelation function G(t), which contains information on the physical processes responsible for the fluctuations, is defined by GðtÞ ¼

hdIðtÞ  dIðt þ tÞi hIðtÞi2

ð7:1Þ

where I(t) is the fluorescence intensity at a time point t and dIðtÞ ¼ IðtÞ  hIðtÞi, that is, the deviation of the signal from the ensemble average value hIðtÞi. The main contributor to this fluctuation is molecular diffusion. When the molecule or the particle of interest is assumed to be freely diffusing and have constant fluorescence intensity, the autocorrelation function of that fluorescent species takes the form

128

Samuel Kim and Richard N. Zare

GðtÞ ¼

    1 t 1 t 1=2 1þ 2 1þ  N tD k tD

ð7:2Þ

 is the average number of molecules in the observation volume, tD where N is the diffusion time, and k is the geometry factor describing the shape of the  and tD can be obtained by fitting the experiobservation volume. Both N mentally determined autocorrelation function with this equation. The size of diffusing molecules can be estimated from the Stokes– Einstein relation: D¼

kB T 6pr

ð7:3Þ

where D is the diffusion coefficient, kB is the Boltzmann constant, T is the absolute temperature,  is the viscosity of the medium, and r is the hydrodynamic radius of a spherical particle. This equation can be converted into a simple relation between r and tD in the following way. The focal spot size (o0) is determined by obtaining FCS curves from dilution series of standard Alexa Fluor 594 solutions (Perroud et al., 2005).  versus The effective focal volume, Veff, which is calculated from a plot of N 3=2 3 concentration, is related to the spot size as Veff ¼ p kw0 . After replacing D with an expression containing tD (w02 ¼ 4DtD , where o0¼ 0.32  0.05 mm) and inserting numerical values for constants (kB ¼ 1.38
10 23 J/(mol K)1; T ¼ 298.15 K;  ¼ 9.46
10 4 Pa s), a simple relation is obtained: r ¼ 9:46
106 tD

ð7:4Þ

Finally, assuming spherical shapes for immunocomplexes, the molecular weight of a protein complex is found to be proportional to the cube of the diffusion time obtained from FCS measurements. For a mixture of Cy5 and A647 fluorophores (Fig. 7.3), the FCS measurements performed on a microchip after CE and capture are compared with those obtained individually using a standard solution chamber. As summarized in Table 7.1, the diffusion times measured within a 50-pl capture region are consistent with the results obtained in a 50-ml sample chamber. It needs to be noted that resolving Cy5 and A647 purely based on fitting with a multiple-species diffusion model will be more difficult and less accurate because the diffusion times of these fluorophores are very similar. Figure 7.4 shows the separation results of the immunocomplexes between bovine serum albumin labeled with tetramethylrhodamine (TMR-BSA) and monoclonal anti-BSA antibody (mAb), using the DDM/SDS micellar system on a standard double-T CE chip. The observed numbers of separated peaks and the changes in their relative abundances are in agreement with the previous experimental results obtained with an

129

Single-Molecule Spectroscopy on a Microchip

acid-modified CE microchip (Luo et al., 2006). The dependence of relative peak heights on the mAb concentration suggests that different peaks correspond to protein complexes formed at different stoichiometries. Table 7.1 FCS parameters obtained for simple dye solutions

tD (ms)  N

Individual dye in a glass-bottom chamber

1:1 mixture separated in a microchip

Cy5

A647

Cy5

A647

128 ( 4) 5.16 ( 0.06)

137 (3) 5.62 ( 0.03)

125 ( 3) 3.7 ( 0.4)

138 ( 2) 3.7 ( 0.3)

Fluorescence intensity (kHz)

A

C 1400

1400

1200

1200

1000 800

800

600

600

2 3

400

400 3

200

1

200 0

0 20

30

40

50

60

70

20

80

B

30

40

50

60

70

80

D 1400

Fluorescence intensity (kHz)

4

1000

1

1400 1

1200

4

1200

1000

1000

800

800

600

600 2

400

3

400

3

200

1 2

200

0

0 20

30

40

50 Time (s)

60

70

80

20

30

40

50

60

70

80

Time (s)

Figure 7.4 The electropherograms for the immunocomplexes between TMR-labeled BSA and anti-BSA antibody. A solution containing 0.09 mg/ml TMR-BSA was supplemented with (A) 0 mg/ml, (B) 0.028 mg/ml, (C) 0.20 mg/ml, and (D) 1.4 mg/ml of anti-BSA antibodies. Peak 3 is thought to be free TMR from incomplete purification of the fluorescently labeled protein sample. The variation in migration times arises from different concentrations of electrolytes in the immunocomplex mixtures, which are made from antibody solutions containing salts.

130

Samuel Kim and Richard N. Zare

We consider two possibilities for immunocomplex formation. If the antigen (i.e., BSA) has more than one recognition site, the order of formation of immunocomplexes upon increasing the antibody concentration will be BSA, BSA-mAb, and BSA-mAb2 (Fig. 7.5A). In contrast, if the antibody has multivalency, which is reasonable recalling the T-shaped structure of the Fab fragments of the antibody (Harris et al., 1992), the order of formation upon the addition of antibody will be BSA, BSA2-mAb, and BSA-mAb (Fig. 7.5B). Since the molecular weights of BSA and mAb are known (69 and 150 kDa, respectively), the expected diffusion times can be calculated. Figure 7.5 shows these two scenarios schematically and tabulates the expected FCS parameters, assuming that the immunocomplexes are spherical in shape and the densities of the proteins are constant. The experimental results agree with the second hypothesis (Fig. 7.5B), as summarized in Table 7.2. When the mixture contains TMR-BSA only, the electropherogram has two peaks (Fig. 7.4A). The stronger and broader peak (Peak 1), which is TMR-BSA, yields the diffusion time of 469  20 ms. The other peak (Peak 3), smaller and sharper, is composed of TMR impurities, which is evident from its shorter diffusion time (116  5 ms) and it is present in all four samples at nearly the same concentration. As the concentration of mAb increases, another peak (Peak 2) starts to appear 2

3

4

MW 69 (kDa)

219

0.5

tD (ms)

677

89

A

1

461*

B

1

2

3

4

369

69

288

0.5

219

806

461*

742

89

677

Figure 7.5 Expected diffusion times of immunocomplexes based on two different assembly hypotheses. (A) Multivalency of the antigen is assumed. (B) Multivalency of the antibody is assumed. The diffusion time of TMR-BSA is estimated to be 461 ms, an average value calculated from three FCS measurements.

Table 7.2 FCS parameters obtained from immunocomplexes separated by microchip CE Mixture TMR-BSA mAb in Fig. 7.4 (mg/ml) (mg/ml)

(A) (B) (C) (D)

0.09 0.09 0.09 0.09

0 0.028 0.20 1.4

Peak 1

2

3

4

469  20  116  5  483  18 849  32    858  25  770  44 432  19   718  11

Single-Molecule Spectroscopy on a Microchip

131

between these two peaks at about a 44-s migration time (Fig. 7.4B) and is found to have a diffusion time longer than that of TMR-BSA (849  32 ms). At even higher mAb concentrations, another peak (Peak 4) appears at the shortest migration time with a diffusion time of 770  44 and 718  11 ms (Fig. 7.4C and D, respectively). The observation that the values of the diffusion times of the chemical species in Peak 2 are larger than those of Peak 4 is consistent with the ‘‘multivalent antibody’’ hypothesis. The discrepancy between the calculated diffusion times and the measured values is partly because (1) the immunocomplexes have nonspherical shapes and (2) the resolved peaks are still overlapping, arising from incomplete separation. We note that the difference in diffusion times of these complexes is not significant and, therefore, the approach of fitting the FCS data with a multiple-species model (i.e., without electrophoretic separation) will produce larger errors in determining diffusion times and relative abundances of distinct molecular aggregates. Once again, the advantages are apparent of being able to carry out single-molecule spectroscopy.

ACKNOWLEDGMENTS We thank Yiqi Luo for helping with the preparation of immunocomplex solutions. S. K. acknowledges the Stanford Bio-X Graduate Fellowships. This work was supported by the National Science Foundation under MCB-0636284 and MCB-0749638.

REFERENCES Anderson, J. R., Chiu, D. T., Jackman, R. J., Cherniavskaya, O., McDonald, J. C., Wu, H. K., Whitesides, S. H., and Whitesides, G. M. (2000). Fabrication of topologically complex three-dimensional microfluidic systems in PDMS by rapid prototyping. Anal. Chem. 72, 3158–3164. Cai, L., Friedman, N., and Xie, X. S. (2006). Stochastic protein expression in individual cells at the single molecule level. Nature 440, 358–362. Cohen, A. E., and Moerner, W. E. (2006). Suppressing Brownian motion of individual biomolecules in solution. Proc. Natl. Acad. Sci. USA 103, 4362–4365. Cui, B. X., Wu, C. B., Chen, L., Ramirez, A., Bearer, E. L., Li, W. P., Mobley, W. C., and Chu, S. (2007). One at a time, live tracking of NGF axonal transport using quantum dots. Proc. Natl. Acad. Sci. USA 104, 13666–13671. Fogarty, K., and Van Orden, A. (2009). Fluorescence correlation spectroscopy for ultrasensitive DNA analysis in continuous flow capillary electrophoresis. Methods 47, 151–158. Harris, L. J., Larson, S. B., Hasel, K. W., Day, J., Greenwood, A., and Mcpherson, A. (1992). The 3-dimensional structure of an intact monoclonal-antibody for canine lymphoma. Nature 360, 369–372. Huang, B., Kim, S., Wu, H., and Zare, R. N. (2007). Use of a mixture of n-dodecyl-beta-Dmaltoside and sodium dodecyl sulfate in poly(dimethylsiloxane) microchips to suppress adhesion and promote separation of proteins. Anal. Chem. 79, 9145–9149. Jorgenson, J. W., and Lukacs, K. D. (1983). Capillary zone electrophoresis. Science 222, 266–272.

132

Samuel Kim and Richard N. Zare

Kim, S., Huang, B., and Zare, R. N. (2007). Microfluidic separation and capture of analytes for single-molecule spectroscopy. Lab Chip 7, 1663–1665. Lipman, E. A., Schuler, B., Bakajin, O., and Eaton, W. A. (2003). Single-molecule measurement of protein folding kinetics. Science 301, 1233–1235. Luo, Y., Huang, B., Wu, H., and Zare, R. N. (2006). Controlling electroosmotic flow in poly(dimethylsiloxane) separation channels by means of prepolymer additives. Anal. Chem. 78, 4588–4592. Melin, J., and Quake, S. R. (2007). Microfluidic large-scale integration: The evolution of design rules for biological automation. Annu. Rev. Biophys. Biomol. Struct. 36, 213–231. Perroud, T. D., Huang, B., and Zare, R. N. (2005). Effect of bin time on the photon counting histogram for one-photon excitation. ChemPhysChem 6, 905–912. St. Claire, R. L. (1996). Capillary electrophoresis. Anal. Chem. 68, R569–R586. Terabe, S., Otsuka, K., Ichikawa, K., Tsuchiya, A., and Ando, T. (1984). Electrokinetic separations with micellar solutions and open-tubular capillaries. Anal. Chem. 56, 111–113. Unger, M. A., Chou, H. P., Thorsen, T., Scherer, A., and Quake, S. R. (2000). Monolithic microfabricated valves and pumps by multilayer soft lithography. Science 288, 113–116. Van Orden, A., and Keller, R. A. (1998). Fluorescence correlation spectroscopy for rapid multicomponent analysis in a capillary electrophoresis system. Anal. Chem. 70, 4463–4471. Wu, H., Wheeler, A., and Zare, R. N. (2004). Chemical cytometry on a picoliter-scale integrated microfluidic chip. Proc. Natl. Acad. Sci. USA 101, 12809–12813.

C H A P T E R

E I G H T

Detection of Protein–Protein Interactions in the Live Cell Plasma Membrane by Quantifying Prey Redistribution upon Bait Micropatterning Julian Weghuber,* Mario Brameshuber,* Stefan Sunzenauer,* ¨bler,* Manuela Lehner,§ Christian Paar,§ Thomas Haselgru Michaela Schwarzenbacher,* Martin Kaltenbrunner,* Clemens Hesch,* Wolfgang Paster,† Bettina Heise,‡ ¨tz* Alois Sonnleitner,§ Hannes Stockinger,† and Gerhard J. Schu Contents 134 136 136 136 136 137 137 137 137 137 139 139 139 139 139

1. Introduction 2. Methodological Requirements 2.1. Detection of weak interactions 2.2. Quantification 2.3. Applicability to living cells 2.4. Applicability to plasma membrane proteins 2.5. Dynamic range/sensitivity 2.6. False negatives/false positives 2.7. High throughput capabilities 3. The Micropatterning Technique 4. Experimental Design 4.1. Cellular expression system 4.2. Capture ligand 4.3. Instrumentation 4.4. Chip production

* Biophysics Institute, Johannes Kepler University Linz, Linz, Austria Department of Molecular Immunology, Center for Physiology, Pathophysiology, and Immunology, Medical University of Vienna, Vienna, Austria { Department of Knowledge-based Mathematical Systems, Johannes Kepler University Linz, Linz, Austria } Center for Biomedical Nanotechnology, Upper Austrian Research GmbH, Linz, Austria {

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72012-7

#

2010 Elsevier Inc. All rights reserved.

133

134

Julian Weghuber et al.

5. Procedure 5.1. Microcontact printing 5.2. Incubation of cells onto the micropatterned surface 5.3. Microscopy 5.4. Data analysis 6. Interpretation of Results 7. Figures of Merit 8. Conclusions Acknowledgments References

140 141 142 143 144 145 147 148 149 149

Abstract Our understanding of complex biological systems is based on high-quality proteomics tools for the parallelized detection and quantification of protein interactions. Current screening platforms, however, rely on measuring protein interactions in rather artificial systems, rendering the results difficult to confer on the in vivo situation. We describe here a detailed protocol for the design and the construction of a system to detect and quantify interactions between a fluorophore-labeled protein (‘‘prey’’) and a membrane protein (‘‘bait’’) in living cells. Cells are plated on micropatterned surfaces functionalized with antibodies to the bait exoplasmic domain. Bait–prey interactions are assayed via the redistribution of the fluorescent prey. The method is characterized by high sensitivity down to the level of single molecules, the capability to detect weak interactions, and high throughput, making it applicable as a screening tool. The proof-of-concept is demonstrated for the interaction between CD4, a major coreceptor in T-cell signaling, and Lck, a protein tyrosine kinase essential for early T-cell signaling.

1. Introduction Unraveling the interaction network of molecules in living cells is key to understanding the mechanisms that regulate cell metabolism and function (Papin et al., 2005). The driving force on our way to obtaining holistic pictures of cell function is a growing repertoire of methodologies that provide qualitative or quantitative data as input for the models. Today, the most straightforward wet-lab approach is based on affinity purification of interaction partners of a bait protein, for example, via coimmunoprecipitation (Barrios-Rodiles et al., 2005). Modifying the bait with a tag— for example, consisting of Protein A and a protease cleavage site (‘‘TAP’’tag)—allows for specific purification of the prey molecules while maintaining the integrity of complexes, which can then be analyzed via, for example, mass spectrometry (Bauch and Superti-Furga, 2006; Puig et al., 2001). Yet, this approach is prone to false positives, since spatial organization is lost

Detection of Protein-Protein Interactions

135

during cell lysis, but also false negatives, since weakly interacting complexes do not endure the purification process. Photocrosslinking of proteins was recently introduced to stabilize interactions prior to cell lysis (Suchanek et al., 2005); whether the global presence of photoactivatable amino acids alters protein function, however, remains to be shown from case to case. Finally, biotinylation of proteins within the same clusters can be achieved using bait-fused horseradish peroxidase, which produces active radical species of arylazide biotin (Kotani et al., 2008); content analysis is performed via Western blotting. In addition, several approaches were designed for the analysis of protein interactions in the living cell. For example, bait and prey can be linked to the nonfunctional fragments of a protein; upon interaction, the two fragments complement each other to reconstitute a functional protein, for example, a transcription factor in yeast two-hybrid screens (Fields and Song, 1989), a signaling molecule in the Ras recruitment system (Broder et al., 1998), a ubiquitin molecule in the split ubiquitin system (Stagljar et al., 1998), or a fluorescent protein (FP) (Ghosh et al., 2000; Hu et al., 2002). While these techniques provide a practical way for detecting the few hits out of a vast number of possible combinations (Uetz et al., 2000), deducing insights into the interaction mechanism appears problematic due to the rather artificial character of the systems, which renders them susceptible to false positives and negatives. For example, for affinity purification, the molecular interactions have to be stable enough to endure the preparation steps. In complementation assays, molecular orientation and distance may hamper the formation of the readout complex. In particular, membrane proteins are difficult to analyze (Stagljar and Fields, 2002). Thus, the verification of the identified interactions by these methods becomes indispensable. Moreover, results are in general of Boolean type, that is, interactions cannot be further quantified. There are a few methods to quantify protein interactions in living cells. The most prominent technique utilizes the energy transfer between a donor and an acceptor dye (Fo¨rster Resonant Energy Transfer, FRET) to sense molecular proximity ( Jares-Erijman and Jovin, 2003; Maurel et al., 2008). While this technique has the advantage of analyzing the molecular interaction in situ, the interpretation of results is complicated by the requirement of a precise knowledge of the respective concentrations and spectral properties of the dyes involved (Valentin et al., 2005). Bioluminescence resonance energy transfer (BRET) may be an attractive alternative to circumvent some practical problems associated with FRET, yet the obtainable signal is rather low, rendering the single cell analysis challenging (Pfleger and Eidne, 2006). Interactions may also be probed via the resulting molecular colocalization using crosscorrelation analysis in time (Bacia et al., 2006; Schwille et al., 1997) or space (Digman et al., 2009). There are a few nonspectroscopic alternative methods for detecting the molecular proximity in living cells. Coimmobilization can be used to detect

136

Julian Weghuber et al.

interactions between a surface-bound membrane protein and its mobile ligand (Dorsch et al., 2009). Finally, protein contact can be detected via proximity ligation (Soderberg et al., 2006). In this technique, bait and prey are labeled with specific antibodies linked to oligonucleotides; upon contact, circular DNA strands are formed which can be detected by rolling circle amplification. In summary, we currently face a multitude of methodological options for addressing molecular interactions in cells. Most of these methods were developed for recording large populations of cells, yet suffer from being rather indirect and therefore hardly quantitative. On the contrary, a few high-end quantitative approaches were introduced, which however are difficult to extend to high throughput. To combine high throughput capabilities with the possibility to extract quantitative information, we recently developed a new concept for identifying protein–protein interactions in situ. The methodological requirements that guided our approach are summarized in the following section.

2. Methodological Requirements 2.1. Detection of weak interactions Up to now, proteomics research mainly focused on detecting a few strong interactions. However, an increasing amount of protein-binding domains with a spectrum of affinities is currently being identified, indicating that weak protein–protein interactions may indeed be the prevalent case.

2.2. Quantification Strong interactions hardly demand for further quantification; for example, the timing of networks may be reproduced solely by the geometry of the circuits, without knowledge on input parameter (Davidich and Bornholdt, 2008). When weak interactions are involved, knowledge on equilibrium and kinetic rate constants becomes critical for describing the system behavior (Li et al., 2004).

2.3. Applicability to living cells In addition to characterizing protein interactions, it would be interesting to know their dependencies on environmental parameters, for example, temperature and the presence of ions or peptides. Experiments on live cells are superior for testing the effect of environmental changes, modulators, or inhibitors on a given bait–prey pair.

Detection of Protein-Protein Interactions

137

2.4. Applicability to plasma membrane proteins The plasma membrane hosts about half of the human proteome, and is the central signaling platform with a multitude of highly controlled interactions. Association of membrane proteins highly depends on the lipid environment (McIntosh and Simon, 2006), making it critical to study such interactions directly in the membrane.

2.5. Dynamic range/sensitivity To account for expression variability between cells, the detection of weakly and highly expressed proteins should be possible at the same readout settings. In particular, weakly expressing cells are analyzable utilizing stateof-the-art ultrasensitive detection capabilities (Schu¨tz et al., 2000).

2.6. False negatives/false positives The method should not miss interaction partners, or lead to positive results in case no interaction occurs.

2.7. High throughput capabilities The assay should enable the screening of large amounts of cells—for example, to test for a library of potential modulators.

3. The Micropatterning Technique Our method is based on the developments of other research groups, who forced membrane proteins into specific patterns within the plasma membrane of living cells (Cavalcanti-Adam et al., 2006; Mossman et al., 2005; Orth et al., 2003; Tanaka et al., 2004; Wu et al., 2004). In these studies, microstructured glass surfaces functionalized with a ligand to the membrane protein of interest were used for specific enrichment. When applied to signaling molecules, the structure of the micropattern was found to influence, for example, cell adhesion (Cavalcanti-Adam et al., 2006) or T-cell activation (Mossman et al., 2005). Triggering the FcE receptor via microstructured lipid bilayers allowed for studying the formation and composition of signaling complexes (Wu et al., 2004). We have extended those approaches in order to identify and quantify interactions between a fluorophore-labeled protein (prey) and a membrane protein (bait) in vivo (Brameshuber et al., 2009; Schwarzenbacher et al., 2008). Figure 8.1 shows the principle of the method. A specific ligand to the

138

Julian Weghuber et al.

exoplasmic domain of the bait is arranged in micropatterns on a glass surface; an example for such a ligand may be an antibody. The intermediate gaps are passivated with BSA. When cells expressing the bait are plated on such surfaces, the bait follows the antibody patterns. To address bait–prey interactions, the lateral distribution of fluorescently tagged prey is analyzed and compared with the antibody/BSA micropatterns. Interaction leads to pronounced copatterning, whereas no interaction yields homogeneous prey-distribution.

BSACy5

Bait

Prey

Biotinylated capture ligand

Streptavidin

Figure 8.1 Schematic illustration of the assay. Grids of BSA-Cy5 are printed on functionalized glass coverslips, and interspaces are filled with streptavidin and biotinylated monoclonal ligands (antibodies) against the membrane protein bait. In cells grown on such microbiochips, the bait will be arranged in the plasma membrane according to the antibody micropattern. Interactions with a second fluorescently labeled protein (prey) are probed by measuring the degree of copatterning (high for specific protein– protein interactions, left; low for noninteracting proteins, right). Image reproduced from (Schwarzenbacher et al., 2008).

Detection of Protein-Protein Interactions

139

4. Experimental Design 4.1. Cellular expression system The simplest approach is to use an adherent cell line for the assay. In principle, the method allows for the analysis of endogenously expressed bait, or proteins overexpressed upon transfection. The basic requirement for choosing the bait is its accessibility to the immobilized capture ligand. The prey has to be labeled with a fluorophore, for example, by fusion to a FP. Alternatively, labeling strategies using tags (Lin et al., 2008; O’Hare et al., 2007) or fluorescent antibodies, but also painting cells with directly labeled proteins or lipids (Legler et al., 2005) could be employed.

4.2. Capture ligand To rearrange exclusively the bait, the interaction of the selected ligand has to be highly specific. An antibody may thus represent the most straightforward choice, but any other specific ligand (e.g., a toxin) can be used. In the case of small ligands, oriented but flexible immobilization may be preferential, which can be mediated, for example, by using a PEG linker (Veronese, 2001).

4.3. Instrumentation Readout is preferentially performed on an epifluorescence microscope operated in total internal reflection (TIR) configuration (Axelrod, 2003). TIR allows for confining the excitation to the ventral plasma membrane, thereby eliminating contributions from prey molecules located on the dorsal membrane or distributed throughout the cytosol. An imaging resolution of at least 1 mm is required to resolve the micropatterns. In addition, scanning capability is required to collect information from many cells. We recently developed a device, that is, perfectly suited for such readout (Hesse et al., 2004).

4.4. Chip production Microcontact printing (Kane et al., 1999) turned out to be a convenient way for producing microstructured surfaces (Fig. 8.2A). We first transfer fluorescent BSA to a reactive glass surface using a microstructured PDMS stamp. The gaps are filled with streptavidin from solution. To have an inherent control on the location and quality of the micropatterns, one may use fluorescently labeled BSA or streptavidin; in our experiments, we used Cy5-labeled BSA. Finally, the surfaces are functionalized using biotinylated antibait ligands. An example for micropatterns is shown in Fig. 8.2B.

140

Julian Weghuber et al.

A

B BSA-Cy5 solution PDMS Dried by nitrogen stream

20 mm

Epoxy-coated glass slide

20 mm

Streptavidin solution

Incubation with biotinylated ligand 20 mm

Figure 8.2 Microcontact-printing for producing structured surfaces. (A) From top to bottom: the PDMS stamp is incubated with BSA-Cy5 solution, dried, and put on a reactive glass surface. After removal of the stamp, the surface is incubated with streptavidin solution, washed, and finally incubated with biotinylated ligand. (B) Two-color image of the obtained micropatterns. Micropatterns were prepared as described and incubated with Alexa555-labeled secondary antibody. The image shows a 270  190 mm detail of the coverslip scanned sequentially at 647 nm for the excitation of BSA-Cy5 (top) and at 514 nm for the excitation of Alexa555 (center). The overlay (bottom) demonstrates the efficient separation of the two reagents, yielding a highcontrast micropattern.

5. Procedure Microcontact printing is performed in close analogy to described protocols (see e.g., Bernard et al., 2000). Polydimethylsiloxane (PDMS) is generated from basic elastomer mixed with starter in a 10:1 ratio, and applied to a silicon master for 30 min at 80
C. The silicon master containing the desired array is generated by standard photolithography. The PDMS stamp is then peeled off the mask and stored at room temperature.

141

Detection of Protein-Protein Interactions

For our experiments, we used epoxy-derivatized glass coverslips prepared as described (Schlapak et al., 2006). In the following paragraphs, we give a quick overview on the printing process and then describe in more detail the handling of the cells, microscopy, and data analysis.

5.1. Microcontact printing Selected steps of microcontact printing are shown in Fig. 8.3. 1. Cut-out field of micropattern of the PDMS stamp (Fig. 8.3A). 2. Wash the field containing the micropattern by flushing it with ethanol (100%) and distilled water. 3. Dry the PDMS stamp with nitrogen or argon gas. 4. Pipette 50 ml BSA-Cy5 work solution (0.67 mg/ml) onto the PDMS stamp so that the whole micropattern field is covered with solution (Fig. 8.3B). Incubate for 30 min at room temperature (Fig. 8.3C). 5. Wash the micropattern field by flushing it with phosphate buffered saline (PBS) and distilled water (Fig. 8.3D). 6. Dry the PDMS with nitrogen or argon gas.

A

E

C

B

F

G

D

H

Figure 8.3 Selected steps of microcontact printing. The micropatterned field is cut out generously of the PDMS stamp and incubated with BSA-Cy5 (A–C). After washing with PBS-buffer and water (D), BSA-Cy5 is transferred onto the epoxy-coated glassslide (E); for convenience, the respective position is labeled on the back (F). A hybridization chamber is placed over the microcontact-printed field (G), which is then incubated sequentially with streptavidin and the bait-antibody. After each incubation step, the slide is washed with PBST-buffer. Finally, adherent cells expressing bait and prey proteins are seeded within the camber and incubated overnight in a humidified petri dish (H).

142

Julian Weghuber et al.

7. Place the PDMS stamp under its own weight onto the middle of one epoxy-coated coverslip and incubate for 30 min at room temperature in a petri dish (Fig. 8.3E). 8. Label the position of the PDMS stamp on the back of the coverslip and strip the stamp from the slide with a forceps (Fig. 8.3F). 9. Stick a Secure Seal Hybridization chamber (Sigma Aldrich article number C0975) over the microcontact-printed field on the coverslip. The label (step 8) helps to localize the center of the microcontactprinted field (Fig. 8.3). 10. Pipette 60 ml streptavidin work solution (50 mg/ml) in the reaction chamber and incubate the sample for 60 min at room temperature (Fig. 8.3H). 11. Wash the sample with 1 ml PBS by adding the buffer into one port of the chamber surface and removing it again at the second port with a pump. 12. Pipette 60 ml biotinylated antibody work solution (10 mg/ml in PBS supplemented with 0.1% Tween 20; i.e., PBST) into the reaction chamber and incubate for 60 min at room temperature. 13. Wash the sample with 1 ml PBST and subsequently, 1 ml PBS by adding the buffer into one port of the chamber surface and removing it again at the second port with a pump. 14. Store the micropatterned surfaces in PBS in the dark at room temperature until cells are ready for seeding.

5.2. Incubation of cells onto the micropatterned surface 1. Grow adherent cells to 50% confluence in a 3-cm tissue culture plate. 2. Detach cells expressing bait and prey proteins of interest with Ethylenediaminetetraacetic acid (EDTA) solution and centrifuge 5 min at 1000 rpm. So far, this protocol has been tested for T24, HEK293, and Chinese Hamster Ovary (CHO) cells. 3. Discard the supernatant and dissolve the cell pellet in the appropriate growth medium. 4. Centrifuge 5 min at 1000 rpm. 5. Discard the supernatant and dissolve the cell pellet in the appropriate growth medium. 6. Remove the PBS from the reaction chamber on the micropatterned coverslip and seed 60 ml of the cell suspension. 7. Create a humid chamber by soaking a sterile pad in distilled water and putting it into the petri dish to prevent the sample from running dry.

143

Detection of Protein-Protein Interactions

8. Incubate the cells overnight at 37
C in a 5% CO2 atmosphere on the micropatterned surfaces. 9. Before analyzing the cells on the microscope, exchange the medium by Hank’s buffered salt solution, including Ca2þ and Mg2þ.

5.3. Microscopy 1. The coverslip is placed on a suitable mount and the cell morphology is checked under white light. 2. The quality of the BSA-Cy5 patterns is checked by excitation at 647 nm. 3. Expression of fluorescent prey protein (GFP or YFP) is analyzed at 488 or 514 nm. If available, the expression of a fluorescent bait protein (CFP) is tested at 405 nm. Importantly, before starting interaction studies, the applied antibody should be tested for its ability to immobilize the respective fluorescently labeled protein. Exemplarily, the redistribution of CD4-YFP on CD4-antibody or CD71-GFP (Transferrin-receptor) on CD71-antibody coated surfaces is shown in Fig. 8.4A and B, respectively. In Fig. 8.4C, we included an image showing the redistribution of CD4-YFP on a micropattern generated by biotinylated murine MHCII (Schwarzenbacher et al., 2008), employing its crossreactivity to human CD4 (Fleury et al., 1996); also using this ligand, the bait could be rearranged in micropatterns, indicating the applicability of alternative monovalent ligands for the assay. 4. Adjust TIR-angle by exciting a cell that expresses the prey protein. Adequate TIRF adjustment is indispensable to visualize patterns due to cytosolic background (Fig. 8.5).

A

B

C

Figure 8.4 Rearrangement of the bait in the live cell plasma membrane. To control the micropatterning of the bait, we transfected cells with fluorescent protein-fusions of the bait and plated them on biochips functionalized with specific bait ligands. TIR images of cells transfected with (A) CD71-GFP on CD71-antibody and (B) CD4-YFP on CD4-antibody microbiochips are shown. (C) Micropatterning of CD4-YFP on a microbiochip functionalized with monovalent mouse MHC class II I-Ek. Scale bars, 20 mm. (B) and (C) are reproduced from Schwarzenbacher et al. (2008).

144

Julian Weghuber et al.

Non-TIRF

TIRF A

B

Figure 8.5 Importance of correct TIR adjustment for the detection of micropatterns. Cells were transfected with CD71-GFP and plated on microbiochips functionalized with CD71-antibody. The images show the same cell using TIR (A) and conventional epi-illumination (B). Scale bars, 20 mm.

5.4. Data analysis For the data evaluation, we applied an in-house developed semiautomated image and data analysis software (Fig. 8.6). The BSA-Cy5 image is used in an automatic gridding algorithm to determine the rotation of the image with respect to the scan direction characterized by the angle ’, and the gridsize. To this end, the user has to first select interesting micropattern structures and to mark them as a region of interest (ROI) in the prey channel of the scan. Second, the intensity of the background (Fbg) has to be determined for normalization. The following stepwise analysis is then performed automatically: 1. Image registration: Green (prey) and red (BSA) channels are mutually aligned (in particular, automatic rotation by the angle ’, correction by projection methods, and entropy maximization). 2. ROI gridding: According to the stamp pattern in the red channel, a grid dividing the ROI into the single grid elements is computed and applied at the green channel image pattern, so that the image pattern structure is split into single spots. 3. Spot masking: Each spot is overlaid by a circular mask, dividing the grid element into the inside and outside mask regions. 4. Feature extraction: For each grid element in the green channel, intensitybased features are computed (mean intensity, variance, entropy, discrepancy according to the inside and outside mask regions, and the resulting contrast). For the examples shown in Fig. 8.6, we made use of the parameters Fþ and F, which specify the mean fluorescence intensity

145

Detection of Protein-Protein Interactions

j

F-

F+

Figure 8.6 Data analysis. We applied an automatic gridding algorithm to calculate the grid-size and the rotation angle w of the image. The grid subdivides the total image into adjacent squares, which were quantified according to the average specific signal within a central circle (Fþ) and the unspecific background outside this circle (F). The information was used to calculate fluorescence and contrast.

within and outside of the circle, respectively. In addition, the background signal Fbg of the glass surface was determined on a part of the chip containing no cells. For each feature, the fluorescence signal F ¼ (Fþ  Fbg) and the contrast C ¼ (Fþ  F)/(Fþ  Fbg) were computed. 5. Data compilation and visualization: Data are compiled in feature matrices; two features (in general, F and C) can be selected and are displayed in 2D histograms (Fig. 8.7). 6. Histogram evaluation: The histograms of different experiments can be compared with respect to, for example, local maxima, spread and mean values.

6. Interpretation of Results Recently, we described the application of the method to the characterization of the interaction between human CD4, the major coreceptor in T-cell activation, and human Lck, the protein tyrosine kinase essential for early T-cell signaling (Brameshuber et al., 2009; Schwarzenbacher et al., 2008). Stable CD4-Lck association is regarded as the basis for Lck recruitment to the immunological synapse, the crucial site for the initiation

146

Before chelation

5

log10F

14 12 10 8 6 4 2 0

= 0.64

4.5 4 3.5 3

2.5 −0.5

C

0.5

1

After chelation 5

12 10

= 0.57

log10F

4.5

C

8 6

4 3.5

4

3

2

2.5 −0.5

0

5

Lck-DN249-mGFP

0

C

0.5

1

8

= 0.16

4.5 log10F

# of events

B

0

6

4 4

3.5

2

3 2.5 −0.5

0

C

0.5

1

# of events

A

# of events

Julian Weghuber et al.

0

Figure 8.7 Selected microbiochip measurements for characterizing the interaction between CD4 and Lck. (A and B) show the influence of zinc chelation on CD4-Lck binding. Cells cotransfected with CD4 and Lck-YFP were plated on CD4-antibody microbiochips in the absence (A) or presence (B) of a zinc chelator. Statistical analysis of multiple cells is shown in a density plot for the fluorescence brightness F and contrast C in the right column. The square indicates a new population at high signal F and low contrast C detected after chelation (B). (C) shows the interaction of CD4 with an Lck-mutant lacking the described interaction sites. Cells were cotransfected with CD4 and Lck-DN249-mGFP and grown on a CD4-antibody coated surface. For this interaction pair, predominantly low contrast values were observed. Scale bars, 20 mm. Images reproduced from Schwarzenbacher et al. (2008).

of T-cell signaling (Li et al., 2004). In our study, we could verify the interaction of these two proteins and additionally found multiple Lck domains contributing to CD4 binding with varying strength. In Fig. 8.7, a few of the main results are summarized. We expressed both CD4 and LckYFP in T24 cells, which were grown on surfaces containing micropatterns of CD4-antibody; the distribution of Lck-YFP followed clearly the immobilized CD4 clusters (Fig. 8.7A). Changes in the obtained patterns upon the addition of external substances (i.e., addition of a chelator, which has been

Detection of Protein-Protein Interactions

147

reported to disrupt the CD4-Lck interaction (Kim et al., 2003)) can be easily analyzed (Fig. 8.7B). The degree of interaction with CD4 was substantially decreased for various truncated Lck mutants. As an example, we show here the results obtained for Lck-DN249-mGFP, which yielded only very weak contrast (Fig. 8.7C). Quantitative discrimination between different patterns can be accomplished by the described algorithm for the analysis of micropatterns using the single-spot fluorescence brightness and contrast. Exemplarily, the disruption of CD4-Lck binding by chelator treatment shifted a prominent population of high fluorescence and high contrast to low contrast, confirming the effect of zinc chelation, seen also by nonautomated methods (Fig. 8.7A and B, right).

7. Figures of Merit In the following, we specify particular figures of merit highlighting the advantages of the micropatterning technique. Weak Interactions. The availability of extremely sensitive tools for automated pattern recognition renders the method applicable to very weak interactions. Direct/indirect interactions. Since the local membrane environment remains unchanged, indirect interactions mediated, for example, by plasma membrane domains, can also be detected. Adjustable invasiveness. The method allows for milder conditions to spatially arrange the bait: in particular, the extent of crosslinking and/or the degree of clustering can be further reduced by using monovalent capture ligands at lower surface density. Sensitivity. The method enables TIR excitation to select for membraneproximal prey molecules, thereby enhancing the sensitivity down to the level of single molecules (see Fig. 8.6 in Schwarzenbacher et al., 2008). No three-dimensional readout required. The sites of interaction are aligned in one focal plane at the interface to the glass coverslip and can therefore be measured rapidly in two-dimensional recordings. No topology artifacts. The membrane lies flat on the coverslip, therefore, membrane topology does not affect the signal per pixel. Spatial resolution. The micropatterns contain multiple adjacent bait enrichment/depletion sites over the cell surface, thereby enabling the local, spatially resolved quantification of molecular interactions. Throughput. In principle, any TIRF-based microscopy platform can be used as a readout system. When high sensitivity is desired, advanced microscopes will be required. For example, we recently developed a readout device which allows for scanning areas of 1 cm2 within 45 min at a pixel size of 129 nm and at the ultimate sensitivity of a single molecule

148

Julian Weghuber et al.

(Hesse et al., 2004). Assuming 1000 mm2 per cell, 130,000 cells could be imaged within 1 h. Further reducing the resolution to 1.5 mm per pixel—which is sufficient for the analysis of 3 mm patterns—would increase the throughput to 1.5 million cells per hour. These capabilities match the high throughput demands of pharmaceutical companies for drug screening (Lang et al., 2006; Ramm, 2005). Combination with other imaging tools. Our method can be combined with alternative techniques to obtain additional information. For example, we have demonstrated a combination with Fluorescence Recovery After Photobleaching to determine the off-rate between the immobilized bait (here CD4) and the mobile prey (here Lck) (Brameshuber et al., 2009; Schwarzenbacher et al., 2008).

8. Conclusions In conclusion, we foresee three particular types of applications for our assay. 1. Resting state analysis. For resting state analysis, one has to ensure that the applied surface does not activate the cell. Depending on the proteins of interest, it may be sufficient to reduce the surface density of capture ligand, for example, by diluting with nonbinding molecules (see Fig. 8.7 in Schwarzenbacher et al., 2008). In some cases, bait dimerization mediated by the capture antibody may already induce signaling; then, a monovalent ligand—for example, a Fab fragment—has to be employed. 2. Analysis of signaling platforms. Frequently, the activation of cells is accompanied by the formation of signaling complexes (Grakoui et al., 1999; Suzuki et al., 2007) that can also be analyzed by the described method. Two scenarios may be envisioned: first, cellular activation may be initiated by the surface itself, for example, by micropatterning the stimulating molecule directly; second, signaling may be initiated by other means (e.g., by employing a soluble ligand, a change in temperature, etc.). In both cases, the platform composition can be addressed by testing fluorescent prey. 3. Modulating bait–prey interactions. Molecular interactions frequently depend on the environment, which include the temperature, the presence of proteins or other molecules (e.g., ions), the lipid composition in the case of membrane proteins, or posttranslational modifications. Our assay allows for the screening modulators of a given interaction pair in the context of live cells.

Detection of Protein-Protein Interactions

149

ACKNOWLEDGMENTS We thank Katharina Strub, University of Geneva, Switzerland, for the hCD71-GFP construct. This work was supported by the Austrian Science Fund (FWF; project Y250-B03), the Competence Center for Biomolecular Therapeutics Research-Vienna and the GEN-AU project of the Austrian Federal Ministry for Science and Research.

REFERENCES Axelrod, D. (2003). Total internal reflection fluorescence microscopy in cell biology. Methods Enzymol. 361, 1–33. Bacia, K., Kim, S. A., and Schwille, P. (2006). Fluorescence cross-correlation spectroscopy in living cells. Nat. Methods 3, 83–89. Barrios-Rodiles, M., Brown, K. R., Ozdamar, B., Bose, R., Liu, Z., Donovan, R. S., Shinjo, F., Liu, Y., Dembowy, J., Taylor, I. W., Luga, V., Przulj, N., et al. (2005). Highthroughput mapping of a dynamic signaling network in mammalian cells. Science 307, 1621–1625. Bauch, A., and Superti-Furga, G. (2006). Charting protein complexes, signaling pathways, and networks in the immune system. Immunol. Rev. 210, 187–207. Bernard, A., Renault, J. P., Michel, B., Bosshard, H. R., and Delamarche, E. (2000). Microcontact printing of proteins. Adv. Mater. 12, 1067–1070. Brameshuber, M., Schwarzenbacher, M., Kaltenbrunner, M., Hesch, C., Paster, W., Weghuber, J., Heise, B., Sonnleitner, A., Stockinger, H., and Schutz, G. J. (2009). Reply to ‘‘Uncoupling diffusion and binding in FRAP experiments’’ Nat. Methods 6, 183–184. Broder, Y. C., Katz, S., and Aronheim, A. (1998). The Ras recruitment system, a novel approach to the study of protein-protein interactions. Curr. Biol. 8, 1121–1124. Cavalcanti-Adam, E. A., Micoulet, A., Blummel, J., Auernheimer, J., Kessler, H., and Spatz, J. P. (2006). Lateral spacing of integrin ligands influences cell spreading and focal adhesion assembly. Eur. J. Cell Biol. 85, 219–224. Davidich, M. I., and Bornholdt, S. (2008). Boolean network model predicts cell cycle sequence of fission yeast. PLoS ONE 3, e1672. Digman, M. A., Wiseman, P. W., Choi, C., Horwitz, A. R., and Gratton, E. (2009). Stoichiometry of molecular complexes at adhesions in living cells. Proc. Natl. Acad. Sci. USA 106, 2170–2175. Dorsch, S., Klotz, K. N., Engelhardt, S., Lohse, M. J., and Bunemann, M. (2009). Analysis of receptor oligomerization by FRAP microscopy. Nat. Methods 6, 225–230. Fields, S., and Song, O. (1989). A novel genetic system to detect protein-protein interactions. Nature 340, 245–246. Fleury, S., Huang, B., Zerbib, A., Croteau, G., Long, E. O., and Sekaly, R. P. (1996). Mutations in human CD4 impair the functional interaction with different human and mouse class II isotypes and alleles. J. Immunol. 156, 1848–1855. Ghosh, I., Hamilton, A. D., and Regan, L. (2000). Antiparallel leucine zipper-directed protein reassembly: Application to the green fluorescent protein. J. Am. Chem. Soc. 122, 5658–5659. Grakoui, A., Bromley, S. K., Sumen, C., Davis, M. M., Shaw, A. S., Allen, P. M., and Dustin, M. L. (1999). The immunological synapse: A molecular machine controlling T cell activation. Science 285, 221–227.

150

Julian Weghuber et al.

Hesse, J., Sonnleitner, M., Sonnleitner, A., Freudenthaler, G., Jacak, J., Hoglinger, O., Schindler, H., and Schutz, G. J. (2004). Single-molecule reader for high-throughput bioanalysis. Anal. Chem. 76, 5960–5964. Hu, C. D., Chinenov, Y., and Kerppola, T. K. (2002). Visualization of interactions among bZIP and Rel family proteins in living cells using bimolecular fluorescence complementation. Mol. Cell. 9, 789–798. Jares-Erijman, E. A., and Jovin, T. M. (2003). FRET imaging. Nat. Biotechnol. 21, 1387–1395. Kane, R. S., Takayama, S., Ostuni, E., Ingber, D. E., and Whitesides, G. M. (1999). Patterning proteins and cells using soft lithography. Biomaterials 20, 2363–2376. Kim, P. W., Sun, Z. Y., Blacklow, S. C., Wagner, G., and Eck, M. J. (2003). A zinc clasp structure tethers Lck to T cell coreceptors CD4 and CD8. Science 301, 1725–1728. Kotani, N., Gu, J., Isaji, T., Udaka, K., Taniguchi, N., and Honke, K. (2008). Biochemical visualization of cell surface molecular clustering in living cells. Proc. Natl. Acad. Sci. USA 105, 7405–7409. Lang, P., Yeow, K., Nichols, A., and Scheer, A. (2006). Cellular imaging in drug discovery. Nat. Rev. Drug Discov. 5, 343–356. Legler, D. F., Doucey, M. A., Schneider, P., Chapatte, L., Bender, F. C., and Bron, C. (2005). Differential insertion of GPI-anchored GFPs into lipid rafts of live cells. FASEB J. 19, 73–75. Li, Q. J., Dinner, A. R., Qi, S., Irvine, D. J., Huppa, J. B., Davis, M. M., and Chakraborty, A. K. (2004). CD4 enhances T cell sensitivity to antigen by coordinating Lck accumulation at the immunological synapse. Nat. Immunol. 5, 791–799. Lin, M. Z., Glenn, J. S., and Tsien, R. Y. (2008). A drug-controllable tag for visualizing newly synthesized proteins in cells and whole animals. Proc. Natl. Acad. Sci. USA 105, 7744–7749. Maurel, D., Comps-Agrar, L., Brock, C., Rives, M. L., Bourrier, E., Ayoub, M. A., Bazin, H., Tinel, N., Durroux, T., Prezeau, L., Trinquet, E., and Pin, J. P. (2008). Cell-surface protein-protein interaction analysis with time-resolved FRET and snap-tag technologies: Application to GPCR oligomerization. Nat. Methods 5, 561–567. McIntosh, T. J., and Simon, S. A. (2006). Roles of bilayer material properties in function and distribution of membrane proteins. Annu. Rev. Biophys. Biomol. Struct. 35, 177–198. Mossman, K. D., Campi, G., Groves, J. T., and Dustin, M. L. (2005). Altered TCR signaling from geometrically repatterned immunological synapses. Science 310, 1191–1193. O’Hare, H. M., Johnsson, K., and Gautier, A. (2007). Chemical probes shed light on protein function. Curr. Opin. Struct. Biol. 17, 488–494. Orth, R. N., Wu, M., Holowka, D., Craighead, H. G., and Baird, B. (2003). Mast cell activation on patterned lipid bilayers of subcellular dimensions. Langmuir 19, 1599–1605. Papin, J. A., Hunter, T., Palsson, B. O., and Subramaniam, S. (2005). Reconstruction of cellular signalling networks and analysis of their properties. Nat. Rev. Mol. Cell. Biol. 6, 99–111. Pfleger, K. D., and Eidne, K. A. (2006). Illuminating insights into protein-protein interactions using bioluminescence resonance energy transfer (BRET). Nat. Methods 3, 165–174. Puig, O., Caspary, F., Rigaut, G., Rutz, B., Bouveret, E., Bragado-Nilsson, E., Wilm, M., and Seraphin, B. (2001). The tandem affinity purification (TAP) method: A general procedure of protein complex purification. Methods 24, 218–229. Ramm, P. (2005). Image-based screening: A technology in transition. Curr. Opin. Biotechnol. 16, 41–48. Schlapak, R., Pammer, P., Armitage, D., Zhu, R., Hinterdorfer, P., Vaupel, M., Fruhwirth, T., and Howorka, S. (2006). Glass surfaces grafted with high-density poly

Detection of Protein-Protein Interactions

151

(ethylene glycol) as substrates for DNA oligonucleotide microarrays. Langmuir 22, 277–285. Schu¨tz, G. J., Hesse, J., Freudenthaler, G., Pastushenko, V. P., Knaus, H.-G., Pragl, B., and Schindler, H. (2000). 3D mapping of individual ion channels on living cells. Single Mol. 1, 153–157. Schwarzenbacher, M., Kaltenbrunner, M., Brameshuber, M., Hesch, C., Paster, W., Weghuber, J., Heise, B., Sonnleitner, A., Stockinger, H., and Schutz, G. J. (2008). Micropatterning for quantitative analysis of protein-protein interactions in living cells. Nat. Methods 5, 1053–1060. Schwille, P., Meyer-Almes, F. J., and Rigler, R. (1997). Dual-color fluorescence crosscorrelation spectroscopy for multicomponent diffusional analysis in solution. Biophys. J. 72, 1878–1886. Soderberg, O., Gullberg, M., Jarvius, M., Ridderstrale, K., Leuchowius, K. J., Jarvius, J., Wester, K., Hydbring, P., Bahram, F., Larsson, L. G., and Landegren, U. (2006). Direct observation of individual endogenous protein complexes in situ by proximity ligation. Nat. Methods 3, 995–1000. Stagljar, I., and Fields, S. (2002). Analysis of membrane protein interactions using yeast-based technologies. Trends Biochem. Sci. 27, 559–563. Stagljar, I., Korostensky, C., Johnsson, N., and te Heesen, S. (1998). A genetic system based on split-ubiquitin for the analysis of interactions between membrane proteins in vivo. Proc. Natl. Acad. Sci. USA 95, 5187–5192. Suchanek, M., Radzikowska, A., and Thiele, C. (2005). Photo-leucine and photo-methionine allow identification of protein-protein interactions in living cells. Nat. Methods 2, 261–267. Suzuki, K. G., Fujiwara, T. K., Edidin, M., and Kusumi, A. (2007). Dynamic recruitment of phospholipase C gamma at transiently immobilized GPI-anchored receptor clusters induces IP3-Ca2þ signaling: Single-molecule tracking study 2. J. Cell Biol. 177, 731–742. Tanaka, M., Wong, A. P., Rehfeldt, F., Tutus, M., and Kaufmann, S. (2004). Selective deposition of native cell membranes on biocompatible micropatterns. J. Am. Chem. Soc. 126, 3257–3260. Uetz, P., Giot, L., Cagney, G., Mansfield, T. A., Judson, R. S., Knight, J. R., Lockshon, D., Narayan, V., Srinivasan, M., Pochart, P., Qureshi-Emili, A., Li, Y., et al. (2000). A comprehensive analysis of protein–protein interactions in Saccharomyces cerevisiae. Nature 403, 623–627. Valentin, G., Verheggen, C., Piolot, T., Neel, H., Coppey-Moisan, M., and Bertrand, E. (2005). Photoconversion of YFP into a CFP-like species during acceptor photobleaching FRET experiments. Nat. Methods 2, 801. Veronese, F. M. (2001). Peptide and protein PEGylation: A review of problems and solutions. Biomaterials 22, 405–417. Wu, M., Holowka, D., Craighead, H. G., and Baird, B. (2004). Visualization of plasma membrane compartmentalization with patterned lipid bilayers. Proc. Natl. Acad. Sci. USA 101, 13798–13803.

C H A P T E R

N I N E

Analysis of Complex Single-Molecule FRET Time Trajectories Mario Blanco*,† and Nils G. Walter* Contents 1. Introduction 2. Analysis of Simple Trajectories 2.1. Selection of trajectories for analysis 2.2. Analysis of FRET state distribution 2.3. Kinetic analysis with thresholding algorithms 3. Analysis of Complex Trajectories 3.1. Hidden Markov analysis of complex FRET trajectories 3.2. Overview of HMM software available for smFRET analysis 3.3. Preprocessing trajectories for analysis by HMM 3.4. Selecting the appropriate number of FRET states 4. Post-HMM Processing and Data Visualization 4.1. Local detection of correlation based on HMM 4.2. Data condensation and visualization 4.3. Applications to single-molecule studies of yeast pre-mRNA splicing 4.4. Summary of a detailed strategy for analysis of complex trajectories with QuB Acknowledgment References

154 156 157 158 158 160 160 163 164 167 168 168 169 171 174 176 176

Abstract Single-molecule methods have given researchers the ability to investigate the structural dynamics of biomolecules at unprecedented resolution and sensitivity. One of the preferred methods of studying single biomolecules is singlemolecule fluorescence resonance energy transfer (smFRET). The popularity of smFRET stems from its ability to report on dynamic, either intra- or intermolecular interactions in real-time. For example, smFRET has been successfully used to characterize the role of dynamics in functional RNAs and their protein * Department of Chemistry, Single Molecule Analysis Group, University of Michigan, Ann Arbor, Michigan, USA Program in Cellular and Molecular Biology, University of Michigan, Ann Arbor, Michigan, USA

{

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72011-5

#

2010 Elsevier Inc. All rights reserved.

153

154

Mario Blanco and Nils G. Walter

complexes, including ribozymes, the ribosome, and more recently the spliceosome. Being able to reliably extract quantitative kinetic and conformational parameters from smFRET experiments is crucial for the interpretation of their results. The need for efficient, unbiased analysis routines becomes more evident as the systems studied become more complex. In this chapter, we focus on the practical utility of statistical algorithms, particularly hidden Markov models, to aid in the objective quantification of complex smFRET trajectories with three or more discrete states, and to extract kinetic information from the trajectories. Additionally, we present a method for systematically eliminating transitions associated with uncorrelated fluorophore behavior that may occur due to dye anisotropy and quenching effects. We also highlight the importance of data condensation through the use of various transition density plots to fully understand the underlying conformational dynamics and kinetic behavior of the biological macromolecule of interest under varying conditions. Finally, the application of these techniques to studies of pre-mRNA conformational changes during eukaryotic splicing is discussed.

1. Introduction One of the most significant advances in the single-molecule field was the advent of measuring fluorescence resonance energy transfer (FRET) between a single FRET fluorophore pair (Ha et al., 1996). FRET relies on the distance-dependent interaction between two fluorophores, a donor and acceptor, so termed because the former ‘‘donates’’ its energy through space to the latter, therefore decreasing in emission intensity while the acceptor’s intensity increases. Due to the nature of the transition dipole interaction between the two fluorophores, energy transfer is more efficient when they are in close proximity than when they are further apart, allowing one to measure relative distances of up to
10 nm between labeled sites on one or more biomolecules (Michalet and Weiss, 2002). A simple FRET ratio on a scale from 0 to 1 as a relative measure of the interfluorophore distance can be calculated as FRET ¼ IA =ðIA þ ID Þ;

ð9:1Þ

where IA and ID are the fluorescence intensities of the acceptor and donor, respectively. Changes in this FRET ratio over time are then used as a measure of the conformational changes of the labeled biomolecule(s) over the course of an experiment. While absolute interfluorophore distances can also be estimated from this FRET ratio (Pereira et al., 2008), purely structural studies of this kind are not the focus of this chapter. FRET measurements have long been performed in the ensemble (Stryer, 1978); however, advances in microscopy and sample preparation

155

Analysis of Complex smFRET Trajectories

have allowed FRET to be performed at the single-molecule level with relative ease and reliability (Roy et al., 2008; Walter et al., 2008). In a commonplace single-molecule FRET (smFRET) experiment, the emissions of the fluorophores attached to an immobilized biomolecule are monitored by wide-field video fluorescence microscopy in real-time (Fig. 9.1A; for a description of the necessary instrumentation, the reader is referred to recent reviews of Roy et al. (2008) and Walter et al. (2008)). Corresponding donor and acceptor spots are identified postacquisition by mapping and pattern recognition of the two appropriately color-filtered video images, their signals integrated, and the FRET ratio calculated at the sampling rate of the detector used to collect the signal (Fig. 9.1B; Roy et al., 2008). The most common fluorophores used for smFRET are Cy3 (donor) and Cy5 (acceptor) because of their relative brightness and photostability when compared to other fluorophores (Aitken et al., 2008; Kapanidis and ˚, Weiss, 2002). The Cy3–Cy5 FRET pair has a Fo¨rster radius of
54 A ˚ allowing it to report on distance changes on the 20–100 A scale (Ishii et al.,

A

Cy3

Cy5

Cy3

B 532 nm Prism

400 200

3⬘

Acceptor

Pre-mRNA

FRET

Flow of cell extract Cy3

5⬘ Cover slip

Water

400 200

Cy5

Donor

FRET

Biotinylated PEG Streptavidin Biotin 2⬘O-methyl RNA

Cy5

Quartz slide

0.9 0.6 0.3 0.0 0

5

10 15 Time (s)

20

25

Objective

Figure 9.1 Capturing the conformational dynamics of single pre-mRNA molecules through smFRET in real-time. (A) A pre-mRNA molecule is immobilized through a 20 -O-methylated capture oligonucleotide, and is bound to a PEG passivated quartz slide through a biotin–streptavidin interaction. The inset depicts a portion of a field of view captured by the I-CCD camera, and how the donor (Cy3) and acceptor (Cy5) fluorophores can be captured simultaneously. Peak-finder algorithms are used to automatically find and match corresponding fluorophores from single molecules (white circles). (B) Exemplary fluorescence intensity and FRET changes of a single molecule.Modified in part from Abelson et al. (2010).

156

Mario Blanco and Nils G. Walter

1999). Detection by smFRET has distinct advantages as it directly observes the kinetics of conformational changes of single molecules at (or away from) equilibrium without the need to synchronize and thus perturb them as ensemble FRET methods require; even highly heterogeneous biomolecular samples with rare and/or transient conformational states can thus be analyzed by smFRET in depth, revealing structural dynamics at unprecedented detail (Roy et al., 2008; Walter et al., 2008). The additional information gathered by smFRET over ensemble FRET approaches depends on the ability to detect subtle changes in inherently noisy data from individual molecules. The contribution of static and dynamic heterogeneity (or disorder) to biological systems, which is typically lost by ensemble-averaging, is being unraveled now by single-molecule detection, contradicting the assumption that all molecules in a population are behaving exactly the same (Ditzler et al., 2008; Fiore et al., 2008; Min et al., 2005). In addition, rare and/or short-lived intermediates often go undetected (or are ignored) in ensemble experiments because of their minimal contribution to the overall signal, yet in smFRET they are readily detectable as long as they are longer lived than the inverse of the rate of collection, usually on the order of 25–100 ms. Through the observation of a large population of molecules during wide-field video fluorescence microscopy, sufficient statistics can be built up quickly to recapitulate the ensemble behavior of molecules, while not masking subpopulations. These features have made smFRET experiments popular, but the ability to extract all possible information from single-molecule time trajectories in an unbiased fashion depends on (semi-)automated data analysis routines, the development of which so far lags behind that of the experimental techniques. We faced a particularly daunting task when analyzing complex smFRET trajectories from a pre-messenger RNA (pre-mRNA) during splicing in vitro (Abelson et al., 2010) and describe here the resulting, practical strategy for the largely unbiased extraction of kinetic information from smFRET trajectories with three or more states. Scripts for implementation of this strategy are available upon request.

2. Analysis of Simple Trajectories The practical value of smFRET experiments depends greatly on the ability to extract kinetic and conformational information about the biomolecule of interest. A priority for smFRET assays should be the design of a system (in terms of composition and fluorophore labeling sites) that leads to changes over time and differences between molecules that are significant enough to discern them in single-molecule trajectories. Once established, such a system allows for the application of FRET distribution analyses and

Analysis of Complex smFRET Trajectories

157

thresholding algorithms that, when incorporated with information about dye position, yield information regarding the structural dynamics of the molecule(s) of interest and how experimental conditions affect them. We will first describe a systematic set of analysis tools that are suitable for simple (two- or three-state) trajectories, and often can be used for the initial characterization of more complex trajectories.

2.1. Selection of trajectories for analysis Even in the case of simple trajectories with only a few (two or three) FRET states, the amount of information that is collected throughout a time series of images from an smFRET experiment requires the use of (semi-)automated scripts (often written for the programs MATLAB or IDL) to process the raw imaging data. First, peak-finder and signal integration algorithms (Roy et al., 2008) are usually performed in IDL to select molecules from a field of view and match peaks collected in different color-filtered channels (Cy3 and Cy5) (inset of Fig. 9.1A). For a wide-field microscope this approach is necessary since an average field of view of perhaps 50  100 mm2 often contains over 50 molecules, and five fields of view or more may be observed per experiment. Second, the initial pool of candidate signals identified by the peak-finder algorithm is filtered to reject background noise and singly labeled molecules and select for bona fide FRETlabeled molecules. This selection is typically done by eye and may be subject to user bias; to minimize such bias, it is helpful to establish a set of defined selection criteria. The more complex the trajectories the more relevant the establishment of objective selection criteria, since the behavior of molecules will often be heterogeneous and a single molecule may go from little or no (FRET  0) to high energy transfer efficiency (FRET  1). Perhaps the two simplest criteria to implement are the presence of photoactive fluorophores and some level of FRET between them. These criteria are satisfied when molecules exhibit significant anticorrelation in donor and acceptor fluorophore intensities (detected by eye or using mathematical correlation analysis), followed by single-step photobleaching of one or both fluorophores. For molecules locked in low-FRET states, anticorrelation may not be easily observed, but the presence of an active acceptor fluorophore can be established upon direct acceptor excitation by a laser pulse at the end of the desired observation window (or intermittently). A photoactive acceptor is characterized by a sharp increase in emission upon such illumination (above a threshold to be established), regardless of the FRET state of the molecule; this feature helps distinguish molecules that are doubly labeled but reside in a low-FRET state from those that are singly labeled. It is also helpful to define as a criterion an expected signal threshold based on the experimental background and signal-to-noise ratio, and to exclude molecules whose donor and acceptor trajectories are significantly positively correlated (Munro et al.,

158

Mario Blanco and Nils G. Walter

2007). The remaining time trajectories will provide the pool of molecules from which quantitative data can be extracted.

2.2. Analysis of FRET state distribution One of the first and most straightforward analysis steps is to create a FRET probability distribution by sampling the FRET values from an ensemble of molecules. Such analysis can be accomplished, for example, by binning all FRET values into a histogram that arise from a 10-s segment of all molecule trajectories observed (sometimes a single long molecule trajectory can be analyzed similarly). This simple procedure yields the relative occupancy of FRET states within the molecule population, as well as the associated relative interfluorophore distances. To this end, fitting with a sum of Gaussian functions as implemented in graphing software such as Microcal Origin is used to model the FRET distribution and obtain quantitative information on the mean FRET values, distribution widths (which is largely shot noise derived), and relative abundances of each molecule conformation (or state) detected (Fig. 9.2A). In Microcal Origin, this is accomplished using the ‘‘Fit MultiPeaks’’ analysis routine and selecting approximate peak centers and widths for initial guesses. For Fig. 9.2, a two-state system was simulated wherein 100 molecules reversibly interconvert between FRET states of 0.2 and 0.8 with rate constants of 0.1 and 0.6 s 1 (inset of Fig. 9.2A) over 1000 data points each. The result in Fig. 9.2A shows that the multipeak fitting procedure is able to identify both FRET states used in the simulation. Even though this procedure works reasonably well, it has been shown to be subject to bias from arbitrary values such as the chosen bin size (Okamoto and Terazima, 2008). Such inaccuracies become more problematic with an increasing number of states and a poor signal-to-noise ratio since the overlap between Gaussian distributions will become more severe. The utility of analyzing FRET distribution histograms is thus limited in the case of complex, multistate single-molecule trajectories.

2.3. Kinetic analysis with thresholding algorithms FRET distribution histograms yield only the identity of FRET states in a population of molecules (or possibly a single long molecule trajectory), but do not extract any kinetic information on their dynamics. For this purpose, one needs to first reliably and efficiently identify the FRET states a molecule is sampling, then calculate the dwell times, or the amount of time spent in a state before transitioning to a specific other state, for all observed state transitions. In systems with FRET states that are easily distinguished by eye and/or histogram analysis, it is possible to perform kinetic analysis with the use of simple thresholding. Thresholding is performed by implementing an algorithm wherein a FRET value that has little to no occupancy

159

Analysis of Complex smFRET Trajectories

A

B 900

-1

0.1 s

800

A

700

-1

FRET

B

0.6 s

500

FRET

Counts

600 400 300

A

200 100

B

0.9 0.6 0.3

0.2

0.4

0.6

B A

0.9 0.6 0.3 0.0 0

0 0.0

0.8

HMM

5

10

15

20

25

Time (s)

1.0

FRET Thresholding 1000

0.12 s-1

A

B

400

B

0.67 s-1

A

200

200

0

0 0 10 20 30 40 50 60 Time (s)

800

800

600

Counts

400

Counts

Counts

600

1000

800

800

HMM

D

600 400

0.11 s-1

A

B

0 2

4 6 8 10 Time (s)

600 400

B

0.56 s-1

A

200

200 0

Counts

C

0 10 20 30 40 50 60 Time (s)

0 0

2

4 6 8 Time (s)

10

Figure 9.2 Histogram distribution analysis, thresholding, and hidden Markov modeling are well suited for simple trajectories. (A) Simulated FRET probability distribution created by simulating 10 s at 100 ms repetition rate (1000 data points) each of 100 molecules based on a two-state model (inset). The distribution was fit using two Gaussians whose sum models the distribution well (red outline). The centers of the Gaussians as determined by the fitting routine are situated at 0.203 and 0.802. (B) A sample trajectory of the simulated two-state system with discrete FRET states determined through either thresholding (top) or hidden Markov modeling (HMM) (bottom). (C) Dwell time analysis of the simulated two-state system after determining the dwell times using the thresholding method or HMM. Measured dwell times were binned and plotted as a cumulative distribution and then fit with single-exponential functions using Microcal Origin.

(as determined by state distribution analysis, Fig. 9.2A) is assigned as the threshold point. The algorithm then scans the FRET values within a trajectory and assigns each to a state based on its position relative to the threshold value (Fig. 9.2B, top panel). To avoid noise-induced assignment of artificial transitions, an additional requirement may be imposed that a new state be only assigned if there are two or more consecutive data points in a new position relative to the threshold value. After each data point within a trajectory is assigned a state, dwell times can be calculated, allowing for the determination of kinetic rate constants for state transitions. Rate constants are calculated directly by plotting dwell times as a cumulative probability distribution (or as a noncumulative probability density function) that is then fit with a single- or, if necessary, multiexponential growth curve (left side of Fig. 9.2C). An example where such analysis has been successful

160

Mario Blanco and Nils G. Walter

is the two- to three-state system of the hairpin ribozyme (Bokinsky et al., 2003; Ditzler et al., 2008; Liu et al., 2007; Rueda et al., 2004; Zhuang et al., 2002). Thresholding can be used reliably only for systems with three or fewer states since with more states threshold values become difficult to assign due to increasing overlap of the Gaussian distributions representing each state.

3. Analysis of Complex Trajectories Due to the complex conformational behavior of many biomolecules, simple two- or three-states FRET systems may be rather the exception than the rule. Complex trajectories may arise from the presence of multiple interconverting structural or chemical states, and/or from transacting factors in solution that interact with the labeled molecule. Limited time resolution of the detector relative to the FRET state dwell times can also complicate trajectories (Lee, 2009), as well as photophysical effects such as blinking and dye anisotropy changes that lead to abrupt changes in FRET signal, often lacking anticorrelation of the donor–acceptor signal. Blinking can be detected with relative ease since the fluorescence intensity of the affected fluorophore drops to background level, and this portion of a trajectory can be ignored in (removed from) the analysis. Dye anisotropy changes due to local fluctuations in dye environment are more subtle effects that are harder to detect and are often disguised if relying solely on the FRET ratio for analysis. Distinction of true FRET changes from these photophysical artifacts becomes increasingly difficult with an increasing number of states that occupy more of the limited FRET value range of 0–1. We therefore found that it is not sufficient to rely solely on the FRET ratio in the analysis of more complex trajectories, and that additional confidence can be gained from concomitantly analyzing the donor and acceptor signals (Abelson et al., 2010).

3.1. Hidden Markov analysis of complex FRET trajectories Hidden Markov modeling (HMM) is a statistical algorithm that has been used for applications as varied as speech recognition, sequence alignment, and now smFRET analysis (Eddy, 2004; McKinney et al., 2006; Poritz, 1988). HMM is well suited for smFRET analysis because of its ability to find discrete states within noisy time series data and reliably find the most probable path through these states (Schuster-Bo¨ckler and Bateman, 2007). A hidden Markov model has three main parameter sets—the probability matrices of transition, emission, and initiation—that are optimized through an iterative process to find the set of parameters that best describe the data. The transition probability matrix contains the probabilities of any one FRET state changing to any other state in the subsequent time step.

Analysis of Complex smFRET Trajectories

161

The emission probability matrix contains the probabilities of a specific FRET signal being emitted by each discrete FRET state. The initiation probability matrix gives the probabilities of starting at each of the possible discrete FRET states. Typically, an HMM is ‘‘trained’’ by using expectation maximization algorithms such as the Baum–Welch algorithm to iteratively modify these three parameter sets to better describe the data. The Viterbi algorithm, for example, can then be used to find the most likely sequence of states (MLSS) based on the trained Markov model parameters. For an indepth discussion of the mathematical foundations underlying HMM, we refer the reader to Lawrence and Rabiner (1989). HMM has been successfully applied to smFRET analysis through the use of programs such as HaMMy (McKinney et al., 2006), QuB (Qin and Li, 2004) and vb-FRET (Bronson et al., 2009). These publicly available programs allow for the application of HMM algorithms to smFRET trajectories, and the extraction of FRET states and rate constants of their interconversion. These HMM programs present the most accessible form of data analysis that produces the most reliable results with minimal a priori assumptions required from the user. Additionally, they perform as well if not better than thresholding algorithms. This point is demonstrated by applying HMM to the simulated system of Fig. 9.2A, resulting in the representative hidden Markov model shown in the lower panel of Fig. 9.2B. Even for this simple two-state system, HMM achieves the best agreement with the raw data in terms of both kinetic rate constants and identity of the underlying FRET states, with less need for user defined (and possibly biased) input (compare Fig. 9.2C and D with inset of Fig. 9.2A). For more complex smFRET trajectories, such as those of the simulated five-state system of Fig. 9.3A (100 molecules, 1000 data points per molecule), it is even clearer that thresholding and distribution analysis do not suffice. A sample trajectory in Fig. 9.3B (top panel) shows occupancy of nearly every point in the FRET range, making choosing a threshold value nearly impossible and largely arbitrary. The data are modeled well, however, with a five-state hidden Markov model that accurately identifies the underlying FRET states despite the data noise and correctly detects when the states are changing (Fig. 9.3B, lower panels). The FRET probability distribution of these same data (Fig. 9.3C) shows what appears to be a simple two-state system, obscuring the three additional states used in the simulation, due to the effects of binning and noise overlap between states. It should be noted that Markov models generally assume that each stochastic transition is governed by a single rate constant with exponentially distributed waiting times. Molecular heterogeneity and limited observation windows often are, however, inherent to smFRET trajectories and cause nonexponentially distributed passage times and thus non-Markovian dynamics (Bokinsky et al., 2003; Ditzler et al., 2008; Fiore et al., 2009; Rueda et al., 2004; Zhuang et al., 2002). In situations where non-Markovian

162

Mario Blanco and Nils G. Walter

A

B FRET

Final state Ff 0.20 0.40 0.65 0.75 0.90

Initial state

0.20

–

0.40 0.50

0.50 0.10 0.50 0.40 –

0.65 0.60 0.50

0.40 0.50 0.50 0.50 3.00

–

0.75 3.00 0.50 2.00

–

0.90 0.70 0.50 0.20 0.80

FRET

Fi

0.80 –

HMM

0.9 0.6 0.3 0.0 0.9 0.6 0.3 0.0 0.9 0.6 0.3 0.0 0

5

C

10 15 Time (s)

20

25

400 350

Counts

300 250 200 150 100 50 0 -0.2 0.0

0.2

0.4 0.6 FRET

0.8

1.0

1.2

Figure 9.3 Complex trajectories are not amenable to thresholding and distribution analysis, but require hidden Markov modeling. (A) Parameters used for simulating a five-state system. FRET states and their transition rate constants are depicted as a transition matrix with the slowest and fastest rate constants (all in s 1) in bold. Fi ¼ initial FRET state; Ff ¼ final FRET state. (B) Raw FRET trajectory of one exemplary molecule simulated with the parameters in (A) (top). The idealized trajectory as determined by HMM executed in QuB (middle) and both trajectories overlaid (bottom). (C) FRET probability distribution of the simulated five-state system. This histogram shows that even in simulated data the five underlying states are obscured by noise, rendering distribution analysis less informative as in simple two-state systems.

behavior is detected, one can still apply HMM as a practical solution for the reliable identification of FRET states and quantification of the associated transition kinetics (Talaga, 2007). The remainder of this chapter is dedicated to the practical aspects of using the programs HaMMy, QuB, and vb-FRET, and presents advances in their use derived from their application to the smFRET characterization of the conformational dynamics of a yeast pre-mRNA during splicing in vitro (Abelson et al., 2010).

Analysis of Complex smFRET Trajectories

163

3.2. Overview of HMM software available for smFRET analysis The program HaMMy by Ha and coworkers presented the first readily available software specifically tailored for HMM analysis of smFRET trajectories (McKinney et al., 2006). The program employs HMM algorithms with a simple graphical user interface that processes the donor and acceptor intensities to determine the underlying discrete FRET states as well as the MLSS through a trajectory. HaMMy’s user interface requires minimal input from the user, mainly guesses as to the number and mean FRET values of the states, and a choice of reestimation algorithm. The simplicity of this program is perhaps its best attribute. One simply loads the single-molecule trajectories formatted as *.dat files with a time column, followed by donor and acceptor intensity columns. The program will then analyze each trajectory and output three files that include the idealized FRET path, the dwell times, and a count of the total number of transitions between FRET states. This information can be utilized with the accompanying transition density plot (TDP) program to extract dwell times and rates for transitions of interest. We have successfully applied HaMMy, for example, to three-state smFRET trajectories of VS ribozyme folding (Pereira et al., 2008). The HaMMy and TDP programs and manuals are freely available at http://bio.physics.illinois.edu/. The program QuB (available at http://www.qub.buffalo.edu/) was not originally designed for smFRET trajectories, but instead for single-ion channel measurements (Qin and Li, 2004). The electrophysiology field has been performing single-molecule measurements on ion channels for over 30 years (Neher and Sakmann, 1976), and the signals acquired from such experiments have the same underlying features as smFRET trajectories with discrete states obscured by noise, and of interest are similarly the values of these states and their rate constants of interconversion. QuB has been successfully applied, for example, to smFRET experiments on the ribosome (Munro et al., 2007) and the spliceosome (Abelson et al., 2010). The graphical user interface is somewhat more challenging to master than that of HaMMy, but it provides the user with greater flexibility and control over the model parameters. For example, it provides the user with a variety of algorithms for reestimation, MLSS calculations, and the ability to impose constraints on a model. One current major advantage of QuB is that it performs HMM much more rapidly than either HaMMy or vb-FRET. The program vb-FRET by Gonzalez, Wiggins, and coworkers (available at http://vbFRET.sourceforge.net) is yet another option for HMM analysis tailored for smFRET trajectories (Bronson et al., 2009). It provides an easy-to-use graphical interface similar to that of HaMMy, but with more customizable options similar to QuB. It is also a MATLAB executable file that for our use has been more stable than other choices. In addition to determining optimal parameters for the hidden Markov model, vb-FRET

164

Mario Blanco and Nils G. Walter

uses an approach known as ‘‘maximum evidence’’ to select the most likely model from a set of models being tested (Bronson et al., 2009). This approach alleviates the common problem of overfitting data as a consequence of the use of a maximum likelihood approach, which often leads to the introduction of uninformative (superfluous) parameters to model the data, as well as the need for postprocessing efforts. Table 9.1 lists the major differences between the three HMM programs. In the following, we provide an example protocol for the use of each program, which may not necessarily be the best choice for all possible datasets but has worked effectively in our hands for both simple and more complex smFRET trajectories.

3.3. Preprocessing trajectories for analysis by HMM 3.3.1. Removal of outliers Often single-molecule trajectories exhibit a state with a FRET value near the boundary of the range of 0–1 that will show transient excursions to FRET values either below 0 (when the acceptor intensity briefly dips into negative territory due to noise in the data) or above 1 (when the donor signal goes into negative territory). For trajectories dwelling extensively in low- or high-FRET states this phenomenon may lead to a significant occupancy of a virtual state below zero or above unity, which in turn may result in it being fit with a state by the HMM software. Consequently, Table 9.1 Summary of features for HMM analysis programs

Program

Model selection

HaMMy QuB

Yes, BIC No, postidealization BIC by user Yes, maximum evidence

vb-FRET

Time to analyze five-state system with 105 data points

Data range


3 h 50 min
15 min

[0, 1] [0, 32,767]


1 h

[0, 1], but can accept beyond this range

A five-state system was modeled to simulate a series of complicated smFRET trajectories. The simulation included 105 total data points (100 molecules with each 1000 frames of data collected) which is on par with what is expected from a series of sm-FRET experiments. This dataset was independently analyzed with the various HMM analysis programs. Model selection was carried out with the indicated techniques, and the total time of analysis was recorded. The column ‘‘Data range’’ indicates the acceptable input for the FRET, donor, and acceptor trajectories for each program. Because HaMMy and vb-FRET were designed to accept input of the range [0, 1] the donor and acceptor trajectories need to be scaled accordingly.

Analysis of Complex smFRET Trajectories

165

the analysis will be complicated by introducing irrelevant states and by interrupting dwell times of more relevant states. The removal of outliers can be accomplished by the implementation of a simple algorithm that finds outliers and scales them back to within the FRET range of 0–1. Generally this algorithm involves identifying the points beyond the allowable FRET range and normalizing them to within the 0–1 range by taking an average of adjacent FRET values. A script containing this algorithm is available upon request. 3.3.2. Smoothing (noise reduction) Single-molecule trajectories are sometimes smoothed to help average out the inherent noise of the data collection process and emphasize the discrete states present. Although this is often useful for simple trajectories, complications arise if a larger number of states with rapid interconversion kinetics are present. One simple method of smoothing, rolling (point) averaging, may obscure transitions with dwell times that are shorter than the averaging window and introduce false FRET states in molecules where two or more states are rapidly interconverting. Rolling point averaging can work well with simple trajectories, but breaks down for more complex trajectories. A nonlinear forward–backward filter introduced by Haran has been used to smooth single-molecule trajectories while minimizing their distortion, offering a clear advantage over rolling point averaging (Haran, 2004). We have applied this nonlinear filter, for example, to trajectories of VS ribozyme folding where it reduces noise and emphasizes conformational changes between the three, clearly separated states (Pereira et al., 2008). One possible problem with this filter is the fact that the noise profile will no longer be Gaussian in shape, which is an assumption for analysis by HMM. In practice, this feature may lead to the fitting of small FRET changes that are within the noise of the raw data. 3.3.3. Formatting for HMM analysis To perform HMM analysis, it is necessary that smFRET trajectories be formatted in a manner amenable for manipulation with the desired analysis program. It should be noted that we are presenting only a limited, most commonly used set of input formats for each program, and for a more extensive list of input and output formats one should refer to the downloadable manuals for HaMMy (http://bio.physics.illinois.edu/HaMMy.html), QuB (http://www.qub.buffalo.edu/wiki/index.php/Main_Page), and vbFRET (http://vbfret.sourceforge.net/). HaMMy loads files in ASCII format (e.g., *.dat) with the tabulated structure ‘‘time, donor intensity, acceptor intensity’’. This structure is convenient as it is commonly the form in which the data are extracted from experiment. They can be loaded individually or in batch. HaMMy expects data values to fall within the range of [0, 1] and the data should be scaled to fit as needed. Based on the standard FRET ratio (9.1), values outside of 0 and 1 do not have any physical

166

Mario Blanco and Nils G. Walter

meaning and should be considered outliers that can be corrected as described above. vb-FRET accepts data in ASCII format as does HaMMy, allowing one to load the same data into both programs for comparison. Although not formally supported, we have tested the program with data scaled outside of [0, 1] and found idealizations to work well. QuB loads data in a variety of formats, including ASCII, without a need for normalization. QuB batch analysis can be performed by loading all molecules into a single *.txt file with each molecule separated from its neighbors by line breaks. QuB allows for the analysis of multiple channels in parallel, so one can visualize the donor, acceptor, and FRET ratio into the same file for observation and analysis. Scripts for the stitching and resegmenting of idealized trajectories are available upon request. We have also written a simple script that converts HaMMy formatted data into a *.txt file for input into QuB and comparison. 3.3.4. Stitching trajectories It is often the case that a single trajectory does not display all possible transitions due to the limitations of photobleaching and the presence of some long-lived states that occupy much of the available time window. We have found that this scenario can limit the effectiveness of HMM programs in determining the discrete states for each molecule because there will be little to no occupancy of some states in some molecules. Additionally, the analysis of trajectories separately leads to HMM algorithms fitting the same conformational states with multiple idealized FRET values since subtle differences in background between trajectories can shift the value of the FRET state up or down. The variability of the values of discrete FRET states found when fitting trajectories separately requires that postprocessing steps are carried out to group similar states before kinetic rates can be calculated for dwell time analysis. Such grouping of similar FRET states, however, requires a user-established criterion that is independent of the HMM analysis, which slows down the analysis routine and diminishes the objectivity of the analysis. A better solution to this problem we found to be to ‘‘stitch’’ trajectories together so that a global analysis can be performed on an entire dataset at once, which provides more data for the HMM algorithm to calculate reliable transition probabilities. This approach can be particularly helpful for shorter trajectories, but care should be taken to resegment trajectories from one another after Markov modeling to prevent the introduction of false transitions at the molecule boundaries. Additionally, it is good practice to independently analyze a subset of molecules, large enough to recapitulate the behavior of the entire population, in both manners and compare the fitting results and HMM rates and states to ensure convergence. Scripts for the stitching and resegmenting of idealized trajectories are available upon

Analysis of Complex smFRET Trajectories

167

request. HaMMy will only analyze segments containing up to 50,000 data points, whereas vb-FRET has been tested for up to 150,000 data points and the stability of the program is not affected. For both of these programs the idealization of such large trajectories will take a long time (Table 9.1), and in our hands has affected the stability of HaMMy. QuB has performed best with the stitched trajectories since it performs Markov modeling much more quickly than either of the currently available versions of HaMMy or vb-FRET (Table 9.1).

3.4. Selecting the appropriate number of FRET states Perhaps the most difficult task in HMM analysis is deciding on the number of states to fit a dataset with, a general problem when using statistical models to approximate an experimental distribution. For simple trajectories the number of states is often easily discernable by eye, but for complex trajectories this task becomes more difficult. Finding an objective and reproducible manner of choosing the appropriate number of states requires the use of a ‘‘rule.’’ The HaMMy manual suggests to allow for two more states than the total numbers of states one assumes to be present in the dataset. The program then has the flexibility to sample higher order models, at the sacrifice of calculation speed. Inherently, a likelihood score based on the total probability calculated will continue to increase with an increase in the number of model parameters (Bronson et al., 2009). Therefore, the goal should not be to maximize the likelihood score, but instead to strive for model parsimony, that is, to maintain the simplest model that best describes the data. HaMMy aims for model parsimony by calculating the total probability for each model tested and then using the Bayesian Information Criterion (BIC) to decide on the most appropriate model. BIC corrects for overfitting, a common problem when using maximum likelihood approaches for determining model parameters, by introducing a penalty for complexity (Wasserman, 2000): BIC  2lnðLLÞ þ klnðN Þ;

ð9:2Þ

where LL is the maximum likelihood reached by the model, k is the number of parameters, and N is the number of data points used in the analysis. QuB in addition outputs a log-likelihood (LL) score for each model that can be used to compare the results of two independent idealizations and, when used to calculate the BIC, can help select the most appropriate model. The BIC calculation has to be performed outside of QuB as there is currently no nested model selection within the QuB algorithm, which speeds it up relative to the other analysis programs. vb-FRET takes a different approach to model selection, mainly the use of maximum evidence instead of maximum likelihood for idealization

168

Mario Blanco and Nils G. Walter

(Bronson et al., 2009). Maximum evidence aims to select the most likely model (and not only the most likely parameter values) through the assumption that the model with the correct number of states has the highest probability of reproducing the experimental data. Maximum evidence can be thought of as the probability of selecting the best parameterized model from the pool of all possible Markov models of a defined order (number of states). In a simple example, it is highly unlikely that a system with three distinct states will ever accurately be reproduced by a two-state Markov model. Conversely, a four-state model can reproduce the three-state system, but is overly complex, thus lowering the confidence in such a higher order model. This leaves the three-state model as the best choice since it can both reproduce the experimental data and is simple enough that the properly parameterized model is more likely to be chosen from the set of all possible outcomes.

4. Post-HMM Processing and Data Visualization Traditionally when analyzing smFRET trajectories, the FRET ratio has been the only metric by which transitions have been detected, under the assumption that the donor and acceptor traces are always anticorrelated. When dealing with more complex trajectories, however, it becomes meaningful to maximize the observables to better understand the underlying dynamics of the molecules being studied. Incorporating the donor and acceptor trajectories, in particular, becomes necessary for systems with more than three states since uncorrelated changes between the dyes will often lead to a FRET value that exhibits some occupancy and therefore may appear as a bona fide FRET state. Such uncorrelated changes may arise from subtle variations in the local environment around the fluorophores such as changes in their rotational diffusion behavior (anisotropy) and fluorescence decay pathways.

4.1. Local detection of correlation based on HMM Although the FRET ratio provides us with a bounded metric for conformational change, and can help correct for problems such as focal drift of the microscope, it is sensitive to unilateral changes in fluorescence intensity in either fluorophore. In complex systems, where cofactors may be acting on the molecule of interest in trans or the fluorophore is placed in a central position relevant for activity but exposed to a complex environment, the local environment around one fluorophore may not remain constant, which in turn may lead to changes in FRET ratio that are not caused by changes in FRET efficiency. Such apparent FRET changes uncorrelated in donor and

Analysis of Complex smFRET Trajectories

169

acceptor signal may be intermixed with true changes in FRET efficiency, making their distinction within an smFRET trajectory nearly impossible. As a practical solution to this problem, we have adopted a strategy that analyzes by HMM the donor and acceptor trajectories alongside the corresponding smFRET trajectory, and subsequently scores each FRET change observed based on the presence or absence of corresponding donor and acceptor signal changes. Implementation of this algorithm requires that first the corresponding donor, acceptor, and FRET trajectories are analyzed as follows. HaMMy requires that the input data be scaled to fit between 0 and 1 so that the donor and acceptor intensities need to be normalized accordingly. In addition, HaMMy was designed to analyze FRET trajectories and therefore computes the FRET ratio (Eq. (9.1)) from its input (‘‘time, donor intensity, acceptor intensity’’). As a consequence, if one is interested in analyzing the donor channel, for example, the data should be formatted as: ‘‘time, 1-donor intensity, donor intensity.’’ For vb-FRET, the same input format as HaMMy can be used, except that no normalization is required. QuB does not require that the data values lie between 0 and 1 so that no normalization is necessary. In addition, QuB’s data input allows for flexibility so that the donor, acceptor, and FRET trajectories can all be loaded at once and analyzed independently by specifying how many channels are present in the dataset. Alternatively, it also allows for each channel to be loaded separately. Once each channel has been idealized using HMM algorithms, we employ a simple algorithm in MATLAB to detect donor–acceptor uncorrelated changes. The algorithm first scans the idealized FRET trajectory for each FRET change, then finds the corresponding time point in both the donor and acceptor trajectories to determine if there was a substantial change in idealized fluorescence intensity, and finally scores the transition using the simple scoring system outlined in Fig. 9.4. Each FRET transition is scored based using the same criteria, and the dwell times of FRET states are corrected after donor–acceptor uncorrelated FRET transitions have been identified and removed from consideration. Dwell time correction is important since uncorrelated changes in FRET otherwise distort (accelerate) the kinetics measured for a system since they shorten the apparent dwell times in specific conformational states.

4.2. Data condensation and visualization HMM techniques provide an efficient method for extracting quantitative measures of FRET states and rate constants of transitions between them from single-molecule trajectories. With increasing complexity of smFRET trajectories, the need to assess these parameters in a condensed and comprehensive representation becomes more evident. Of particular interest are

170

Mario Blanco and Nils G. Walter

transitions between pairs of sequential FRET states. Along with the release of HaMMy, Ha and coworkers released a program that visualizes FRET transitions as so-called TDPs, where the number of times a transition occurs

A

Transition scoring classification Score

Event

1

Single transition in both trajectories in interval

2

Multiple transitions in one trajectories, single transition in other trajectories in interval

3

Multiple transitions in both trajectories in interval

4

Transitions in one trajectory, but not in the other in interval

5

No transitions in either trajectory HMM

Cy5

Cy3

B 400 200 600 400 200

FRET

0.9 0.6 0.3 0.0

5 X 0

4 X

5 Time (s)

1 10

Figure 9.4 Local correlation analysis utilizing HMM algorithms and transition scoring. (A) Transition quality scoring used to exclude artificial FRET transitions caused by unilateral changes in one fluorophore. In our studies we are interested in only transitions with scores 1–3, since these transitions exhibit anticorrelated changes in the fluorescence intensity of both fluorophores simultaneously, as indicated. The time window used to search for transitions in the donor and acceptor trajectories is set by examining the FRET trajectory. The size of the time window we chose to relate to the dwell time immediately before the FRET transition being scored. (B) Experimental trajectory and the scores of highlighted transitions after idealizing the donor signal, acceptor signal, and FRET using hidden Markov modeling (black lines). Transitions marked with an x are not characterized by donor–acceptor anticorrelation and not considered in any further analysis. Modified in part from Abelson et al. (2010).

Analysis of Complex smFRET Trajectories

171

is plotted as a heat map on a two-dimensional grid of final versus initial FRET state (Fig. 9.5; McKinney et al., 2006). TDPs highlight the most prevalent transitions within a population of molecules and require that the transition rate constants be represented in a secondary plot ( Joo et al., 2006; Pereira et al., 2008). The mirror symmetry relative to the main diagonal often observed in this type of plot is a result of reversible conformational changes (Fig. 9.5). It is important to note that the number of possible transitions that can be mapped onto a TDP is N*(N  1), where N is the number of discrete states found in the HMM analysis. For example, a trajectory that contains five states can have up to 20 distinct positions on the TDP, leading to a possibly quite complex plot. Importantly, transitions with slow kinetics will show up only infrequently in the trajectories due to their long dwell times relative to the limited observation window imposed by the photobleaching rate of the fluorophores. To help minimize this relative underrepresentation of slow compared to fast transitions in a TDP, we have developed complementary transitional occupancy density plots (TODPs) that scale transitions based on the fraction of molecules that exhibit them at least once, rather than the number of times they are observed over all molecules. This approach makes slow transitions that many molecules exhibit more visible on the heat map of a TODP than a TDP, and thus guards against a visual overrepresentation of unrepresentative fast transitions only few molecules exhibit (Fig. 9.5). To also incorporate the kinetics of each transition into one and the same plot, we have further developed POpulation-weighted and Kinetically Indexed Transition density (POKIT) plots. POKIT plots thus provide two additional, comprehensive pieces of information compared to TDPs. First, they present as a number of concentric circles the fraction of molecules in the entire dataset that exhibits a specific FRET transition at least once (the information represented in TODPs as heat map). Transitions that are common in a majority of molecules can help identify, for example, conformational changes that are important in a reaction with intermediates leading to products, even if this reaction is irreversible and the transition occurs only once per molecule. Second, POKIT plots provide the average dwell time for each transition in the form of circle colors, facilitating the rapid visual comparison of the kinetics of various datasets (Fig. 9.5).

4.3. Applications to single-molecule studies of yeast pre-mRNA splicing Eukaryotic pre-mRNA splicing is a dynamic process that involves the precise recognition and excision of intervening sequences from in between coding regions (exons) of transcribed pre-mRNAs. Splicing is carried out by a multimega Dalton ribonucleoprotein complex known as the spliceosome. Although rivaling the ribosome in size, the spliceosome is unique in

Number of events

1

1

2000

0.6 1000

0.4

500

0.2 0

Final FRET

Final FRET

0.8 1500

0.8

0.8

0.6

0.6

0.4

0.4

0.2

0.2

0

0 0

0.2

0.4 0.6 0.8 Initial FRET

Fraction of molecules 1

TODP

Average dwell times:

0.1–0.3 s

Transition probability:

10.0–35.0 %

0.8

0.2

0.4 0.6 0.8 Initial FRET

0.3–0.5 s 35.0–65.0 %

0.6 0.4 0.2 0

0 0

1

POKIT 1.0 Final FRET

TDP

0

1

0.5–1.0 s 65.0–90.0 %

1.0–5.0 s

0.2 0.4 0.6 0.8 1.0 Initial FRET >5.0 s

90.0–100.0 %

Figure 9.5 Data visualization for complex trajectories, including TDP and POKIT plot analysis. A side-by-side comparison of the same dataset using three representations. Traditional TDPs are scaled by the number of times a transition is observed over all molecules, regardless of whether in only a small subpopulation of molecules with rapid transitions or commonly in all molecules. TODP and POKIT plots are scaled by the fraction of all molecules within a population that exhibits a particular transition. POKIT plots additionally provide kinetic information encoded in the color of the concentric circles. Modified from Abelson et al. (2010).

Analysis of Complex smFRET Trajectories

173

that it lacks a preformed catalytic core. Instead, assembly proceeds in a stepwise manner, and is influenced by the ATPase activity of several RNA-dependent ATPases. Some of the rearrangements involved include the disruption of RNA–protein, RNA–RNA interactions, and the binding of recognition sequences and accessory proteins that lead to the formation of the catalytically competent complex necessary for the two transesterification reactions of intron excision and exon ligation. Yeast genetics and the development of in vitro yeast splicing assays have allowed for the biochemical characterization of the components involved in splicing as well as the key assembly steps required for splicing. The introduction of smFRET has now allowed us to directly observe the conformational dynamics of the pre-mRNA substrate during spliceosome assembly and catalysis in real-time (Fig. 9.1A; Abelson et al., 2010). The complexity of smFRET trajectories from these experiments (Fig. 9.1B) reflects the overall complexity of the rearrangements needed for splicing activity, as well as the asynchronous behavior of splicing in total yeast cell extract. These features lead to a diversity of FRET states, heterogeneous kinetics, and many occupied transitions between the states as a reflection of the range of conformations through which the pre-mRNA substrate is shuttling. To highlight those FRET transitions that are relevant to splicing it was necessary to compare the wild-type substrate with mutant substrates that are blocked at different stages of splicing (Abelson et al., 2010). Spliceosome assembly on a branchpoint mutant (BP) is impaired, leading to a complete lack of splicing activity. In the 30 splice site mutant (30 SS) the second step of splicing and thus exon ligation is blocked. A detailed kinetic and conformational analysis of the substrates in ATP depleted or ATPsupplemented extract allowed us to identify conformational states that are required for splicing activity (Abelson et al., 2010). The analysis of the smFRET data goes as follows: First, trajectories to be studied are prefiltered by searching for the presence of any substantial anticorrelation by visual inspection. The raw donor, acceptor, and FRET trajectories are independently Markov modeled to determine transition boundaries, using a global fitting routine (stitched trajectories) that simultaneously analyzes all data points taken under a given experimental condition. The entire dataset for each condition is analyzed by the iterative application of the Forward–Viterbi and Baum–Welch algorithms in the QuB program to generate idealized trajectories. The number of states assumed in the idealization is varied from 5 to 11, with all states initially being assigned equal probabilities and rate constants that then are iteratively optimized; the resulting fits are evaluated using the BIC. The model that results in the best BIC score is selected for further analysis. This process is performed on all three corresponding trajectories (donor, acceptor, and FRET). After idealization, a post-HMM processing algorithm (coded and executed in MATLAB) classifies each FRET transition by counting the number and

174

Mario Blanco and Nils G. Walter

direction of transitions found in the donor and acceptor trajectories within a time window defined to begin one quarter of the immediately preceding dwell time before the FRET transition in question and ending one quarter of the following dwell time after this FRET transition, with a minimum set to 0.3 s in either direction from the transition. Each transition is scored based on the metric shown in Fig. 9.4A, and transitions with scores of one, two, or three are used for TDPs. Additionally, FRET transitions with a FRET change smaller than 0.1 are not considered significant and consequently removed. The idealized scored trajectories are used to create TDPs and POKIT plots to examine the conformational and kinetic differences between the mutants under the various experimental conditions. Both TDPs and POKIT plots show substantial differences between the mutants and various conditions, with the POKIT plots highlighting even more subtle differences. The POKIT plots show, for example, that after extended incubation of only the WT substrate in ATP-supplemented extract a population of molecules with a relatively stable high-FRET conformation arises that resembles that of the mature mRNA after splicing (Abelson et al., 2010).

4.4. Summary of a detailed strategy for analysis of complex trajectories with QuB Molecule selection; requirements for an accepted trajectory are (1) donor–acceptor anticorrelation; and (2) presence of each a single donor and acceptor fluorophore as verified by the observation of: (I) single-step photobleaching and (II) emission of Cy5 upon direct excitation. 4.4.1. Preparing data for distribution analysis and HMM (1) Background correction: A straightforward method of background correction involves subtracting the mean value of signal after photobleaching from each value in the trajectory. (2) Remove blinking events, and truncate the trajectory at the point of photobleaching. (3) Output raw data in *.dat format (format ‘‘time, donor signal, acceptor signal’’). (4) Sample 10 s worth of signal from each molecule and create a histogram from binned data using Microcal Origin. (5) Stitch (concatenate) all trajectories into a single trajectory, and use the frequency count function in Microcal Origin with a bin size of 0.05 or less.

Analysis of Complex smFRET Trajectories

175

4.4.2. Preprocessing (1) Outlier removal: Trajectories are inspected for data points outside of the FRET range of [0, 1]. These points are corrected based on the values of the adjacent points in the trajectory. If the background correction is performed properly there should be few outliers. If the number of outliers is roughly
10% or more of the data points of a trajectory this hints at incorrect background correction, or that blinking or photobleaching events have not been properly removed. (2) Stitching: All FRET trajectories of a single experimental condition are stitched together, and formatted for QuB. This procedure is also performed for the donor and acceptor trajectories. The QuB input format is a *.txt file with the data columns ‘‘time, donor intensity, acceptor intensity, and FRET’’ that can be truncated to two columns such as ‘‘time, donor intensity’’ to read out the donor or acceptor trajectories.

4.4.3. HMM analysis with QuB (1) Import data into QuB: Specify the number of columns in the data file and whether or not there is a time column. We typically choose to analyze the donor signal, acceptor signal, and FRET ratio as separate files to allow for better visualization of idealized trajectories that are overlaid on the raw data. (2) Create a hidden Markov model: In the modeling window of QuB begin by creating a model with the lowest number of states assumed to be possible (typically we start our models for complex trajectories with five states). (3) Idealization: The amplitudes and standard deviations of the states are then estimated by using the ‘‘Amps’’ function in QuB. This will initiate the model, and then the Baum–Welch and Forward–Viterbi algorithms are used under the ‘‘Idl/Base’’ menu for idealization. There are several other algorithms available for idealization; SKM has been used successfully, for example, to model smFRET data from ribosomes (Munro et al., 2007) and perform idealizations faster than the Baum–Welch, Forward–Viterbi algorithms presented here. SKM did not fit our smFRET data well, perhaps because of the lower signal-to-noise ratio due to the use of crude yeast cell extract. After each idealization add a state to the model and repeat the amplitude initiation and idealization procedures. Save each idealization as a *.dwt file (QuB file format with FRET states and dwell times), and then copy the LL score for the model from the reports panel and paste it into a spreadsheet for the calculation of BIC.

176

Mario Blanco and Nils G. Walter

4.4.4. Postprocessing (1) Model selection: The LL score output from QuB is used to calculate the BIC for each model tested, and the model with the lowest BIC score is selected for further analysis. (2) Parsing idealized data: The *.dwt file for the selected model is read and the idealized FRET states are then matched with the raw data to create a path file that has the format ‘‘time, donor signal, acceptor signal, FRET, idealized FRET.’’ This path file is then segmented back into the individual molecule trajectories that were initially used to generate the stitched data. (3) Transition scoring: The transition scoring routine is run by loading the path files into a MATLAB script that then finds transitions in the FRET channel, and takes note of the directionality and number of transitions at the corresponding time point in the donor and acceptor trajectories. It then scores each transition based on the scale in Fig. 9.4A, and a scored path file is created. Transitions of scores 1–3 are considered true FRET transitions and thus chosen for further analysis. 4.4.5. Data visualization The scored path file is then input into MATLAB scripts written to recognize the score of the transition, and create TDPs, TODPs, or POKIT plots based on this information. From the TDP and TODPs, dwell time information can be extracted from our scripts by boxing a region within MATLAB around the transition(s) of interest. For POKIT plots, the dwell times are automatically extracted for each transition, and an average is calculated and encoded by the color scheme in Fig. 9.5. The percent of molecules exhibiting a transition is calculated by counting how many molecules within the dataset exhibit that particular transition at least once, and then dividing by the total number of molecules used in the analysis.

ACKNOWLEDGMENT The authors would like to acknowledge the work of Franklin Fuller in writing and developing several of the MATLAB scripts used for transition scoring.

REFERENCES Abelson, J., Blanco, M., Ditzler, M. A., Fuller, F., Aravamudhan, P., Wood, M., Villa, T., Ryan, D. E., Pleiss, J. A., Maeder, C., Guthrie, C., and Walter, N. G. (2010). Conformational dynamics of single pre-mRNA molecules during in vitro splicing. Nat. Struct. Mol. Biol. doi:10.1038/nsmb.1767.

Analysis of Complex smFRET Trajectories

177

Aitken, C. E., Marshall, R. A., and Puglisi, J. D. (2008). An oxygen scavenging system for improvement of dye stability in single-molecule fluorescence experiments. Biophys. J. 94, 1826–1835. Bokinsky, G., Rueda, D., Misra, V. K., Rhodes, M. M., Gordus, A., Babcock, H. P., Walter, N. G., and Zhuang, X. (2003). Single-molecule transition-state analysis of RNA folding. Proc. Natl. Acad. Sci. USA 100, 9302–9307. Bronson, J. E., Fei, J., Hofman, J. M., Gonzalez, R. L. Jr., and Wiggins, C. H. (2009). Learning rates and states from biophysical time series: A Bayesian approach to model selection and single-molecule FRET data. Biophys. J. 97, 3196–3205. Ditzler, M. A., Rueda, D., Mo, J., Hakansson, K., and Walter, N. G. (2008). A rugged free energy landscape separates multiple functional RNA folds throughout denaturation. Nucleic Acids Res. 36, 7088–7099. Eddy, S. R. (2004). What is a hidden Markov model? Nat. Biotechnol. 22, 1315–1316. Fiore, J. L., Hodak, J. H., Piestert, O., Downey, C. D., and Nesbitt, D. J. (2008). Monovalent and divalent promoted GAAA tetraloop-receptor tertiary interactions from freely diffusing single-molecule studies. Biophys. J. 95, 3892–3905. Fiore, J., Kraemer, B., Koberling, F., Erdmann, R., and Nesbitt, D. (2009). Enthalpy-driven RNA folding: Single-molecule thermodynamics of tetraloop-receptor tertiary interaction. Biochemistry 48, 2550–2558. Ha, T., Enderle, T., Ogletree, D. F., Chemla, D. S., Selvin, P. R., and Weiss, S. (1996). Probing the interaction between two single molecules: Fluorescence resonance energy transfer between a single donor and a single acceptor. Proc. Natl. Acad. Sci. USA 93, 6264–6268. Haran, G. (2004). Noise reduction in single-molecule fluorescence trajectories of folding proteins. Chem. Phys. 307, 137–145. Ishii, Y., Yoshida, T., Funatsu, T., Wazawa, T., and Yanagida, T. (1999). Fluorescence resonance energy transfer between single fluorophores attached to a coiled-coil protein in aqueous solution. Chem. Phys. 247, 163–173. Joo, C., McKinney, S. A., Nakamura, M., Rasnik, I., Myong, S., and Ha, T. (2006). Realtime observation of RecA filament dynamics with single monomer resolution. Cell 126, 515–527. Kapanidis, A. N., and Weiss, S. (2002). Fluorescent probes and bioconjugation chemistries for single-molecule fluorescence analysis of biomolecules. J. Chem. Phys. 117, 10953–10964. Lawrence, R., and Rabiner, A. (1989). A tutorial on hidden Markov models and selected applications in speech recognition. Proc. IEEE 77, 257–286. Lee, T. (2009). Extracting kinetics information from single-molecule fluorescence resonance energy transfer data using hidden Markov models. J. Phys. Chem. B 113, 11535–11542. Liu, S., Bokinsky, G., Walter, N. G., and Zhuang, X. (2007). Dissecting the multistep reaction pathway of an RNA enzyme by single-molecule kinetic ‘‘fingerprinting’’ Proc. Natl. Acad. Sci. USA 104, 12634–12639. McKinney, S. A., Joo, C., and Ha, T. (2006). Analysis of single-molecule FRET trajectories using hidden Markov modeling. Biophys. J. 91, 1941–1951. Michalet, X., and Weiss, S. (2002). Single-molecule spectroscopy and microscopy. Compt. Rend. Phys. 3, 619–644. Min, W., English, B. P., Luo, G., Cherayil, B. J., Kou, S. C., and Xie, X. S. (2005). Fluctuating enzymes: Lessons from single-molecule studies. Acc. Chem. Res. 38, 923–931. Munro, J. B., Altman, R. B., O’Connor, N., and Blanchard, S. C. (2007). Identification of two distinct hybrid state intermediates on the ribosome. Mol. Cell 25, 505–517. Neher, E., and Sakmann, B. (1976). Single-channel currents recorded from membrane of denervated frog muscle fibres. Nature 260, 799–802.

178

Mario Blanco and Nils G. Walter

Okamoto, K., and Terazima, M. (2008). Distribution analysis for single molecule FRET measurement. J. Phys. Chem. B 112, 7308–7314. Pereira, M. J. B., Nikolova, E. N., Hiley, S. L., Jaikaran, D., Collins, R. A., and Walter, N. G. (2008). Single VS ribozyme molecules reveal dynamic and hierarchical folding toward catalysis. J. Mol. Biol. 382, 496–509. Poritz, A. B. (1988). Hidden Markov models: A guided tour. In ‘‘International Conference on Acoustics, Speech, and Signal Processing.’’ ICASSP-88, Vol. 1. IEEE Press, pp. 7–13. Qin, F., and Li, L. (2004). Model-based fitting of single-channel dwell-time distributions. Biophys. J. 87, 1657–1671. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Rueda, D., Bokinsky, G., Rhodes, M. M., Rust, M. J., Zhuang, X., and Walter, N. G. (2004). Single-molecule enzymology of RNA: Essential functional groups impact catalysis from a distance. Proc. Natl. Acad. Sci. USA 101, 10066–10071. Schuster-Bo¨ckler, B., and Bateman, A. (2007). An introduction to hidden Markov models. Curr. Prot. Bioinform. A.3A.1, A.3A.9 Appendix 3A. Stryer, L. (1978). Fluorescence energy transfer as a spectroscopic ruler. Annu. Rev. Biochem. 47, 819–846. Talaga, D. S. (2007). COCIS: Markov processes in single molecule fluorescence. Curr. Opin. Colloid Interface Sci. 12, 285–296. Walter, N. G., Huang, C.-Y., Manzo, A. J., and Sobhy, M. A. (2008). Do-it-yourself guide: How to use the modern single-molecule toolkit. Nat. Methods 5, 475–489. Wasserman, L. (2000). Bayesian Model Selection and Model Averaging. J. Math. Psychol. 44, 92–107. Zhuang, X., Kim, H., Pereira, M. J. B., Babcock, H. P., Walter, N. G., and Chu, S. (2002). Correlating structural dynamics and function in single ribozyme molecules. Science 296, 1473–1476.

C H A P T E R

T E N

Single-Molecule Fluorescence Studies of Intrinsically Disordered Proteins Allan Chris M. Ferreon, Crystal R. Moran, Yann Gambin, and Ashok A. Deniz Contents 1. Introduction 2. Single-Molecule Fluorescence Methods 2.1. Single-molecule fluorescence resonance energy transfer 2.2. Dual-color single-molecule coincidence 2.3. Fluorescence correlation spectroscopy 3. Site-Specific Labeling of Intrinsically Disordered Proteins 3.1. Common chemistries for protein labeling with fluorescent dyes 3.2. Advanced techniques for protein dual-labeling for smFRET 4. Examples of SMF Characterization of IDP Structure and Dynamics 4.1. Application 1: Structural properties and dynamics of the priondetermining region of the yeast prion protein Sup35 4.2. Application 2: Denaturant-induced expansion of the a-synuclein disordered state 4.3. Application 3: Coupled binding and folding of a-synuclein 5. Concluding Remarks Acknowledgments References

180 181 182 185 186 187 187 189 190 191 193 195 198 200 200

Abstract Intrinsically disordered proteins (IDPs) (also referred to as natively unfolded proteins) play critical roles in a variety of cellular processes such as transcription and translation and also are linked to several human diseases. Biophysical studies of IDPs present unusual experimental challenges due in part to their broad conformational heterogeneity and potentially complex binding-induced folding behavior. By minimizing the averaging over an ensemble (which is typical of most conventional experiments), single-molecule fluorescence (SMF) Department of Molecular Biology, The Scripps Research Institute, La Jolla, California, USA Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72010-3

#

2010 Elsevier Inc. All rights reserved.

179

180

Allan Chris M. Ferreon et al.

techniques have recently begun to add advanced capabilities for structural studies to the experimental arsenal of IDP investigators. Here, we briefly discuss a few common SMF methods that are particularly useful for IDP studies, including SMF resonance energy transfer and fluorescence correlation spectroscopy, along with site-specific protein-labeling methods that are essential for application of these methods to IDPs. We then present an overview of a few studies in this area, highlighting how SMF methods are being used to gain valuable information about two amyloidogenic IDPs, the Parkinson’s disease-linked asynuclein and the NM domain of the yeast prion protein Sup 35. SMF experiments provided new information about the proteins’ rapidly fluctuating IDP forms, and the complex a-synuclein folding behavior upon its binding to lipid and membrane mimics. We anticipate that SMF and single-molecule methods, in general, will find broad application for structural and mechanistic studies of a wide variety of IDPs, both of their disordered conformations, and their ordered ensembles relevant for function and disease.

1. Introduction Intrinsically disordered proteins (IDPs), usually characterized by a combination of low overall hydrophobicity and large net charge (Uversky et al., 2000), represent a considerable proportion (>30%) of the eukaryotic proteome (Dunker et al., 2001). These proteins function in vivo either in their disordered conformations or as ordered structures induced by binding to cellular partners. They are associated with a wide range of biological functions such as in cell signaling and regulation, and linked to diseases such as neurodegenerative disorders and cancer (Tompa, 2005; Wright and Dyson, 1999). During the past several years, the application of computational methods and conventional ensemble techniques such as NMR, CD, and fluorescence spectroscopy has brought about significant advancement in the understanding of IDP structural and functional properties (Wright and Dyson, 2009). More recently, single-molecule methods have begun to be applied in the study of IDP systems. With the ability to directly detect molecular processes without the loss of information due to ensemble averaging and the capacity to resolve complex structural distributions and dynamics in a straightforward manner (Deniz et al., 2008; Joo et al., 2008; Michalet et al., 2006; Moerner and Orrit, 1999; Schuler and Eaton, 2008; Walter et al., 2008; Zhuang, 2005), combined with the extraordinary sensitivity of fluorescence detection, the use of single-molecule fluorescence (SMF) spectroscopy in the study of IDPs can provide important new insights into the conformational properties of the disordered ensemble, and how these features are altered by binding to cellular partners. More detailed

Single-Molecule Fluorescence of IDPs

181

insights into the structural landscapes of IDPs should prove very important for understanding the link between conformation and IDP functional activities and disease roles. Single-molecule methods offer several unique capabilities that are particularly well suited to studies of IDP structure and function. Single-molecule fluorescence resonance energy transfer (smFRET) measurements (Deniz et al., 2008; Michalet et al., 2006; Roy et al., 2008; Schuler and Eaton, 2008) can provide detailed information about long-range structural properties of IDPs, and the coupling between folding and ligand binding of these proteins (Ferreon et al., 2009; Trexler and Rhoades, 2009; Veldhuis et al., 2009). Single-molecule methods can also provide information about folding intermediates and pathways that are very difficult to extract from ensembleaveraged data, as for example in recent studies of the IDP a-synuclein where a complex binding–folding landscape was uncovered by a combination of smFRET and supporting ensemble CD measurements (Ferreon and Deniz, 2007; Ferreon et al., 2009). In addition, correlation-type measurements can be applied to uncover dynamic properties of disordered or other states of IDPs (Ferreon et al., 2009; Mukhopadhyay et al., 2007; Torres and Levitus, 2007). Because several amyloidogenic proteins are also IDPs, structural studies of these aggregation-prone proteins are hindered at high concentrations. The low concentrations required for single-molecule studies drastically reduce the chances of aggregation, which can be directly tested for by single-molecule coincidence analysis (Mukhopadhyay et al., 2007). Molecular pulling experiments can also shed new light on the folding landscapes of this complex protein class (Brucale et al., 2009; Sandal et al., 2008; Yu et al., 2008); however, these will not be discussed further in this chapter. Finally, the amyloid formation mechanisms of these proteins are themselves very complex and of great interest to the scientific community, and could benefit enormously from single-molecule investigations. In the following section, we discuss a few common and useful SMF and protein-labeling methods, followed by more specific descriptions of their application to a couple of important IDP systems.

2. Single-Molecule Fluorescence Methods A variety of single-molecule methods, both fluorescence and force based, can be used to uncover novel information in IDP systems. Here, we limit our discussion to selected SMF methods while providing references to a few important AFM studies on IDPs. Fluorescence-based single-molecule methods (Deniz et al., 2008) can provide information about conformational properties and subpopulations (smFRET), rapid conformational fluctuations (autocorrelation and FRET-correlation), and induced folding and slow

182

Allan Chris M. Ferreon et al.

interconversions between folded structures (smFRET). Additionally, dualcolor coincidence analysis can be used to verify the monomeric state of proteins being studied, as well as to probe aggregation states, particularly for amyloidogenic IDPs. In Section 2.1, we provide a practical outline of basic experimental details of some of these SMF methods. Examples of how the methods were applied to IDP studies can be found in the final section.

2.1. Single-molecule fluorescence resonance energy transfer Fluorescence (or Fo¨rster) resonance energy transfer (FRET) is the nonradiative transfer of singlet excitation energy from a donor molecule to an acceptor molecule via a dipole–dipole coupling mechanism. The transfer efficiency (EFRET) exhibits a strong dependence on the donor–acceptor intermolecular distance (Fig. 10.1), and is given by EFRET ¼

1 ; 1 þ ðr=R0 Þ6

ð10:1Þ

where r is the interdye distance and R0 is the Fo¨rster distance, which depends on the dye pair and experimental conditions. The intramolecular distance dependence of this phenomenon was first tested by ensemble experiments in 1967 (Stryer and Haugland, 1967). The first single-molecule demonstration of FRET under dry conditions was performed in 1996 (Ha et al., 1996), and the initial testing of the distance dependence of smFRET was reported in 1999 (Deniz et al., 1999). Following these initial demonstrations, smFRET has been extensively used for measurements of biomolecular structure and

FRET efficiency (EFRET)

1.0 0.8 0.6 0.4 0.2

R0

0.0 0

20

40 60 Interdye distance (Å)

80

100

Figure 10.1 Distance dependence of the transfer efficiency. Simulated curve was ˚ , which is in calculated using Eq. (10.1) and assuming a Fo¨rster distance (R0) of 50 A the range typical for dye pairs used for single-molecule FRET studies.

183

Single-Molecule Fluorescence of IDPs

dynamics (Deniz et al., 2008; Michalet et al., 2006; Roy et al., 2008; Schuler and Eaton, 2008; Walter et al., 2008; Zhuang, 2005). The method can typically provide distance information in the 30–70 A˚ range for commonly used dye pairs. Measurements of smFRET are typically performed in a ratiometric manner, with EFRET derived from the relative intensities of donor and acceptor photon emission. In doing so under single-molecule conditions, an increased signal-to-noise ratio can be achieved; for example, errors due to alterations in excitation intensity (due to diffusion) can be minimized. For a typical smFRET experiment, a protein molecule is site-specifically labeled with donor and acceptor dyes across a region of interest in the protein using one of the labeling methods discussed later (Section 3). Experiments are then carried out in one of two formats, either with freely diffusing proteins in a confocal format (Fig. 10.2A) or with immobilized proteins in a total internal reflection fluorescence (TIRF) format (Fig. 10.2B). Although immobilization experiments can provide long time trajectories of fluctuations of single molecules, they can result in substantial perturbations caused by surface interactions for many protein systems. Such perturbations could be especially problematic for dynamic systems such as IDPs because their energy landscapes are flatter (and hence more easily deformed) than for native states of folded proteins; hence, most SMF experiments on IDPs to date have been performed on freely diffusing molecules (Ferreon et al., 2009; Mukhopadhyay et al., 2007; Trexler and Rhoades, 2009; Veldhuis et al., 2009). For experiments on freely diffusing molecules (Mukhopadhyay and Deniz, 2007), a laser beam is focused into a dilute solution (50–100 pM) of dual-labeled protein using a high-numerical aperture (NA) objective (typically a 1.2 NA water immersion objective) in order to excite the donor. The emitted fluorescence from both donor and acceptor dyes is collected using the same objective, then filtered through a pinhole to achieve a diffraction limited sub-fL detection volume. The fluorescence is then separated into donor and acceptor components, and further optically filtered to reduce excitation light and background signals, followed by detection using highsensitivity avalanche photodiode detectors (APDs). Bursts of photons are detected in the donor and acceptor channels when molecules diffuse through the focal volume. These bursts are sorted from background signals by applying a threshold or similar algorithm, and used to determine FRET efficiencies (EFRET) of individual molecules using the relation EFRET ¼

IA ; IA þ gID

ð10:2Þ

where IA and ID are the acceptor and donor signals (photon counts), respectively, and g is a factor that corrects for unequal quantum yields

A

B Freely diffusing molecules

Cuvette

Quartz slide Laser

Immobilized molecules Prism

Laser

High NA objective

High NA objective Inverted

microscope

Inverted

Images

Pinhole

APD (donor)

Filters

APD (acceptor)

Donor

Donor

Dichroic mirror

Acceptor

Dichroic mirror

Acceptor # Bursts

Lens

microscope

Filters

EFRET High-sensitivity CCD camera

Figure 10.2 Schematics of SMF (specifically smFRET) experimental setups for freely diffusing (A) and immobilized (B) molecules. (A) Laser excitation via a high-NA microscope objective lens is followed by fluorescence collection by the same objective, separation, and filtering of FRET components, and detection via avalanche photodiodes (APDs). (B) Prism-type TIRF setup. A quartz prism is used to couple excitation light into a quartz slide with immobilized molecules at an angle resulting in total internal reflection and excitation of molecules by the evanescent wave. Light collection occurs via a high-NA objective, followed by filtering, separation of FRET components, and detection via a high-sensitivity CCD camera.

Single-Molecule Fluorescence of IDPs

185

and/or detection efficiencies for the dyes. EFRET values for the sample are plotted in the form of a histogram, which can then be used to detect conformational subpopulations and monitor changes in their properties (i.e., interdye distance, relative population, distance distribution, and dynamics based on peak position, area, width, and shape) over varied conditions. If the work requires resolution of small shifts in EFRET or determination of absolute distances from the data, it becomes necessary to carry out a careful quantification of the correction factor g and the Fo¨rster distance R0 as a function of labeling construct (protein and labeling position) and solution condition (see supplementary information of Ferreon et al. (2009) for detailed information). Additionally, it is preferable to use constructs where the donor–acceptor dyes are rotationally mobile, which simplifies interpretation of the data. In a more advanced version of the FRET experiment dubbed ALEX (alternating-laser excitation; Kapanidis et al., 2005; Lee et al., 2005), both donor and acceptor dyes are alternately excited, using rapid switching between two laser beams. By analyzing the fluorescence emitted for each excitation during the molecule’s diffusion through the focal volume, this approach provides enhanced capabilities for identifying low-FRET populations by allowing the experimenter to distinguish them from FRET-incapable species (as for acceptor-bleached molecules). Alternatively, experiments can be carried out on immobilized molecules in a TIRF format (Roy et al., 2008). A popular setup utilizes prism-based TIRF, where a quartz prism is used to couple laser light into a quartz slide, and total internal reflection occurs at the slide–sample interface (Fig. 10.2B). Molecules immobilized on the slide at this interface are selectively excited by the evanescent field, which only extends on the order of 100 nm into the sample solution. Fluorescence collection is accomplished via a high-NA objective, similar to the case of freely diffusing molecules. The light is split into donor and acceptor components, and usually imaged onto two regions of a high-sensitivity CCD camera, allowing parallel detection of FRET properties for multiple molecules as a function of time, albeit with slower time resolution than for confocal experiments. Although immobilization can result in artifacts, use of protein-friendly surfaces in combination with careful controls for artifacts will no doubt soon result in many SMF studies in an immobilized format performed on this interesting class of proteins.

2.2. Dual-color single-molecule coincidence Dual-color coincidence is a method that can report on whether two or more IDPs interact to form complexes or aggregates (Mukhopadhyay et al., 2007; Orte et al., 2006). This technique requires two separate single-labeled protein samples (generally labeled at the same location) using two different dyes.

186

Allan Chris M. Ferreon et al.

The experimental setup is the same as that described for smFRET measurements on freely diffusing molecules, except that the two dyes used do not need to exhibit FRET, and are simultaneously excited during the experiment. As described in the discussion below of experiments on the yeast prion protein Sup35, one application of this method is to rule out oligomerization in amyloidogenic IDPs, thus confirming study of conformational properties of the monomeric proteins. Perhaps more importantly, dual-color coincidence can also be used to study oligomerization itself in amyloidogenic IDPs by illustrating the pathway to amyloid formation.

2.3. Fluorescence correlation spectroscopy By using the same experimental setup described for the freely diffusing smFRET, one can record the summed fluorescence fluctuations from a small number of molecules. These fluctuating signals can then be analyzed via a correlation analysis to reveal fluctuation timescales and amplitudes. The basic idea behind fluorescence correlation spectroscopy (FCS) is that a fluorescence fluctuation due to a molecular event (which can be indicative of important structural rearrangements) is associated with a characteristic time duration. Such an event would lead to a signal level that is correlated (e.g., continuously high) for the duration of the event. This is not the case for random noise, where each sampled data point will contain a different ‘‘random’’ value of the random noise, and would be observed as a lack of correlation. A correlation function of the data in effect tests for the time duration over which the signal is correlated; hence, a molecular fluctuation should produce a correlation function decay (or rise in some cases), while random noise does not. Thus, by recording fluctuating signals from a small number of molecules, correlation analysis can extract fluctuation characteristics even in the presence of significant noise. This method has been used extensively for studying molecular fluctuations (Bacia and Schwille, 2007; Magde et al., 1974; Webb, 2001). In the case of IDPs, correlation analyses can be used in conjunction with quenching or FRET (FRET–FCS, Torres and Levitus, 2007) to measure conformational fluctuation timescales (Ferreon et al., 2009; Mukhopadhyay et al., 2007), to gain subtle but important structural information, as discussed below in the examples with the yeast prion and a-synuclein. Autocorrelation analysis of fluorescence fluctuations due to protein diffusion into and out of the confocal volume can provide information about IDP diffusion constants and hence size. This feature has been used in an elegant study of the scaling behavior of the dimensions of a series of polyglutamine peptides (Crick et al., 2006). Proteins need to be singly labeled for simple FCS analysis, while FRET– FCS requires dual-labeling as for FRET experiments.

187

Single-Molecule Fluorescence of IDPs

3. Site-Specific Labeling of Intrinsically Disordered Proteins Before any SMF experiment can be performed on an IDP, it is a requirement that one or more fluorescent dyes be attached to the protein. For FCS measurements, single-site labeling is adequate. Because dye fluorescence is affected by the properties of its immediate environment, it is best to specifically label the protein at a unique site to minimize complications in the resulting data. This site-specific labeling is most often achieved by taking advantage of the chemistries of naturally occurring amino acids. Several common and generally facile methods are presented below, with the decision of which to use depending on the particular characteristics of the IDP of interest and experimental design details. A general principle mechanism applies to all of these labeling schemes: a particular reactive group in the IDP of interest (e.g., a thiol in a cysteine residue) is used to target a complementary reactive group appended to the dye (e.g., maleimide, which reacts specifically with thiol groups) in an irreversible chemical reaction. The key for site-specific labeling is that the reaction with the particular IDP-reactive group should be orthogonal to the chemistries of all other reactive groups present. We list some convenient labeling chemistries below and also display them in Fig. 10.3.

3.1. Common chemistries for protein labeling with fluorescent dyes 3.1.1. Cysteine Cysteine labeling is by far the most common method for labeling proteins for SMF experiments, due in large part to the naturally low frequency of cysteine residues in proteins. Cysteine-reactive dyes (containing maleimide Thiol (Cys)

Maleimide

Amine

NHS-ester

Ketone

Hydroxylamine

O

O HN

O

+ O

N

SH

NH2

N

+

O

HN O

O

O

+

O

H2N

O

O

N

HN

H N

HN

N

O

S O

O

O

O

Figure 10.3 A few selected chemistries for protein labeling. Displayed are thiol (cysteine), amine (lysine or N-terminus), and ketone (unnatural amino-acid) along with the corresponding reactive dye (dyes shown as spheres), and the products formed following reaction.

188

Allan Chris M. Ferreon et al.

or iodoacetamide groups) are commercially available and generally yield high reaction-efficiencies and low off-target reactivities. Multiple cysteines in the native protein structure present an obvious roadblock in achieving site-specificity, raising the need for additional steps to remove (i.e., by sitedirected mutagenesis) or sterically block competing sites ( Jager et al., 2005). These methods are also used if the native positioning of cysteine residues is not optimal for experimentation. However, even minimal changes in the wild-type amino acid sequence could cause structural perturbations. Mutagenesis should thus be applied sparingly, and it should be verified that any structural or functional perturbations for the resulting cysteine mutants are negligible. 3.1.2. Amine groups Amine groups are found abundantly in protein structures (in lysine side chains and the N-termini) and provide a reactive target group for labeling reactions. For SMF experiments, the high occurrence of this target functionality in proteins generally presents more of a problem than an advantage. In theory, the variation in pKa values of the N-terminus and lysine side chains is sufficiently different to allow for selective reaction of at least the N-terminus. However, in practice, labeling specificity is often difficult to achieve and must be carefully ascertained (if necessary, through postlabeling separation of labeling isomers) on a case-by-case basis; hence, site-specific labeling using amine chemistry (Amir and Haas, 1986; Amir et al., 1986) has not been a generally used method for SMF. 3.1.3. Unnatural amino acid functional groups By providing experimenters the ability to introduce biologically novel chemical functionalities, incorporation of unnatural amino acids into proteins of interest significantly expands the repertoire of labeling chemistries. Unnatural amino acids can be introduced to the protein structure in vitro via protein synthesis (Dawson and Kent, 2000) or in vivo using engineered organisms that contain the necessary machinery (e.g., functional tRNA and tRNA-synthetase molecules) to specifically incorporate a particular unnatural amino acid in response to, for example, the amber stop codon (‘‘UAG’’ in an mRNA transcript) (Xie and Schultz, 2006). In either case, insertion of unnatural amino acids creates singular reaction locations for labeling, thus offering maximal site-selectivity. This technique is most useful when incorporating unnatural amino acids that react with widely available fluorophore conjugates, as shown in a recent paper from the Deniz and Schultz labs (Brustad et al., 2008). In this work, a ketone-bearing unnatural amino acid was incorporated into the T4 lysozyme sequence in place of D72 and labeled with an alkoxyamine derivative of the Alexa Fluor 488 dye (Invitrogen, see below for more information). Some disadvantages of using unnatural amino acids include the following: (1) difficulty in generating

Single-Molecule Fluorescence of IDPs

189

large quantities of material; (2) potential for structural perturbations; (3) limitations on size of protein (generally <100 amino acids) for synthetic production; and (4) complexity in engineering organisms capable of incorporating unnatural amino acids for in vivo expression. However, this has already proved to be a powerful labeling technology, and with ongoing improvements, it should see increasing use in the future.

3.2. Advanced techniques for protein dual-labeling for smFRET Precise placement of fluorescent probes increases the level of difficulty in smFRET work because the experimenter must achieve site-specific labeling twice on a single molecule. Typically, cysteine reactivity is used for labeling both positions, and experimental conditions (e.g., dye ratios and order of addition) are used to maximize yield of the desired product. However, this approach inherently leads to a mixed population of labeled species consisting of the proper label combination (one each of donor (D) and acceptor (A) dyes that are FRET-capable), dual-acceptor-labeled (that appear invisible under donor-excitation) and dual-donor-labeled (that lead to increased "zero peak" area and compromised ability to detect low-EFRET structures). Even after excluding dual-donor- and dual-acceptor-labeled proteins (e.g., by chromatography), the experimenter is still left with two different FRET-capable molecular species: D–A and A–D. While the calculated distance values should not be significantly affected by this difference in positioning, variations in local environments at the two unique positions on the protein can introduce additional error and reduce resolution. To address this issue, SMF experimenters can employ a number of more sophisticated techniques for enhancing siteselectivity beyond the inherent limitations of dual-cysteine labeling and the basic reactivities described above. Given the scope of this chapter, we present a limited discussion of orthogonal chemistries and protein ligation, two of the most promising and widely applicable methods for labeling. 3.2.1. Concurrent orthogonal chemistries Orthogonal chemical reactions allow attachment of two fluorescent probes to a single molecule in a simple and highly selective fashion because each of the probes has only one suitable attachment point. A particularly attractive possibility is to combine site-specific cysteine and unnatural amino-acid chemistries. This technique was recently demonstrated, using an expressed T4 lysozyme mutant containing a ketone-bearing unnatural amino acid with a single cysteine residue to attach Alexa Fluor 488-alkoxyamine derivative and Alexa Fluor 594-maleimide, respectively, yielding a high level of the desired hetero-labeled species (Brustad et al., 2008). Moreover, if the reaction chemistry conditions are compatible, both reactions can be done simultaneously in a single step, minimizing intermediate purification steps.

190

Allan Chris M. Ferreon et al.

3.2.2. Protein ligation Chemical reactions that result in an amide bond, like the one naturally formed during ribosomal translation, have been employed in ligating unprotected polypeptide fragments together to create a final, cohesive protein (Dawson and Kent, 2000). The most successful and widely used of these methods has been native chemical ligation (NCL), in which an athioester on the N-terminal fragment reacts with an N-terminal cysteine on the C-terminal fragment (Dirksen and Dawson, 2008; Muralidharan and Muir, 2006). These polypeptides can be produced through standard protein synthesis methods or expressed recombinantly, using fragment-intein fusion to generate the cysteine-reactive species (expressed protein ligation (EPL)). Alternatively, Staudinger ligation (Dirksen and Dawson, 2008) can be used, which applies a different chemistry to produce the amide bond. Protein ligation can be particularly helpful in SMF experiments by significantly increasing the potential for site-specific labeling through the exclusion of portions of the protein from labeling reactions, such as those containing competing reactive sites. As an example, work by Deniz et al. (2000) demonstrated the use of ligation together with cysteine and on-resin N-terminal labeling chemistries to produce dual-labeled chymotrypsin inhibitor 2 for early smFRET protein folding experiments. It is particularly noteworthy that by utilizing protein ligation, the inherent limitations on protein size associated with synthetic production can be circumnavigated, making this direct method of unnatural amino acid incorporation accessible for larger proteins. The disadvantages of using protein ligation methods include specific sequence limitations (a necessary function of the chemistry employed to couple the fragments, such as the need for cysteine at the junction in NCL), and the increase in experimental steps required to produce a FRET-capable labeled molecule.

4. Examples of SMF Characterization of IDP Structure and Dynamics In this section, we illustrate the utility of single-molecule methods for IDP studies, using two example systems, the NM domain of the yeast prion protein Sup35 and a-synuclein. Both proteins share the ability to form amyloid, which in yeast (in the case of Sup35) is believed to have a beneficial function, but is implicated in neurodegenerative diseases such as Parkinson’s and Alzheimer’s diseases in the case of human a-synuclein. Using these examples, we illustrate a number of the methods we have discussed above, highlighting the different kinds of structural information available from each. For NM, a combination of smFRET, coincidence and FCS experiments revealed a compact and rapidly fluctuating ensemble of

Single-Molecule Fluorescence of IDPs

191

structures, while smFRET and FRET–FCS experiments showed a complex binding–folding landscape for a-synuclein, features that could have strong implications for the functional and/or disease roles of these proteins. These methods could be used in similar experiments either individually or in combination to study a wide range of other IDPs.

4.1. Application 1: Structural properties and dynamics of the prion-determining region of the yeast prion protein Sup35 Yeast prions (Patino et al., 1996; Wickner, 1994), which are structurally and functionally different from mammalian prions, act as protein-only elements of inheritance in yeast (Shorter and Lindquist, 2005). Their ‘‘prion’’ structures can produce new beneficial phenotypes by altering processes such as translation termination, nitrogen metabolism, and heterokaryon formation (Uptain and Lindquist, 2002). The Saccharomyces cerevisiae translation termination factor Sup35 (discussed here) is one such prion protein capable of switching into a self-perpetuating prion state from the normal nonprion state (Aguzzi and Polymenidou, 2004; Serio and Lindquist, 2000). In the prion state, the protein is sequestered into an amyloid structure within aggregates, reducing the efficiency of translation termination. This effect results in ribosomes reading through stop codons, changing several phenotypes. The NM segment (250 residues) of Sup35 determines the prion state and comprises two distinct regions, an N-terminal region (N; residues 1– 123) that forms the major amyloid core, and a highly charged middle region (M; residues 124–250), which confers solubility (Krishnan and Lindquist, 2005; Serio et al., 2000). This prion-determining NM region shows some characteristics of an IDP. Here, we discuss how single-molecule studies of the NM protein performed in our lab (Mukhopadhyay et al., 2007) have revealed new insights into the protein’s structural properties and dynamics. For FRET studies, a dual-cysteine mutant (21/121C) of the NM protein was labeled with the Alexa Fluor 488/Alexa Fluor 594 donor–acceptor dye pair, wherein the labeling positions flank the amyloid core of the protein within the N region. smFRET experiments on native NM resulted in histograms with a single high-FRET peak (EFRET  0.8), which could either mean that the native monomeric NM is relatively compact or that the high-FRET peak arises due to oligomerization. To confirm that monomeric proteins were being monitored at the dilute experimental protein concentrations used (100 pM), dual-color single-molecule coincidence measurements (Li et al., 2003) were carried out using the same confocal setup. Two protein samples that had been individually labeled with green and red dyes were mixed and assayed. In an experiment where both dyes were simultaneously excited and detected using Ar-ion and He–Ne lasers,

192

Allan Chris M. Ferreon et al.

we observed that most SMF bursts showed signal on either the green or red channels, but not both (Fig. 10.4A, right). Because dimerization or other types of oligomerization would have resulted in a significant fraction of coincident bursts in the two channels due to complexes with both red and green dyes (Fig. 10.4A, left; control experiment), this experiment directly demonstrated that the protein was predominantly monomeric in the experimental conditions used. We then determined the approximate dimensions of the native state of the protein using the derived EFRET values. Interestingly, the probed region of NM appeared to be quite compact in comparison with the denatured states of other proteins. Next, we proceeded to carry out a guanidine hydrochloride (GdnHCl) titration study of the protein to understand if

Dual-color single-molecule coincidence Dual-labeled DNA

150

Number of events

Number of events

A

100 50 0

0.0

B 100

300 200

60

100

40

Singly labeled NM mixture

20 0

0.0

0.2 0.4 0.6 0.8 1.0 Stoichiometry factor (s)

0

Number of events

0.25 Conformational fluctuation 0.2

Diffusion

0.20

0.0

0.2

0.4

0.6

0.8

1.0

0.8

1.0

80

Fluorescence correlation spectroscopy

Fast component

0 M GdnHCI

80

0.2 0.4 0.6 0.8 1.0 Stoichiometry factor (s)

C

Normalized autocorrelation

smFRET

0.1

Slower component

0.5 M GdnHCI

60 40 20 0

0.0

0.2

0.4

0.6

0.0 1E-5

1E-3

0.1

10

1000

Delay time (ms)

0.15

250

2 M GdnHCI

200 150 100 50

0.10 0

2

6 4 Delay time (ms)

8

10

0 0.0

0.2 0.4 0.6 0.8 FRET efficiency (EFRET)

1.0

Figure 10.4 Single-molecule fluorescence characterization of the NM region of Sup35. (A) Two-color single-molecule fluorescence coincidence for a dual-labeled DNA standard (left) and a mixture of NM singly labeled with either of two dyes (right). (B) Denaturant-induced expansion of NM monitored by single-molecule fluorescence resonance energy transfer (smFRET). (C) Observation of nanosecond conformational fluctuations in NM using fluorescence correlation spectroscopy (FCS). Adapted from Mukhopadhyay et al. (2007). Proc. Natl. Acad. Sci., USA, 104:2649.

Single-Molecule Fluorescence of IDPs

193

the collapsed native state represented a specific stable structure. We observed that the FRET peak showed a gradual nonsigmoidal transition from high to low EFRET (Fig. 10.4B; Mukhopadhyay et al., 2007), in sharp contrast to previously observed sigmoidal transitions for other proteins that fold via a two-state mechanism (Deniz et al., 2000; Schuler et al., 2002; Sherman and Haran, 2006), representing the transition between folded and unfolded populations. This result showed that the native state of this protein does not populate a single stable structure, but rather an ensemble of conformations. Furthermore, the relatively narrow shape of the FRET peak (as compared to a simulated peak, Mukhopadhyay et al., 2007) whose width is limited by statistical or shot noise showed that such an ensemble must be rapidly interconverting on the 0.5 ms timescale of the experiment. In order to directly test for the presence of fast structural fluctuations, we next carried out FCS experiments, taking advantage of quenching of Alexa 488 by tyrosines, and using a confocal setup and a hardware correlator. We uncovered the existence of conformational fluctuations on the 30–300 ns timescale as decays in the autocorrelation functions (Fig. 10.4C), whose amplitude and timescale qualitatively correlated with the number of nearby tyrosines in the sequence. Overall, our results showed that NM is a natively unfolded protein, whose amyloid core region exists as a compact and rapidly fluctuating ensemble of structures, features which may be important in its function and mechanism of aggregation. In particular, the observation of rapid structural fluctuations may provide a link to an aggregation mechanism where misfolding to an amyloid-like conformation occurs within the context of an oligomer rather than in the monomer. Similar analysis can be applied to study other IDPs, as illustrated next for the denaturation of a-synuclein.

4.2. Application 2: Denaturant-induced expansion of the a-synuclein disordered state a-Synuclein is a negatively charged 140-amino acid human IDP expressed mainly in the brain and concentrated in presynaptic nerve terminals ( Jakes et al., 1994; Weinreb et al., 1996). The protein has been linked to the pathology of both sporadic and familial Parkinson’s disease (Kruger et al., 1998; Polymeropoulos et al., 1997; Spillantini et al., 1997). To characterize the unfolded ensemble of a-synuclein, smFRET experiments were carried out and GdnHCl-induced changes in protein dimension as reported by FRET were monitored. Single-molecule measurements were performed in varying denaturant concentrations using 100 pM FRET-labeled G7,84C a-synuclein variant in 0.2 M NaCl, 10 mM sodium acetate, 10 mM NaH2PO4, and 10 mM glycine, pH 7.50  0.05 at room temperature, in the presence of 20 mM wild-type protein. More details on the preparation of the dual-labeled protein are given in the next section.

194

Allan Chris M. Ferreon et al.

B 0.5 48

2 FRET efficiency

k area Total events/U pea

3

1 0 −1 −2 0.1 0.2 FR 0.3 0.4 ET effi 0.5 0.6 cien 0.7 cy

3

[

5

52 0.3

54

6 )

M l] (

2 C nH Gd

1 0

4

50

0.4

56

0.2 0

1

2 3 4 [GdnHCl] (M)

5

6

Approximate interdye distance (Å)

A

Figure 10.5 Single-molecule FRET-characterization of the a-synuclein disordered state. (A) smFRET GdnHCl titration data of a-synuclein. smFRET histograms normalized to the Gaussian function-fitted nonzero peak area are plotted against denaturant concentration, and projected to the EFRET vs. [GdnHCl] plane, where the fitted EFRET parameters are also presented. (B) Measured EFRET versus denaturant concentration. Symbols and error bars are for the fitted EFRET and fitting error values, respectively. The approximate interdye distance equivalents of EFRET values are given as a second y-axis coordinate. The curves in A and B were drawn to show the trends.

Fig. 10.5A shows a summary of the collected smFRET histograms, along with the derived EFRET values. Fig. 10.5B presents an additional representation of the [GdnHCl] dependence of EFRET and the corresponding interdye distances, determined by measuring appropriate correction factors as a function of [GdnHCl]. A detailed description of the general method and instruments used for determining interdye distances is given in Ferreon et al. (2009). Briefly, a series of experiments are carried out to determine dye spectral properties and relative detection efficiencies as a function of experimental condition or molecular state. These data allowed the required g correction factors and Fo¨rster distances to be determined. In the case of Fig. 10.5 data, for quantum yield determinations, the indices of refraction (nD25) of GdnHCl solutions were estimated using the equation [GdnHCl] ¼ 57.147Dn þ 38.68Dn2 – 91.6Dn3, where Dn is the difference between the refractive index of the denaturant solution and the buffer solution, solving numerically for Dn using Mathematica (Wolfram Research, Inc.), and calculating the refractive index for the denaturant solutions from Dn and the measured refractive index of the buffer in the absence of denaturant (Ferreon and Bolen, 2004; Pace, 1986). In the 0–6 M GdnHCl concentration range, the donor and acceptor quantum yields were determined to be practically constant at 0.29  0.01 and 0.23  0.01, respectively. The experimentally derived [GdnHCl] dependence of the Fo¨rster distance (R0) for the Alexa Fluor 488-Alexa Fluor 594 donor–acceptor dye

Single-Molecule Fluorescence of IDPs

195

pair is given by the linear equation R0 ¼ (45.9  0.2) – (0.29  0.05) ˚ [GdnHCl]. For the free dyes in the solution conditions used, R0 is 55 A (Ferreon et al., 2009). The smFRET data presented in Fig. 10.5B clearly shows a noncooperative transition of the protein from high to low EFRET upon GdnHCl titration. Additionally, sub-ms fluctuations were observed in the absence of denaturant using FRET–FCS (see below). Hence, as in the case of NM, the intrinsically disordered state of a-synuclein shows the characteristics of an unfolded but compact state under nondenaturing conditions. Therefore, in addition to providing specific structural information about the complex disordered states, these features in the single-molecule data can also be used as a signature of structural disorder for novel IDPs. We next explore the ability of single-molecule methods to shed light on how binding to partner molecules can induce disorder-to-order transitions and control conformation and hence function for the case of a-synuclein.

4.3. Application 3: Coupled binding and folding of a-synuclein Although the precise biological function of a-synuclein remains elusive, it is very likely that the physiological activity of this IDP involves lipid membrane interaction ( Jensen et al., 1998; Sharon et al., 2001) with concomitant transition from the disordered state to structures rich in a-helical content (Bussell and Eliezer, 2003; Davidson et al., 1998; Jo et al., 2000). Using sodium dodecyl sulfate (SDS), a well-studied lipid mimetic (Helenius et al., 1979; Reynolds and Tanford, 1970), we previously reported a detailed thermodynamic characterization of the coupled binding and multistate folding of a-synuclein as a function of ligand concentration, temperature, and pH (Ferreon and Deniz, 2007), employing circular dichroism spectroscopy and protein phase diagram analysis (Ferreon et al., 2007; Rosgen and Hinz, 2003). More recently, we applied SMF techniques (i.e., smFRET, FCS–FRET, and ALEX–smFRET) to study the structural dynamics of asynuclein interaction with SDS and small unilamellar lipid vesicles (Ferreon et al., 2009), and we discuss these studies below. To directly observe and quantify the different protein conformational and ligand binding states, smFRET was used to monitor the SDS-induced folding of a-synuclein (Ferreon et al., 2009). For these measurements, a dual-cysteine variant (G7,84C) of the IDP was labeled with donor (Alexa Fluor 488) and acceptor (Alexa Fluor 594) dyes (Fig. 10.6A). Labeling reactions were performed in 6 M GdnHCl, 50 mM Tris, pH 7.2, 4  C, overnight and in the dark. Partial reactions were first carried out with the donor dye at equimolar ratio with the protein. Singly labeled proteins were separated and HPLC-purified from the unlabeled and double-labeled proteins using reverse phase chromatography, lyophilized in the dark, and checked by mass spectrometry for degree of labeling and purity. Confirmed

196

Allan Chris M. Ferreon et al.

A

N

NAC

C

B Fm

F

Im

U I events % Total

10

Fraction population

C 1.0 0.8 0.6 0.4 0.2 0.0 0.01 0.1 1 10 100 1000 [SDS] (mM) U

I

F

Im

5

0 0.2 FR ET 0.4 eff icie 0.6 ncy 0.8 (E FR

ET )

100 10 0.1 1.0 0.01

m ](

S

D

[S

)

M

1

Fm

Figure 10.6 smFRET-detection of a-synuclein coupled binding and folding. (A) a-Synuclein primary and micelle-bound tertiary structures (1XQ8) (Ulmer et al., 2005). The N-terminal, NAC, and C-terminal regions of the protein are color-coded as blue, purple, and red, respectively. Green and red spheres indicate residue positions that were labeled with donor and acceptor fluorescent probes. (B) smFRET SDS-binding data of a-synuclein. U, I, F, Im, and Fm refer to the different protein conformations observed. (C) a-Synuclein population distributions as modulated by ligand concentration. Adapted from Ferreon et al. (2009). Proc. Natl. Acad. Sci., USA, 106:5645

singly labeled proteins were then reacted with fivefold excess concentration of the acceptor dye to yield a mixture of labeling isomers (A–D and D–A) of the dual-labeled protein that was then HPLC-purified and checked by MALDI-TOF mass spectrometry. smFRET histograms were collected for the dual-labeled protein at different SDS concentrations using 100 pM dual-labeled protein in 0.2 M NaCl, 10 mM sodium acetate, 10 mM NaH2PO4 and 10 mM glycine, pH 7.50  0.05 at room temperature, in the presence of 20 mM wild-type a-synuclein. Raw smFRET data are summarized in Fig. 10.6B as a threedimensional map. Projection of the data to the EFRET–[SDS] plane results in an empirical, model-free protein phase diagram. Protein conformations are easily identified as peaks exhibiting characteristic EFRET. Consistent with our previous ensemble data (Ferreon and Deniz, 2007), a total of five distinct a-synuclein SDS binding modes were resolved. The five conformations can be grouped into three thermodynamic states (U, I-type, and Ftype states) or distinguished as either SDS monomer- or micelle-binding

197

Single-Molecule Fluorescence of IDPs

(I and F, or Im and Fm, respectively). Nonlinear least-squares fitting of individual smFRET histograms to Gaussian functions derived structural distributions at the different experimental conditions. The calculated fractional populations for the different a-synuclein conformations as a function of SDS concentration are shown in Fig. 10.6C. Interestingly, practically the same isothermal SDS titrations as those described above performed in the absence of background unlabeled a-synuclein resulted in smFRET data wherein the F state was missing (Ferreon et al., 2009; Veldhuis et al., 2009), likely due to competitive binding of the micelle and monomer forms of SDS and the use of very dilute FRET-labeled samples. To probe the structural dynamics of the different a-synuclein SDS binding modes, correlation measurements were performed using the FRET donor and acceptor signals and the previously described FCS– FRET method (Torres and Levitus, 2007). Conformational fluctuations were detected on the ns–ms timescales (Fig. 10.7). Rapid fluctuations were revealed for the IDP state, similar to the case of Sup35 (Mukhopadhyay et al., 2007). Strikingly, although states I and Im are structurally very similar

FRET fluctuations

I

2.5 2.0

Fast

Fast

Slow

1.5 1.0 10−6 10−4 10−2 100 Time (ms)

FRET fluctuations

3.0

102 10−6 10−4 10−2 100 Time (ms) F

Im

Intermediate

1.5 1.0 10−6 10−4 10−2 100 Time (ms)

102 10−6 10−4 10−2 100 Time (ms)

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

102

2.5 2.0

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

G(t)

U

G(t)

3.0

102

Figure 10.7 a-Synuclein structural dynamics quantified by correlation analysis of FRET fluctuations (FCS–FRET). FCS–FRET experiments were performed under conditions where the indicated protein conformations (U, I, F, or Im) were preferentially populated. Fast, intermediate, and slow FRET fluctuations (yellow symbols) were observed. Pseudo-cross-correlation of donor intensities, and crosscorrelation of donor and acceptor intensities are represented as green and red symbols, respectively. Adapted from Ferreon et al. (2009). Proc. Natl. Acad. Sci., USA, 106:5645.

198

Allan Chris M. Ferreon et al.

on the basis of their near identical helicity and similar EFRET, and can be considered as belonging to the same thermodynamic states (Ferreon and Deniz, 2007), a clear distinction between the two conformations is observed in terms of their dynamics. Signals from low-EFRET species like the F conformation of a-synuclein, usually very difficult to differentiate from the zero peak signals in conventional smFRET measurements, are easily distinguished using ALEX– smFRET (Fig. 10.8A and B). Low-EFRET F-like conformations are assumed by a-synuclein in the presence and absence of interacting surfaces. The same is true for the high-EFRET I-like structures. This maintenance of the protein’s multistate character at different interaction regimes implies a sequence-dictated propensity to form the observed structures. smFRET data from our lab (Ferreon et al., 2009) and the Rhoades group (Trexler and Rhoades, 2009), consistent with pulse ESR data by Georgieva et al. (2008), showed the dominant conformation for a-synuclein bound to the investigated lipid membranes to be the low-EFRET F-like type.

5. Concluding Remarks We have outlined basic ideas of how SMF methods can be used to study IDP structural and dynamic properties, and briefly discussed some of the handful of studies that have recently been published in this area. While still in its infancy, we anticipate that this area of research will grow rapidly. As an example, future studies will no doubt further probe the functional link between a-synuclein’s three-state behavior (Ferreon and Deniz, 2007) and its soluble and membrane-bound structures in cells, as well as the corresponding links to its regulation and function. Furthermore, while only a few protein systems have been studied thus far, the presented methods should be applicable to a broad range of IDPs. In the case of amyloidogenic IDPs, a combination of smFRET and other single-molecule methods (such as polarization or coincidence) could revolutionize our understanding of the extremely complex mechanisms by which structures of these IDPs evolve and contribute to amyloid formation. In addition to the methods discussed in this chapter, other and more novel SMF techniques that are currently under active development could also be used; for example, three-color or multicolor smFRET (Clamme and Deniz, 2005; Hohng et al., 2004; Lee et al., 2007) can be applied for measuring more global structural properties, while the incorporation of microfluidics would allow more facile studies of the evolution of single-molecule properties with higher time resolution (Hamadani and Weiss, 2008; Lemke et al., 2009; Lipman et al., 2003) or over a large range of conditions (Vandelinder et al., 2009). Moreover, single-molecule manipulation studies, for example, using

199

Single-Molecule Fluorescence of IDPs

Number of events

A 80 60 40 20 0

6

1.2

Zero peak

1.0

5

Stoichiometry

0.8 4

0.6 3

0.4 0.2

F 2

Im 0.0

1

–0.2 0.0

Number of events

B 80 60 40 20 0

0.4 EFRET

0.8

Zero peak

1.2

6

1.0 Stoichiometry

40 80 1.2 0 Number of events

5

0.8 4

0.6 3

0.4 Im

0.2

2

0.0 1

–0.2 0.0

0.4 EFRET

0.8

50 1.2 0 Number of events

Figure 10.8 a-Synuclein extended conformation resolved using alternating-laser excitation single-molecule FRET (ALEX–smFRET). ALEX–smFRET measurements were performed using conditions that favor predominant population of the SDS monomer-bound F (A) or spherical micelle-bound Im (B) conformations. Adapted from Ferreon et al. (2009). Proc. Natl. Acad. Sci., USA, 106:5645

200

Allan Chris M. Ferreon et al.

AFM or optical tweezers will also continue to provide new insights about IDP structure and interactions, such as shown with the IDP a-synuclein (Brucale et al., 2009; Sandal et al., 2008; Yu et al., 2008). Finally, it is particularly noteworthy that the special characteristics of single-molecule experiments allow more direct comparisons and interactions with predictions from simulation and theory, and this powerful capability will no doubt be leveraged to achieve an enhanced understanding of these complex molecules.

ACKNOWLEDGMENTS We thank Dr. Nelson B. Cole and Dr. Robert L. Nussbaum for providing us the plasmid construct for wild-type a-synuclein. We also thank the several colleagues and collaborators who contributed to the authors’ work reviewed here.

REFERENCES Aguzzi, A., and Polymenidou, M. (2004). Mammalian prion biology: One century of evolving concepts. Cell 116, 313–327. Amir, D., and Haas, E. (1986). A series of site-specific fluorescently labeled BPTI derivatives prepared by nonselective acylation and chromatographic separations. Int. J. Pept. Protein Res. 27, 7–17. Amir, D., Levy, D. P., Levin, Y., and Haas, E. (1986). Selective fluorescent labeling of amino groups of bovine pancreatic trypsin inhibitor by reductive alkylation. Biopolymers 25, 1645–1658. Bacia, K., and Schwille, P. (2007). Fluorescence correlation spectroscopy. Methods Mol. Biol. 398, 73–84. Brucale, M., Sandal, M., Di Maio, S., Rampioni, A., Tessari, I., Tosatto, L., Bisaglia, M., Bubacco, L., and Samori, B. (2009). Pathogenic mutations shift the equilibria of asynuclein single molecules towards structured conformers. Chembiochem 10, 176–183. Brustad, E. M., Lemke, E. A., Schultz, P. G., and Deniz, A. A. (2008). A general and efficient method for the site-specific dual-labeling of proteins for single molecule fluorescence resonance energy transfer. J. Am. Chem. Soc. 130, 17664–17665. Bussell, R. Jr., and Eliezer, D. (2003). A structural and functional role for 11-mer repeats in a-synuclein and other exchangeable lipid binding proteins. J. Mol. Biol. 329, 763–778. Clamme, J. P., and Deniz, A. A. (2005). Three-color single-molecule fluorescence resonance energy transfer. Chemphyschem 6, 74–77. Crick, S. L., Jayaraman, M., Frieden, C., Wetzel, R., and Pappu, R. V. (2006). Fluorescence correlation spectroscopy shows that monomeric polyglutamine molecules form collapsed structures in aqueous solutions. Proc. Natl. Acad. Sci. USA 103, 16764–16769. Davidson, W. S., Jonas, A., Clayton, D. F., and George, J. M. (1998). Stabilization of asynuclein secondary structure upon binding to synthetic membranes. J. Biol. Chem. 273, 9443–9449. Dawson, P. E., and Kent, S. B. (2000). Synthesis of native proteins by chemical ligation. Annu. Rev. Biochem. 69, 923–960. Deniz, A. A., Dahan, M., Grunwell, J. R., Ha, T., Faulhaber, A. E., Chemla, D. S., Weiss, S., and Schultz, P. G. (1999). Single-pair fluorescence resonance energy transfer on freely diffusing molecules: Observation of Forster distance dependence and subpopulations. Proc. Natl. Acad. Sci. USA 96, 3670–3675.

Single-Molecule Fluorescence of IDPs

201

Deniz, A. A., Laurence, T. A., Beligere, G. S., Dahan, M., Martin, A. B., Chemla, D. S., Dawson, P. E., Schultz, P. G., and Weiss, S. (2000). Single-molecule protein folding: Diffusion fluorescence resonance energy transfer studies of the denaturation of chymotrypsin inhibitor 2. Proc. Natl. Acad. Sci. USA 97, 5179–5184. Deniz, A. A., Mukhopadhyay, S., and Lemke, E. A. (2008). Single-molecule biophysics: At the interface of biology, physics and chemistry. J. R. Soc. Interface 5, 15–45. Dirksen, A., and Dawson, P. E. (2008). Expanding the scope of chemoselective peptide ligations in chemical biology. Curr. Opin. Chem. Biol. 12, 760–766. Dunker, A. K., Lawson, J. D., Brown, C. J., Williams, R. M., Romero, P., Oh, J. S., Oldfield, C. J., Campen, A. M., Ratliff, C. M., Hipps, K. W., Ausio, J., Nissen, M. S., et al. (2001). Intrinsically disordered protein. J. Mol. Graph. Model. 19, 26–59. Ferreon, A. C., and Bolen, D. W. (2004). Thermodynamics of denaturant-induced unfolding of a protein that exhibits variable two-state denaturation. Biochemistry 43, 13357–13369. Ferreon, A. C., and Deniz, A. A. (2007). a-Synuclein multistate folding thermodynamics: Implications for protein misfolding and aggregation. Biochemistry 46, 4499–4509. Ferreon, A. C., Ferreon, J. C., Bolen, D. W., and Rosgen, J. (2007). Protein phase diagrams II: Nonideal behavior of biochemical reactions in the presence of osmolytes. Biophys. J. 92, 245–256. Ferreon, A. C., Gambin, Y., Lemke, E. A., and Deniz, A. A. (2009). Interplay of a-synuclein binding and conformational switching probed by single-molecule fluorescence. Proc. Natl. Acad. Sci. USA 106, 5645–5650. Georgieva, E. R., Ramlall, T. F., Borbat, P. P., Freed, J. H., and Eliezer, D. (2008). Membrane-bound alpha-synuclein forms an extended helix: Long-distance pulsed ESR measurements using vesicles, bicelles, and rodlike micelles. J. Am. Chem. Soc. 130, 12856–12857. Ha, T., Enderle, T., Ogletree, D. F., Chemla, D. S., Selvin, P. R., and Weiss, S. (1996). Probing the interaction between two single molecules: Fluorescence resonance energy transfer between a single donor and a single acceptor. Proc. Natl. Acad. Sci. USA 93, 6264–6268. Hamadani, K. M., and Weiss, S. (2008). Nonequilibrium single molecule protein folding in a coaxial mixer. Biophys. J. 95, 352–365. Helenius, A., McCaslin, D. R., Fries, E., and Tanford, C. (1979). Properties of detergents. Methods Enzymol. 56, 734–749. Hohng, S., Joo, C., and Ha, T. (2004). Single-molecule three-color FRET. Biophys. J. 87, 1328–1337. Jager, M., Michalet, X., and Weiss, S. (2005). Protein–protein interactions as a tool for sitespecific labeling of proteins. Protein Sci. 14, 2059–2068. Jakes, R., Spillantini, M. G., and Goedert, M. (1994). Identification of two distinct synucleins from human brain. FEBS Lett. 345, 27–32. Jensen, P. H., Nielsen, M. S., Jakes, R., Dotti, C. G., and Goedert, M. (1998). Binding of asynuclein to brain vesicles is abolished by familial Parkinson’s disease mutation. J. Biol. Chem. 273, 26292–26294. Jo, E., McLaurin, J., Yip, C. M., St George-Hyslop, P., and Fraser, P. E. (2000). a-Synuclein membrane interactions and lipid specificity. J. Biol. Chem. 275, 34328–34334. Joo, C., Balci, H., Ishitsuka, Y., Buranachai, C., and Ha, T. (2008). Advances in singlemolecule fluorescence methods for molecular biology. Annu. Rev. Biochem. 77, 51–76. Kapanidis, A. N., Laurence, T. A., Lee, N. K., Margeat, E., Kong, X., and Weiss, S. (2005). Alternating-laser excitation of single molecules. Acc. Chem. Res. 38, 523–533. Krishnan, R., and Lindquist, S. L. (2005). Structural insights into a yeast prion illuminate nucleation and strain diversity. Nature 435, 765–772.

202

Allan Chris M. Ferreon et al.

Kruger, R., Kuhn, W., Muller, T., Woitalla, D., Graeber, M., Kosel, S., Przuntek, H., Epplen, J. T., Schols, L., and Riess, O. (1998). Ala30Pro mutation in the gene encoding a-synuclein in Parkinson’s disease. Nat. Genet. 18, 106–108. Lee, N. K., Kapanidis, A. N., Wang, Y., Michalet, X., Mukhopadhyay, J., Ebright, R. H., and Weiss, S. (2005). Accurate FRET measurements within single diffusing biomolecules using alternating-laser excitation. Biophys. J. 88, 2939–2953. Lee, N. K., Kapanidis, A. N., Koh, H. R., Korlann, Y., Ho, S. O., Kim, Y., Gassman, N., Kim, S. K., and Weiss, S. (2007). Three-color alternating-laser excitation of single molecules: Monitoring multiple interactions and distances. Biophys. J. 92, 303–312. Lemke, E. A., Gambin, Y., Vandelinder, V., Brustad, E. M., Liu, H. W., Schultz, P. G., Groisman, A., and Deniz, A. A. (2009). Microfluidic device for single-molecule experiments with enhanced photostability. J. Am. Chem. Soc. 131, 13610–13612. Li, H. T., Ying, L. M., Green, J. J., Balasubramanian, S., and Klenerman, D. (2003). Ultrasensitive coincidence fluorescence detection of single DNA molecules. Anal. Chem. 75, 1664–1670. Lipman, E. A., Schuler, B., Bakajin, O., and Eaton, W. A. (2003). Single-molecule measurement of protein folding kinetics. Science 301, 1233–1235. Magde, D., Elson, E. L., and Webb, W. W. (1974). Fluorescence correlation spectroscopy II. An experimental realization. Biopolymers 13, 29–61. Michalet, X., Weiss, S., and Jager, M. (2006). Single-molecule fluorescence studies of protein folding and conformational dynamics. Chem. Rev. 106, 1785–1813. Moerner, W. E., and Orrit, M. (1999). Illuminating single molecules in condensed matter. Science 283, 1670–1676. Mukhopadhyay, S., and Deniz, A. A. (2007). Fluorescence from diffusing single molecules illuminates biomolecular structure and dynamics. J. Fluoresc. 17, 775–783. Mukhopadhyay, S., Krishnan, R., Lemke, E. A., Lindquist, S., and Deniz, A. A. (2007). A natively unfolded yeast prion monomer adopts an ensemble of collapsed and rapidly fluctuating structures. Proc. Natl. Acad. Sci. USA 104, 2649–2654. Muralidharan, V., and Muir, T. W. (2006). Protein ligation: An enabling technology for the biophysical analysis of proteins. Nat. Methods 3, 429–438. Orte, A., Clarke, R., Balasubramanian, S., and Klenerman, D. (2006). Determination of the fraction and stoichiometry of femtomolar levels of biomolecular complexes in an excess of monomer using single-molecule, two-color coincidence detection. Anal. Chem. 78, 7707–7715. Pace, C. N. (1986). Determination and analysis of urea and guanidine hydrochloride denaturation curves. Methods Enzymol. 131, 266–280. Patino, M. M., Liu, J. J., Glover, J. R., and Lindquist, S. (1996). Support for the prion hypothesis for inheritance of a phenotypic trait in yeast. Science 273, 622–626. Polymeropoulos, M. H., Lavedan, C., Leroy, E., Ide, S. E., Dehejia, A., Dutra, A., Pike, B., Root, H., Rubenstein, J., Boyer, R., Stenroos, E. S., Chandrasekharappa, S., et al. (1997). Mutation in the a-synuclein gene identified in families with Parkinson’s disease. Science 276, 2045–2047. Reynolds, J. A., and Tanford, C. (1970). Binding of dodecyl sulfate to proteins at high binding ratios. Possible implications for the state of proteins in biological membranes. Proc. Natl. Acad. Sci. USA 66, 1002–1007. Rosgen, J., and Hinz, H. J. (2003). Phase diagrams: A graphical representation of linkage relations. J. Mol. Biol. 328, 255–271. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Sandal, M., Valle, F., Tessari, I., Mammi, S., Bergantino, E., Musiani, F., Brucale, M., Bubacco, L., and Samori, B. (2008). Conformational equilibria in monomeric a-synuclein at the single-molecule level. PLoS Biol. 6, e6.

Single-Molecule Fluorescence of IDPs

203

Schuler, B., and Eaton, W. A. (2008). Protein folding studied by single-molecule FRET. Curr. Opin. Struct. Biol. 18, 16–26. Schuler, B., Lipman, E. A., and Eaton, W. A. (2002). Probing the free-energy surface for protein folding with single-molecule fluorescence spectroscopy. Nature 419, 743–747. Serio, T. R., and Lindquist, S. L. (2000). Protein-only inheritance in yeast: Something to get [PSIþ]-ched about. Trends Cell Biol. 10, 98–105. Serio, T. R., Cashikar, A. G., Kowal, A. S., Sawicki, G. J., Moslehi, J. J., Serpell, L., Arnsdorf, M. F., and Lindquist, S. L. (2000). Nucleated conformational conversion and the replication of conformational information by a prion determinant. Science 289, 1317–1321. Sharon, R., Goldberg, M. S., Bar-Josef, I., Betensky, R. A., Shen, J., and Selkoe, D. J. (2001). a-Synuclein occurs in lipid-rich high molecular weight complexes, binds fatty acids, and shows homology to the fatty acid-binding proteins. Proc. Natl. Acad. Sci. USA 98, 9110–9115. Sherman, E., and Haran, G. (2006). Coil-globule transition in the denatured state of a small protein. Proc. Natl. Acad. Sci. USA 103, 11539–11543. Shorter, J., and Lindquist, S. (2005). Prions as adaptive conduits of memory and inheritance. Nat. Rev. Genet. 6, 435–450. Spillantini, M. G., Schmidt, M. L., Lee, V. M., Trojanowski, J. Q., Jakes, R., and Goedert, M. (1997). a-Synuclein in Lewy bodies. Nature 388, 839–840. Stryer, L., and Haugland, R. P. (1967). Energy transfer: A spectroscopic ruler. Proc. Natl. Acad. Sci. USA 58, 719–726. Tompa, P. (2005). The interplay between structure and function in intrinsically unstructured proteins. FEBS Lett. 579, 3346–3354. Torres, T., and Levitus, M. (2007). Measuring conformational dynamics: A new FCS– FRET approach. J. Phys. Chem. B 111, 7392–7400. Trexler, A. J., and Rhoades, E. (2009). a-Synuclein binds large unilamellar vesicles as an extended helix. Biochemistry 48, 2304–2306. Ulmer, T. S., Bax, A., Cole, N. B., and Nussbaum, R. L. (2005). Structure and dynamics of micelle-bound human a-synuclein. J. Biol. Chem. 280, 9595–9603. Uptain, S. M., and Lindquist, S. (2002). Prions as protein-based genetic elements. Annu. Rev. Microbiol. 56, 703–741. Uversky, V. N., Gillespie, J. R., and Fink, A. L. (2000). Why are ‘‘natively unfolded’’ proteins unstructured under physiologic conditions? Proteins 41, 415–427. Vandelinder, V., Ferreon, A. C., Gambin, Y., Deniz, A. A., and Groisman, A. (2009). Highresolution temperature-concentration diagram of alpha-synuclein conformation obtained from a single Forster resonance energy transfer image in a microfluidic device. Anal. Chem. 81, 6929. Veldhuis, G., Segers-Nolten, I., Ferlemann, E., and Subramaniam, V. (2009). Singlemolecule FRET reveals structural heterogeneity of SDS-bound a-synuclein. Chembiochem 10, 436–439. Walter, N. G., Huang, C. Y., Manzo, A. J., and Sobhy, M. A. (2008). Do-it-yourself guide: How to use the modern single-molecule toolkit. Nat. Methods 5, 475–489. Webb, W. W. (2001). Fluorescence correlation spectroscopy: Inception, biophysical experimentations, and prospectus. Appl. Opt. 40, 3969–3983. Weinreb, P. H., Zhen, W., Poon, A. W., Conway, K. A., and Lansbury, P. T., Jr. (1996). NACP, a protein implicated in Alzheimer’s disease and learning, is natively unfolded. Biochemistry 35, 13709–13715. Wickner, R. B. (1994). [URE3] as an altered URE2 protein: Evidence for a prion analog in Saccharomyces cerevisiae. Science 264, 566–569. Wright, P. E., and Dyson, H. J. (1999). Intrinsically unstructured proteins: Re-assessing the protein structure–function paradigm. J. Mol. Biol. 293, 321–331.

204

Allan Chris M. Ferreon et al.

Wright, P. E., and Dyson, H. J. (2009). Linking folding and binding. Curr. Opin. Struct. Biol. 19, 31–38. Xie, J., and Schultz, P. G. (2006). A chemical toolkit for proteins—an expanded genetic code. Nat. Rev. Mol. Cell Biol. 7, 775–782. Yu, J., Malkova, S., and Lyubchenko, Y. L. (2008). a-Synuclein misfolding: Single molecule AFM force spectroscopy study. J. Mol. Biol. 384, 992–1001. Zhuang, X. (2005). Single-molecule RNA science. Annu. Rev. Biophys. Biomol. Struct. 34, 399–414.

C H A P T E R

E L E V E N

Measuring the Energetic Coupling of Tertiary Contacts in RNA Folding using Single Molecule Fluorescence Resonance Energy Transfer Max Greenfeld*,† and Daniel Herschlag*,† Contents 206 207 209 210 210 212 212 215 217 217 219 219

1. 2. 3. 4.

Introduction Thermodynamic Cooperativity Overview Measuring Folding Equilibrium in RNA Designing an smFRET Experiment to Measure Cooperativity 4.1. Identification of tertiary contacts 4.2. Knocking out tertiary contacts 4.3. Designing single molecule constructs 4.4. Validating a new single-molecule construct 4.5. Measurement of cooperativity 5. Additional Comments Acknowledgments References

Abstract Tertiary contacts are critical to stabilizing the folded conformations of structured RNAs. In some cases, these contacts have been shown to interact with positive cooperativity. Measuring the energetic coupling of tertiary contact formation is among the most basic physical characterizations of a structured RNA. With proper experimental design, single-molecule fluorescence resonance energy transfer (smFRET) allows the rigorous determination of the energetic coupling. This chapter aims to provide a general experimental approach to measuring the energetic coupling of tertiary contacts, using smFRET.

* Department of Chemical Engineering, Stanford University, Stanford, California, USA Department of Biochemistry, Stanford University, Stanford, California, USA

{

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72009-7

#

2010 Elsevier Inc. All rights reserved.

205

206

Max Greenfeld and Daniel Herschlag

1. Introduction Many RNAs have functions that require the formation of a welldefined tertiary structure, and new RNA structures are being solved at an exciting pace. These structures provide windows into the intricate conformations that RNAs can adopt. Dissecting the energetics of how a limited set of functional folded conformations is stabilized over the vast ensemble of unfolded or misfolded conformations remains central to understanding the fundamental physical properties of RNA (Cruz and Westhof, 2009; Li et al., 2008; Noller, 2005). Tertiary contacts are critical to maintaining the overall folds of structured RNAs. Yet dissecting the energetic coupling between the multiple tertiary contacts present in large structured RNAs has been difficult. While the thermodynamic contributions of individual tertiary contacts in stabilizing larger structured RNAs can have purely additive contributions or cooperative contributions, few measurements accurately determine these energetics. Measuring energetic coupling in RNA has been particularly challenging. This situation results from both experimental and conceptual difficulties that arise from RNA being a polyelectrolyte (Das et al., 2005; Draper, 2004). In the limited case where an accurate quantitative measure of tertiary energetic coupling was determined, positive cooperativity was shown to provide a significant energetic contribution to folding (Sattin et al., 2008). Single-molecule fluorescence resonance energy transfer (smFRET) has been used to study conformations of a multitude of macromolecules, and some of the first applications were to structured RNAs (Ha et al., 1999; Zhuang et al., 2000). Currently, smFRET is a powerful and widely used tool in modern biophysics. Among the great strengths of smFRET is its ability to directly measure equilibria over a large dynamic range. This property enables RNA molecules with substantial differences in stabilities to be compared in solutions with identical ionic composition. This ability circumvents spurious assumptions made when using a Hill analysis to assess RNA energetics (Das et al., 2005; Draper, 2004). The application of smFRET for measuring tertiary contact cooperativity in the P4-P6 domain of the Tetrahymena Group I intron (P4-P6) provided the first accurate measurement of positive cooperativity in P4-P6 folding and revealed that cooperativity depends on the ionic conditions (Sattin et al., 2008). Although this was the first measurement of RNA cooperativity using smFRET, the approach taken does not rely on the details of P4-P6 as a model system. This chapter provides a general approach to measuring the energetic coupling of tertiary contacts in structured RNAs. We first introduce the types of energetic couplings that can occur and explain why most techniques do not provide accurate thermodynamic measures of these energetics.

Measuring the Energetic Coupling of Tertiary Contacts in RNA Folding

207

The use of smFRET as a general technique for accurately measuring the coupling in tertiary contacts is then presented. As there are multiple methods available for making RNA constructs suitable for smFRET and multiple strategies for measuring smFRET, references to other resources are provided for those details. Discussion instead focuses on the steps that are required for making meaningful energetic comparisons.

2. Thermodynamic Cooperativity Overview Among the most basic characterization of any macromolecular binding or conformational change is determination of the equilibrium constant of that event. However, a single-equilibrium constant does not give mechanistic insight into the complex reaction being studied. A more complete thermodynamic understanding of a complex equilibrium can often be obtained by dividing the reaction into smaller discrete equilibria. The parsing of a complex equilibrium has two primary aims: (1) to identify the energetic importance of observed structural interactions and (2) to evaluate how smaller discrete energetic components combine to contribute to the overall equilibrium of a reaction. The latter aim is the focus of this chapter. When the free energy of a complex chemical equilibrium is evaluated by measuring the free energy of simpler components, the individual components might or might not sum to the overall free energy. The question of whether the energetic components of a particular equilibrium are additive or cooperative is a recurring question in macromolecule studies. A particularly lucid discussion of how energetic components relate to the overall energetics is presented in the classic paper by Jencks (1981). The basic approach to measuring the energetic coupling of a reaction is shown in Fig. 11.1. In this generic thermodynamic diagram, Koverall is the

KA

KB

Koverall

KA

KB

Figure 11.1 Generic thermodynamic cycle for assessing cooperativity of two binding sites. In the case of RNA folding, the square and oval would represent two distinct tertiary contacts, filled shapes are indicative of native tertiary contacts, while the outlines only are indicative of ablated tertiary contacts. Figure adapted from Jencks (1981).

208

Max Greenfeld and Daniel Herschlag

overall equilibrium constant. The overall reaction can be broken into four distinct equilibria, KA, K 0 A, KB, and K 0 B, representing the piecewise completion of the overall reaction. The question becomes, how do the free energies of the individual reaction steps compare to one another. As shown in Fig. 11.2, there are three possible ways in which the thermodynamic cycle of Fig. 11.1 can be completed. In the case of no cooperativity (i.e., energetic additivity), there is no free energy difference depending on the order in which the tertiary contacts are formed (DGa ¼ DG 0 a and DGb ¼ DG 0 b). In the case of positive cooperativity, the free energy gained upon formation of the second contact is greater than if the same contact were formed first (DGa > DG 0 a and DGb > DG 0 b). And in the case of negative cooperativity, the free energy gained upon formation of the second contact is less than if the same contact were formed first (DGa < DG 0 a and DGb < DG 0 b). It is convenient to express the relationships depicted in the free energy diagram of Fig. 11.2 by Eq. (11.1). 0

DGcoop ¼ RT ln

0

KA KB ¼ RT ln : KA KB

ð11:1Þ

This representation highlights that DGcoop is a function solely arising from the order of contact formation. Whereas DGcoop is a measurement that can be made for diverse systems, the actual value and mechanistic origins can vary greatly (Mammen et al., 1998; Williamson, 2008). In practice, there are many considerations that arise when trying to experimentally realize the thermodynamic diagram in Fig. 11.1. Most basically, I.

No cooperativity

II. Positive cooperativity

DGA

DGA

DG B

DG

DGB DGB

 DGA  D GA

DGoverall = DGA + DGB DGA = DGA; DGB = DGB DGcoop = 0

III. Negative cooperativity

DGA DGB

DGB

DGB

DGoverall

 DGA

DGoverall = DGA + DGB + DGcoop

DGoverall = DGA + DGB + DGcoop DGcoop = DGA – DGA = DGB - DGB < 0 DGcoop = DGA – DGA = DGB - DGB > 0

Figure 11.2 Free energy diagram indicating the possible solutions of the thermodynamic cycle in Fig. 11.1. (I) No cooperativity arises when the sum of the free energies of each individual contact is equal to the overall free energy. (II) Positive cooperativity arises when the overall free energy is greater than the sum of the individual components. (III) Negative cooperativity arises when the overall free energy is less than the sum of the individual components.

Measuring the Energetic Coupling of Tertiary Contacts in RNA Folding

209

cleanly ablating contacts that maintain well-defined intermediate conformations can be difficult, and measurements of K 0 A and K 0 B can be difficult or impossible, so DGcoop is usually calculated in terms of Koverall, as shown in Eq. (11.2). DGcoop ¼ RT ln

Koverall : KA KB

ð11:2Þ

Even rigorous measurement of Koverall, KA, and KB can be difficult. Dealing with each of these measurements is critical to successfully measuring cooperativity in RNA folding.

3. Measuring Folding Equilibrium in RNA Each RNA residue has a formal negative charge so that any RNA more than a few residues in length is a polyelectrolyte. It has been estimated that the columbic repulsion, in the absence of counterions, upon folding of the approximately 400-nucleotide Tetrahymena Group I intron is 103 kcal/ mol (Bai et al., 2005). However, RNA in solution always exists in the presence of counterions that form an ion atmosphere surrounding the molecule (Bai et al., 2007). These counterions attenuate the electrostatic repulsion arising from the phosphate backbone and consequently affect the stability of the folded and unfolded states (Das et al., 2005; Draper, 2004; Grilley et al., 2006). Making correct energetic measurements of RNA requires comparisons of equilibria under identical ionic conditions. Models that assume only specific ion-binding stoichiometries are not energetically correct, as is extensively discussed elsewhere (Das et al., 2005; Draper, 2004). It is possible to measure the equilibrium of an RNA molecule by titrating a solution component (e.g., Mg2þ), monitoring a variable such as the radius of gyration or the protection of a tertiary contact from chemical modification, and fitting the results assuming a two-state model. However, RNA folding is not, in general, two state in nature, and there is limited accuracy of empirical extrapolations away from the well-determined midpoints of such sigmoidal curves. These limitations prevent accurate energetic comparisons of molecules with significantly different folding midpoints. The limitations associated with typical methods of determining an RNA folding equilibrium can be overcome with smFRET. Many RNAs that have been examined with smFRET have a FRET signal that is dominated by a single high-FRET and single-low-FRET state. This clear distinction of states allows the direct measurement of the equilibrium constant of a molecule without extrapolation or assumptions of a two-state model. Moreover, smFRET is accurate over a wide dynamic range and is hence

210

Max Greenfeld and Daniel Herschlag

suitable for comparing molecules with large stability differences. Despite these advantages, the time required to set up and conduct smFRET measurements should be considered before initiating the measurements highlighted in this chapter.

4. Designing an smFRET Experiment to Measure Cooperativity The precipitous rise in smFRET publications is inevitably related to the relative ease with which smFRET experiments can be designed and measurements carried out ( Joo et al., 2008). Indeed, it is possible for an investigator without a specialized interest in smFRET to design and carry out an enlightening set of experiments. And although interpreting smFRET measurements requires careful experimental design at multiple steps and successful completion of these steps can be difficult, the added insight of smFRET experiments can be well worth the effort. Experimental measurement of the energetic coupling of tertiary contacts in RNA requires realizing the thermodynamic scheme in Fig. 11.1 for an actual RNA. Figure 11.3 highlights the five major steps that were required for measuring tertiary contact cooperativity in the P4-P6 domain of the Tetrahymena Group I ribozyme. The discrete structure of RNA tertiary contacts suggests that there is nothing unique to P4-P6 that enabled the measurements by Sattin et al. (2008). This section details the experimental approaches and decisions that must be made during each of the five steps in Fig. 11.3. There were a number of key controls and design choices that were necessary for the measurements made by Sattin et al. (2008) that will be required for most if not all structured RNAs studied in an analogous fashion. The methods presented are biased by the experimental decisions that were successful in the approaches taken to measure P4-P6 tertiary contact cooperativity, but the aim of this section is to present the general experimental design process. Wherever possible, alternative approaches, controls, and methods are discussed.

4.1. Identification of tertiary contacts There are multiple ways in which tertiary contacts can be identified in a structured RNA. Phylogenetic analysis, chemical footprinting, site-directed mutagenesis, and crystallography have all been used to identify or confirm the presence and location of tertiary contacts. In practice, the first two methods are synergistic and common for identifying tertiary contacts. The increased proficiency with which RNAs can be crystallized can more commonly provide aid in the design of smFRET experiments. However, high-resolution structural

2) Knock out 3˚ 3) Design single molecule 4) Validate single contacts constructs molecule constructs

1) Identify 3˚contacts of interest Structure

A A

G U G U G C A G AA

U G U G C G G U CU

Cy3

1.0

0.8

0.8

0.6

0.6

0.4

0.4

0.2

0.2

0.0 0.1

1

10

0.0 100

2+

Footprinting Mg2+

U C G AA U A AU

A G U A

U C U UU U U U

[Mg ] (mM)

Cy5

Compare bulk and single molecule results

Normalized frequency

AA

1) Mutagenesis 2) Verify with footprinting

1.0

Fraction protected

GC

Fraction high FRET

1.0

Biotin

Phylogeny

A

5) Measure equilibrium with smFRET ΔTetraloop

0.5 0 1.0

WT

0.5 0 1.0

ΔMetal core

0.5 0 0.0 0.5 1.0 FRET level

Figure 11.3 Key steps required to measure tertiary contact cooperativity in RNA using smFRET. (1) Use the available structural information to identify the key tertiary contacts of interest. (2) Make constructs that knock out the tertiary contacts of interest. (3) Construct the wild type and two tertiary contact mutant molecules with FRET pairs and surface attachment. (4) Verify the single molecule constructs recapitulate the expected behavior of the unlabeled molecules. (5) Directly determine the equilibrium of the molecules under identical ionic conditions.

212

Max Greenfeld and Daniel Herschlag

information is not required. Additionally, it is not necessary to identify all of the tertiary contacts in an RNA. The smFRET technique can be used for the rigorous thermodynamic characterization of any two contacts that influence the same conformational change.

4.2. Knocking out tertiary contacts Constructing mutant RNAs that have the tertiary contacts of interest knocked out is straightforward. Point mutations of highly conserved residues or key contacts observed in crystal structures are particularly good targets to mutate. Alternatively, simply changing a few residues to uridines, base pairing, or deleting bulged residues in a tertiary contact is usually sufficient to knock out the contact. It is important to assess the effects of mutations on the overall fold of the RNA. It is possible to make tertiary contact mutations that weaken and do not completely ablate the tertiary contacts. Alternatively, it is possible that the mutation is so severe that multiple tertiary contacts are affected simultaneously. Therefore, it is important to verify with a technique like chemical footprinting that the contacts have been locally removed (Takamoto et al., 2004). It is also valuable to compare the effects of multiple mutations aimed to knock out the same interaction. The limitations of site-directed mutagenesis will affect the interpretation of downstream energetic measurements, so it is important to have multiple strains of evidence that define the behavior of all molecules being studied.

4.3. Designing single molecule constructs Nucleic acids have been among the most amenable molecules for study by single-molecule spectroscopy. This fact arises for a number of reasons. There are vast and accessible synthetic resources for constructing modified nucleotides. Watson/Crick base pairing rules at times allow for the rational engineering of simple molecular structure. And an underappreciated advantage is that nucleic acids are less prone to nonspecific surface absorption than are proteins. Taken together, these properties provide significant flexibility and creativity with the design and construction of RNAs for single-molecule spectroscopy. When designing an RNA construct for study with smFRET, there are three major design considerations: (1) the location of the donor and acceptor fluorophore; (2) the specific FRET pair to be used (e.g., Cy3-Cy5) and the method of fluorophore incorporation; and (3) the location of the surface tether and the choice of attachment method. There is no a priori method for sorting through all the possibilities of the three steps. Indeed at the beginning of a project, it is advisable to try multiple strategies with the intent of

213

Measuring the Energetic Coupling of Tertiary Contacts in RNA Folding

narrowing the possibilities once the utility of each is tested in a specific system. The choice of dye location is most critical to downstream quantification and interpretation of the measured FRET signal. The FRET pairs should be placed in locations that monitor a conformational change mediated by the two tertiary contacts being studied. Although techniques sensitive to the unfolded structure of an RNA such as SAXS or native gels can be used to infer the major conformational changes, they do not identify positions for labeling that are far apart in the unfolded state and nonperturbing in the folded state. Crystal structures and footprinting studies provide useful information for this design. Ideally, the locations should be chosen to maximize the difference between the high- and low-FRET values of the conformational change being monitored and to not interfere with important structural interactions. In the case of P4-P6, detailed structural information provided significant constraints on the location of FRET pairs ( just above the tetraloop and just below the tetraloop receptor). This placement monitors a large conformational change mediated by a hinge region distal to the FRET pairs and the two tertiary contacts in P4-P6. The specific location of the dyes, at the level of a single nucleotide, was determined by the structural information and the constraints of the labeling scheme used. In most cases, fluorophores are attached to an RNA by derivatizing a primary amino group. This approach requires specifically incorporating a modified base or a terminal amino modified linker. There are a number of commercially available fluorophores that are commonly used in smFRET studies. A current summary of the commonly used dyes is included in Table 11.1 of Roy et al. (2008). As new fluorophores become commercially available this list will grow. Currently, there are four methods available for site-specifically modifying RNAs, and these methods and their strengths and limitations are listed in Table 11.1 herein. Table 11.1 Strategies for incorporating FRET pairs into RNA

Fully synthetic RNA Splinted ligation Base pairing 20 -OH ligation

Minimal perturbation

Length limitation

Covalent attachment

Many steps

Yes

 50 nt

Yes

No

Yes

No

Yes

Yes

No No

No No

No Yes

No No

214

Max Greenfeld and Daniel Herschlag

Using fully synthetic RNA is the easiest method for making singlemolecule constructs (Hodak et al., 2005; Pereira et al., 2008; Tan et al., 2003; Zhuang et al., 2002). Indeed if dyes are to be placed on the 50 - or 30 -end of an RNA, it is possible to purchase molecules with the dyed attached during synthesis, eliminating the need to postsynthetically label the RNA. However, there are considerable length limitations with fully synthetic RNAs. The use of ligation to introduce short synthetic RNAs, which can be derivatized with a fluorophore prior to incorporation into the complete RNA, removes the length limitation of using fully synthetic RNAs. This technique is gaining in prevalence, although it can be time consuming to implement (Akiyama and Stone, 2009; Sattin et al., 2008; Solomatin and Herschlag, 2009; Stone et al., 2007). The use of base-pairing rules to hybridize fluorescently labeled oligonucleotides to 50 - or 30 -extensions of an RNA or to loop structures have been used successfully in a number of cases (Dorywalska et al., 2005; Smith et al., 2005, 2008; Zhuang et al., 2000). Despite the potential ease of this approach, extensions are restricted to the ends of RNAs and hybridization to internal loops can perturb the RNA’s structure. Finally, a method has recently been introduced for site-specifically ligating an RNA oligonucleotide to the 20 -OH of adenosines in an RNA sequence (Baum and Silverman, 2007). This technique has not yet been used for smFRET studies. If the rather bulky labels tend not to perturb the molecules, as suggested by the initial study, this technique could have broad applications. The choice of labeling methods will largely depend on the molecule being studied. For instance, the design of P4-P6 for smFRET studies required all three of the available techniques to be considered at the time. More information is provided in recent Methods in Enzymology chapters (Akiyama and Stone, 2009; Solomatin and Herschlag, 2009). Compared with the first two design steps, there are a limited set of choices in the position and location of the surface tether. To the best of the authors’ knowledge, a biotin/streptavidin linkage has been used universally as the final surface attachment in smFRET studies of nucleic acids. Alternative schemes such as incorporating an RNA-binding protein motif into an RNA and using a surface derivatized with the RNA-binding protein could be practical in some situations, such as very large RNAs but has yet to be implemented. To rule out the potential of surface affects, molecules have been confined in tethered lipid vesicles (Boukobza et al., 2001; Okumus et al., 2004). This approach is typically used as a secondary attachment scheme for a limited set of control experiments. However, it could be used to avoid the need to design a covalently attached surface tether in an RNA construct. Surface tethers are typically placed at the 50 - or 30 -end of an RNA. If synthetic RNA is used on either the 50 - or 30 -end of an RNA, a terminal biotin-derivatized base can be incorporated for surface attachment. Alternatively, tails of extra bases, typically 20–30 nucleotides, can be incorporated.

Measuring the Energetic Coupling of Tertiary Contacts in RNA Folding

215

In this situation, a biotin-derivatized oligonucleotide complementary to the tail can be used for surface attachment. A potential advantage of using a 50 - or 30 -tail is that there is space between the fluctuating RNA and the site of attachment. However, the ease of direct covalent attachment of the biotin has been used without trouble by a number of investigators. It should also be noted that RNA can be internally labeled with biotin, either during synthesis or via ligation to the 20 -OH, although these strategies have not seen use in smFRET studies.

4.4. Validating a new single-molecule construct Gaining expertise with the instrumentation required for carrying out smFRET measurements is an involved process and has been discussed elsewhere ( Joo and Ha, 2008; Roy et al., 2008). The confidence with which a new construct can be validated is dependent on past successes. It is important for an experimentalist to validate their protocols by redetermining kinetic and thermodynamic properties of molecules that have been previously studied. There are many variables that affect the quality of an smFRET measurement, only one of which is the intrinsic behavior of the molecule being studied. Trouble in reproducing values could result from technical differences such as poor signal-to-noise, short trace length or different criteria for analyzing molecules. If differences arise, it is necessary to resolve the discrepancies. There is significant variability in the behavior of single-molecule constructs. Figure 11.4A shows a trace of wild-type P4-P6, which by standard measures is a well-behaved molecule. P4-P6 has stable high- and lowFRET states that do not change with time. The FRET pairs monitor a large conformational change, which provides significant separation between the high- and low-FRET states. And under the correct experimental conditions (Sattin et al., 2008), P4-P6 produces traces with high signal-to-noise and long lifetimes. However, this behavior cannot be expected of all molecules. Mutations that significantly destabilize the folded conformation are likely to produce poorer quality traces. This is the case with the tetraloop knockout of P4-P6, which has the lowest stability of the three constructs being compared and has a different high-FRET state value than the wild type and metal core knockout (Fig. 11.4B). In the case of P4-P6, the FRET difference was consistent with conformational changes known to occur in the P5abc region of the molecule. For P4-P6, having an understanding of why the high-FRET value changed for one of the constructs was important for ensuring that the correct conformations were being monitored. Alternatively, concerns that the mutations were significantly altering the folded conformations would persist, undermining the assumptions required of the thermodynamic diagram in Fig. 11.1.

216

Max Greenfeld and Daniel Herschlag

B 1.0

DTetraloop

Normalized frequency

0.5 A

FRET

1 0.5 0 10

20

30 40 Time (s)

50

0 1.0

WT

0.5 0 1.0

DMetal core

60 0.5 0 0.0

0.5

1.0

FRET level

Figure 11.4 Characteristic data from the measurement of P4-P6 cooperativity. (A) Representative trace for WT P4-P6. P4-P6 has well-defined high- and lowFRET states. Using wide field imaging on a custom built prism based total internal reflection fluorescence microscope (see Bartley et al., 2003; Zhuang et al., 2000 for description), a high signal-to-noise of 4 and stable donor and acceptor intensities are obtainable for traces well over 1 min. (B) Cumulative FRET histograms for hundreds of molecules of each construct. Histograms indicate the molecules can be thermodynamically described with two states. As such the equilibrium is the ratio of the areas under the two peaks in the FRET histograms. Panel (B) reprinted with permission from Sattin et al. (2008).

The most basic quantitative analysis of smFRET measurements is the determination of a cumulative FRET histogram. Examples of these are shown for the wild type and two mutant forms of P4-P6 in Fig. 11.4B. This analysis simply involves making a histogram of FRET traces from many molecules (typically hundreds). In most cases, the histogram consists of two peaks that are well fitted by a Gaussian distribution (an approximation since FRET values outside the range of 0–1 have no physical meaning—the actual peak shapes are governed by the b-distribution (McKinney et al., 2006)). If this is not the case and there are more than two equilibrium states of the molecule being studied, the analysis of cooperativity is valid only if changes in the distributions are limited to two of the peaks. The equilibrium between any two peaks is simply the ratio of the area of the two peaks as given by Eq. (11.3) High

Koverall ¼

AFRET ; ALow FRET High

ð11:3Þ

where Koverall is the equilibrium constant, AFRET area of the high-FRET peak, and ALow FRET is the area of the low-FRET peak.

Measuring the Energetic Coupling of Tertiary Contacts in RNA Folding

217

Analysis of single-molecule data can be immensely complex because of the heterogeneity that is revealed by removing ensemble averaging. Almost every RNA that has been studied with smFRET has demonstrated this complex underlying behavior, making it quite likely that any new constructs will behave analogously. Although the molecular origins of this type of behavior remain largely unexplained (Ditzler et al., 2008; Huang et al., 2009; Korennykh et al., 2007), the analysis of heterogeneity has recently become experimentally tractable (Elenko et al., 2009; Solomatin et al., 2010). Fortunately, the measurement of cooperativity is a strictly thermodynamic measurement that assumes that all molecules are covalently identical and unperturbed by the surface or dyes. In this instance, heterogeneity should not affect the results. For P4-P6, it was possible to test this assumption by comparing the equilibrium behavior of P4-P6 determined by smFRET to that determined by bulk hydroxyl radical footprinting (see the supplemental information of Sattin et al. (2008)). The fact that the two measurements matched well was an important validation of the accuracy of the cooperativity measurement.

4.5. Measurement of cooperativity At the successful completion of the first four steps, the actual equilibrium measurements are straightforward. The only strict requirement is that the three-way comparisons between the two mutant and wild-type molecules be made under identical ionic conditions. For P4-P6, a number of different ionic conditions were attempted before finding a single condition where the equilibrium constant for each of the molecules could be accurately determined. Some searching for conditions where accurate measurements are made can be expected for other constructs. Once measurements are made, the thermodynamic cycle shown in Fig. 11.1 can be completed usingEqs. (11.1) and (11.2). The results of the previous P4-P6 analysis are shown in Fig. 11.5. The measured DGcoop of 2.3 kcal/mol (note that the sign is consistent with the convention used here) likely plays a significant role in stabilizing the folded structure of P4-P6. This value differs significantly from less rigorous measurements, affirming the importance of smFRET as a technique for correctly measuring cooperativity in RNA tertiary structure.

5. Additional Comments The generality with which smFRET can be used to measure the energetic coupling of RNA tertiary contacts should provide an important tool for the study of RNA structure. In addition to extending cooperativity measurements to RNAs other than P4-P6, future work will hopefully give

218

Max Greenfeld and Daniel Herschlag

KMC = 0.1

KTL = 120

Cy3

Cy5

Mg2+

FRET = 0.5 Kfold = 12

Cy3

Mg2+

Cy5

KTL = 0.5

Cy3

KMC = 24 Mg

2+

Cy5

FRET = 0.1 FRET = 0.8

Cy3 Cy5 FRET = 0.8

Figure 11.5 Thermodynamic cycle for tertiary contact cooperativity in P4-P6. The cooperativity was measured to be  3.2  0.2 kcal/mol in 10 mM Mg2þ, 200 mM NaCl, 50 mM Na-MOPS, pH 7.0, and 22  C. Reprinted with permission from Sattin et al. (2008).

insight to the influence of the ionic conditions and the mechanistic implications of heterogeneity on cooperativity. As further measurements are made on different RNAs and under increasingly diverse ionic conditions, it is likely that interpretable trends will arise and new properties suggested. Already with P4-P6, it is clear that cooperativity depends on the ionic conditions, which is an additional energetic coupling that is not yet understood. Heterogeneity as observed by smFRET suggests the presence of longlived conformational differences among RNA molecules. The possibility of measuring cooperativity differences among conformationally distinct structures is an exciting and uniquely single-molecule measurement. It will be a worthy experimental challenge to develop a model system where the thermodynamic diagram in Fig. 11.1 can be recapitulated for a single molecule in a distinct conformational state. We hope that this chapter provides some insight into the experimental details important for a thorough analysis of smFRET cooperativity measurements. There are certainly many additional opportunities for smFRET to provide detailed biophysical insight into RNA structure. Considering the significant insights already provided, it seems likely that this technique will expand our current understanding of RNA biophysics and contribute insights into the complex functions carried out by RNA.

Measuring the Energetic Coupling of Tertiary Contacts in RNA Folding

219

ACKNOWLEDGMENTS We thank Bernie D. Sattin for help in preparing the figures. Funding was provided by NIH program project grant P01-GM-66275 to D. H.

REFERENCES Akiyama, B. M., and Stone, M. D. (2009). Assembly of complex RNAs by splinted ligation. Methods Enzymol. 469, 27–46. Bai, Y., Das, R., Millett, I., Herschlag, D., and Doniach, S. (2005). Probing counterion modulated repulsion and attraction between nucleic acid duplexes in solution. Proc. Natl. Acad. Sci. USA 102, 1035–1040. Bai, Y., Greenfeld, M., Travers, K. J., Chu, V. B., Lipfert, J., Doniach, S., and Herschlag, D. (2007). Quantitative and comprehensive decomposition of the ion atmosphere around nucleic acids. J. Am. Chem. Soc. 129, 14981–14988. Bartley, L. E., Zhuang, X. W., Das, R., Chu, S., and Herschlag, D. (2003). Exploration of the transition state for tertiary structure formation between an RNA helix and a large structured RNA. J. Mol. Biol. 328, 1011–1026. Baum, D. A., and Silverman, S. K. (2007). Deoxyribozyme-catalyzed labeling of RNA. Angew. Chem. Int. Ed. 46, 3502–3504. Boukobza, E., Sonnenfeld, A., and Haran, G. (2001). Immobilization in surface-tethered lipid vesicles as a new tool for single biomolecule spectroscopy. J. Phys. Chem. B 105, 12165–12170. Cruz, J. A., and Westhof, E. (2009). The dynamic landscapes of RNA architecture. Cell 136, 604–609. Das, R., Travers, K. J., Bai, Y., and Herschlag, D. (2005). Determining the Mg2þ stoichiometry for folding an RNA metal ion core. J. Am. Chem. Soc. 127, 8272–8273. Ditzler, M. A., Rueda, D., Mo, J. J., Hakansson, K., and Walter, N. G. (2008). A rugged free energy landscape separates multiple functional RNA folds throughout denaturation. Nucleic Acids Res. 36, 7088–7099. Dorywalska, M., Blanchard, S. C., Gonzalez, R. L., Kim, H. D., Chu, S., and Puglisi, J. D. (2005). Site-specific labeling of the ribosome for single-molecule spectroscopy. Nucleic Acids Res. 33, 182–189. Draper, D. E. (2004). A guide to ions and RNA structure. RNA 10, 335–343. Elenko, M. P., Szostak, J. W., and van Oijen, A. M. (2009). Single-molecule imaging of an in vitro-evolved RNA aptamer reveals homogeneous ligand binding kinetics. J. Am. Chem. Soc. 131, 9866–9867. Grilley, D., Soto, A. M., and Draper, D. E. (2006). Mg2þ-RNA interaction free energies and their relationship to the folding of RNA tertiary structures. Proc. Natl. Acad. Sci. USA 103, 14003–14008. Ha, T., Zhuang, X. W., Kim, H. D., Orr, J. W., Williamson, J. R., and Chu, S. (1999). Ligand-induced conformational changes observed in single RNA molecules. Proc. Natl. Acad. Sci. USA 96, 9077–9082. Hodak, J. H., Fiore, J. L., Nesbitt, D. J., Downey, C. D., and Pardi, A. (2005). Docking kinetics and equilibrium of a GAAA tetra loop-receptor motif probed by single-molecule FRET. Proc. Natl. Acad. Sci. USA 102, 10505–10510. Huang, Z., Pei, W. M., Han, Y., Jayaseelan, S., Shekhtman, A., Shi, H., and Niu, L. (2009). One RNA aptamer sequence, two structures: A collaborating pair that inhibits ampa receptors. Nucleic Acids Res. 37, 4022–4032. Jencks, W. P. (1981). On the attribution and additivity of binding-energies. Proc. Natl. Acad. Sci. USA 78, 4046–4050.

220

Max Greenfeld and Daniel Herschlag

Joo, C., and Ha, T. (2008). Single-molecule FRET with total internal reflection microscopy. Single Molecule Techniques: A Laboratory Manual pp. 3–36. Joo, C., Balci, H., Ishitsuka, Y., Buranachai, C., and Ha, T. (2008). Advances in singlemolecule fluorescence methods for molecular biology. Annu. Rev. Biochem. 77, 51–76. Korennykh, A. V., Plantinga, M. J., Correll, C. C., and Piccirilli, J. A. (2007). Linkage between substrate recognition and catalysis during cleavage of sarcin/ricin loop RNA by restrictocin. Biochemistry 46, 12744–12756. Li, P. T. X., Vieregg, J., and Tinoco, I. (2008). How RNA unfolds and refolds. Annu. Rev. Biochem. 77, 77–100. Mammen, M., Choi, S. K., and Whitesides, G. M. (1998). Polyvalent interactions in biological systems: Implications for design and use of multivalent ligands and inhibitors. Angew. Chem. Int. Ed. 37, 2755–2794. McKinney, S. A., Joo, C., and Ha, T. (2006). Analysis of single-molecule FRET trajectories using hidden Markov modeling. Biophys. J. 91, 1941–1951. Noller, H. F. (2005). RNA structure: Reading the ribosome. Science 309, 1508–1514. Okumus, B., Wilson, T. J., Lilley, D. M. J., and Ha, T. (2004). Vesicle encapsulation studies reveal that single molecule ribozyme heterogeneities are intrinsic. J. Phys. Chem. B 87, 2798–2806. Pereira, M. J. B., Nikolova, E. N., Hiley, S. L., Jaikaran, D., Collins, R. A., and Walter, N. G. (2008). Single vs ribozyme molecules reveal dynamic and hierarchical folding toward catalysis. J. Mol. Biol. 382, 496–509. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Sattin, B., Zhao, W., Travers, K., Chut, S., and Herschlag, D. (2008). Direct measurement of tertiary contact cooperativity in RNA folding. J. Am. Chem. Soc. 130, 6085–6087. Smith, G. J., Sosnick, T. R., Scherer, N. F., and Pan, T. (2005). Efficient fluorescence labeling of a large RNA through oligonucleotide hybridization. RNA 11, 234–239. Smith, G. J., Lee, K. T., Qu, X. H., Xie, Z., Pesic, J., Sosnick, T. R., Pan, T., and Scherer, N. F. (2008). A large collapsed-state RNA can exhibit simple exponential single-molecule dynamics. J. Mol. Biol. 378, 943–953. Solomatin, S., and Herschlag, D. (2009). Methods of site-specific labeling of RNA with fluorescent dyes. Methods Enzymol. 469, 47–68. Solomatin, S. V., Greenfeld, M., Chu, S., and Herschlag, D. (2010). Multiple native states reveal persistent ruggedness of an RNA folding landscape. Nature 463, 681–684. Stone, M. D., Mihalusova, M., O’Connor, C. M., Prathapam, R., Collins, K., and Zhuang, X. W. (2007). Stepwise protein-mediated RNA folding directs assembly of telomerase ribonucleoprotein. Nature 446, 458–461. Takamoto, K., Das, R., He, Q., Doniach, S., Brenowitz, M., Herschlag, D., and Chance, M. R. (2004). Principles of RNA compaction: Insights from the equilibrium folding pathway of the P4-P6 RNA domain in monovalent cations. J. Mol. Biol. 343, 1195–1206. Tan, E., Wilson, T. J., Nahas, M. K., Clegg, R. M., Lilley, D. M. J., and Ha, T. (2003). A four-way junction accelerates hairpin ribozyme folding via a discrete intermediate. Proc. Natl. Acad. Sci. USA 100, 9308–9313. Williamson, J. R. (2008). Cooperativity in macromolecular assembly. Nat. Chem. Biol. 4, 458–465. Zhuang, X. W., Bartley, L. E., Babcock, H. P., Russell, R., Ha, T. J., Herschlag, D., and Chu, S. (2000). A single-molecule study of RNA catalysis and folding. Science 288, 2048–2051. Zhuang, X. W., Kim, H., Pereira, M. J. B., Babcock, H. P., Walter, N. G., and Chu, S. (2002). Correlating structural dynamics and function in single ribozyme molecules. Science 296, 1473–1476.

C H A P T E R

T W E LV E

A Highly Purified, Fluorescently Labeled In Vitro Translation System for Single-Molecule Studies of Protein Synthesis Jingyi Fei,* Jiangning Wang,* Samuel H. Sternberg,*,1 Daniel D. MacDougall,* Margaret M. Elvekrog,* Dileep K. Pulukkunat,* Michael T. Englander,*,† and Ruben L. Gonzalez Jr.* Contents 1. Introduction 2. A Highly Purified, Escherichia coli-Based In Vitro Translation System 2.1. Tris–polymix buffer system 2.2. Preparation and purification of ribosomes and ribosomal subunits 2.3. Preparation of mRNAs 2.4. Preparation and purification of fMet-tRNAfMet, Phe-tRNAPhe, and Lys-tRNALys 2.5. Preparation and purification of translation factors 3. Biochemical Assays 3.1. Initiation assays 3.2. Elongation assays 3.3. Termination assays 3.4. Ribosome recycling assays 4. Preparation of Fluorescently Labeled Translation Components 4.1. Phylogenetic analysis/structural modeling 4.2. Ribosome labeling 4.3. tRNA labeling 4.4. Translation factor labeling

222 225 225 226 226 227 228 233 233 237 239 241 242 243 244 249 252

* Department of Chemistry, Columbia University, New York, USA Integrated Program in Cellular, Molecular, and Biomedical Sciences, Columbia University, New York, USA 1 Current address: Department of Chemistry, University of California at Berkeley, Berkeley, California, USA {

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72008-5

#

2010 Elsevier Inc. All rights reserved.

221

222

Jingyi Fei et al.

5. Conclusions and Future Perspectives Acknowledgments References

253 254 255

Abstract Single-molecule fluorescence resonance energy transfer (smFRET) has emerged as a powerful tool for mechanistic investigations of increasingly complex biochemical systems. Recently, we and others have successfully used smFRET to directly investigate the role of structural dynamics in the function and regulation of the cellular protein synthesis machinery. A significant challenge to these experiments, and to analogous experiments in similarly complex cellular machineries, is the need for specific and efficient fluorescent labeling of the biochemical system at locations that are both mechanistically informative and minimally perturbative to the biological activity. Here, we describe the development of a highly purified, fluorescently labeled in vitro translation system that we have successfully designed for smFRET studies of protein synthesis. The general approaches we outline should be amenable to single-molecule fluorescence studies of other complex biochemical systems.

1. Introduction Rapid and accurate translation of messenger RNA (mRNA) into the encoded protein product comprises a vital step in gene expression within all living cells. The central component of translation is the ribosome, a twosubunit, ribonucleoprotein-based molecular machine (Fig. 12.1A) which translocates along an mRNA template and synthesizes a polypeptide chain through the repetitive, mRNA-directed binding and incorporation of aminoacyl-transfer RNA (aa-tRNA) substrates (Fig. 12.1B). Throughout translation, a number of essential protein factors, termed initiation (IF), elongation (EF), release (RF), and ribosome recycling (RRF) factors interact with the ribosome, catalyzing many of the individual steps of translation and helping to ensure the overall speed and accuracy of protein synthesis (Liljas, 2004; Wilson et al., 2002). Recently, single-molecule fluorescence resonance energy transfer (smFRET) (Ha, 2001; Roy et al., 2008) has emerged as a powerful tool in mechanistic studies of protein synthesis (Frank and Gonzalez, 2010; Marshall et al., 2008a). By combining the ability to monitor single molecules with a time-resolved, biophysical signal that is exquisitely sensitive to conformational changes, smFRET complements static structural and ensemble biochemical/biophysical studies by revealing the conformational trajectories of individual molecules in real time. Thus, smFRET studies often provide mechanistically important dynamic data that are unavailable

223

Single-Molecule Fluorescence Studies of Protein Synthesis

A

B

GDP

G

RRF RRF

E

P

GTP

G

3

Ribosome recycling

A 3

E P A GTP

1/2

Termination

2 1

GDP

3

Initiation GDP

2

1/2

3

GTP

3 1 GDP GTP

GDP

Tu

Ts

G GTP GTP

aa-tRNA selection

Translocation

G

Tu

GDP

Elongation cycle E P A

3

50S 30S E P A

1

Tu

2

Ts 1/2

mRNA tRNA

RRF

Initiation factors

G

Elongation factors

3

Release factors

Peptidyl transfer

Ribosome recycling factor

Figure 12.1 Ribosome structure and protein synthesis. (A) X-ray crystallographic structure of the ribosome and its tRNA substrates (Selmer et al., 2006). The 50S ribosomal subunit is shown in lavender and the 30S ribosomal subunit in tan. The mRNA (cartooned as a gray curve) binds to the 30S subunit where the sequence of mRNA codons specifies the amino acid sequence of the protein to be synthesized. There are three tRNA binding sites on the ribosome specific for aa-tRNA (purple tRNA, A site), peptidyl-tRNA (red tRNA, P site), and deacylated tRNA (orange tRNA, E site). (B) Cartoon representation of the translation cycle. During the initiation stage of translation, assembly of the 70S initiation complex from the 30S and 50S subunits, mRNA, and initiator tRNA, fMet-tRNAfMet, is mediated by IF1, 2, and 3. During the elongation stage of translation, the 70S ribosomal complex undergoes multiple rounds through the elongation cycle, with each cycle involving EF-Tu-catalyzed incorporation of the mRNA-encoded aa-tRNA, ribosome-catalyzed peptidyl transfer, and EF-G-catalyzed translocation of the mRNA–tRNA complex by one codon. Translocation of a stop codon into the A site triggers the termination stage of translation, during which RF1 or 2 hydrolyzes the newly synthesized polypeptide chain followed by RF3-catalyzed dissociation of RF1/2. Finally, the posttermination ribosomal complex is disassembled during the ribosome recycling stage of translation by the action of RRF, EF-G, and IF3.

224

Jingyi Fei et al.

from static X-ray crystallographic and cryogenic electron microscopic structures and are obscured by the signal averaging inherent to biochemical/biophysical studies of asynchronous molecular ensembles. smFRET studies of the mechanism through which the translating ribosome selects the correct, mRNA-encoded (i.e., cognate) aa-tRNA while discriminating against nearly correct (i.e., near-cognate) aa-tRNAs (aatRNA selection step in Fig. 12.1B) provide one example of the type of mechanistic detail that is uniquely accessible to this approach (Blanchard et al., 2004a; Gonzalez et al., 2007; Lee et al., 2007). These smFRET studies revealed that incoming aa-tRNAs, delivered as a ternary complex with EFTu and GTP, sample a short-lived intermediate configuration on the ribosome that is decisive in discriminating cognate from near-cognate aa-tRNAs. This intermediate configuration of the ternary complex represents a critical branchpoint during aa-tRNA selection, at which the ribosome selectively permits a cognate ternary complex to progress forward in the reaction pathway but rapidly dissociates near-cognate ternary complexes. In this example, the asynchronous nature of ternary complex binding events among the ensemble of ribosomes, combined with the energetically unstable and short-lived nature of this ribosome-bound ternary complex configuration, yields an intermediate that is rarely populated and nonaccumulating. As a result, this critical intermediate during aa-tRNA selection had gone unobserved and uncharacterized in ensemble biochemical/biophysical (Daviter et al., 2006) and static structural (Li et al., 2008; Ogle and Ramakrishnan, 2005; Schuette et al., 2009; Stark et al., 2002; Valle et al., 2003; Villa et al., 2009) studies. Additional examples of the contributions that smFRET studies have made to our mechanistic understanding of protein synthesis have been recently reviewed (Frank and Gonzalez, 2010; Marshall et al., 2008a). One of the most significant challenges to smFRET studies of complex biochemical systems such as the cellular protein synthesis machinery is the labeling of system components with the donor and acceptor fluorophores that are required to generate the smFRET signal. Fluorescent labeling for smFRET studies must be (1) efficient, such that a large population of the observed molecules contain both a donor and an acceptor fluorophore; (2) specific, such that any heterogeneity detected over the entire population of observed molecules reflects the conformational heterogeneity of the molecular ensemble rather than heterogeneity in the positions of the donor or acceptor fluorophores; (3) mechanistically informative, such that the conformational change of interest yields a distance change between the donor and acceptor pair that generates a detectable change in FRET value; (4) minimally perturbative, such that the presence of the donor or acceptor fluorophore does not block or significantly interfere with the biochemical reaction under investigation.

Single-Molecule Fluorescence Studies of Protein Synthesis

225

Here, we describe a highly purified in vitro translation system which, in combination with a series of standard biochemical assays, has allowed us to develop and validate numerous fluorescence labeling strategies for smFRET studies of protein synthesis. We present a general strategy for the design of fluorophore labeling positions and describe the procedures used to generate site-specifically labeled ribosomes, translation factors, and tRNA constructs. These fluorescently labeled translation components can then be tested using the biochemical assays described below in order to assess their compatibility with our in vitro translation system and thus their suitability for use in smFRET experiments. Many of the protocols we describe here are adaptations or modifications of protocols previously developed by numerous groups working on structural and mechanistic studies of protein synthesis. Thus, throughout this chapter, we only briefly describe and provide references for those protocols that are used essentially as previously reported and describe in detail only those protocols that we have significantly modified or developed de novo. It is our hope that the general approaches we outline here will be applicable to smFRET investigations of other complex biochemical systems such as DNA replication, transcription, and pre-mRNA splicing.

2. A Highly Purified, Escherichia coli-Based In Vitro Translation System 2.1. Tris–polymix buffer system The Tris–polymix buffer used in our experiments is primarily based on the polymix buffer originally described by Jelenc and Kurland (1979) and further elaborated upon by Pavlov and Ehrenberg (1996) and Wagner et al. (1982). We further optimized this polymix buffer by testing the protein synthesis activity of purified ribosomes (see below) within a partially purified, fractionated in vitro translation system as described by Chambliss et al. (1983). The mRNA template used for these buffer optimization experiments was an in vitro transcribed mRNA (McKenna et al., 2007; Milligan et al., 1987; Wyatt et al., 1991) encoding a C-terminal truncated variant of gene product 32 from bacteriophage T4, where the UUC codon encoding phenylalanine at position 225 was mutated to a UAA stop codon (hereafter referred to as T4gp321–224 mRNA). In these experiments, the yield and rate of T4gp321–224 synthesis was monitored by analyzing [35S]-methionine-labeled translation products by SDS-PAGE (Gallagher, 2006). The optimal buffer conditions, which were adopted for all of our biochemical and single-molecule experiments, are 50 mM Tris–acetate (Tris–HOAc) (pH25
C ¼ 7.5), 100 mM KCl, 3.5–15 mM Mg(OAc)2 (exact concentration depends on the nature of the experiment), 5 mM

226

Jingyi Fei et al.

NH4OAc, 0.5 mM Ca(OAc)2, 6 mM b-mercaptoethanol (BME), 5 mM putrescine–HCl, and 1 mM spermidine-free base (Blanchard et al., 2004b).

2.2. Preparation and purification of ribosomes and ribosomal subunits Highly active, tightly coupled E. coli 70S ribosomes are purified by preparative sucrose density gradient ultracentrifugation of S30 cleared lysates of E. coli strain MRE600 using a combination of the protocols reported by Blanchard et al. (2004b), Powers and Noller (1991), and Robertson and Wintermeyer (1981). The use of strain MRE600, which lacks the gene encoding the ribosomal RNA (rRNA)-active RNase I (Cammack and Wade, 1965), helps ensure the integrity of 70S ribosomes during purification. 70S ribosomes are distinguished by their sedimentation as intact 70S ribosomes, rather than as dissociated small (30S) and large (50S) ribosomal subunits, when centrifuged through sucrose density gradients containing a specified, low Mg2þ concentration (Hapke and Noll, 1976). The specific concentration of Mg2þ used to define tightly coupled 70S ribosomes varies depending on the E. coli strain used (5.25 mM for MRE600; Robertson and Wintermeyer, 1981). Highly active 30S and 50S subunits can be obtained by dissociating purified, tightly coupled 70S ribosomes into their constituent 30S and 50S subunits via dialysis against buffer containing 1 mM Mg2þ and subsequently purifying the 30S and 50S subunits by preparative sucrose density gradient ultracentrifugation in buffer containing 1 mM Mg2þ (Powers and Noller, 1991; Recht et al., 1999).

2.3. Preparation of mRNAs The mRNAs used for biochemical and smFRET studies in our laboratory are either chemically synthesized (Dharmacon, Inc.) or in vitro transcribed using well-established protocols (McKenna et al., 2007; Milligan et al., 1987; Wyatt et al., 1991). Chemically synthesized mRNAs are purified by the manufacturer using high-performance liquid chromatography and are resuspended in mRNA buffer (10 mM Tris–HOAc (pH25
C ¼ 7.5), 10 mM KCl, and 0.1 mM EDTA) prior to use. In vitro transcription reactions are quenched by addition of 0.1 reaction volume of 500 mM EDTA, and the mRNA product is extensively buffer exchanged into mRNA buffer and concentrated using a molecular weight cutoff (MWCO) ¼ 10,000 centrifugal filtration device (Amicon Ultra, Millipore). The mRNAs used in all of our studies are variants of the T4gp321–224 mRNA (Section 2.1) and are based on the following general sequence construct: 50 -[GG]CAACCUAAAACUUACACAGGGCCCUAAGGAAAUAAAAAUG(XYZ)n-30 , where nucleotides that facilitate in vitro transcription are bracketed, nucleotides that serve as a target sequence for

Single-Molecule Fluorescence Studies of Protein Synthesis

227

hybridizing a complementary, 30 -biotinylated DNA oligonucleotide (Integrated DNA Technologies; 50 -TGTGTAAGTTTTAGGTTGATT TG-Biotin-30 ), to enable surface-immobilization for smFRET studies (Zhuang et al., 2000) are underlined, the core Shine-Dalgarno ribosome binding site is underlined and in bold, the AUG start codon encoding initiator fMet-tRNAfMet is underlined and in italics, and the number of codons that are appended to the end of the general construct, which is variable depending on the study, are denoted by (XYZ)n.

2.4. Preparation and purification of fMet-tRNAfMet, Phe-tRNAPhe, and Lys-tRNALys Overexpression vectors for E. coli methionyl tRNA synthetase and E. coli formylmethionyl-tRNA formyltransferase were provided by Prof. Sylvain Blanquet (CNRS-Ecole Polytechnique, Palaiseau Cedex, France), for phenylalanyl tRNA synthetase by Prof. David Tirrell (California Institute of Technology, Pasadena, CA, USA), and for lysyl tRNA synthetase by Prof. Takuya Ueda (University of Tokyo, Japan). Methionyl tRNA synthetase was prepared as reported in Fourmy et al. (1991), formylmethionyl-tRNA formyltransferase as reported in Schmitt et al. (1999), and phenylalanyl tRNA synthetase and lysyl tRNA synthetase as reported in Shimizu et al. (2001). The formyl donor substrate for formylmethionyl-tRNA formyltransferase, 10-formyltetrahydrofolate, is chemically prepared starting from the calcium salt of folinic acid (Acros Organics) as previously described (Dubnoff et al., 1971). Aminoacylation and formylation of tRNAfMet (Sigma or MP Biomedicals) is achieved simultaneously by incubating 20 mM tRNAfMet with 25 mM Tris–HCl (pH37
C ¼ 7.5), 7 mM MgCl2, 150 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol (DTT), 2.5 mM ATP, 300 mM 10-formyltetrahydrofolate, 80 mM methionine, 0.02 mM methionyl tRNA synthetase, and 0.2 mM formylmethionyl-tRNA formyltransferase for 10 min at 37
C. Aminoacylation of tRNAPhe (Sigma) is achieved by incubating 15 mM tRNAPhe (Sigma) with 200 mM Tris–HCl (pH37
C ¼ 7.5), 15 mM MgCl2, 25 mM KCl, 2 mM BME, 5 mM ATP, 10 mM phosphoenolpyruvate (PEP), 30 U ml 1 pyruvate kinase, 55 mM phenylalanine, and 0.75 mM phenylalanyl tRNA synthetase for 10 min at 37
C. Aminoacylation of tRNALys (Sigma) is achieved by incubating 20 mM tRNALys (Sigma) with 50 mM Tris–HCl (pH37
C ¼ 7.5), 7 mM MgCl2, 150 mM KCl, 0.1 mM EDTA, 1 mM DTT, 2.5 mM ATP, 80 mM lysine, and 1.1 mM lysyl tRNA synthetase for 10 min at 37
C. All formylation and/or aminoacylation reactions are quenched by addition of 0.1 reaction volume of 3 M NaOAc (pH ¼ 5.2), extracted twice with 1 reaction volume of phenol, and extracted twice with 1 reaction volume of chloroform. tRNAs are then ethanol precipitated by addition of

228

Jingyi Fei et al.

3 reaction volume of –20
C ethanol and incubation for a minimum of 1 h at –80
C, followed by centrifugation for 15 min at 18,000  g at 4
C. Pellets are resuspended in ice-cold 10 mM KOAc (pH ¼ 5.0), passed through a Micro Bio-Spin 6 gel filtration spin column (Bio-Rad) equilibrated against ice-cold 10 mM KOAc (pH ¼ 5.0), rapidly aliquoted, flashfrozen in liquid nitrogen, and stored at –80
C. One aliquot is used to measure the final tRNA concentration using ultraviolet absorbance at 260 nm; the extinction coefficient at 260 nm for a particular species of purified tRNA can be estimated based on the amino acid acceptor activity of 1 A260 Unit of the purified tRNA, a value which is typically provided by the supplier. One A260 Unit is the amount of tRNA per 1 ml that yields an absorbance of 1 in a 1 cm path length cuvette at 260 nm. tRNAfMet aminoacylation/formylation yields are assessed by hydrophobic interaction chromatography (HIC) on a TSKgel Phenyl-5PW column (8.0 mm (ID)  7.5 cm (L)) (Tosoh Bioscience) operating at 4
C using a previously described protocol (Schmitt et al., 1999). An aliquot from the aminoacylation/formylation reaction (0.05 nmol of tRNA) is diluted 10fold into ice-cold tRNA HIC Buffer A (1.7 M NH4SO4, 10 mM NH4OAc (pH ¼ 6.3); note that the pH of the stock NH4OAc solution, rather than of the final tRNA HIC Buffer A, should be adjusted to 6.3), injected onto the Phenyl-5PW column preequilibrated against tRNA HIC Buffer A, and eluted using a linear gradient of 0–100% tRNA HIC Buffer B (10 mM NH4OAc (pH ¼ 6.3), 10% CH3OH; note that the pH of the stock NH4OAc solution, rather than of the final tRNA HIC Buffer B, should be adjusted to 6.3) over 25 column volumes. Due to the increasing hydrophobicity of deacylated tRNAfMet, Met-tRNAfMet, and fMet-tRNAfMet, these species elute from the Phenyl-5PW column at 15.5%, 18.5%, and 24% tRNA HIC Buffer B, respectively, providing an effective means of assessing the yields of the aminoacylation/formylation reactions. This same protocol can be used to assess the yields of tRNAPhe and tRNALys aminoacylation reactions. In line with its increased hydrophobicity, Phe-tRNAPhe exhibits an increased retention volume relative to deacylated tRNAPhe, whereas the positively charged N e of Lys-tRNALys generates a decreased retention volume relative to deacylated tRNALys. Based on this assessment, we are routinely able to achieve >90% aminoacylation/formylation of tRNAfMet, >90% aminoacylation of tRNAPhe, and  60% aminoacylation of tRNALys.

2.5. Preparation and purification of translation factors Genes encoding the 10 canonical translation factors: IF1, 2 (g isoform), and 3; EF-Tu, Ts, and G; RF1, 2, and 3; and RRF (Fig. 12.1B) were PCRamplified from E. coli K12 genomic DNA prepared as described (Wilson, 1987) or purchased from the American Type Culture Collection (ATCC

Single-Molecule Fluorescence Studies of Protein Synthesis

229

#10798D-5). The PCR primers targeting each factor gene introduce appropriate restriction sites for cloning into the pProEX-HTb plasmid expression vector system (Invitrogen). Translation factor genes cloned into pProEX-HTb are placed under the control of an isopropyl b-D1-thiogalactopyranoside (IPTG)-inducible pTrc promoter. In addition, the pProEX-HTb vector introduces a six-histidine (6xHis) affinity tag followed by a highly specific tobacco etch virus (TEV) protease cleavage site at the amino terminus of the expressed factor. The 6xHis tag allows affinity purification of each factor using Ni2þ-nitrilotriacetic acid (Ni2þNTA) resin (Qiagen), and the TEV protease cleavage site allows subsequent removal of the 6xHis tag from the purified factor. Due to the sequence recognition and cleavage requirements of TEV protease as well as limitations in the restriction enzymes which can be used to clone the individual factor genes into pProEX-HTb, the N-terminus of each purified factor includes 1–5 additional, non-wild-type amino acids which precede the wild-type amino acid sequence. Thus, the N-terminal ends of each of our specific clones are: G-A-M1 (IF1), G-A-Q-D-D-M1 (IF2g), G-A-M-AK2 (IF3), G-A-M-G-S2 (EF-Tu), G-A-M1 (EF-Ts), G-A-M-G-S-A2 (EF-G), G-A-M1 (RF1), G-A-M1 (RF2), G-A-M1 (RF3), and G-A-M1 (RRF), where the underlined amino acid and sequence position denote the beginning of the wild-type gene sequence. We have developed a general translation factor purification strategy based on standard Ni2þ-NTA affinity purification procedures (Hoffmann and Roeder, 1991), which can be applied to all 10 translation factors. For several factors this general strategy must be slightly modified to meet special conditions or expanded to include additional chromatographic steps in order to achieve high purity. Thus, in this paragraph, we describe our general strategy and in the paragraphs that follow we describe special considerations specific to several factors. Each factor is overexpressed in BL21(DE3) cells in 1–2 L Terrific Broth (Difco) (Elbing and Brent, 2002) supplemented with 100 mg ml 1 a-carboxybenzylpenicillin (Sigma) (Raleigh et al., 2002). IPTG is added to a final concentration of 1 mM when the cell cultures reach an optical density of 0.8–1.0 at 600 nm. Overexpressing cells are grown for an additional 2–4 h at 37
C (IFs, RFs, RRF) or overnight at 30
C (EFs) and subsequently harvested by centrifugation at 5000  g for 15 min at 4
C. All subsequent steps are performed at 4
C. The resulting cell pellet is resuspended into TF Buffer A (20 mM Tris–HCl (pH4
C ¼ 7.5), 300 mM NaCl, 10 mM imidazole, 0.2 mM phenylmethanesulphonyl fluoride (PMSF), and 2 mM BME) and lysed by passing through a French Press at an internal cell pressure of 1200 psi. The resulting lysate is cleared by centrifugation at 20,000  g for 30 min. The cleared lysate is added to 2–3 ml Ni2þ-NTA resin that has been preequilibrated with 5 column volumes of TF Buffer A, and the mixture is slowly stirred in a disposable polypropylene tube (BD Biosciences) for

230

Jingyi Fei et al.

30 min in order to allow binding of the 6xHis-tagged factor to the Ni2þNTA resin. The Ni2þ-NTA resin is then poured into a disposable polypropylene column (Pierce) and washed with 10 column volumes of TF Buffer B (TF Buffer A containing 30 mM imidazole). Bound 6xHis-tagged factor is eluted with 4 column volumes of TF Buffer C (Buffer A containing 500 mM NaCl and 250 mM imidazole) and collected over 4–10 fractions. Factor-containing fractions are identified and initial purity is assessed by SDS-PAGE and Coomassie staining (Sasse and Gallagher, 2009). Factorcontaining fractions are combined, 0.05 mg 6xHis-tagged TEV protease (Promega) is added per 1 mg of factor (as measured by the Bradford assay; Simonian and Smith, 2006), and the reaction mixture is dialyzed against TF Buffer D (20 mM Tris–HCl (pH4
C ¼ 7.5), 200 mM NaCl, 0.1% TritonX, and 2 mM BME). The cleavage reaction is monitored by the change in molecular weight of the cleaved versus uncleaved factor using SDS-PAGE and Coomassie staining. Depending on the activity of the TEV protease and on the specific factor being purified, cleavage may require 12–48 h to go to completion. After TEV cleavage is complete, cleaved factor is separated from uncleaved factor, cleaved 6xHis-tag fragments, and the 6xHis-tagged TEV protease by adding the cleavage reaction to 2–3 ml Ni2þ-NTA resin preequilibrated against TF Buffer D supplemented with 30 mM imidazole. If the volume of the cleavage reaction is significantly increased during the dialysis/cleavage procedure, the cleavage reaction may be concentrated prior to mixing with the Ni2þ-NTA resin using a centrifugal filtration device (Millipore) with an appropriate MWCO. The cleavage reaction/ Ni2þ-NTA resin mixture is slowly stirred in a disposable polypropylene tube for 1 h, poured into a disposable polypropylene column, and the flowthrough containing the cleaved, purified factor is collected. The column is washed with 2 column volumes of TF Buffer D supplemented with 30 mM imidazole to collect any remaining cleaved, purified factor. The cleaved, purified factor is then buffer exchanged into 2 TF Buffer E (20 mM Tris– HOAc (pH4
C ¼ 7.5), 100 mM KCl, 10 mM BME) and concentrated using a centrifugal filtration device, diluted to 1 TF Buffer E by addition of 100% glycerol, and stored at –20
C. Final concentrations of all translation factors are typically determined using the Bradford assay, with the exception of IF2g and EF-G (see below). Approximate final protein yields are 0.5 mg L 1 culture for IF1, 8 mg L 1 for IF2g, 1 mg L 1 for IF3, 10–20 mg L 1 for EF-Tu, 25–50 mg L 1 for EF-Ts, 40 mg L 1 for EF-G, 1.5 mg l 1 for RF1/2, 50 mg L 1 for RF3, and 10 mg L 1 for RRF. 2.5.1. Special considerations for IF1 IF1 Buffer A (10 mM Tris–HCl (pH4
C ¼ 7.5), 60 mM NH4Cl, 10 mM MgCl2, 5 mM BME, 0.1 mM PMSF, and 10 mM imidazole) replaces TF Buffer A. Cells are lysed by three passes through a French Press at an internal cell pressure of 1200 psi. After batch binding of 6xHis-tagged IF1 to the

Single-Molecule Fluorescence Studies of Protein Synthesis

231

Ni2þ-NTA resin and transfer to a disposable polypropylene column, the resin is washed with 10 column volumes of IF1 Buffer B (IF1 Buffer A lacking NH4Cl and containing 30 mM imidazole) to remove nonspecifically bound proteins. Bound 6xHis-tagged IF1 is eluted with IF1 Buffer C (IF1 Buffer B containing 250 mM imidazole). 6xHis-tagged IF1 containing fractions are identified using a Tris–tricine gradient gel (10–20%) (Gallagher, 2006) with Coomassie staining. Fractions containing 6xHistagged IF1 are pooled and dialyzed against TF Buffer D overnight. TEV cleavage proceeds as described in the general protocol above, and the cleavage reaction is monitored by Tris–tricine gradient gel (10–20%) with Coomassie staining. Removal of the cleaved 6xHis-tag fragments and the 6xHis-tagged TEV protease is achieved by mixing the cleavage reaction with Ni2þ-NTA resin that has been preequilibrated against TF Buffer D supplemented with 30 mM imidazole. The flow-through and wash containing cleaved, purified IF1 is passed through a HiLoad 16/60 Superdex 75 prep grade (GE Biosciences) gel filtration column using TF Buffer E as a column preequilibration and running buffer. IF1 elutes at a retention volume of 87 ml. The fractions containing IF1 are pooled, buffer exchanged into 2 TF Buffer E and concentrated using a centrifugal filtration device, diluted to 1 TF Buffer E by addition of 100% glycerol, and stored at –20
C. 2.5.2. Special considerations for IF2g TF Buffer A is supplemented with 0.22 U ml 1 DNase I (New England BioLabs). Prior to addition of TEV protease, the 6xHis-tagged IF2g eluted from the Ni2þ-NTA resin is diluted to a final concentration of 0.25 mg ml 1 using IF2g Buffer D (50 mM Tris–HCl (pH4
C ¼ 7.5), 50 mM KCl, 0.1% Triton-X, and 2 mM BME), before dialyzing against TF Buffer D. Prior to mixing the cleavage reaction with the Ni2þ-NTA resin, the cleaved IF2g is concentrated using a MWCO ¼ 10,000 centrifugal filtration device. Removal of the cleaved 6xHis-tag fragments and the 6xHis-tagged TEV protease is achieved by mixing the cleavage reaction with Ni2þ-NTA resin that has been preequilibrated against TF Buffer D. The flow-through and two washes of 1 column volume each are collected, and the cleaved IF2g is loaded onto a HiTrap SP HP cation exchange column (5 ml column volume) (GE Biosciences) preequilibrated against IF2g Buffer IEX1 (40 mM Tris–HCl (pH4
C ¼ 7.5), 30 mM NaCl, 40 mM NH4Cl, 5 mM MgCl2, 2 mM BME). The column is washed with 5–10 column volumes of IF2g Buffer IEX1 and IF2g is eluted with a linear gradient of 0–75% IF2g Buffer IEX2 (IF2g Buffer IEX1 containing 750 mM NaCl) over 30 column volumes (Antoun et al., 2004). IF2g elutes at 33% IF2g Buffer IEX2. Fractions containing purified IF2g are pooled, buffer exchanged into 2 IF2g Buffer E (20 mM Tris–HOAc (pH4
C ¼ 7.5), 100 mM KCl, 20 mM Mg(OAc)2, 10 mM BME) and concentrated

232

Jingyi Fei et al.

using a centrifugal filtration device, diluted to 1 IF2g Buffer E by addition of 100% glycerol, and stored at –20
C. The final concentration of IF2g is measured using ultraviolet absorbance at 280 nm and a molar extinction coefficient of 27,390 M 1 cm 1, calculated using the ProtParam tool on the ExPASy Proteomics Server (http://ca.expasy.org/tools/protparam. html), which bases its calculation on protein amino acid composition in conjunction with the molar extinction coefficients of tyrosine, tryptophan, and cystine. 2.5.3. Special considerations for IF3 Purification of IF3 is identical to the procedure described above for IF1 up through collection of cleaved, purified factor from the second Ni2þ-NTA column. At this point the cleaved IF3 is loaded onto a HiTrap SP HP cation exchange column (5 ml column volume) (GE Biosciences) preequilibrated against 5 column volumes of IF2g Buffer IEX1. The column is washed with 3 column volumes of IF2g Buffer IEX1 and IF3 is eluted with a linear gradient of 0–100% of IF2g Buffer IEX2 over 20 column volumes. IF3 elutes at 65% IF2g Buffer IEX2. Fractions containing purified IF3 are pooled, buffer exchanged into 2 TF Buffer E and concentrated using a centrifugal filtration device, diluted to 1 TF Buffer E by addition of 100% glycerol, and stored at –20
C. 2.5.4. Special considerations for EF-Tu TF Buffers A–E are supplemented with 0.2 mM GDP and 0.5 mM MgCl2. These supplements help to maintain the integrity of EF-Tu throughout the purification procedure and during storage at –20
C. 2.5.5. Special considerations for EF-G The final concentration of EF-G is measured using ultraviolet absorbance at 280 nm and a molar extinction coefficient of 61,310 M 1 cm 1, calculated using the ProtParam tool on the ExPASy Proteomics Server (http://ca. expasy.org/tools/protparam.html) as described in Section 2.5.2. 2.5.6. Special considerations for RF1 and 2 RF1 and 2 (RF1/2) are posttranslationally modified through methylation at residue Q235 (RF1) or Q252 (RF2) by an N5-glutamine methyltransferase encoded by the PrmC gene, and defects in the efficiency of translation termination have been clearly correlated with incomplete modification (Dincbas-Renqvist et al., 2000; Heurgue-Hamard et al., 2002; Mora et al., 2007). Therefore, to prepare fully modified RF1/2, we have cotransformed BL21(DE3) strains for RF1/2 overexpression with a plasmid-encoded copy of the PrmC gene, and RF1/2 are co-overexpressed together with their methyltransferase.

Single-Molecule Fluorescence Studies of Protein Synthesis

233

3. Biochemical Assays 3.1. Initiation assays During translation initiation, IF1, 2, and 3 promote the formation of a 30S initiation complex that contains initiator fMet-tRNAfMet and the correct AUG start codon at the 30S P site. Docking of the 50S subunit onto the 30S initiation complex is then catalyzed by IF2 in its GTP-bound form, an event that stimulates GTP hydrolysis by IF2. Subsequent dissociation of the IFs yields a 70S initiation complex that is competent for formation of the first peptide bond (Fig. 12.1B). We typically use the standard assays described below to test the biochemical activities of initiation components. 3.1.1. Primer-extension inhibition assay The activities of ribosomes, fMet-tRNAfMet, and IFs in initiation are tested using a well-established primer-extension inhibition, or ‘‘toeprinting,’’ assay (Hartz et al., 1988; Hartz et al., 1989). Briefly, initiation reactions are carried out on an mRNA that has been preannealed with a 50 [32P]-labeled DNA primer. Subsequent reverse transcription of the primer-annealed, initiated mRNA is strongly blocked when the reverse transcriptase encounters an mRNA-bound ribosome, thereby producing a 50 [32P]-labeled cDNA of defined length, or ‘‘toeprint.’’ Analysis of the cDNA products on a 9% sequencing PAGE gel (Slatko and Albright, 1992) therefore reports the position of the ribosome on the mRNA with single-nucleotide resolution. Three distinct toeprinting assays, described below, are used to test the individual activities of IF1, IF2g, and IF3. All toeprinting assays are performed using T4gp321–224 mRNA (Section 2.1) preannealed with a 50 [32P]-labeled DNA primer of sequence TATTGCCATTCAGTTTAG (Integrated DNA Technologies). The Primer Labeling Reaction is performed by mixing 70 pmol DNA primer, 42 pmol [g-32P]ATP (6000 Ci mmol 1, Perkin Elmer), and 14 Units T4 polynucleotide kinase (New England Biolabs) in a final reaction volume of 30 ml, prepared in 1 T4 polynucleotide kinase buffer (New England Biolabs) and incubating for 30 min at 37
C. The labeling reaction is subsequently incubated for 10 min at 75
C to inactivate the T4 polynucleotide kinase and unincorporated [g-32P]ATP is removed using a G25 Sephadex gel filtration spin column (GE Healthcare). The Primer Annealing Reaction is performed by mixing 4 ml of the Primer Labeling Reaction with 100 pmol of T4gp321–224 in a final reaction volume of 40 ml, prepared in 25 mM Tris–HOAc (pH25
C ¼ 7.0), incubating in a dry block heater for 1.5 min at 90
C, and slowly cooling to room temperature by transferring the dry block from the heater to the bench top.

234

Jingyi Fei et al.

The IF2g assay tests the ability of IF2g to direct the selection of fMettRNAfMet over elongator tRNA during initiation. The T4gp321–224 mRNA’s AUG start codon, encoding tRNAfMet, is followed by a UUU triplet at the second codon position, encoding tRNAPhe. Binding of fMettRNAfMet to the AUG start codon at the 30S P site generates a toeprint at a position that is 15 nucleotides 30 to the A nucleotide of the AUG start codon (i.e., a þ 15 toeprint), whereas binding of tRNAPhe to the UUU codon at the 30S P site generates a toeprint at a position that is 18 nucleotides 30 to the A nucleotide of the AUG start codon (i.e., a þ18 toeprint). Thus, selection of fMet-tRNAfMet over tRNAPhe using the T4gp321–224 mRNA can be easily observed by monitoring the intensity of the þ15 toeprint relative to the intensity of the þ18 toeprint. Each Initiation Reaction is performed in three steps: 1. A mixture of 10 pmol 30S subunits, 100 pmol IF2g, and 16 nmol GTP is incubated for 10 min at 37
C. 2. 2 ml of the Primer Annealing Reaction is added to the reaction, followed by an additional 10 min incubation at 37
C. 3. 16 pmol each of fMet-tRNAfMet and tRNAPhe, prepared as an equimolar mixture, are added to the reaction, followed by an additional 10 min incubation at 37
C. The final reaction volume is 20 ml, prepared in Tris–polymix buffer (3 mM Mg2þ). Initiation Reactions are placed on ice until ready for use in Primer-Extension Reactions. Each Primer-Extension Reaction is performed by mixing 5 ml of an Initiation Reaction with 30 nmol ATP, 12.5 nmol each of dATP, dGTP, dCTP, and dTTP, and 6 Units AMV reverse transcriptase (Promega) in a final reaction volume of 25 ml, prepared in Tris–polymix buffer (10 mM Mg2þ), and incubating for 15 min at 37
C. Primer-Extension Reactions are extracted twice with 1 reaction volume of phenol and twice with 1 reaction volume of chloroform. cDNA products are ethanol precipitated by mixing Primer-Extension Reactions with 0.1 reaction volume of 3 M Na(OAc) (pH ¼ 5.5) and 3 reaction volume of 100% ethanol, followed by incubation for 10 min at room temperature and centrifugation at 18,000  g for 10 min. The resulting cDNA pellets are washed once with 70% ethanol. The cDNA pellets are scintillation counted and 5000– 10,000 counts per minute (cpm) are loaded into each lane of a 9% sequencing PAGE gel (40 cm  20 cm, 0.2–0.4 mm thickness), which is run at a constant power of 55 W in 1 TBE (Tris/borate/EDTA) electrophoresis buffer (Moore, 2000). The gel is then dried and phosphorimaged using a STORM PhosphorImager (GE Healthcare). Five control Primer-Extension Reactions are typically performed with all toeprinting assays. The first four control Primer-Extension Reactions are performed by mixing 3.5 ml of diluted Primer Annealing Reaction (diluted

Single-Molecule Fluorescence Studies of Protein Synthesis

235

2.5-fold with Tris–polymix buffer (10 mM Mg2þ)), 50 nmol ATP, 20 nmol each of dATP, dGTP, dCTP, and dTTP, 10 nmol of either dideoxy ATP, GTP, CTP, or TTP, and 10 Units AMV reverse transcriptase in a final volume of 40 ml, prepared in Tris–polymix buffer (10 mM Mg2þ), and incubating for 30 min at 37
C. These mRNA sequencing reactions allow the þ15 and þ 18 toeprint positions to be located within the T4gp321–224 mRNA. The fifth control Primer-Extension Reaction is performed as described in the previous paragraph, but in the absence of added Initiation Reaction and incubated for only 15 min at 37
C in order to detect intrinsic sites of reverse transcriptase stops caused by local secondary structures within the mRNA. The intensities of the bands corresponding to the þ15 and þ18 cDNA products in this control reaction are used to background correct the intensities of all þ15 and þ18 toeprints. Four reactions are typically performed to test the activity of IF2g. The first and second reactions are performed in the absence of IF2g but in the presence of either fMet-tRNAfMet or tRNAPhe, in order to demonstrate that both tRNAs can actively bind to the 30S P site and generate strong þ15 and þ18 toeprints, respectively. The third and fourth reactions are run in the absence or presence of IF2g and equimolar amounts of fMettRNAfMet and tRNAPhe. In the absence of IF2g, one observes þ15 and þ18 toeprints of equal intensity, consistent with the inability of the mRNA-bound 30S subunit to discriminate between fMet-tRNAfMet and tRNAPhe in the absence of IF2g. In the presence of IF2g, however, one observes a very strong þ15 toeprint and a missing or very weak þ18 toeprint, demonstrating the ability of IF2g to direct the selection of fMettRNAfMet over tRNAPhe. The IF1 assay tests the ability of IF1 to enhance the formation of a correctly initiated 70S initiation complex in the presence of IF2 and IF3 (Hartz et al., 1989). Each Initiation Reaction is prepared in four steps: 1. A mixture of 12 pmol each of 30S and 50S subunits are incubated for 10 min at 37
C. 2. 12 pmol IF3, 48 pmol IF2g, and 48 pmol IF1 are added to the reaction, followed by an additional 10 min incubation at 37
C. 3. 2.4 ml of Primer Annealing Reaction is added to the reaction, followed by an additional 10 min incubation at 37
C. 4. 35 pmol each of fMet-tRNAfMet and tRNAPhe, prepared as an equimolar mixture, are added to the reaction, followed by an additional 10 min incubation at 37
C. The final reaction volume is 26 ml, prepared in Tris–polymix buffer (5 mM Mg2þ). Primer-Extension Reactions and all subsequent steps are performed as in the IF2g assay. Reactions in the absence and presence of IF1 are typically performed. An approximately threefold increase in the intensity of the þ15 toeprint is observed in the presence versus the absence of IF1,

236

Jingyi Fei et al.

demonstrating IF1’s ability to enhance the formation of a correctly initiated 70S initiation complex. The IF3 assay demonstrates the ability of IF3 to regulate fMet-tRNAfMet selection on 30S subunits. Initiation Reactions are prepared in two steps: 1. A mixture of 2 pmol 30S subunits, 2 ml of diluted Primer Annealing Reaction (diluted fivefold into Tris–polymix buffer (5 mM Mg2þ)), 20 pmol tRNAfMet, and 200 pmol tRNAPhe are incubated for 10 min at 37
C. 2. 24 pmol IF3 is added to the reaction, followed by an additional 10 min incubation at 37
C. The final reaction volume is 20 ml, prepared in Tris–polymix buffer (5 mM Mg2þ). Primer-Extension Reactions and all subsequent steps are performed as in the IF2g assay. Reactions in the absence and presence of IF3 are typically performed. In the absence of IF3, the 10-fold molar excess of tRNAPhe produces a strong þ18 toeprint relative to the þ15 toeprint. In the presence of IF3, a strong þ15 toeprint, relative to the þ18 toeprint, is observed despite the 10-fold molar excess of tRNAPhe; this result demonstrates the ability of IF3 to regulate the binding of tRNAs to the 30S P site (Hartz et al., 1988, 1989; Maar et al., 2008). 3.1.2. GTP hydrolysis assay Ribosome-dependent, multiple-turnover GTP hydrolysis by IF2g is assayed using [a-32P]GTP and thin layer chromatography (TLC) as described by Brandi et al. (2004), with several modifications. The reaction is performed in three steps: 1. A GTP/[a-32P]GTP Mix is prepared by mixing 200 nmol of GTP and 2 pmol [a-32P]GTP (3000 Ci mmol 1, PerkinElmer) in a final volume of 1 ml, prepared in Barnstead NANOpure (Thermo Scientific) purified water and adjusted to pH ¼ 7.0 with 1 M KOH. 2. A 70S/IF2g Mix is prepared by mixing 6 pmol 70S ribosomes (or the equivalent amounts of 30S and 50S subunits) with 18 pmol IF2g in a final volume of 13 ml, prepared in Tris–polymix buffer (5 mM Mg2þ). 3. 2 ml of the GTP/[a-32P]GTP Mix is added to 13 ml of the 70S/IF2g Mix and the reaction is incubated for 10 min at 37
C. The reaction is quenched by addition of 5 ml 100 mM EDTA (pH ¼ 9.5), heated at 95
C for 1 min, and centrifuged for 5 min at 18,000  g. Two microliters of the supernatant is spotted onto a PEI-F cellulose TLC plate (EMD Chemicals), and separation of [a-32P]GTP and [a-32P]GDP is achieved using 0.9 M guanidine HCl as solvent (Bochner and Ames, 1982; Liu et al., 1998). The TLC plates are dried, phosphorimaged, and the extent of GTP hydrolysis is quantified by calculating the percentage of [a-32P]GTP hydrolyzed to [a-32P]GDP. Reactions in the

Single-Molecule Fluorescence Studies of Protein Synthesis

237

absence of 70S ribosomes, IF2g, or both, typically exhibit a basal level of 1% hydrolysis. Reactions in the presence of all reaction components exhibit 30% hydrolysis. An analogous assay is available for testing the GTPase activity of EF-G (Mohr et al., 2002).

3.2. Elongation assays During each elongation cycle, aa-tRNA, in a ternary complex with EF-Tu and GTP, is selected and incorporated into the A site. Peptidyl transfer from the P-site peptidyl-tRNA to the newly incorporated A-site aa-tRNA results in deacylation of the P-site tRNA and formation of a peptidyl-tRNA at the A site that has been elongated by one amino acid. Following peptidyl transfer, EF-G promotes translocation of the mRNA–tRNA complex by precisely one codon (Fig. 12.1B). We typically use the standard assays described below to test the biochemical activities of elongation components. 3.2.1. Primer-extension inhibition assay The toeprinting assay used to test initiation components (Section 3.1.1) (Hartz et al., 1988; Hartz et al., 1989) can be easily adapted for testing the activities of ribosomes, aa-tRNAs, and EFs in elongation (Fredrick and Noller, 2003; Joseph and Noller, 1998). An Initiation Reaction is performed in three steps: 1. A mixture of 35 pmol 70S ribosomes (or equivalent amounts of 30S and 50S subunits), 45 pmol IF1, 45 pmol IF2g, 45pmol IF3, and 40 nmol GTP is incubated for 10 min at 37
C. 2. 6.4 ml Primer Annealing Reaction (see Section 3.1.1) is added to the reaction, followed by a 10 min incubation at 37
C. 3. 45 pmol fMet-tRNAfMet is added to the reaction, followed by a 10 min incubation at 37
C. The final Initiation Reaction volume is 20 ml, prepared in Tris–polymix buffer (3 mM Mg2þ). The Initiation Reaction is then placed on ice until use. A Phe-tRNAPhe Ternary Complex is formed in three steps: 1. A GTP Charging Mix is prepared by mixing 200 nmol GTP, 600 nmol PEP, and 0.25 Units pyruvate kinase in a final volume of 20 ml, prepared in TC buffer (50 mM Tris–HOAc (pHRT ¼ 7.5), 100 mM KCl, 50 mM NH4OAc, 1 mM Ca(OAc)2, 0.1 mM EDTA, 5 mM Mg(OAc)2 and 6 mM BME). 2. An EF-Tu(GTP)/EF-Ts Mix is prepared by mixing 320 pmol of EF-Tu, 240 pmol of EF-Ts, and 2.2 ml GTP Charging Mix in a final volume of 20 ml, prepared in TC buffer, and incubating for 3 min at 37
C.

238

Jingyi Fei et al.

3. 30 pmol Phe-tRNAPhe is added to 15 ml EF-Tu(GTP)/EF-Ts Mix in a final volume of 20 ml, prepared in TC buffer, and the reaction is incubated for another 3 min at 37
C. 4. Phe-tRNAPhe Ternary Complex is then placed on ice until use. EF-G(GTP) is prepared by mixing 260 pmol EF-G with 2 ml GTP Charging Mix in a final reaction volume of 20 ml, prepared in Tris–polymix buffer (10 mM Mg2þ) and incubating for 3 min at 37
C. EF-G(GTP) is then placed on ice until use. Each Elongation Reaction is performed by mixing 12 ml of Initiation Reaction, 11.5 ml of Phe-tRNAPhe Ternary Complex, and 2 ml of EF-G (GTP) and incubating for 5 min at 37
C. Elongation Reactions are quenched by addition of 0.1 reaction volume 10 mM viomycin ( Joseph and Noller, 1998), a ribosome-targeting antibiotic that strongly inhibits EF-G-promoted translocation. Primer-Extension Reactions and all subsequent steps are performed as described in Section 3.1.1. Reactions in the absence and presence of Phe-tRNAPhe Ternary Complex and/or EF-G(GTP) are typically performed. In the absence of Phe-tRNAPhe Ternary Complex and EF-G(GTP), a strong þ15 toeprint corresponding to the initiated ribosomal complex is observed. In the absence of EF-G(GTP), Phe-tRNAPhe binding at the A site of the initiated ribosomal complex shifts the strong þ 15 toeprint to þ16. In the presence of EF-G(GTP), Phe-tRNAPhe binding at the A site of the initiated ribosomal complex followed by EF-G-catalyzed translocation further shifts the strong þ16 toeprint to þ18. Translocation efficiency is estimated by dividing the intensity of the þ18 toeprint by the sum of the intensities of the þ15, þ16, and þ 18 toeprints; we typically achieve 90% translocation efficiency in the first round of elongation. Toeprinting assays to assess two rounds of elongation (generating a þ21 toeprint) can be achieved by performing all reactions as outlined above, with the exception that a second ternary complex, Lys-tRNALys Ternary Complex (decoding the third codon, AAA, in the T4gp321–224 mRNA), is formed following the same procedure as that for Phe-tRNAPhe Ternary Complex formation above. Elongation Reactions are performed by mixing 12 ml of Initiation Reaction, 11.5 ml of Phe-tRNAPhe Ternary Complex, and 2.5 ml of EF-G(GTP), and incubating the reaction for 5 min at 37
C. This is followed by the addition of 11.5 ml of Lys-tRNALys Ternary Complex to the reaction and an additional incubation for 5 min at 37
C. Under these conditions, we typically achieve 90% and 70% translocation efficiencies in the first and second rounds of elongation, respectively. 3.2.2. Polypeptide synthesis assay In addition to the primer-extension inhibition assay, the activities of ribosomes, aa-tRNAs, and EFs in elongation can be independently assayed using a well-established polypeptide synthesis assay (Weinger et al., 2004). Each Elongation Reaction is performed in four steps:

Single-Molecule Fluorescence Studies of Protein Synthesis

239

1. An Initiation Reaction is prepared as described in Section 3.2.1, with the exception that the Primer Annealing Reaction is replaced with 9.2 pmol T4gp321-224 mRNA and the fMet-tRNAfMet is replaced with 0.3 pmol of f-[35S]Met-tRNAfMet (prepared by aminoacylating/formylating tRNAfMet as described in Section 2.4, with the exception that the 80 mM methionine is replaced with 16 mM methionine and 4 mM [35S] methionine (1175 Ci mmol 1, Perkin Elmer)). 2. Ternary Complexes are formed as described in Section 3.2.1, with the exception that the 30 pmol of Phe-tRNAPhe and, if included, 30 pmol Lys-tRNALys are decreased to 4.5 pmol each. 3. EF-G(GTP) is prepared as described in Section 3.2.1. 4. Elongation Reactions are performed as described in Section 3.2.1. Elongation Reactions are quenched by addition of 0.5 M KOH to a final concentration of 150 mM. Quenched reactions are spotted onto precoated, plastic-backed cellulose TLC plates (EMD Chemicals) and f-[35S]Met, f-[35S]Met-Phe, and f-[35S]Met-Phe-Lys products are separated using electrophoretic TLC (eTLC) as described in Youngman et al. (2004) using a 0.5% pyridine/20% glacial acetic acid buffer. eTLCs are run for 30 min at 1200 V, air-dried, phosphorimaged, and quantified in order to determine the percentage of f-[35S]Met that is converted to f-[35S]Met-Phe and the percentage of f-[35S]Met-Phe converted to f-[35S]Met-Phe-Lys. We typically achieve  70% conversion of f-[35S]Met to f-[35S]Met-Phe and 75% conversion of f-[35S]-Met-Phe to f-[35S]-Met-Phe-Lys using wild-type translation components.

3.3. Termination assays Once translocated into the ribosomal A site, stop codons are decoded by the class I release factors, RF1 or RF2. In response to a stop codon, RF1/2 binds at the A site and catalyzes hydrolysis of the nascent polypeptide chain from the P-site peptidyl-tRNA. Subsequently, the GTPase class II release factor, RF3, binds to the posthydrolysis, RF1/2-bound ribosomal complex in its GDP form, couples GDP-to-GTP exchange with the dissociation of RF1/2, and couples ribosome-stimulated GTP hydrolysis by RF3 with the dissociation of RF3 from the ribosomal complex. We typically use a standard polypeptide release assay, previously developed by Freistroffer et al. (1997), to test the biochemical activities of termination components. 3.3.1. Polypeptide release assay The activity of RF1/2 in polypeptide release is determined by performing a single-round fMet-[14C]Phe dipeptide release assay in the presence of excess RF3 without any guanine nucleotide (Zavialov et al., 2001). An Elongation Reaction is performed in four steps:

240

Jingyi Fei et al.

1. An Initiation Reaction is prepared as described in Section 3.2.1, with the exception that the Primer Annealing Reaction is replaced with 40 pmol of a variant T4gp321–224 mRNA containing AUG-UUU-UAA as the first three codons (i.e., encoding fMet-Phe-STOP). 2. Phe-tRNAPhe Ternary Complex is formed as described in Section 3.2.1, with the exception that the 30 pmol Phe-tRNAPhe is replaced with 15 pmol of [14C]Phe-tRNAPhe, prepared by aminoacylating tRNAPhe as described in Section 2.4, with the exception that the 55 mM phenylalanine is replaced with 55 mM [14C]phenylalanine (450 mCi mmol 1, Perkin Elmer). 3. EF-G(GTP) is prepared as described in Section 3.2.1. 4. An Elongation Reaction is performed by mixing 20 ml of Initiation Reaction, 20 ml of Phe-tRNAPhe Ternary Complex, and 4 ml EF-G (GTP) and incubating for 5 min at room temperature. An Elongation Reaction prepared in this way is stalled such that the stop codon at the third codon position of the mRNA resides at the A site. Free GTP and GDP are removed from the Elongation Reaction by buffer exchanging into Tris–polymix buffer (5 mM Mg2þ) using two successive Micro Bio-Spin 30 gel filtration spin columns. The stalled Elongation Reaction is then aliquoted, flash-frozen in liquid nitrogen, and stored at –80
C. Release Reactions are performed in two steps: 1. An (RF1/2)/RF3 Mix is prepared by mixing 0.05 pmol RF1/2 and 2 pmol RF3 in a final volume of 5 ml, prepared in Tris–polymix buffer (5 mM Mg2þ). 2. 5 ml Elongation Reaction and 5 ml (RF1/2)/RF3 Mix are preincubated separately for 1 min at 37
C, mixed together, and incubated for an additional 1 min at 37
C. Release Reactions are quenched and ribosomal complexes are precipitated by addition of 1 reaction volume of ice-cold 25% formic acid, incubation for 15 min on ice, and centrifugation at 14,000  g. The amount of [14C]Phe in the resulting pellet (containing unreacted ribosomal complexes still carrying P-site fMet-[14C]Phe-tRNAPhe as well as any free [14C] Phe-tRNA) and in the supernatant (containing released fMet-[14C]Phe dipeptide) is measured by scintillation counting and a calibration curve is used to convert the resulting cpm into molar amount of dipeptide released. Typically three reactions are performed to determine the activity of RF1/2 in polypeptide release. In the first reaction, the 5 ml (RF1/2)/RF3 Mix is replaced with 5 ml of 0.2 mM puromycin. Puromycin is a ribosometargeting antibiotic that mimics the aminoacyl-end of an aa-tRNA and quantitatively deacylates the P-site peptidyl-tRNA via peptidyl transfer; thus, the puromycin reaction reports on the total amount of P-site fMet-[14C]Phe-tRNAPhe that is competent for hydrolysis by RF1/2 (typically 85%).

Single-Molecule Fluorescence Studies of Protein Synthesis

241

The second and third reactions are performed in the absence and presence of (RF1/2)/RF3 Mix. The reaction in the absence of (RF1/2)/RF3 Mix reports the amount of uncatalyzed, background fMet-[14C]Phe dipeptide release, which is subtracted from the amount of fMet-[14C]Phe dipeptide released from the reaction in the presence of (RF1/2)/RF3. Dividing this corrected amount of fMet-[14C]Phe dipeptide released from the reaction in the presence of (RF1/2)/RF3 Mix by the amount of RF1/2 in the reaction yields the percent activity of RF1/2. Typically, wild-type RF1/2 exhibits a percent activity of 30–40%, in line with previous measurements (Zavialov et al., 2001). The stop-codon dependence of RF1/2-catalyzed peptide release is tested by replacing the mRNA in the Elongation Reactions such that ribosomes become stalled at a lysine sense codon (AAA) instead of at a stop codon (UAA); in this case, wild-type RF1/2 exhibits an undetected level of polypeptide release activity. RF3 catalyzes the dissociation of RF1/2 from the ribosome following polypeptide release, and is itself dependent on GTP hydrolysis for recycling off the ribosome. We therefore test RF3 activity by following the extent of polypeptide release in cases where RF1 is limiting and RF3 is required to actively recycle RF1, thereby enabling multiple turnover (Zavialov et al., 2001). All reactions are performed identically as above, with two major exceptions: when present, 2 nmol guanine nucleotide is added to the (RF1/ 2)/RF3 Mix and, upon adding the (RF1/2)/RF3/Nucleotide Mix to the Elongation Reaction, reactions are incubated for 10 min instead of 1 min. Typically reactions are performed without any nucleotide, with GDP, and with GTP; the dependence of multiple-turnover fMet-[14C]Phe dipeptide release on RF3 and GTP can be readily observed.

3.4. Ribosome recycling assays Following termination and the dissociation of both class I and II release factors, the resulting 70S posttermination complex, which contains just the mRNA and deacylated P-site tRNA, is dissociated into its respective 30S and 50S subunits through the joint action of RRF and EF-G in a GTP-dependent reaction (Hirokawa et al., 2005) (Fig. 12.1B). While the precise role of IF3 during recycling is still debated (Seshadri and Varshney, 2006), it has been suggested that IF3 is dispensable for actual subunit splitting, but plays a critical role by binding to the 30S subunit and both preventing dissociated subunits from reassociating and promoting the ejection of deacylated tRNA and mRNA (Peske et al., 2005; Zavialov et al., 2005) (Fig. 12.1B). Here, we describe a general assay, previously developed by Hirokawa et al., (2005), to monitor subunit dissociation by sucrose density gradient ultracentrifugation.

242

Jingyi Fei et al.

3.4.1. Subunit dissociation assay Recycling Reactions are performed by mixing 8 pmol 70S ribosomes, 800 pmol RRF, 800 pmol EF-G, 200 pmol IF3, and 20 nmol GTP in a final reaction volume of 40 ml, prepared in Tris–polymix buffer (6 mM Mg2þ), and incubating for 20 min at 37
C. After a brief incubation on ice, Recycling Reactions are loaded onto a 10–40% sucrose density gradient in the same Tris–polymix buffer, and 30S and 50S subunits are separated from 70S ribosomes by ultracentrifugation in an SW40 rotor (Beckman Coulter) at 25,000 rpm for 12 h at 4
C. Gradients are analyzed by monitoring the absorbance at 254 nm with a density gradient fractionator (Brandel) and 70S ribosome dissociation is qualitatively assessed by comparing the area of absorbance peaks corresponding to the dissociated 30S and 50S subunits with that corresponding to intact 70S ribosomes. Reactions are typically performed in the absence of all factors (i.e., with only 70S ribosomes), in the absence of just RRF, in the absence of just IF3, and with all of the factors present. The experiment performed in the absence of all factors reports on the extent of intrinsic subunit dissociation and typically yields predominantly intact 70S ribosomes. Similarly, predominantly intact 70S ribosomes are obtained in the absence of just RRF (since RRF is required for optimal subunit dissociation) or in the absence of just IF3 (since IF3 is required to prevent dissociated subunits from reassociating). The reaction in the presence of all factors, however, yields a significant population of dissociated 30S and 50S subunits, thereby demonstrating RRF’s subunit dissociation activity (Sternberg et al., 2009).

4. Preparation of Fluorescently Labeled Translation Components The spectroscopic properties of the Cy3 and Cy5 cyanine fluorophores make them an excellent donor (Cy3) and acceptor (Cy5) pair for smFRET studies of biomolecular systems. The efficiency of FRET between ˚ (Bastiaens Cy3 and Cy5, characterized by a Fo¨rster distance (R0) of 55 A and Jovin, 1996; Hohng et al., 2004), is most sensitive to the distance ˚ , a length between Cy3 and Cy5 within a distance range of 35–75 A scale that is ideal for probing conformational changes within the transla˚ tional machinery (the E. coli ribosome has maximum dimensions of 250 A in each direction (Schuwirth et al., 2005) and tRNAs are expected to move ˚ each, along a total through the ribosome in a series of steps that are tens of A ˚ path of length >100 A (Korostelev et al., 2008)). Thus, we have made extensive use of the Cy3/Cy5 FRET pair in our smFRET studies of protein synthesis. We routinely make use of amine-, thiol-, and aldehyde-/ketonereactive derivatives of Cy3 and Cy5, which are commercially available from

Single-Molecule Fluorescence Studies of Protein Synthesis

243

GE Healthcare. In our studies, we primarily use N-hydroxysuccimidyl (NHS) ester or maleimide derivatives of Cy3/5 to specifically label molecular constructs containing a single, unique amine or thiol group, respectively. In the sections below, we provide general protocols for designing labeling schemes and for specifically labeling ribosomes, tRNAs, and translation factors for smFRET studies of protein synthesis.

4.1. Phylogenetic analysis/structural modeling Generally speaking, the choice of labeling positions is guided by two criteria: (1) labeling positions should not be located within active sites or other highly conserved regions in order to minimize the risk of interfering with biological activity; (2) the distance between Cy3 and Cy5 should be close to R0, where the FRET efficiency will be most sensitive to changes in distance (Lakowicz, 1999). In order to achieve these criteria, phylogenetic analysis and structural modeling of the target molecules and/or complexes is usually necessary. For our smFRET studies of protein synthesis, we typically perform phylogenetic analysis using multiple sequence alignments of protein or RNA sequences from a variety of bacterial species (20–50 species) using BLAST (http://www.ncbi.nlm.nih.gov/BLAST) (Altschul et al., 1990) and CLUSTAL-W (http://www.ebi.ac.uk/clustalw) (Thompson et al., 1994). Poorly conserved amino acid residues or nucleotides that are distal from active sites can be identified based on the alignments and selected as candidate positions for labeling. Structural modeling usually involves the comparison of coordinates derived from cryo-EM reconstructions and/or X-ray crystal structures of relevant ribosomal complexes using molecular visualization software such as PyMOL (http://pymol.sourceforge.net) (DeLano, 2008) or Swiss PDB Viewer (http://spdbv.vital-it.ch) (Guex and Peitsch, 1997). Typically, superpositions of various functionally related complexes are performed in order to identify Cy3 and Cy5 labeling positions where the conformational rearrangement of interest is expected to result in a relative distance change between the two fluorophores that corresponds to a maximal change in FRET (bearing in mind that the FRET efficiency of the Cy3/Cy5 FRET pair is most sensitive to changes in distance for inter-fluorophore distances in the range of 35–75 A˚). Based on the results of phylogenetic analysis and structural modeling, we typically design a minimum of three candidate labeling constructs which are generated, fluorescently labeled, biochemically assayed, and used for preliminary smFRET experiments. Based on the results of these experiments, the optimal construct is identified and chosen for detailed biochemical characterization and smFRET data collection.

244

Jingyi Fei et al.

4.2. Ribosome labeling Ribosomes can be fluorescently labeled at either rRNA or ribosomal proteins (r-proteins), depending on the specific experiment. An approach for labeling the ribosome based on hybridization of fluorescently labeled oligonucleotides to helical extensions engineered into surface-exposed rRNA hairpins has been described (Dorywalska et al., 2005) and one such construct has been recently used to conduct smFRET studies of ribosome dynamics (Marshall et al., 2008b, 2009). Here, we describe a method for labeling ribosomes that involves reconstitution of fluorescently labeled rproteins into mutant ribosomes lacking the target r-proteins. In the sections below we describe our general approach, developed using r-proteins L1 and L9 as targets (Fei et al., 2008, 2009; Sternberg et al., 2009). 4.2.1. Preparation of mutant ribosomes lacking target r-proteins Ribosomes lacking a single r-protein (Fei et al., 2008) are obtained from single-deletion E. coli strains generated using a one-step gene deletion technique originally developed by Baba et al. (2006) and Datsenko and Wanner (2000) (Fig. 12.2A). Briefly, strain BW25113, a recombinationproficient derivative of E. coli K12, is transformed with the Red helper plasmid pKD46 encoding the l Red recombination system under the control of the arabinose-inducible, ParaB promoter. A linear DNA fragment targeting the r-protein gene of interest is constructed by PCR amplification using plasmid pKD13 (carrying a kanamycin resistance cassette) or pKD3 (carrying a chloramphenicol resistance cassette) as a template. The 30 -ends of the PCR primers used to generate the linear DNA fragment contain 20 nucleotides complementary to the sequences flanking the antibiotic resistance genes in pKD13 or pKD3 while the 50 -ends contain 50 nucleotide extensions homologous to E. coli chromosomal sequences immediately upstream and downstream of the gene encoding the target r-protein. A 500–800 ng of linear DNA fragment is electroporated into electrocompetent BW25113(pKD46) cells. Cells are grown in antibiotic-free SOC media supplemented with 1 mM of L-arabinose for 2 h at 37
C, spread onto agarose plates supplemented with the appropriate antibiotic (30 mg ml 1), and incubated at 37
C. Antibiotic resistant colonies are selected and grown in Luria-Bertani (LB) media (Difco) and gene deletion is verified by PCR amplification of the targeted region of the chromosome and DNA sequencing. Ribosomes lacking two r-proteins (Fei et al., 2009) are obtained from double-deletion E. coli strains generated by P1 vir phage transduction of a donor single-deletion strain into a recipient single-deletion strain (Goldberg et al., 1974; Moore and Sauer, 2009; Wall and Harriman, 1974). In our case, we construct a chloramphenicol-resistant donor single-deletion strain using pKD3 and a kanamycin-resistant recipient single-deletion strain using

A L1

Plasmid encoding wild-type L1

Wild-type strain L1

Mutagenesis

Ka n r

L1 ΔL1 strain

Plasmid encoding single-cys L1 mutant

Gene deletion Kanr

Protein purification

Ribosome purification

cys-L1

(-)L1 ribosome

Fluorescent labeling (Cy5)L1

Reconstitution

(Cy5)L1 ribosome

B d5

0S

Wild-type ribosomes

ute

ute

fMet

con

stit

fMet-tRNA Phe Phe-tRNA Phe Phe-(Cy3)tRNA EF-G(GTP) Overlay

Fluorescence

Re

stit con

ld-t Wi Coomassie

Re

ype

50S

d3 0S

C − − − −

+ − − −

(Cy5)L1 ribosomes

+ + − −

+ + − +

+ − − −

+ + − −

3

4

5

6

+ + − +

+ − − −

+ − + −

+ − + +

8

9

10

+15 +16 +18 L1

1

Figure 12.2

2

7

Preparation and characterization of (Cy5)L1 ribosomes. (A) Generation of (Cy5)L1 ribosomes. Ribosomes lacking r-protein L1

246

Jingyi Fei et al.

pKD13. Using these donor and recipient single-deletion strains, we then follow a variation of the phage P1 vir transduction protocols developed by Sauer and coworkers (http://openwetware.org/wiki/Sauer:P1vir_phage_ transduction). Briefly, a 2.5 ml culture of the donor single-deletion strain is infected with P1 vir phage and grown for 1–3 h until the culture becomes clear, indicating that the cells have been completely lysed. The resulting lysate contains transducing particles which carry random fragments of the donor single-deletion strain genome, including the kanamycin resistance gene; this lysate is used to infect a liquid culture of the recipient singledeletion strain. Double-deletion mutants are selected on agarose plates supplemented with both kanamycin and chloramphenicol. Antibiotic resistant colonies are selected and grown in LB media, and gene deletion is verified by PCR amplification and DNA sequencing. Single- and double-deletion strains may exhibit a slow-growth phenotype whose severity will depend on the specific r-protein(s). In our case, the growth rate of the L9 single-deletion strain was comparable to that of the wild-type BW25113 strain, while the doubling times of the L1 singledeletion strain and L1/L9 double-deletion strain were approximately twoand sixfold slower than the wild-type strain, respectively. Tightly coupled

((–)L1 ribosomes) are purified from an E. coli strain in which the gene encoding r-protein L1 has been deleted by an in-frame knock out (DL1). In parallel, r-protein L1 is cloned and mutagenized to generate a single-cysteine variant. The single-cysteine mutant L1 is purified and labeled with Cy5-maleimide. (Cy5)L1 is then in vitro reconstituted with (–)L1 ribosomes in order to generate (Cy5)L1-labeled ribosomes. (B) Incorporation of (Cy5)L1 into (–)L1 ribosomes. Coomassie staining (left), fluorescence scanning (middle), and overlay (right) of an SDS-PAGE gel containing ribosomal proteins extracted from wild-type and reconstituted ribosomal subunits. (C) Elongation toeprinting assay. The activities of unlabeled and (Cy3/5)-labeled translation elongation components are tested by a standard toeprinting assay. cDNA bands corresponding to mRNA positions þ 15, þ 16, and þ 18, relative to the A of the AUG start codon, report on the formation of a 70S initiation complex (þ 15), the incorporation of the first A-site aa-tRNA (Phe-tRNAPhe) (þ 16), and a single translocation step (þ 18). Lane 1 is a control primer extension of the mRNA in the absence of any translation components that is used to detect sites of reverse transcriptase inhibition caused by local secondary structures within the mRNA. The intensities of the bands corresponding to the þ15, þ 16, and þ 18 toeprints in this lane are used to correct the raw intensities of the þ15, þ 16, and þ 18 toeprints in Lanes 2–10. The activities of unlabeled ribosomes with unlabeled Phe-tRNAPhe (compare Lane 2 with Lanes 3 and 4), (Cy5)L1 ribosomes with unlabeled Phe-tRNAPhe (compare Lane 5 with Lanes 6 and 7), and (Cy5)L1 ribosomes with Phe-(Cy3)tRNAPhe (compare Lane 8 with Lanes 9 and 10) are indistinguishable. Comparison of the corrected intensities of the þ15 and þ 18 toeprints in Lanes 4, 7, and 10 suggests that for all combinations of unlabeled and labeled components, 70S initiation complexes are  90% active in the first round of elongation.

Single-Molecule Fluorescence Studies of Protein Synthesis

247

70S ribosomes lacking one or two r-proteins are purified from single- or double-deletion BW25113 strains, respectively, using the protocol described in Section 2.2. 4.2.2. Preparation of fluorescently labeled r-proteins Fluorescently labeled r-proteins are prepared in four steps: 1. The target r-protein genes are PCR-amplified from C600 genomic DNA and cloned into the pProEX-HTb plasmid system (Section 2.5). 2. Cloned r-protein genes are mutagenized using the QuickChange Mutagenesis Kit (Stratagene) to mutate wild-type cysteine residues to nonreactive amino acids (serine is a typical structurally and chemically conservative choice) and to introduce a unique cysteine residue at a position selected through phylogenetic analysis and structural modeling (Section 4.1). 3. Single-cysteine r-protein mutants are overexpressed and purified under denaturing conditions (described below). 4. Single-cysteine r-protein mutants are labeled with maleimide derivatives of Cy3/5 (described below). Overexpression and purification of r-proteins follows the protocol for translation factor purification presented in Section 2.5, with the following modifications. Cells from a 500 ml culture are lysed in r-Protein Buffer A (50 mM Tris–HCl (pH4
C ¼ 8), 5 mM MgCl2, 0.1 mM PMSF, and 5 mM BME) and the resulting lysate is cleared by centrifugation at 10,000  g for 45 min at 4
C. An SDS-PAGE gel is used to determine whether the majority of the overexpressed r-protein partitions into the supernatant or into insoluble inclusion bodies that co-sediment with the cell pellet. For r-proteins that primarily partition into inclusion bodies, such as L1 and L9, the pellet is resuspended in r-Protein Buffer B (10 mM Tris–HCl (pH4
C ¼ 8), 100 mM NaH2PO4 (pH ¼ 8), 6 M urea, 0.1 mM PMSF, and 5 mM BME) by gently stirring overnight at 4
C. For r-proteins that primarily partition into the supernatant, the supernatant is dialyzed against r-Protein Buffer B overnight at 4
C. The resulting r-protein mixture is cleared again by centrifugation at 12,000  g for 30 min at 4
C. 6xHistagged r-proteins are purified as described in Section 2.5 with the exception that the Ni2þ-NTA column is washed with 8 column volumes of r-Protein Buffer C (r-Protein Buffer B adjusted to pH4
C ¼ 6.7) and r-proteins are eluted with r-Protein Buffer D (r-Protein Buffer B adjusted to pH4
C ¼ 5.5). r-Protein-containing fractions are combined, diluted to an r-protein concentration of 0.1–0.2 mg ml 1 (as measured by the Bradford assay), and dialyzed extensively against r-Protein Buffer E (50 mM Na2HPO4 (pH ¼ 7.0), 100 mM NaCl, and 2 mM BME) to remove urea and renature the r-protein. Renatured r-protein is concentrated to 0.5–1 mg ml 1, 6xHis-tagged TEV protease is added, and dialysis against

248

Jingyi Fei et al.

r-Protein Buffer E is continued. Cleaved r-protein is separated from uncleaved r-protein, 6xHis-tag fragments, and 6xHis-tagged TEV protease using a second Ni2þ-NTA column as described in Section 2.5 with the exception that the Ni2þ-NTA resin is preequilibrated against r-Protein Buffer E. The cleaved, purified r-protein is dialyzed or gel filtered into 2 r-Protein Buffer F (50 mM Na2HPO4 (pH ¼ 7.0), 200 mM NaCl, and 2 mM BME), concentrated using a centrifugal filtration device, diluted to 1 r-Protein Buffer F by addition of 100% glycerol, and stored at –20
C. Final yields of 10–20 mg of r-protein per liter culture are typically obtained. Fluorescent labeling of r-proteins is generally performed in a Tris- or phosphate-based labeling buffer at pH ¼ 7.0–7.5, with the exact composition varying depending on the specific r-protein. As examples, L1 labeling buffer is composed of 100 mM Na2HPO4 (pH ¼ 7.2), 100 mM NaCl, and a 100-fold molar excess of tris(2-carboxyethyl)phosphine hydrochloride (TCEP, a nonthiol-containing reducing agent which selectively reduces disulfides) over L1, while L9 labeling buffer is composed of 50 mM Tris– HCl (pHRT ¼ 7.2), 200 mM KCl, 4 M urea, and a 100-fold excess of TCEP over L9. r-Protein is buffer exchanged into labeling buffer and concentrated to 40 mM using a centrifugal filtration device, and the resulting solution is incubated for 30 min at room temperature in order to fully reduce r-protein disulfide bonds. A 20-fold molar excess of Cy3/5maleimide, predissolved in a minimum volume (typically less than 5% of the total reaction volume) of anhydrous dimethyl sulfoxide (DMSO), is added to the r-protein solution and the labeling reaction is incubated for 2 h at room temperature followed by a minimum of 5 h at 4
C. The reaction is quenched by adding BME to a final concentration of 6 mM. Labeled proteins are separated from unreacted, free Cy3/Cy5 using a HiLoad 16/60 Superdex 75 prep grade gel filtration column (GE Healthcare) preequilibrated against gel filtration buffer. Again, the exact composition of the gel filtration buffer will vary depending on the r-protein; L1 gel filtration buffer is 20 mM Tris–HCl (pHRT ¼ 7.8), 200 mM NaCl, 2 mM MgCl2, and 6 mM BME and L9 gel filtration buffer is 20 mM Tris–HCl (pHRT ¼ 7.8), 400 mM NH4Cl, 4 mM MgCl2, 4 M urea and 6 mM BME. The labeling efficiencies are typically 65–100% for L1 and 50% for L9. 4.2.3. Reconstitution of fluorescently labeled r-proteins into mutant ribosomes lacking target r-proteins Reconstitution generally involves incubation of mutant ribosomes lacking the target r-protein(s) with a molar excess of the purified r-protein(s). The specific concentrations of ribosomes and r-protein(s), as well as the buffer conditions, incubation time, and temperature, will generally need to be optimized for specific r-protein(s). As a starting point, here we provide references and protocols for reconstituting (Cy3/5)L1 and (Cy3/5)L9 into 50S subunits lacking L1 ((–)L1), L9 ((–)L9), or both L1 and L9 ((–)L1/L9).

Single-Molecule Fluorescence Studies of Protein Synthesis

249

(Cy3/5)L1 is reconstituted into (–)L1 50S subunits by incubating 1.8 nmol (Cy3/5)L1 and 1.2 nmol (–)L1 50S subunits in 300 ml of L1 reconstitution buffer (10 mM Tris–HCl (pH37
C ¼ 7.5), 8 mM Mg(OAc)2, 150 mM NH4Cl, and 5 mM BME) for 10 min at 35
C (Odom et al., 1990) (Fig. 12.2A). (Cy3/5)L9 is reconstituted into (–)L9 50S subunits by incubating 1.8 nmol (Cy3/5)L9 and 1.2 nmol (–)L9 50S subunits in 300 ml of L9 reconstitution buffer (50 mM HEPES(KOH) (pH37
C ¼ 7.5), 4 mM MgCl2, 400 mM NH4Cl, 6 mM BME, and 0.1% Nikkol) for 15 min at 37
C (Ermolenko et al., 2007). (Cy3/5)L1 and (Cy3/5)L9 are reconstituted into (–)L1/L9 50S subunits by incubating 1.8 nmol (Cy3/5)L1 and 1.2 nmol (–)L1/L9 50S subunits in 300 ml of L1/L9 reconstitution buffer (20 mM Tris– HCl (pHRT ¼ 7.85), 4 mM MgCl2, 400 mM NH4Cl, and 6 mM BME) for 15 min at 37
C followed by addition of 1.8 nmol (Cy3/5)L9 and an additional 10 min incubation at 37
C. Reconstituted, fluorescently labeled 50S subunits are purified from unincorporated (Cy3/5)L1 and/or (Cy3/5)L9 using sucrose density gradient ultracentrifugation (Section 2.2). Under these conditions we achieve reconstitution efficiencies of 100% for (Cy3/5)L1 (Fig. 12.2B) and 60% for (Cy3/5)L9 (Fei et al., 2008, 2009). Reconstituted, fluorescently labeled 50S subunits are fully active in the elongation toeprinting assay described in Section 3.2.1 (Fig. 12.2C) (Fei et al., 2008, 2009.

4.3. tRNA labeling 4.3.1. tRNAfMet labeling Fluorescent labeling of initiator tRNAfMet at the 4-thiouridine at nucleotide position 8 (s4U8) is achieved via reaction with Cy3/5-maleimide using slight modifications of a previously published protocol (Carbon and David, 1968). Labeling is achieved by incubating 13 nmol tRNAfMet and 650 nmol Cy3/5maleimide in 150 ml tRNAfMet labeling buffer (50 mM Tris–HCl (pH37
C ¼ 7.8)) for 5 h at 37
C. The labeling reaction is quenched with 0.1 reaction volume of 3 M NaOAc (pH ¼ 5.5). Multiple extractions with 1 reaction volume phenol are performed until unreacted Cy3/5 is no longer visibly extracted (this typically requires approximately six phenol extractions). Phenol phases are saved and back-extracted with 0.25 volume of 0.4 M NaOAc (pH ¼ 5.5) and the back-extracted aqueous phase is combined with the original aqueous phase. The pooled sample is extracted twice with 1 reaction volume chloroform, and ethanol precipitated by addition of 3 reaction volume of –20
C ethanol and overnight incubation at –20
C, and finally centrifuged at 18,000  g for 20 min at 4
C. The tRNAfMet pellet is resuspended in tRNA HIC Buffer A and (Cy3/5) tRNAfMet is separated from unlabeled tRNAfMet using HIC as described in Section 2.4. (Cy3)tRNAfMet elutes from the Phenyl-5PW column at 34.5% tRNA HIC Buffer B whereas (Cy5)tRNAfMet typically elutes as two peaks at 36.5% and 45% tRNA HIC Buffer B. While it is currently not known

250

Jingyi Fei et al.

why (Cy5)tRNAfMet elutes as two peaks, it is possible that the two peaks arise from the distinct hydrophobicities of two interconverting isomers of (Cy5) tRNAfMet; in support of this possibility, when the peak eluting at 45% tRNA HIC Buffer B is collected, incubated at 37
C for 10 min, and reinjected onto the Phenyl-5PW column, two peaks are again eluted with the same retention times as before. Using this protocol, a labeling efficiency of 2–5% is typically achieved. HIC fractions containing the 95–98% of unlabeled tRNAfMet can be relabeled as described above, yielding a similar, 2–5% labeling efficiency; this observation suggests that the degree of s4U8 modification within the tRNAfMet sample is not limiting the reaction. Instead, it is likely that hydrolysis of the Cy3/5-maleimide to Cy3/5-maleamic acid effectively outcompetes reaction of Cy3/5-maleimide with the thione group of s4U8 (Carbon and David, 1968). Attempts to further optimize reaction conditions in order to obtain labeling efficiencies above 5% have not been successful. The eluted (Cy3/5)tRNAfMet is buffer exchanged and concentrated into Barnstead NANOpure water using a centrifugal filter device (MWCO ¼ 10,000). We routinely achieve >90% aminoacylation/formylation efficiency of (Cy3/5) tRNAfMet using the procedures described in Section 2.4, with the exception that the concentrations of methionyl tRNA synthetase and formylmethionyltRNA formyltransferase are increased to 0.2 and 2 mM, respectively. fMet(Cy3)tRNAfMet elutes from the Phenyl-5PW column at  41.5% tRNA HIC Buffer B and the two fMet-(Cy5)tRNAfMet peaks elute at 43% and 51% tRNA HIC Buffer B. fMet-(Cy3/5)tRNAfMet are fully active in the IF2g toeprinting assay described in Section 3.1.1 ( Jiangning Wang and Ruben L. Gonzalez, unpublished data) as well as the elongation toeprinting assay described in Section 3.2.1 (Blanchard et al., 2004b). 4.3.2. tRNAPhe labeling Fluorescent labeling of E. coli tRNAPhe (Sigma) at the primary aliphatic amino group of the 3-(3-amino-3-carboxypropyl)-uridine at position 47 (acp3U47) is achieved by reaction with Cy3/5-NHS esters (Fig. 12.3A) using slight modifications of a previously published protocol (Plumbridge et al., 1980). Labeling is achieved by incubating 10 nmol of tRNAPhe and 200 nmol Cy3/5-NHS ester in 75 ml tRNAPhe labeling buffer (50 mM HEPES (pH ¼ 8.0), 0.9 M NaCl) for 8 h at 30
C, followed by overnight incubation at 4
C. The reaction is quenched, phenol extracted, chloroform extracted, and ethanol precipitated as described above for (Cy3/5)tRNAfMet. (Cy3/5) tRNAPhe is separated from unlabeled tRNAPhe using HIC (Fig. 12.3B) as described in Section 2.4. (Cy3)tRNAPhe and (Cy5)tRNAPhe elute from the Phenyl-5PW column at 55% and 61% tRNA HIC Buffer B, respectively. Using this protocol, a labeling efficiency of 30% is routinely achieved. (Cy3/5)tRNAPhe can be aminoacylated with >90% efficiency (Fig. 12.3C) using the method described in Section 2.4 and is fully active in the elongation

A

-O S

tRNAPhe

SO

3

N

SO3-

-O S 3

N+

+

Cy3-NHS ester

acp3U47

3

O O N

O

+

NH

H

N

O

O

COOH

O

2

HOOC

HO

N

O

H

H

O

H N

O HO

OH OH

O N

N

H

O

OH OH

acp3U47 B

C

Purification of (Cy3)tRNAPhe

100

100

Phe-(Cy3)tRNA

40

80

80

tRNAPhe (Cy3)tRNAPhe

Phe-(Cy3)tRNAPhe

Phe

60 % tRNA HIC buffer B

Co-injection of (Cy3)tRNAPhe and Phe-(Cy3)tRNAPhe

Aminoacylation of (Cy3)tRNAPhe

(Cy3)tRNAPhe

60

60 40

40

20

20

20

0

0 0

20 40 60 Retention volume (ml)

0 0

20 40 60 80 Retention volume (ml)

tRNAPhe (260 nm)

Cy3 (550 nm)

0

20 40 60 80 Retention volume (ml)

Conc B (%)

Figure 12.3 Preparation of Phe-(Cy3/5)tRNAPhe. (A) Left panel: Structure of tRNAPhe, indicating the 3-(3-amino-3-carboxypropyl)-uridine residue at position 47 (acp3U47) (indicated in green) whose primary aliphatic amino group is reacted with Cy3-NHS ester. Right panel: Reaction chemistry involved in labeling acp3U47 with Cy3-NHS ester. (B) HIC chromatogram demonstrating the separation of (Cy3)tRNAPhe from

252

Jingyi Fei et al.

toeprinting assay described in Section 3.2.1 (Fig. 12.2C) (Blanchard et al., 2004b; Fei et al., 2008).

4.4. Translation factor labeling Single-cysteine translation factor mutants are designed and constructed as described above for single-cysteine r-proteins and are overexpressed and purified as described in Section 2.5. The general labeling procedure described in the following paragraph was developed using a single-cysteine RF1 mutant (Sternberg et al., 2009), but can be easily extended to any translation factor containing a unique cysteine. Fluorescent labeling of translation factors generally follows the procedures described above for fluorescent labeling of r-proteins (Section 4.2.2). Briefly, labeling is performed in TF labeling buffer (100 mM Tris–HOAc (pH25
C ¼ 7.0), 50 mM KCl, and a 10–20-fold molar excess of TCEP over translation factor). The translation factor is buffer exchanged into TF labeling buffer, concentrated to 50–100 mM, and incubated for 15–30 min at room temperature in order to fully reduce translation factor disulfide bonds. A 10–20-fold molar excess of Cy3/5-maleimide, predissolved in a minimum amount of anhydrous DMSO, is added to the translation factor solution and the labeling reaction is incubated for 1–2 h at room temperature with occasional mixing followed by an overnight incubation at 4
C. The reaction is quenched with BME and the translation factor is separated from free Cy3/5 on an appropriately sized gel filtration column (a HiLoad 16/60 Superdex 75 prep grade column is appropriate for all translation factors except IF2g, EF-G, and RF3, where a HiLoad 16/60 Superdex 200 prep grade should be used instead) using TF Buffer E (Section 2.5) as a column preequilibration and running buffer. Translation factor-containing fractions are pooled, buffer exchanged and concentrated into 2 TF Buffer E using a centrifugal filtration device, diluted to 1 TF Buffer E by addition of 100% glycerol, and stored at –20
C. The labeling efficiency can be estimated from the gel filtration chromatogram using the translation factor and Cy3/5 extinction coefficients and the integrated absorbance at 280 nm (translation factor) and 550 nm (Cy3) or 650 nm (Cy5) of the protein peak. Using these protocols, the labeling efficiency of a single-cysteine RF1 mutant is 40–60% (Sternberg et al., 2009). unlabeled tRNAPhe. Peaks corresponding to (Cy3)tRNAPhe and tRNAPhe are labeled. (C) Left panel: HIC chromatogram demonstrating that (Cy3)tRNAPhe can be aminoacylated with phenylalanine with an efficiency of > 90%. Right panel: HIC chromatogram of a coinjection of equimolar amounts of (Cy3)tRNAPhe and Phe-(Cy3)tRNAPhe, demonstrating the shift in elution position of Phe-(Cy3)tRNAPhe relative to (Cy3) tRNAPhe, thereby confirming the assignment of the peak in the left panel as Phe-(Cy3)tRNAPhe and confirming the > 90% aminoacylation efficiency.

Single-Molecule Fluorescence Studies of Protein Synthesis

253

Depending on the labeling efficiency, it may be necessary or desirable to further separate (Cy3/5)translation factor from unlabeled translation factor. Experiments that involve the binding of (Cy3/5)translation factor to (Cy3/5)ribosomes will suffer from a low detection of smFRET events when the ratio of unlabeled translation factor to (Cy3/5)translation factor is high. Additionally, the interpretation of biochemical activity assays may be complicated when unlabeled translation factor may compete with (Cy3/ 5)translation factor. Taking advantage of the added hydrophobicity that Cy3/5 imparts onto the translation factor, (Cy3/5)translation factor can be efficiently separated from unlabeled translation factor using HIC. Translation factor-containing fractions from the gel filtration purification are buffer exchanged into TF HIC Buffer A (100 mM Na2HPO4 (pH25
C ¼ 7.0), 1 M (NH4)2SO4), concentrated using a centrifugal filtration device and subsequently injected onto the Phenyl-5PW column (Section 2.4) preequilibrated against TF HIC Buffer A. (Cy3/5)translation factor is then separated from unlabeled translation factor by elution with a linear gradient of 0–100% TF HIC Buffer B (100 mM Na2HPO4, pH25
C ¼ 7.0) over 16 column volumes. This purification procedure yields 100% homogenously labeled (Cy3/5)translation factor, which is buffer exchanged into 2 TF Buffer E (Section 2.5) and concentrated using a centrifugal filtration device, diluted to 1 TF Buffer E with 100% glycerol, and stored at –20
C. (Cy5)RF1, prepared and purified as described here, is fully active in the polypeptide release assay described in Section 3.3.1 (Sternberg et al., 2009).

5. Conclusions and Future Perspectives Over the past few years, the highly purified, fluorescently labeled in vitro translation system described here, and similar systems in our colleagues’ laboratories, have been successfully used in numerous smFRET studies of protein synthesis (reviewed in Blanchard, 2009; Frank and Gonzalez, 2010; Marshall et al., 2008a). Going forward, it is imperative that the basic experimental system described here be expanded in several important directions. For example, smFRET studies of functionally disrupted mutant ribosomes, tRNAs, and translation factors will be crucial for determining the molecular basis through which ribosome, tRNA, and translation factor dynamics are coupled to each other and to the mechanisms underlying protein synthesis; to date only one such smFRET study of mutant ribosomes has been performed (Munro et al., 2007). In addition to mutant translation components, relatively straightforward expansions of the experimental system described here should allow smFRET studies to test the hypothesis that regulatory factors might exert

254

Jingyi Fei et al.

their control over protein synthesis by specifically altering the dynamics of the translating ribosome. To name a few candidate factors: the reverse translocase LepA (EF-4), which promotes reverse translocation of the ribosome along its mRNA template by precisely one codon (Qin et al., 2006); elongation factor P (EF-P), which stimulates formation of the first peptide bond at the interface between the initiation and elongation stages of protein synthesis (Aoki et al., 1997); the ribosomal protection protein Tet (O), which catalyzes the dissociation of the ribosome-targeting antibiotic tetracycline from ribosomes, thereby protecting ribosomes from inhibition by tetracycline (Spahn et al., 2001); and the specialized, EF-Tu-like elongation factor SelB, which, in response to a cis-acting mRNA structural element, recodes a stop codon with a selenocysteine tRNA in order to incorporate selenocysteine into selenoproteins (Forchhammer et al., 1989). Perhaps the most important and challenging extension of our experimental platform is the establishment of a eukaryotic-based, highly purified, fluorescently labeled in vitro translation system. Development of such a system would open the door to detailed mechanistic studies of eukaryotic translation initiation and its regulation. This is a critical area of ongoing mechanistic research that is driven by the increased complexity of the eukaryotic translation initiation machinery relative to its bacterial counterpart (Kapp and Lorsch, 2004), the prominent role of translation initiation in the translational control of eukaryotic gene expression (Sonenberg and Hinnebusch, 2009), and the increasingly apparent correlation between the deregulation of translation initiation and human diseases such as cancer (Clemens, 2004) and viral infections (Schneider and Mohr, 2003).

ACKNOWLEDGMENTS R. L. G. would like to acknowledge scientific and financial support from Profs. Joseph D. Puglisi (Stanford University School of Medicine) and Steven Chu (formerly at Stanford University, currently at the U.S. Department of Energy), in whose laboratories early work on the development of the highly purified, fluorescently labeled in vitro translation system described here was performed. R. L. G. would also like to acknowledge Scott C. Blanchard (formerly at the Stanford University School of Medicine, currently at Weill Cornell Medical School) and Harold D. Kim (formerly at Stanford University, currently at the Georgia Institute of Technology), with whom the early work in the Puglisi and Chu laboratories was collaboratively performed. R. L. G.’s postdoctoral work in the Puglisi and Chu laboratories was funded by postdoctoral fellowships from the American Cancer Society (PF-01-119-01-GMC) and the Burroughs Wellcome Fund (CABS 1004856). Work in the Gonzalez laboratory is supported by a Burroughs Wellcome Fund CABS Award (CABS 1004856), an NSF CAREER Award (MCB 0644262), an NIH-NIGMS R01 grant (R01 GM084288), an American Cancer Society Research Scholar Grant (RSG GMC-117152), and a Columbia University Research Initiatives in Science and Engineering (RISE) Award to R.L.G. S.H.S. was supported by the Columbia University Langmuir Scholars Program. M. M. E. and M. T. E. are supported by Columbia University’s NIH Training Program in Molecular Biophysics (T32-GM008281).

Single-Molecule Fluorescence Studies of Protein Synthesis

255

REFERENCES Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990). Basic local alignment search tool. J. Mol. Biol. 215, 403–410. Antoun, A., Pavlov, Michael Y., Tenson, Tanel, and Ehrenberg, M. (2004). Ribosome formation from subunits studied by stopped-flow and Rayleigh light scattering. Biol. Proced. Online 6, 35–54. Aoki, H., Adams, S. L., Turner, M. A., and Ganoza, M. C. (1997). Molecular characterization of the prokaryotic efp gene product involved in a peptidyltransferase reaction. Biochimie 79, 7–11. Baba, T., Ara, T., Hasegawa, M., Takai, Y., Okumura, Y., Baba, M., Datsenko, K. A., Tomita, M., Wanner, B. L., and Mori, H. (2006). Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: The Keio collection. Mol. Syst. Biol. 2, 0008. Bastiaens, P. I., and Jovin, T. M. (1996). Microspectroscopic imaging tracks the intracellular processing of a signal transduction protein: Fluorescent-labeled protein kinase C beta I. Proc. Natl. Acad. Sci. USA 93, 8407–8412. Blanchard, S. C. (2009). Single-molecule observations of ribosome function. Curr. Opin. Struct. Biol. 19, 103–109. Blanchard, S. C., Gonzalez, R. L., Kim, H. D., Chu, S., and Puglisi, J. D. (2004a). tRNA selection and kinetic proofreading in translation. Nat. Struct. Mol. Biol. 11, 1008–1014. Blanchard, S. C., Kim, H. D., Gonzalez, R. L. Jr., Puglisi, J. D., and Chu, S. (2004b). tRNA dynamics on the ribosome during translation. Proc. Natl. Acad. Sci. USA 101, 12893–12898. Bochner, B. R., and Ames, B. N. (1982). Complete analysis of cellular nucleotides by twodimensional thin layer chromatography. J. Biol. Chem. 257, 9759–9769. Brandi, L., Marzi, S., Fabbretti, A., Fleischer, C., Hill, W. E., Gualerzi, C. O., and Lodmell, J. S. (2004). The translation initiation functions of IF2: Targets for thiostrepton inhibition. J. Mol. Biol. 335, 881–894. Cammack, K. A., and Wade, H. E. (1965). The sedimentation behaviour of ribonucleaseactive and -inactive ribosomes from bacteria. Biochem. J. 96, 671–680. Carbon, J., and David, H. (1968). Studies on the thionucleotides in transfer ribonucleic acid. Addition of N-ethylmaleimide and formation of mixed disulfides with thiol compounds. Biochemistry 7, 3851–3858. Chambliss, G. H., Henkin, T. M., and Leventhal, J. M. (1983). Bacterial in vitro proteinsynthesizing systems. Methods Enzymol. 101, 598–605. Clemens, M. J. (2004). Targets and mechanisms for the regulation of translation in malignant transformation. Oncogene 23, 3180–3188. Datsenko, K. A., and Wanner, B. L. (2000). One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97, 6640–6645. Daviter, T., Gromadski, K. B., and Rodnina, M. V. (2006). The ribosome’s response to codon-anticodon mismatches. Biochimie 88, 1001–1011. DeLano, W. (2008). The PyMOL Molecular Graphics System. DeLano Scientific LLC, Palo Alto, CA, http://www.pymol.org. Dincbas-Renqvist, V., Engstrom, A., Mora, L., Heurgue-Hamard, V., Buckingham, R., and Ehrenberg, M. (2000). A post-translational modification in the GGQ motif of RF2 from Escherichia coli stimulates termination of translation. EMBO J. 19, 6900–6907. Dorywalska, M., Blanchard, S. C., Gonzalez, R. L., Kim, H. D., Chu, S., and Puglisi, J. D. (2005). Site-specific labeling of the ribosome for single-molecule spectroscopy. Nucleic Acids Res. 33, 182–189. Dubnoff, J. S., Maitra, U., Moldave, Kivie, and Lawrence, G. (1971). Isolation and properties of protein factors involved in polypeptide chain initiation in Escherichia coli. Methods Enzymol. 20, 248–261.

256

Jingyi Fei et al.

Elbing, K., and Brent, R. (2002). Escherichia coli, plasmids, and bacteriophages. Curr. Protoc. Mol. Biol. 59, 1.1.1–1.1.7. Ermolenko, D. N., Majumdar, Z. K., Hickerson, R. P., Spiegel, P. C., Clegg, R. M., and Noller, H. F. (2007). Observation of intersubunit movement of the ribosome in solution using FRET. J. Mol. Biol. 370, 530–540. Fei, J., Kosuri, P., MacDougall, D. D., and Gonzalez, R. L., Jr. (2008). Coupling of ribosomal L1 stalk and tRNA dynamics during translation elongation. Mol. Cell 30, 348–359. Fei, J., Bronson, J. E., Hofman, J. M., Srinivas, R. L., Wiggins, C. H., and Gonzalez, R. L. J. (2009). Allosteric collaboration between elongation factor G and the ribosomal L1 stalk directs tRNA movements during translation. Proc. Natl. Acad. Sci. USA 106, 15702–15707. Forchhammer, K., Leinfelder, W., and Bock, A. (1989). Identification of a novel translation factor necessary for the incorporation of selenocysteine into protein. Nature 342, 453–456. Fourmy, D., Mechulam, Y., Brunie, S., Blanquet, S., and Fayat, G. (1991). Identification of residues involved in the binding of methionine by Escherichia coli methionyl-tRNA synthetase. FEBS Lett. 292, 259–263. Frank, J., and Gonzalez, R. L., Jr. (2010). Structure and dynamics of a processive Brownian motor: The translating ribosome. Annu. Rev. Biochem. doi: 10.1146/annurev-biochem060408-173330. Fredrick, K., and Noller, H. (2003). Catalysis of ribosomal translocation by sparsomycin. Science 300, 1159–1162. Freistroffer, D. V., Pavlov, M. Y., MacDougall, J., Buckingham, R. H., and Ehrenberg, M. (1997). Release factor RF3 in E.coli accelerates the dissociation of release factors RF1 and RF2 from the ribosome in a GTP-dependent manner. EMBO J. 16, 4126–4133. Gallagher, S. R. (2006). One-dimensional SDS gel electrophoresis of proteins. Curr. Protoc. Mol. Biol. 75, 10-2A-1, 10-2A-37. Goldberg, R. B., Bender, R. A., and Streicher, S. L. (1974). Direct selection for P1-sensitive mutants of enteric bacteria. J. Bacteriol. 118, 810–814. Gonzalez, R. L. Jr., Chu, S., and Puglisi, J. D. (2007). Thiostrepton inhibition of tRNA delivery to the ribosome. RNA 13, 2091–2097. Guex, N., and Peitsch, M. C. (1997). SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modeling. Electrophoresis 18, 2714–2723. Ha, T. (2001). Single-molecule fluorescence resonance energy transfer. Methods 25, 78–86. Hapke, B., and Noll, H. (1976). Structural dynamics of bacterial ribosomes. IV. Classification of ribosomes by subunit interaction. J. Mol. Biol. 105, 97–109. Hartz, D., McPheeters, D. S., Traut, R., and Gold, L. (1988). Extension inhibition analysis of translation initiation complexes. Methods Enzymol. 164, 419–425. Hartz, D., McPheeters, D. S., and Gold, L. (1989). Selection of the initiator tRNA by Escherichia coli initiation factors. Genes Dev. 3, 1899–1912. Heurgue-Hamard, V., Champ, S., Engstrom, A., Ehrenberg, M., and Buckingham, R. H. (2002). The hemK gene in Escherichia coli encodes the N(5)-glutamine methyltransferase that modifies peptide release factors. EMBO J. 21, 769–778. Hirokawa, G., Nijman, R. M., Raj, V. S., Kaji, H., Igarashi, K., and Kaji, A. (2005). The role of ribosome recycling factor in dissociation of 70S ribosomes into subunits. RNA 11, 1317–1328. Hoffmann, A., and Roeder, R. G. (1991). Purification of his-tagged proteins in nondenaturing conditions suggests a convenient method for protein interaction studies. Nucleic Acids Res. 19, 6337–6338. Hohng, S., Joo, C., and Ha, T. (2004). Single-molecule three-color FRET. Biophys. J. 87, 1328–1337.

Single-Molecule Fluorescence Studies of Protein Synthesis

257

Jelenc, P. C., and Kurland, C. G. (1979). Nucleoside triphosphate regeneration decreases the frequency of translation errors. Proc. Natl. Acad. Sci. USA 76, 3174–3178. Joseph, S., and Noller, H. F. (1998). EF-G-catalyzed translocation of anticodon stem-loop analogs of transfer RNA in the ribosome. EMBO J. 17, 3478–3483. Kapp, L. D., and Lorsch, J. R. (2004). The molecular mechanics of eukaryotic translation. Annu. Rev. Biochem. 73, 657–704. Korostelev, A., Ermolenko, D. N., and Noller, H. F. (2008). Structural dynamics of the ribosome. Curr. Opin. Chem. Biol. 12, 674–683. Lakowicz, J. R. (1999). Principles of Fluorescence Spectroscopy. Springer, New York, NY. Lee, T. H., Blanchard, S. C., Kim, H. D., Puglisi, J. D., and Chu, S. (2007). The role of fluctuations in tRNA selection by the ribosome. Proc. Natl. Acad. Sci. USA 104, 13661–13665. Li, W., Agirrezabala, X., Lei, J., Bouakaz, L., Brunelle, J. L., Ortiz-Meoz, R. F., Green, R., Sanyal, S., Ehrenberg, M., and Frank, J. (2008). Recognition of aminoacyl-tRNA: A common molecular mechanism revealed by cryo-EM. EMBO J. 27, 3322–3331. Liljas, A. (2004). Structural Aspects of Protein Synthesis. World Scientific Publishing Co. Pte. Ltd., Singapore. Liu, C., Wang, R., Varlamova, O., and Leyh, T. S. (1998). Regulating energy transfer in the ATP sulfurylase-GTPase system. Biochemistry 37, 3886–3892. Maar, D., Liveris, D., Sussman, J. K., Ringquist, S., Moll, I., Heredia, N., Kil, A., Blasi, U., Schwartz, I., and Simons, R. W. (2008). A single mutation in the IF3 N-terminal domain perturbs the fidelity of translation initiation at three levels. J. Mol. Biol. 383, 937–944. Marshall, R. A., Aitken, C. E., Dorywalska, M., and Puglisi, J. D. (2008a). Translation at the single-molecule level. Annu. Rev. Biochem. 77, 177–203. Marshall, R. A., Dorywalska, M., and Puglisi, J. D. (2008b). Irreversible chemical steps control intersubunit dynamics during translation. Proc. Natl. Acad. Sci. USA 105, 15364–15369. Marshall, R. A., Aitken, C. E., and Puglisi, J. D. (2009). GTP hydrolysis by IF2 guides progression of the ribosome into elongation. Mol. Cell 35, 37–47. McKenna, S. A., Kim, I., Puglisi, E. V., Lindhout, D. A., Aitken, C. E., Marshall, R. A., and Puglisi, J. D. (2007). Purification and characterization of transcribed RNAs using gel filtration chromatography. Nat. Protoc. 2, 3270–3277. Milligan, J. F., Groebe, D. R., Witherell, G. W., and Uhlenbeck, O. C. (1987). Oligoribonucleotide synthesis using T7 RNA polymerase and synthetic DNA templates. Nucleic Acids Res. 15, 8783–8798. Mohr, D., Wintermeyer, W., and Rodnina, M. V. (2002). GTPase activation of elongation factors Tu and G on the ribosome. Biochemistry 41, 12520–12528. Moore, D. D. (2000). Commonly used reagents and equipment. Curr. Protoc. Mol. Biol. 35, A.2.5. Moore, S., and Sauer, R.T. (2009). P1vir phage transduction. http://openwetware.org/ wiki/Sauer:P1vir_phage_transduction. Mora, L., Heurgue-Hamard, V., de Zamaroczy, M., Kervestin, S., and Buckingham, R. H. (2007). Methylation of bacterial release factors RF1 and RF2 is required for normal translation termination in vivo. J. Biol. Chem. 282, 35638–35645. Munro, J. B., Altman, R. B., O’Connor, N., and Blanchard, S. C. (2007). Identification of two distinct hybrid state intermediates on the ribosome. Mol. Cell 25, 505–517. Odom, O. W., Picking, W. D., and Hardesty, B. (1990). Movement of tRNA but not the nascent peptide during peptide bond formation on ribosomes. Biochemistry 29, 10734–10744. Ogle, J. M., and Ramakrishnan, V. (2005). Structural insights into translational fidelity. Annu. Rev. Biochem. 74, 129–177.

258

Jingyi Fei et al.

Pavlov, M. Y., and Ehrenberg, M. (1996). Rate of translation of natural mRNAs in an optimized in vitro system. Arch. Biochem. Biophys. 328, 9–16. Peske, F., Rodnina, M. V., and Wintermeyer, W. (2005). Sequence of steps in ribosome recycling as defined by kinetic analysis. Mol. Cell 18, 403–412. Plumbridge, J. A., Baumert, H. G., Ehrenberg, M., and Rigler, R. (1980). Characterisation of a new, fully active fluorescent derivative of E. coli tRNA Phe. Nucleic Acids Res. 8, 827–843. Powers, T., and Noller, H. F. (1991). A functional pseudoknot in 16S ribosomal RNA. EMBO J. 10, 2203–2214. Qin, Y., Polacek, N., Vesper, O., Staub, E., Einfeldt, E., Wilson, D. N., and Nierhaus, K. H. (2006). The highly conserved LepA is a ribosomal elongation factor that back-translocates the ribosome. Cell 127, 721–733. Raleigh, E. A., Elbing, K., and Brent, R. (2002). Escherichia coli, plasmids, and bacteriophages. Curr. Protoc. Mol. Biol. 59, 1.4.1–1.4.14. Recht, M. I., Douthwaite, S., Dahlquist, K. D., and Puglisi, J. D. (1999). Effect of mutations in the A site of 16S rRNA on aminoglycoside antibiotic-ribosome interaction. J. Mol. Biol. 286, 33–43. Robertson, J. M., and Wintermeyer, W. (1981). Effect of translocation on topology and conformation of anticodon and D loops of tRNAPhe. J. Mol. Biol. 151, 57–79. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Sasse, J., and Gallagher, S. R. (2009). Staining proteins in gels. Curr. Protoc. Mol. Biol. 85, 10.6.1–10.6.27. Schmitt, E., Blanquet, S., and Mechulam, Y. (1999). Crystallization and preliminary X-ray analysis of Escherichia coli methionyl-tRNAMet(f) formyltransferase complexed with formyl-methionyl-tRNAMet(f). Acta Crystallogr. Sect. D Biol. Crystallogr. 55(Pt 1), 332–334. Schneider, R. J., and Mohr, I. (2003). Translation initiation and viral tricks. Trends Biochem. Sci. 28, 130–136. Schuette, J. C., Murphy, F. V.t., Kelley, A. C., Weir, J. R., Giesebrecht, J., Connell, S. R., Loerke, J., Mielke, T., Zhang, W., Penczek, P. A., Ramakrishnan, V., and Spahn, C. M. (2009). GTPase activation of elongation factor EF-Tu by the ribosome during decoding. EMBO J. 28, 755–765. Schuwirth, B. S., Borovinskaya, M. A., Hau, C. W., Zhang, W., Vila-Sanjurjo, A., Holton, J. M., and Cate, J. H. D. (2005). Structures of the bacterial ribosome at 3.5 A resolution. Science 310, 827–834. Selmer, M., Dunham, C. M., Murphy, F. V.t., Weixlbaumer, A., Petry, S., Kelley, A. C., Weir, J. R., and Ramakrishnan, V. (2006). Structure of the 70S ribosome complexed with mRNA and tRNA. Science 313, 1935–1942. Seshadri, A., and Varshney, U. (2006). Mechanism of recycling of post-termination ribosomal complexes in eubacteria: A new role of initiation factor 3. J. Biosci. 31, 281–289. Shimizu, Y., Inoue, A., Tomari, Y., Suzuki, T., Yokogawa, T., Nishikawa, K., and Ueda, T. (2001). Cell-free translation reconstituted with purified components. Nat. Biotechnol. 19, 751–755. Simonian, M. H., and Smith, J. A. (2006). Spectrophotometric and colorimetric determination of protein concentration. Curr. Protoc. Mol. Biol. 76, 10-1A-2, 10-1A-9. Slatko, B. E., and Albright, L. M. (1992). Denaturing gel electrophoresis for sequencing. Curr. Protoc. Mol. Biol. 16, 7.6.1–7.6.13. Sonenberg, N., and Hinnebusch, A. G. (2009). Regulation of translation initiation in eukaryotes: Mechanisms and biological targets. Cell 136, 731–745. Spahn, C. M., Blaha, G., Agrawal, R. K., Penczek, P., Grassucci, R. A., Trieber, C. A., Connell, S. R., Taylor, D. E., Nierhaus, K. H., and Frank, J. (2001). Localization of the

Single-Molecule Fluorescence Studies of Protein Synthesis

259

ribosomal protection protein tet(o) on the ribosome and the mechanism of tetracycline resistance. Mol. Cell 7, 1037–1045. Stark, H., Rodnina, M. V., Wieden, H. J., Zemlin, F., Wintermeyer, W., and van Heel, M. (2002). Ribosome interactions of aminoacyl-tRNA and elongation factor Tu in the codon-recognition complex. Nat. Struct. Biol. 9, 849–854. Sternberg, S. H., Fei, J., Prywes, N., McGrath, K. A., and Gonzalez, R. L., Jr. (2009). Translation factors direct intrinsic ribosome dynamics during translation termination and ribosome recycling. Nat. Struct. Mol. Biol. 16, 861–868. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994). CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22, 4673–4680. Valle, M., Zavialov, A., Li, W., Stagg, S. M., Sengupta, J., Nielsen, R. C., Nissen, P., Harvey, S. C., Ehrenberg, M., and Frank, J. (2003). Incorporation of aminoacyl-tRNA into the ribosome as seen by cryo-electron microscopy. Nat. Struct. Mol. Biol. 10, 899–906. Villa, E., Sengupta, J., Trabuco, L. G., LeBarron, J., Baxter, W. T., Shaikh, T. R., Grassucci, R. A., Nissen, P., Ehrenberg, M., Schulten, K., and Frank, J. (2009). Ribosome-induced changes in elongation factor Tu conformation control GTP hydrolysis. Proc. Natl. Acad. Sci. USA 106, 1063–1068. Wagner, E. G., Jelenc, P. C., Ehrenberg, M., and Kurland, C. G. (1982). Rate of elongation of polyphenylalanine in vitro. Eur. J. Biochem. 122, 193–197. Wall, J. D., and Harriman, P. D. (1974). Phage P1 mutants with altered transducing abilities for Escherichia coli. Virology 59, 532–544. Weinger, J. S., Parnell, K. M., Dorner, S., Green, R., and Strobel, S. A. (2004). Substrateassisted catalysis of peptide bond formation by the ribosome. Nat. Struct. Mol. Biol. 11, 1101–1106. Wilson, K. (1987). Preparation of genomic DNA from bacteria. Curr. Protoc. Mol. Biol. 2.4.1–2.4.2. Wilson, D. N., Blaha, G., Connell, S. R., Ivanov, P. V., Jenke, H., Stelzl, U., Teraoka, Y., and Nierhaus, K. H. (2002). Protein synthesis at atomic resolution: Mechanistics of translation in the light of highly resolved structures for the ribosome. Curr. Protein Pept. Sci. 3, 1–53. Wyatt, J. R., Chastain, M., and Puglisi, J. D. (1991). Synthesis and purification of large amounts of RNA oligonucleotides. BioTechniques 11, 764–769. Youngman, E. M., Brunelle, J. L., Kochaniak, A. B., and Green, R. (2004). The active site of the ribosome is composed of two layers of conserved nucleotides with distinct roles in peptide bond formation and peptide release. Cell 117, 589–599. Zavialov, A. V., Buckingham, R. H., and Ehrenberg, M. (2001). A posttermination ribosomal complex is the guanine nucleotide exchange factor for peptide release factor RF3. Cell 107, 115–124. Zavialov, A. V., Hauryliuk, V. V., and Ehrenberg, M. (2005). Splitting of the posttermination ribosome into subunits by the concerted action of RRF and EF-G. Mol. Cell 18, 675–686. Zhuang, X., Bartley, L. E., Babcock, H. P., Russell, R., Ha, T., Herschlag, D., and Chu, S. (2000). A single-molecule study of RNA catalysis and folding. Science 288, 2048–2051.

C H A P T E R

T H I R T E E N

Watching Individual Proteins Acting on Single Molecules of DNA Ichiro Amitani,*,†,1 Bian Liu,*,†,‡,1 Christopher C. Dombrowski,*,† Ronald J. Baskin,† and Stephen C. Kowalczykowski*,†,‡ Contents 1. Introduction 2. Preparation of DNA Substrates 2.1. Preparation of biotinylated l DNA 2.2. Preparation of DNA–bead complexes 2.3. Preparation of DNA–bead complexes end-labeled with Cy3-labeled antibody 3. Preparation of Fluorescent Proteins 3.1. RecBCD labeled with a fluorescent nanoparticle (RecBCD–nanoparticle) 3.2. Rad54/Tid1 labeled with a fluorescent antibody (FITC–Rad54/Tid1) 3.3. Chemically modified fluorescent RecA or Rad51 proteins (RecAFAM/Rad51FAM) 4. Instrument 4.1. Flow cell design 4.2. Flow cell fabrication 4.3. Microscope with laser trap and microfluidic system 4.4. Temperature determination and control 5. Single-Molecule Imaging of Proteins on DNA 5.1. Unwinding of DNA by a single RecBCD enzyme 5.2. Direct observation of RecBCD–nanoparticle translocation 5.3. Rad54/Tid1 translocation 5.4. Real-time Rad51 assembly 5.5. Real-time Rad51 disassembly 5.6. Visualization of RecAFAM/RecA-RFP/Rad51FAM filament formation

262 265 265 265 266 268 268 268 269 270 270 271 273 276 280 281 281 283 284 285 286

* Department of Microbiology, University of California, Davis, California, USA Department of Molecular and Cellular Biology, University of California, Davis, California, USA Biophysics Graduate Group, University of California, Davis, California USA 1 These authors contributed equally to this work. { {

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72007-3

#

2010 Elsevier Inc. All rights reserved.

261

262

Ichiro Amitani et al.

6. Data Analysis Methods 6.1. Two-dimensional Gaussian fitting 6.2. Automatic DNA length measurement Acknowledgments References

287 288 288 289 289

Abstract In traditional biochemical experiments, the behavior of individual proteins is obscured by ensemble averaging. To better understand the behavior of proteins that bind to and/or translocate on DNA, we have developed instrumentation that uses optical trapping, microfluidic solution delivery, and fluorescent microscopy to visualize either individual proteins or assemblies of proteins acting on single molecules of DNA. The general experimental design involves attaching a single DNA molecule to a polystyrene microsphere that is then used as a microscopic handle to manipulate individual DNA molecules with a laser trap. Visualization is achieved by fluorescently labeling either the DNA or the protein of interest, followed by direct imaging using high-sensitivity fluorescence microscopy. We describe the sample preparation and instrumentation used to visualize the interaction of individual proteins with single molecules of DNA. As examples, we describe the application of these methods to the study of proteins involved in recombination-mediated DNA repair, a process essential for the maintenance of genomic integrity.

1. Introduction In traditional ensemble experiments, the behavior of individual proteins is averaged by the obligatory need to study a population of molecules. However, it has become increasingly evident that the analysis of single molecules is not only possible, but that it can reveal novel information about the behavior and function of enzymes (see, e.g., Amitani et al., 2006; Bianco et al., 2001; Galletto et al., 2006; Handa et al., 2005; Nimonkar et al., 2007; Spies et al., 2003, 2007). To better understand the molecular behavior of individual proteins, we have used optical trapping to capture and visualize the action of individual proteins on single molecules of DNA (Bianco et al., 2001). The general experimental design involves attaching a single DNA molecule to a polystyrene microsphere. The microsphere is then used as a handle to manipulate the DNA molecule. Visualization is achieved by using a fluorescence microscope to image fluorescently labeled DNA or protein (Amitani et al., 2006; Bianco et al., 2001; Galletto et al., 2006; Handa et al., 2005, 2009; Hilario et al., 2009). To both extend the DNA and exchange solutions rapidly, we designed and fabricated multichannel microfluidic flow cells that provide parallel paths for different solutions that remain separated by laminar flow (Fig. 13.1) (Bianco et al., 2001; Brewer and

263

Watching Individual Proteins Acting on Single Molecules of DNA

Divider 100 mm

A

750 mm

B

200 mm

Region inside the mean diffusion boundary

10 mm 4.5 mm

Figure 13.1 Illustration of the three-channel flow cells used in the single-molecule experiments. (A) Photograph of a three-channel flow cell. The flow cell is fabricated using the process described in Section 4.2. To demonstrate the flow path, green dye flows through channels I and III, whereas yellow dye flows through channel 2. (B) Schematic (drawing not to scale) of a three-channel flow cell showing typical dimensions; magnification shows the detail at the end of channel divider. The divider is 100 mm wide with a semicircular end of radius of  50 mm. The gray area to the right of the divider illustrates the region inside the mean diffusion length boundary. Experiments are conducted at a point  200 mm downstream of the divider,  750 mm into each channel, and  35 mm from the surface where the effects of diffusion are minimal.

Bianco, 2008). These flow paths are used to introduce the optically trapped DNA to solutions that contain the proteins of interest, or that permit the controlled initiation of enzymatic reactions. We have applied these methods to the study of proteins involved in recombinational DNA repair, a conserved biological process responsible for the repair of DNA breaks. A DNA double-strand break (DSB) is a lethal type of DNA damage. These breaks are constantly created by many endogenous and exogenous sources in cells. Because an unrepaired DSB often leads to cell death, all organisms have evolved various methods to repair broken DNA. Among them, homologous recombination (HR) is the most accurate method for DSB repair (Kowalczykowski, 2000). The process of HR consists of three stages. First, the end of a broken double-stranded DNA (dsDNA) molecule is processed by helicase and nuclease to generate a 30 -ended, single-stranded DNA (ssDNA) tail onto which a DNA strand exchange protein self-assembles. Second, this protein– ssDNA complex searches for homology on a donor dsDNA molecule and then catalyzes the pairing and exchange of DNA strands. Finally, the heteroduplex DNA product is resolved (Kowalczykowski, 2000).

264

Ichiro Amitani et al.

In Escherichia coli, the RecBCD helicase/nuclease is responsible for the resection of dsDNA ends (Spies and Kowalczykowski, 2005). RecBCD is a bipolar DNA helicase and nuclease (Dillingham and Kowalczykowski, 2008). It unwinds and degrades dsDNA rapidly and processively (Bianco et al., 2001; Roman and Kowalczykowski, 1989). Its biological activities are regulated by an octameric DNA sequence called Chi (Crossover hotspot instigator, Chi: 50 -GCTGGTGG-30 ) (Dillingham and Kowalczykowski, 2008). Single-molecule analysis uniquely revealed that, upon interaction with Chi, the enzyme pauses for a few seconds, and then it translocates at a reduced rate due to a switch in motor usage (Spies et al., 2003, 2007). The interaction also downregulates the nuclease activity (Dixon and Kowalczykowski, 1993) and switches the polarity of DNA degradation (Anderson and Kowalczykowski, 1997a). These alterations of nuclease activity generate a processed dsDNA ending with a 30 -ssDNA tail (Taylor and Smith, 1995) onto which RecA is loaded by RecBCD to form a nucleoprotein filament (Anderson and Kowalczykowski, 1997b). DNA strand exchange is catalyzed by RecA in bacteria and Rad51 in eukaryotes (Bianco et al., 1998). Both RecA and Rad51 form a helical nucleoprotein filament on either ssDNA or dsDNA in the presence of ATP. In the filament, RecA/Rad51 occupies 3 nucleotides or base pairs (depending on whether ssDNA or dsDNA is used), and it stretches DNA to 150% of its B-form DNA length (Benson et al., 1994; Chen et al., 2008; Conway et al., 2004; Ogawa et al., 1993; Stasiak et al., 1981). Although the typical active form of RecA/Rad51 is the ssDNA–RecA/Rad51 complex, when assembled on dsDNA, RecA can also promote DNA pairing with ssDNA (Zaitsev and Kowalczykowski, 2000). However, when RecA/Rad51 forms a complex with dsDNA, DNA strand exchange with a RecA/Rad51– ssDNA complex is impeded, resulting in defective recombination (Campbell and Davis, 1999; Sung and Robberson, 1995). Assembly of RecA/Rad51 nucleoprotein filaments occurs by nucleation and growth, a process that was imaged at the single-molecule level (Galletto et al., 2006; Handa et al., 2009; Hilario et al., 2009; Modesti et al., 2007; Prasad et al., 2006; Robertson et al., 2009; van der Heijden et al., 2007). In eukaryotes, the inhibitory Rad51 bound to chromosomes is removed by Rad54 (Solinger et al., 2002), a chromatin-remodeling protein (Alexeev et al., 2003). Rad54 and Tid1, a Rad54 homolog with an important role in meiosis (Klein, 1997; Shinohara et al., 1997), work together with Rad51 (Mazin et al., 2000, 2003; Petukhova et al., 1998; Solinger and Heyer, 2001; Solinger et al., 2001, 2002) and Dmc1, the meiotic Rad51 homolog (Holzen et al., 2006; Shinohara et al., 2000), respectively. Both Rad54 and Tid1 are dsDNA translocases as defined by the direct visualization of their movements on individual DNA molecules (Amitani et al., 2006; Nimonkar et al., 2007; Prasad et al., 2007).

Watching Individual Proteins Acting on Single Molecules of DNA

265

In this chapter, we detail the sample preparation and instrumentation used to visualize the interaction of individual proteins with single molecules of DNA.

2. Preparation of DNA Substrates 2.1. Preparation of biotinylated l DNA Bacteriophage l DNA (New England Biolabs, Ipswich, MA) is biotinylated by ligation to a 30 -biotinylated 12-mer oligonucleotide (50 -GGGCGGCG ACCT-30 or 50 -AGGTCGCCGCCC-30 , Operon Technologies, Huntsville, AL) that is complementary to one of the cohesive ends of l DNA (Bianco et al., 2001). In all the subsequent protocols, the pipetting of solutions containing l DNA should be performed with cut pipette tips to minimize shearing of the DNA. 1. Phosphorylate the oligonucleotide by incubating the oligonucleotide (5 mM) in 50 ml of polynucleotide kinase (PNK) buffer (5 mM dithiothreitol (DTT)), 10 mM MgCl2, 70 mM Tris–HCl (pH 7.6), 1 mM ATP, and 0.2 U/ml PNK at 37  C for 1 h. 2. Stop the reaction by incubation at 75  C for 10 min. 3. Anneal the phosphorylated oligonucleotide and the l DNA by preparing a reaction (90 ml) containing 28 ng/ml of l DNA, 0.56 mM phosphorylated oligonucleotide, and 100 mM NaCl. 4. Incubate the reaction at 75  C for 20 min in a heat block to denature the annealed cohesive ends of the l DNA. 5. Remove the heat block and place it on the bench to slowly cool the reaction to room temperature (2–3 h), and then chill the reaction on ice. 6. Ligate the phosphorylated oligonucleotide to the l DNA by adding 10 ml of 10 T4 DNA ligase buffer (10 mM ATP, 100 mM DTT, 100 mM MgCl2, 500 mM Tris–HCl, pH 7.5) and 1 ml of T4 ligase (400 Units) to the annealing reaction from the previous step. 7. Incubate the reaction at 16  C overnight or at room temperature for 1 h. 8. Inactivate the ligase by incubating at 75  C for 10 min. 9. Remove excess oligonucleotide and ATP by filtration through a spin column (MicroSpin S-400 HR, GE Healthcare, Piscataway, NJ).

2.2. Preparation of DNA–bead complexes DNA–bead complexes are prepared by incubating 1 ml of 35 pM streptavidin-coated polystyrene beads (1.0 mm, Bangs Laboratories, Fishers, IN), 1 ml of 100 mM NaHCO3 (pH 8.3), and 2 ml of 100 pM biotinylated

266

Ichiro Amitani et al.

l DNA for 1 h at 37  C. The ratio of beads to DNA may be varied and optimized for different experiments. To fluorescently stain the DNA, add 500 ml of sample buffer (see below for experiment-specific recipes) containing 20 nM YOYO-1 (Invitrogen, Carlsbad, CA) to the DNA–bead complex and stain in the dark at room temperature for at least 1 h. The dye to DNA (in base pairs) ratio can be altered to vary from 1:1 to 1:5. The sample buffer is degassed for at least 1 h to remove oxygen and to reduce oxygen-mediated photobleaching and cleavage of DNA.

2.3. Preparation of DNA–bead complexes end-labeled with Cy3-labeled antibody 2.3.1. Fluorescent secondary antibody To visualize the end of a DNA molecule in order to measure its length without the use of a nonspecifically binding dye such as YOYO-1, we attach a fluorescent tag at the free end of the DNA–bead complex (Hilario et al., 2009). To obtain a strong signal for imaging, we fluorescently label a secondary antibody and bind it to a primary antibody that is bound to the end of DNA, which is labeled with digoxigenin (DIG). 1. Exchange the storage solution of donkey antisheep IgG antibody (50 ml, 2 mg/ml, Millipore, Billerica, MA) to a buffer lacking primary amines by using a P30 spin column (850  g for 4 min; Bio-Rad, Hercules, CA) equilibrated with labeling buffer (50 mM sodium borate (pH 9.3), 140 mM NaCl, and 2.7 mM KCl). 2. Add a 20-fold molar excess of Cy3 succinimidyl ester (Cy3-NHS, GE Healthcare) and incubate at room temperature for 1 h in the dark. 3. Remove the unreacted Cy3-NHS with a P30 spin column equilibrated with phosphate buffered saline (PBS; 10 mM Na2HPO4, 1.8 mM KH2PO4, (pH 7.4), 137 mM NaCl, and 2.7 mM KCl). 4. Determine the Cy3 and antibody concentrations by using the extinction coefficients e552 ¼ 1.5105 M 1cm 1 for Cy3, and e280 ¼ 1.7  105 M 1cm 1 for the antibody. The effect of absorption by Cy3 at 280 nm is corrected by: [antibody] ¼ (A280 – (0.08  A552))/ 1.7  105 (GE Healthcare, Amersham product booklet, ‘‘CyDyeTM monoreactive NHS Esters’’). 5. Determine the degree of labeling by calculating the ratio of Cy3 and antibody concentrations. A typical degree of labeling is 6–8 dyes/ protein. 6. Store Cy3-antisheep antibody at 4  C in the dark and use within a few days.

Watching Individual Proteins Acting on Single Molecules of DNA

267

2.3.2. DNA labeled with biotin and digoxigenin (Biotin-l DNA–DIG) Bacteriophage l DNA that is labeled with biotin at one end and DIG at the other end is prepared by attaching a biotin-labeled oligonucleotide and DIG-labeled oligonucleotide to opposite cohesive ends of l DNA in successive steps. 1. Incubate 750 pM of l DNA (molecules) with 375 nM of DIG-labeled oligonucleotide (Operon Technologies) in 90 ml of 100 mM NaCl at 75  C for 15 min in a heat block. 2. Remove the heat block and place it on the bench to slowly cool the reaction mixture to room temperature (2–3 h). 3. Add 10 ml of 10 T4 DNA ligase buffer and T4 DNA ligase to a final concentration of 4 U/ml. 4. Incubate at room temperature for 1 h. 5. Inactivate the DNA ligase at 65  C for 10 min. 6. Remove unreacted DIG-oligonucleotide with an S-400 spin column (850  g for 5 min) equilibrated with TE buffer (10 mM Tris–HCl (pH 7.5), 1 mM EDTA). 7. Add 50-fold molar excess of biotinylated-oligonucleotide and 4 U/ml T4 DNA ligase to DIG-labeled l DNA. 8. Incubate at room temperature for 1 h. 9. Inactivate the DNA ligase at 65  C for 10 min.

2.3.3. Binding Cy3-labeled antibody to the DNA–bead complex DNA–bead complexes that are end-labeled with Cy3-labeled antibody are prepared by binding the sheep anti-DIG antibody to DNA–bead complex, and then binding the Cy3-antisheep secondary antibody to the anti-DIG antibody. 1. Attach the biotin-l DNA–DIG to streptavidin-coated beads as described in Section 2.2. 2. Add bovine serum albumin (BSA; stock 10 mg/ml) and sheep anti-DIG antibody (stock 200
g/ml) to final concentrations of 1 mg/ml and 20
g/ml, respectively. 3. Incubate at room temperature for 2 min. 4. Add Cy3-antisheep IgG antibody to a final concentration of 60
g/ml. The final volume is 7
l. 5. Incubate at room temperature for 2 min. 6. Immediately dilute the Cy3-antibody end-labeled DNA–bead complex into 400
l of single-molecule buffer (SMB; 40 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 15% (w/v) sucrose). The final concentration of streptavidin-beads is about 90 fM.

268

Ichiro Amitani et al.

3. Preparation of Fluorescent Proteins 3.1. RecBCD labeled with a fluorescent nanoparticle (RecBCD–nanoparticle) Translocation by individual RecBCD enzyme molecules can be directly visualized by labeling the protein with a fluorescent nanoparticle, which provides a strong and stable fluorescence signal (Handa et al., 2005). Biotinylated RecBCD was purified from an E. coli strain that expresses RecD with an N-terminal hexahistidine tag, followed by an amino-acid sequence that directs the biotinylation in vivo of a single lysine residue (Handa et al., 2005; Schatz, 1993). To attach the fluorescent nanoparticle: 1. Mix 4.8 ml of the biotinylated RecBCD enzyme (1.22 mM in storage buffer: 20 mM Tris–HCl (pH 7.5), 0.1 mM EDTA, 0.1 mM DTT, 100 mM NaCl, and 50% (v/v) glycerol) with 3 ml of a streptavidincoated fluorescent nanoparticle (0.5% solids in 50 mM sodium phosphate (pH 7.5), 50 mM NaCl, and 5 mM sodium azide; 40 nm TransFluoSpheres; excitation 488 nm; emission 645 nm; Molecular Probes, Carlsbad, CA). 2. Incubate for 10 min at 37  C. The RecBCD–nanoparticle is subsequently bound to the DNA–bead complex (see below, Section 5.2).

3.2. Rad54/Tid1 labeled with a fluorescent antibody (FITC–Rad54/Tid1) As an alternative to biotinylation, proteins can be prepared as fusion products; the choice of using a biotinylation tag versus a fusion protein depends on a number of empirical factors, including the efficiency of biotinylation in the organism used for protein expression versus the expression, solubility, and activity of the modified protein. Rather than attaching a streptavidincoated nanoparticle to the biotin, a fluorescent antibody can be used to label the fusion protein. Yeast Rad54 and Tid1 proteins are purified as a GST fusion product. Consequently, one can directly visualize the translocation of Rad54 or Tid1 by binding a fluorescent antibody to the GST moiety of Rad54 or Tid1 (Amitani et al., 2006; Nimonkar et al., 2007). 1. Prepare DNA–bead complexes as described in Section 2.2. 2. Add DNA translocase to a final concentration of 10 nM. 3. Add 670 nM FITC-anti-GST antibody (an average degree of labeling of six fluorophores/antibody; RGST-45F-Z, Immunology Consultants Laboratory, Newberg, OR) in PBS containing 0.2% (w/v) BSA. The final volume is  5 ml.

Watching Individual Proteins Acting on Single Molecules of DNA

269

4. Incubate the mixture at room temperature for 10 min. 5. Immediately dilute the complex into 400
l of degassed buffer containing 40 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 15% (w/v) sucrose. The final bead concentration is 90 fM.

3.3. Chemically modified fluorescent RecA or Rad51 proteins (RecAFAM/Rad51FAM) RecA and Rad51 can be imaged by the covalent addition of a fluorescent adduct to the N-terminus of the protein (Galletto et al., 2006; Hilario et al., 2009). Chemical modification is performed by coupling 5(6)-carboxyfluorescein succinimidyl ester (FAM-SE, Invitrogen) to the N-terminal amine. Because of the difference in pKa between N-terminal a-amino group (pKa7) and the e-amino group of lysine (pKa  10–11), by a judicious adjustment of dye concentration and incubation time, the N-terminal amine of the protein can be relatively specifically labeled ( 103-fold over other primary amino groups) at near neutral pH to yield products with typically 1 dye/protein monomer. Here, we describe the protocol to label RecA. The reaction conditions need to be optimized for the protein of interest. 1. The protein is first dialyzed extensively against a solution lacking primary amines (50 mM K2HPO4/KH2PO4 (pH 7.0), 1 M NaCl, 0.1 mM DTT, and 10% glycerol). 2. Dissolve the FAM-SE in dry DMSO to a stock concentration of  50– 75 mM. The precise concentration of the stock is determined spectroscopically by making a 1:10,000 dilution into 10 mM Tris–HCl (pH 9.0) and using an extinction coefficient of 7.8  104 M 1 cm 1 at 492 nm. 3. Add a 12-fold molar excess of FAM-SE to typically 80–100 mM RecA (500 ml) and incubate at 4  C for 4 h in the dark. 4. Stop the reaction by adding Tris–HCl (pH 7.5) to a final concentration of 50 mM. 5. Remove unreacted fluorescein (FAM) by using a Bio-Gel P10 (Bio-Rad) column (1 cm  16 cm). 6. Dialyze the sample against storage buffer (20 mM Tris–HCl (pH7.5), 0.1 mM EDTA, 0.5 mM DTT, 10% (v/v) glycerol). 7. Determine the RecA and FAM concentrations by measuring the absorption at 280 and 492 nm using the extinction coefficients of e280 ¼ 2.7  104 M 1 cm 1 for RecA and e492 ¼ 7.8  104 M 1 cm 1 for fluorescein measured at pH 9. 8. Determine the degree of labeling by calculating the ratio of FAM and RecA concentrations. A correction factor (CF ¼ A280/A492) of 0.32 measured for the free dye in the absence of protein is used to account for the absorption of FAM at 280 nm using the following calculation:

270

Ichiro Amitani et al.

ARecA ¼ A280 – CF  A492. The correction factor may be sensitive to buffer conditions, specifically pH. It is recommended that the CF provided by the manufacturer is verified experimentally. The chemical modification of human Rad51 (at 50–60
M) is performed in buffer containing 50 mM KH2PO4 (pH 7.1), 200 mM KCl, 0.1 mM DTT, and 25% (v/v) glycerol, using 20-fold molar excess of FAM, reacted for 6 h at 4  C. The labeled protein is stored in storage buffer (50 mM Tris–HOAc (pH 7.5), 200 mM KCl, 1 mM DTT, 0.1 mM EDTA, 50% glycerol) (Hilario et al., 2009).

4. Instrument 4.1. Flow cell design Single-molecule reactions are carried out in multi-channel flow cells. The photograph of a three-channel flow cell is shown in Fig. 13.1A. Flow cells are designed to ensure laminar flow and to minimize mixing of solutions from different channels (Figs. 13.1 and 13.2). Channel dividers are 100 mm in width with an approximately semicircular end of 10–100 mm, depending on the manufacturing process. Each flow channel is 1.5 mm in width and 70
m in depth. Fluid flow in this geometry is at a very low Reynolds number (<1) and laminar. The velocity field is a Poiseuille flow with a parabolic profile; maximum velocity is midway between the top and bottom of the flow cell, and zero velocity at the top and bottom surfaces. To ensure that experiments are conducted in regions where diffusion from adjacent channels is minimal, trapping must occur at a position that is downstream of the confluence and away from the boundary between flow channels. The mean diffusion distance, x, of a solute can be calculated as hx2 i ¼ 2Dt

ð13:1Þ

kB T 6pa

ð13:2Þ

D¼

Here, D is the diffusion rate of the solute, kB is the Boltzmann constant, T is the absolute temperature,  is the viscosity of the solution, and a is the radius of the diffusing particle. The mean diffusion distance provides a good approximation of the boundary where little diffusion between the channels has occurred. However, the absolute concentration will contain solute or ligand from adjacent channels some distance past the mean diffusion boundary. Fig. 13.2 shows a simulation of a solute concentration for two typical solutes (Mg2þ and ATP) in our flow cells under nominal conditions. The

271

Watching Individual Proteins Acting on Single Molecules of DNA

Mg2+

100

100%

I -200

Concentration

Distance across (mm)

-400

0 200

Concentration (%)

A

II 400

100

200

300

400

80 60 40 20 0 -400

0%

Distance along flow (mm)

B

ATP

0 200 II 100

200

300

400

Distance along flow (mm)

0%

Concentration (%)

-200

400

100

100% I Concentration

Distance across (mm)

-400

50 mm 150 mm 250 mm 350 mm 450 mm

80 60 40

-200 0 200 Distance across (mm)

400

50 mm 150 mm 250 mm 350 mm 450 mm

20 0 -400

-200 0 200 Distance across (mm)

400

Figure 13.2 The calculated concentrations of Mg2þ, D ¼ 10 5 cm2/s (A, left panel), and ATP, D ¼ 10 6 cm2/s (B, left panel) as a function of position for diffusion from channel I into channel II. The flow cell has the same dimensions as described in Fig. 13.1, and the flow rate, v, is 50 mm/s from left to right; the end of the divider is at the origin of the plot. The cross section of concentration as a function of the distance downstream of the flow cell is also shown (right panels). The solid white line in the left panels indicates the mean diffusion distance of the solute from and into each channel. The calculations were performed in MATLAB (MathWorks).

simulations are based on the exact solution of Fick’s equation for one-dimensional diffusion in a pipe. Typically, a flow velocity of 100–200
m/s is used (Fig. 13.1B). Experiments are conducted by optically trapping 200
m downstream of the dividers into the flow cell, halfway between the top and bottom surfaces ( 35 mm), and halfway into the channel ( 750
m). This position ensures that experiments are conducted in a region where the local solution concentration is identical to the bulk concentration within a channel.

4.2. Flow cell fabrication Several methods have been used to make multichannel flow cells (for a review, see Brewer and Bianco, 2008). MMR Technologies employs a dry etching technique to create channels on a glass slide; the coverslip is attached to the slide by melting powdered glass at 660  C (MMR Technologies,

272

Ichiro Amitani et al.

Mountain View, CA). Another way to construct flow cells is to use chemically amplified, negative-tone, epoxy-based photoresists. A multichannel flow cell  70 mm deep can be made using the following process (Fig. 13.3). 1. Prepare a mask on a chrome borosilicate photomask using standard lithography techniques. 2. Drill inlet and outlet holes on a glass coverslip or slide for either an upright or an inverted microscope using diamond grinding bit (model 750, Dremel, WI). The glass should be submerged under water during grinding. 3. Clean the coverslip and slide with hot piranha treatment (96% H2SO4:30% H2O2 ¼ 3:1 (v/v)). 4. Spin coat KMPR 1050 photoresist (MicroChem Corp., Newton, MA) onto the slide at 2000 rpm for 30 s. 5. Soft bake at 100  C for 20 min. 6. Expose the coated slide with 365 nm light using the prepared mask on a Karl-Suss MA4 Mask Aligner (Karl Suss America, Inc., Waterbury Center, VT). 7. Bake at 100  C for 4 min. 8. Remove unexposed photoresist in SU-8 developer (MicroChem Corp.) for 4 min with slow shaking. Slide

Coverslip

Spin-coat photoresist

Exposure KMPR 1050

Develop KMPR 1050

Flip and assemble

Expose KMPR 1005

Develop KMPR 1005

Figure 13.3 Flow diagram for the microfabrication of a three-channel flow cell (see text for details).

Watching Individual Proteins Acting on Single Molecules of DNA

273

9. Spin coat KMPR 1005 photoresist (MicroChem Corp.) onto the coverslip at 1600 rpm for 30 s. 10. Put both the coverslip and the slide on a hotplate at 60  C with the KMPR-coated side facing up. 11. Flip the coverslip onto the slide. Carefully align the coverslip and the slide before dropping the coverslip onto the slide. 12. Increase the hotplate temperature to 90  C. When the temperature reaches 90  C, gently touch the coverslip so that the coverslip and the slide bond together, then decrease the temperature to 60  C. 13. Expose the coverslip and slide assembly to 365 nm light using the mask prepared in step 1 on a Karl-Suss MA4 Mask Aligner. 14. Bake at 100  C on a hotplate for 1 min. 15. Remove unexposed photoresist in SU-8 developer by running the developer through the channels with vacuum. 16. Attach machined connectors (P-770-01, Upchurch Scientific, Oak Harbor, WA) to the coverslip at the holes using epoxy. An alternative way to make flow cells is to use thermobond film or Parafilm: 1. Drill inlet and outlet holes (1 mm diameter) on the coverslip or slide, depending on the type of microscope (upright or inverted) used. 2. Place either thermobond film (Thermobond film 668EG, 2.5 mil (62
m), 3 M, St. Paul, MN) or Parafilm on the slide. Cut the desired pattern using a razor blade. 3. Place a coverslip on top of the spacer. 4. Place the assembly on a heat block at 150  C for 20–30 s. Gently press the coverslip so that the slide and the coverslip bond evenly. 5. Using either a handheld grinding tool or a razor blade, create a V-shape at one end of a short piece (1–2 cm) of PEEK tubing (Upchurch Scientific). 6. Attach the V-shaped end of the PEEK tubing into the holes and glue it using epoxy. Care must be taken not to block the channels.

4.3. Microscope with laser trap and microfluidic system The laser-trap systems are constructed around a Nikon Eclipse E400 or a Nikon TE2000U microscope (Nikon, Tokyo, Japan). Schematics of the instruments are shown in Figs. 13.4 and 13.5. A brief description of the components follows. 4.3.1. Single optical trap imaging system A high-pressure mercury lamp (USHIO America, Inc., Cypress, CA) and Y-FL 4-cube Epi-Fluorescence (Nikon) attachment are used for illumination. Images are captured using a high-sensitivity electron bombardment

274

Ichiro Amitani et al.

Monitor

Camera

Lamp Camera controller Syringe+ pump

DM2 L2

S L1

Beam expander

IR laser

DM1 OBJ

Stage

PC

Figure 13.4 Schematic diagram of the microscope, optical trap, and flow cell. The trapping IR laser initially passes through a 20 beam expander, and is then further collimated and steered by lenses L1 and L2 an electronic shutter (S) is in-between. A high-pass IR dichroic mirror (DM1) directs laser beam into the objective (OBJ). The flow cell is mounted on an x–y translocation stage that is controlled by a computer (PC) solutions are delivered to the flow cell using a multisyringe pump. A high-pressure mercury arc lamp is used for illumination (fluorescence and bright field). A second dichroic mirror (DM2) is used to image the fluorescent protein–DNA–bead complex onto an electron bombardment camera; the real-time image is displayed on a monitor.

CCD camera (EB-CCD C7190, Hamamatsu Photonics, Hamamatsu City, Japan), recorded on video tape, and subsequently digitalized using an LG-3 frame grabber at 30 frame/s (Scion Corporation, Frederick, MD). The optical trap is created by focusing a 1064 nm laser (Nd:YVO4, 6 W max, J-series power supply, Spectra Physics, Mountain View, CA) through a high numerical aperture (NA) objective (100/1.3 oil DICH, Nikon). A high NA objective is necessary to create an intensity gradient sufficiently large to form the trap (Neuman and Block, 2004). The laser is expanded with a 20 beam expander (HB-20XAR.33, Newport, Irvine, CA) to fill the back aperture of the objective. The laser is collimated and aligned using two lenses with the same focal length forming a 1 telescope. The laser is reflected along the optical axis of the microscope by means of a low-pass dichroic mirror (DM) placed between the objective and the fluorescence cube. Experiments are carried out in a multichannel microfluidic flow cell held on a computer controlled motorized stage (MS-2000, Applied Scientific Instruments, Eugene, OR) mounted on the microscope. The solutions are introduced into the flow cell by a syringe pump with multiple syringes

275

Watching Individual Proteins Acting on Single Molecules of DNA

Xe lamp BS

L9

M

PC

M

Stage L5 L4 L3 AOM

L8 L7 L6

Objective lens Camera M

BS M

HWP

M

ck ba

L1

Fe

QPD L11 HM L10

L2 ed

DM

Amplifier+ low-pass filter

IR laser PC

Figure 13.5 Schematic diagram for a dual laser-trap microscope. Lenses L1 and L2 initially collimate and expand the laser. The first beam path (black line) passes through an AOM which is imaged on the back aperture of the objective lens by lenses L3, L4, L5, and L9. The second beam path (gray line) is reflected off a movable mirror which is imaged onto the back aperture of the objective by lenses L6, L7, L8, and L9. The image from the objective (dashed line) is split between a camera that images the fluorescent protein–DNA–bead complex, and a quadrant photodiode (QPD) for position detection of the bead in the first trap by means of a half-mirror (HM). The signal from the QPD passes through an amplifier and a low-pass filter before being processed by a PC which uses the information to control the AOM, thus providing feedback control on the position of the first trap with nm resolution. Mirrors (M) and beam splitters (BS) serve to direct the beam path. A dichroic mirror (DM) is used to direct the trapping lasers into the objective and to pass light from the Xenon lamp to the camera and QPD; the realtime image is displayed on a monitor. Lenses L10 and L11 image the trapped bead onto the camera and QPD.

(KD Scientific, Hollston, MA). PEEK tubing (Upchurch Scientific) is used to connect the syringes to the flow cell. The microfluidic system permits the imaging of protein–DNA complexes on a single molecule of flow-stretched DNA; it also enables the rapid movement of the sample to the different buffers in the channels of the flow cell. The position of the stage and hence the flow cell, is controlled using a custom-built program. Because the translation speed of the motorized sample stage is typically 0.5–1 mm/s, and the distance being moved to the adjacent flow channel is 0.7–1.5 mm, the time required to move between solution channels is 1–2 s. 4.3.2. Dual optical trap imaging system For force measurements, a double laser-trap system is constructed around a TE2000U microscope (Fig. 13.5). An infrared laser (Nd:YVO4, 6 W max, J-series power supply, Spectra Physics) beam is passed through lenses L1 and

276

Ichiro Amitani et al.

L2 ( f ¼ 6, 12 mm, respectively) serving as a collimator. The beam is then passed through a half-wave plate (HWP) and a polarizing beam splitter (BS) creating two separate beam paths. The first beam is steered with an acoustic optical modulator (AOM) for force feedback. The first beam path is then magnified with lenses L3, L4, L5, and L9 ( f ¼ 75, 25, 25, 800 mm, respectively) to fill the back aperture of the microscope objective. Similarly, the second path is magnified with lenses L6, L7, L8, and L9 ( f ¼ 75, 25, 25, 800 mm, respectively) to fill the back aperture of the microscope objective. The laser-trap strength can be up to 0.4 pN/nm, but heating at high-power settings can be a problem. A dichroic mirror (DM) is used to direct the trapping lasers into the objective and to pass light from the Xenon lamp to the camera and quadrant photodiode (QPD). The florescent image of the trapped protein–DNA–bead complex is focused onto a CCD camera (iXonþ, Andor) via the objective through another BS. One of the trapped beads is also imaged onto the QPD (S1557-03, Hamamatsu) to provide precise high-bandwidth information about its position. Signals are digitized (PCI6052E, National Instruments) and processed with software written in LabView (LabView 6.1, National Instruments). The position data from the QPD controls the deflection angle of the AOM allowing for feedback between the bead position and the trap position. This arrangement permits the movement of one optical trap relative to the other. The total moveable range of the AOM is 2.4
m, but the linear range is limited to 200 nm.

4.4. Temperature determination and control To achieve reliable trapping in a flow field, laser power in the range of several hundred milliwatts (mW) is used. Water has a measurable absorption at the near-infrared wavelength typically used for an optical trap (l ¼ 1064 nm). Consequently, the effect of local heating on sample temperature is an important consideration. Different ways of estimating the temperature in an optical trap have been reported (Celliers and Conia, 2000; Liu et al., 1995; Peterman et al., 2003). We adapted the methods using fluorescence, and we measured temperature based on the thermal quenching of rhodamine B (RhB) fluorescence (Karstens and Kobs, 1980; Romano et al., 1989). The temperature measurement is carried out using a customized sample chamber (Fig. 13.6). This chamber is constructed as described in Section 4.3. A thermocouple (model CHAL-002, Omega Engineering, Stamford, CT) is placed in the middle of the channel before assembling the coverslip and slide. 4.4.1. Temperature determination The following procedure is used to measure temperature based on the fluorescence intensity measurements of RhB relative to Alexa-488.

277

Watching Individual Proteins Acting on Single Molecules of DNA

Thermocouple

Slide

Parafilm

Coverslip Thermistor

Brass jacket

Water circulation

Objective

Copper tubing

Figure 13.6 Schematic illustration of the components used for temperature measurement and control. Top: the flow cell used for temperature determination is made of a glass slide, a coverslip, and a layer of Parafilm sandwiched in between. A thermocouple is placed inside the channel. Bottom: an illustration of the temperature controller (side view; not to scale).

1. Construct a flow cell as described in Figure 13.6, Section 4.3. Place a thermocouple (model CHAL-002, Omega Engineering, Stamford, CT) in the middle of the channel before assembling the coverslip and slide. 2. Fill the flow cell with 10 dye solution (10
M RhB (Wako Pure Chemical Industries, Ltd.) and 15
M Alexa-488 (Invitrogen) in TE buffer) and incubate at 40  C overnight to coat the flow cell. 3. Replace the 10 dye solution with degassed 1 dye solution. Seal the outlets using Parafilm. 4. Set the flow cell temperature using either a thermoelectric microscope slide temperature controller (BC-100, 20/20 technologies, Wilmington, NC) or an objective jacket (see next Section 4.4.3). Wait for at least 30 min for the system to equilibrate. 5. Select a region (5 mm  10
m) around the trapping position and record the fluorescence images of RhB and Alexa-488 with the appropriate filter sets (Ethidium Bromide set 41006 and Blue set 11001v2; Chroma Technology Corp., Rockingham, VT). 6. Repeat steps 4 and 5 for at least four different temperatures. These measurements are used to relate the fluorescence intensity to temperature (see below). 7. Turn on the IR laser, set the desired power, and wait for about 30 min. 8. Select the same region as in step 4. Record the fluorescence images of RhB and Alexa-488. 9. Change the IR laser power and repeat steps 7 and 8. 10. Replace the dye solution with TE buffer. Record the background fluorescence images of RhB and Alexa-488.

278

Ichiro Amitani et al.

11. Remove the flow cell and the objective lens. Measure the laser power at the back aperture of the objective. Calculate the power delivered to the focus using the infrared transmission coefficient of the objective used (60% for the objective used) (Neuman and Block, 2004). To determine the average temperature around the focus, the fluorescence images are analyzed with ImageJ (NIH; http://rsb.info.nih.gov/ij/). A calibration curve of relative fluorescent intensity versus temperature is generated for the data acquired in the absence of trapping laser power. The temperature at any given laser power setting is then obtained from the calibration curve: 1. Calculate the relative fluorescence intensity, Ir(T), at each temperature, T, measured above in the absence of IR irradiation using Ir ðTÞ ¼ ðIR  IR;BG Þ=ðIA  IA;BG Þ

ð13:3Þ

where IR and IA are the average fluorescence intensity of RhB and Alexa-488 at temperature T, respectively; IR,BG and IA,BG are the average background fluorescence intensities using RhB and Alexa-488 filter sets, respectively. 2. Normalize the relative intensity Ir(T) to an arbitrary reference temperature, T0 (e.g., 25  C), using rðT Þ ¼ Ir ðT Þ=Ir ðT0 Þ

ð13:4Þ

where r(T) is the normalized intensity ratio. 3. Determine the constant, a, in the empirical linear relationship (Kato et al., 1999) (Fig. 13.7A): rðT Þ ¼ 1  aðT  T0 Þ

ð13:5Þ

4. The temperature at any given IR laser power is given by T¼

1r þ T0 a

ð13:6Þ

where r is the observed intensity ratio normalized to the reference temperature at that given IR laser power. 5. The temperature change for any given IR laser power is DT ¼ T  T0 ¼

1r a

ð13:7Þ

For our instrument, the IR laser-induced temperature change is about 1.2  C/100 mW of laser power delivered at the focus (Fig. 13.7B).

279

Watching Individual Proteins Acting on Single Molecules of DNA

Normalized fluorescence intensity ratio, r(T)

A

a = -0.021 ± 0.001 per ⬚C

1.0

0.9

0.8 25

30 Temperature (⬚C)

35

B Temperature increase (⬚C)

20 ΔT = 1.20 ± 0.02 ⬚C/100 mW

15 10 5 0 0

500 1000 Laser power (mW)

1500

Figure 13.7 Measurement of IR laser-induced temperature changes. (A) Relative fluorescence as a function of temperature. The ratio of the background-corrected fluorescence of RhB and Alexa-488 is normalized to that at 25  C. The relative fluorescence intensity decreases 2% per  C increase. (B) Relationship between induced temperature change and the IR laser power delivered at the focus, at a starting temperature of 25  C.

4.4.2. Temperature gradient around the trap center In an optical trap, hundreds of mW of laser power are focused on a micrometer-sized spot. The temperature gradient caused by such localized heating is another concern. To experimentally determine this gradient, the fluorescence images of RhB and Alexa-488 are recorded as described in Section 4.4.1. The images are then analyzed using the same procedure as in Section 4.4.1 except that the images are analyzed on a pixel-by-pixel basis, using the intensity measured at each pixel, instead of the average fluorescence intensity of the region. The constant a (Eq. (13.5)) for each pixel is then calculated from the normalized intensity ratios r(T) (Kato et al., 1999; Romano et al., 1989). The temperature distribution in a selected region of our instrument is shown in Fig. 13.8A. The temperature gradient in the absence of flow is about 0.06  C/mm (Fig. 13.8B).

280

Ichiro Amitani et al.

A

B 35 36 °C T 10 mm

Trap position

31 °C

The region used to calculate the temperature gradient shown in Fig. B on the right

Temperature (°C)

Temperature distribution at an IR power of 587 mW

0.061 ± 0.002 ⬚C/mm

34

33

32

-5

0 5 10 15 Distance from the trap center (mm)

20

Figure 13.8 Thermal gradient in a flow cell due to infrared heating by an optical trap. (A) Example showing the measured temperature distribution in an optical trap with an IR laser power of 587 mW when the temperature is set at a starting temperature of 24  C using only the microscope slide temperature controller. The circle indicates the position of the trap. The rectangle indicates the region where the temperature shown in B is measured. (B) Plot of the temperature distribution in the horizontal direction of the region shown in (A). Black: measured temperature; gray: linear fitting from 0.5 to 14.5 mm away from the trap center. Linear fitting gives a temperature gradient of 0.06  C/mm.

4.4.3. Temperature control Due to the need to use an oil-immersion lens, we discovered that the temperature of the flow cell is largely determined by the heat transfer from the objective to the sample. Although a thermoelectric microscope slide temperature controller (BC-100, 20/20 Technologies, Wilmington, NC) can be used for temperatures within a few degrees of ambient and for low power IR laser settings, a thermostated objective lens is a more effective regulator of sample temperature (Mao et al., 2005). A water circulation system is used to control the temperature of the oil-immersion objective (Fig. 13.6). The temperature control module consists of a brass jacket that fits onto the objective lens and copper tubing that is soldered for several turns around the jacket. The temperature of the objective is controlled by circulating temperature-controlled water from a water bath (Isotemp Refrigerated Circulator, Model 910, Fisher Scientific, Pittsburgh, PA) through the tubing. Water circulation does not perturb the optical trapping. This temperature control module can control the temperature of the sample from 15 to 45  C and reduces laser-induced heating by 60%.

5. Single-Molecule Imaging of Proteins on DNA A general single-molecule experiment includes the following steps: (1) prepare DNA attached to the polystyrene bead, either with or without a bound protein of interest; (2) introduce appropriate solutions into different

Watching Individual Proteins Acting on Single Molecules of DNA

281

channels of the multichannel flow cell; (3) capture a DNA–bead complex in the optical trap; and (4) move it into other channels containing ligands or proteins of interest.

5.1. Unwinding of DNA by a single RecBCD enzyme To visualize DNA unwinding by an individual RecBCD enzyme (Bianco et al., 2001; Spies et al., 2003), a complex of RecBCD enzyme bound to YOYO-1-stained DNA is optically trapped in the absence of ATP; the complex is then moved into the reaction channel, which contains ATP, to initiate DNA unwinding. DNA unwinding and RecBCD enzyme translocation are monitored as a shortening of the DNA length. 1. Prepare a sample buffer containing 45 mM NaHCO3 (pH 8.3), 20% sucrose (w/v), and 50 mM DTT; degas for at least 1 h. 2. Wash all syringes, tubing, and flow cell with 500 ml of 0.5% (v/v) of blocking reagent (B-10710, Molecular Probes, Carlsbad, CA) in sample buffer using a flow rate of 800 ml/h. 3. Prepare DNA–bead complexes as described in Section 2.2. 4. Add 500 ml of 20 nM YOYO-1 in sample buffer to the DNA–bead reaction. 5. Incubate in the dark at room temperature for at least 1 h. 6. Add Mg(OAc)2 and RecBCD to the stained DNA–bead complex at final concentrations of 2 mM and 50 nM; immediately transfer to the sample syringe (first channel). 7. Prepare 500 ml reaction solution containing the sample buffer supplemented with 2 mM Mg(OAc)2 and ATP at various concentrations; load the reaction syringe (second channel). 8. Trap a RecBCD–DNA–bead complex in the first channel. 9. Immediately move the trapped complex to the second channel to start the reaction. The unwinding of dsDNA is manifested by the shorting of the YOYO-1-labeled DNA (Fig. 13.9A and B).

5.2. Direct observation of RecBCD–nanoparticle translocation Another way to visualize translocation by individual RecBCD enzyme molecules is to attach a fluorescent nanoparticle to RecBCD (Handa et al., 2005). 1. Prepare a sample buffer containing 45 mM NaHCO3 (pH 8.3), 20% (w/v) sucrose, and 50 mM DTT; degas for at least 1 h. 2. Wash all syringes, tubing, and flow cell with 500 ml of 0.5% (v/v) of blocking reagent (B-10710, Molecular Probes) in sample buffer using a flow rate of 800 ml/h.

282

Ichiro Amitani et al.

A

B

RecBCD DNA length (bp)

60,000 50,000

1300 ± 4 bp/s

40,000 30,000 20,000 10,000 0 0

0⬙

10

20 Time (s)

40⬙

30

40

RecBCD

0⬙

35⬙

Nanoparticle position (bp)

D

C

60,000 50,000

1199 ± 5 bp/s

40,000 30,000 20,000 10,000 0 0

10

20 Time (s)

30

Figure 13.9 RecBCD translocating through, and unwinding, an individual l DNA molecule. (A) Kymograph showing a YOYO-1 stained l dsDNA molecule being unwound by a RecBCD molecule bound to the free DNA end. The drawing to the left of the kymograph depicts the optically trapped bead–YOYO-1-DNA–RecBCD complex. (B) Plot of DNA length versus time. Black line shows the fit to a straight line. (C) Kymograph showing translocation by a fluorescent nanoparticle-labeled RecBCD molecule on l dsDNA. The drawing on the left side of the kymograph depicts the optically trapped bead–DNA–RecBCD–nanoparticle complex. (D) Plot of the position of the RecBCD molecule, indicated by the nanoparticle, versus time. The black line shows the fit to a straight line. Note that the difference in unwinding rates in (B) and (D) is not due to a difference in the techniques, but rather reflects the intrinsic heterogeneity of individual RecBCD enzyme behavior.

3. Prepare DNA–bead complexes as described in Section 2.2. 4. Label biotinylated RecBCD using a fluorescent nanoparticle as described in Section 3.1. 5. Add the DNA–bead complex and 2 mM Mg(OAc)2 to the biotinylated RecBCD–nanoparticle complex, and incubate the resulting mixture for 2 min. 6. Dilute the nanoparticle–RecBCD–DNA–bead complex with 400 ml of degassed sample buffer supplemented with 2 mM Mg(OAc)2 and 0.5% (v/v) blocking solution (B-10710, Molecular Probes); transfer to the sample syringe (first channel).

283

Watching Individual Proteins Acting on Single Molecules of DNA

7. Load the reaction syringe with the reaction buffer consisting of 1 mM ATP, 2 mM Mg(OAc)2, and 0.5% (v/v) blocking solution in sample buffer (second channel). 8. Trap a nanoparticle–RecBCD–DNA–bead complex in the first channel. 9. Immediately move the trapped complex to the second channel. The fluorescent particle is seen to move toward the trapped bead as RecBCD translocates on the DNA (Fig. 13.9C and D).

5.3. Rad54/Tid1 translocation To observe Rad54/Tid1 translocation (Amitani et al., 2006; Nimonkar et al., 2007), a two-channel flow cell is used. Figure 13.10A shows a schematic illustration of the translocation assay, a kymograph showing A

FITC-Rad54 Downstream l DNA

Optical trap Upstream

Flow

B Start

End

Bead 0⬙

Rad54 position from bead (bp)

C

370⬙ 50,000 40,000 30,000 20,000 10,000 0 0

100

200

300

400

Time (s)

Figure 13.10 Rad54 translocating on a single dsDNA molecule. (A) Schematic illustration of the optically trapped l DNA–bead complex with a bound FITC–Rad54 complex. (B) Kymographs depicting upstream translocation (in the direction opposite to flow) of Rad54 on the dsDNA. (C) Plot of FITC–Rad54 position relative to the bead versus time.

284

Ichiro Amitani et al.

translocation by Rad54 labeled with FITC-antibody (Fig. 13.10B), and a graph of FITC–Rad54 position as a function of time (Fig. 13.10C). 1. To reduce nonspecific binding, the syringe, tubing, and flow cell are incubated with 0.5 mg/ml of BSA in 50 mM Tris–HOAc (pH 7.5) at room temperature for 15 min, then rinsed with a 10-fold volume of 50 mM Tris–HOAc (pH 7.5). 2. Prepare FITC–Rad54/Tid1–DNA–bead complexes as described in Section 3.2. 3. Load 400 ml of the FITC–Rad51/Tid1–DNA–bead complex (typically 90 fM) in 40 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 15% (w/v) sucrose into the first channel. 4. Load 400 ml of solution containing 1 mM ATP, 2 mM Mg(OAc)2, 40 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 15% (w/v) sucrose into the second channel. 5. Trap a FITC–Rad54/Tid1–DNA–bead complex in the optical trap in the first channel. 6. Move the complex to the second channel containing the ATP to initiate translocation. The fluorescently tagged Rad54 is seen to translocate toward the trapped bead (upstream) (Fig. 13.10B and C) or away from the trapped bead (downstream).

5.4. Real-time Rad51 assembly To detect the assembly of Rad51 in real time, a two-channel flow cell is used (Hilario et al., 2009). Rad51 assembly can be measured by monitoring the increase in the length of fluorescently end-labeled DNA. A kymograph is shown in Fig. 13.11A; Fig. 13.11B shows a graph of DNA length versus time. 1. To reduce nonspecific binding, the syringe, tubing, and flow cell are washed at 800 ml/h with 0.5 mg/ml of BSA and 0.5 mg/ml casein in 50 mM Tris–HOAc (pH 7.5) at room temperature for 1 h, followed by a rinse with sample buffer for 1 h. 2. Prepare Cy3–DNA–bead complexes as described in Section 2.3. 3. Load 400 ml of solution containing Cy3–DNA–bead complexes (about 90 fM), 40 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 15% (w/v) sucrose into the first channel. 4. Load 400 ml of solution containing Rad51, 10 mM Mg(OAc)2, 2 mM ATP, 40 mM Tris–HOAc (pH 7.5), 30 mM DTT, and 15% (w/v) sucrose into the second channel. The Rad51 concentration can be varied from 50 nM to 1 mM. 5. Trap a Cy3–DNA–bead complex in the first channel. 6. Move the complex to the second channel to initiate Rad51 assembly.

285

Watching Individual Proteins Acting on Single Molecules of DNA

A

B

C

20 15 10 0

Rad51 nucleoprotein filament disassembly

0⬙

25

50⬙

60⬙

10

D DNA extension (mm)

0⬙

DNA extension (mm)

Rad51 nucleoprotein filament assembly

30 20 Time (s)

40

50

25 20 15 10 0

20

40 60 Time (s)

80

Figure 13.11 Rad51 assembling onto, and dissociating from, a single dsDNA molecule. (A) Kymograph of Rad51 assembly on Cy3-end-labeled l DNA. The schematic on the left side of kymograph depicts the optically trapped bead, initial position of Cy3end-label of the DNA (star), and DNA (solid line); the schematic on the right side depicts the extended Rad51 nucleoprotein filament. DNA length is measured from the center of the bead to the Cy3-end-label. (B) Plot of DNA length versus time for the assembly of Rad51 on DNA analyzed using two-dimensional Gaussian fitting of the end-label position. (C) Kymograph of Rad51 disassembly from Cy3-end-labeled l DNA. The schematic on the left side of kymograph depicts the optically trapped bead, initial position of Cy3-end-label of the DNA (star), Rad51 (filled circles), and DNA (solid line). (D) Plot of DNA length versus time for the disassembly of Rad51 from DNA.

5.5. Real-time Rad51 disassembly To visualize the disassembly of a Rad51 nucleoprotein filament in real time, a three-channel flow cell is used (Hilario et al., 2009). Rad51 disassembly can be measured by monitoring the decrease in the length of fluorescently end-labeled DNA onto which Rad51 is assembled. A kymograph of Rad51 disassembly is shown in Fig. 13.11C; Fig. 13.11D is the graph of DNA length versus time. 1. The syringe, tubing, and flow cell are washed at 800 ml/h with 0.5 mg/ml of BSA and 0.5 mg/ml casein in 50 mM Tris–HOAc (pH 7.5) at room temperature for 1 h, followed by a rinse with sample buffer for 1 h. 2. Prepare Cy3–DNA–bead complexes as described in Section 2.3.

286

Ichiro Amitani et al.

3. Load 400 ml of solution containing Cy3–DNA–bead complexes (typically about 90 fM), 40 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 15% (w/v) sucrose into the first channel. 4. Load 400 ml of solution containing 40 mM Tris–HOAc (pH 7.5), 10 mM Mg(OAc)2, 30 mM DTT, and 15% (w/v) sucrose into the second channel; the ATP concentration can be varied from 0 to 2 mM. 5. Load 400 ml of solution containing 1 mM Rad51, 40 mM Tris–HOAc (pH 7.5), 10 mM Mg(OAc)2, 2 mM ATP, 30 mM DTT, and 15% (w/v) sucrose into the third channel. 6. Trap a Cy3–DNA–bead complex in the first channel. 7. Move the complex to the third channel to assemble Rad51 onto the DNA. 8. Move the complex to the second channel to initiate Rad51 disassembly from the DNA.

5.6. Visualization of RecAFAM/RecA-RFP/Rad51FAM filament formation To directly visualize nucleation (cluster formation) by fluorescent RecA or Rad51 (Galletto et al., 2006; Handa et al., 2009; Hilario et al., 2009), a threechannel flow cell is used. To confirm that a single molecule of DNA is present, YOYO-1-stained DNA is used in the initial trapping. YOYO-1 is removed by washing the DNA in a buffer containing 5–10 mM Mg(OAc)2 before starting the assay. A time course of RecA-RFP cluster formation is shown in Fig. 13.12. Here, we describe the protocol for imaging RecARFP cluster formation. 1. The syringe, tubing, and flow cell are incubated with 0.5 mg/ml of BSA in 50 mM Tris–HOAc (pH 7.5) at room temperature for 15 min, then rinsed with a 10-fold volume of 50 mM Tris–HOAc (pH 7.5). 2. Prepare YOYO-1-stained DNA–bead complexes as described in Section 2.2. 3. Load 400 ml of solution containing YOYO-1-stained DNA–bead complex, 20 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 20% (w/v) sucrose into the first channel. 4. Load 400 ml of solution containing 0.5 mM ATPgS, 5 mM Mg(OAc)2, 20 mM Tris–HOAc (pH 8.2), 30 mM DTT, and 20% (w/v) sucrose into the second channel. 5. Load 400 ml of solution containing RecA-RFP, 0.5 mM ATPgS, 1 mM Mg(OAc)2, 20 mM MES (pH 6.2), 30 mM DTT, and 20% (w/v) sucrose into the third channel. The RecA-RFP concentration is varied from 150 to 400 nM. 6. Trap a YOYO-1-stained DNA–bead complex in the first channel.

287

Watching Individual Proteins Acting on Single Molecules of DNA

A

RecA-RFP cluster

Bead

l DNA B 110 nM RecA-RFP

140 nM RecA-RFP

170 nM RecA-RFP

200 nM RecA-RFP

10 s

10 s

10 s

5s

20 s

20 s

20 s

10 s

30 s

30 s

30 s

15 s

40 s

40 s

40 s

20 s

Bead

Figure 13.12 Direct visualization of nucleation and growth of RecA-RFP nucleoprotein filaments on individual molecules of dsDNA. (A) Illustration of the nucleation of RecA-RFP clusters on DNA. (B) Representative video frames showing nucleation at four different RecA-RFP concentrations. Flow is left to right. Each vertical strip represents the same DNA molecule repeatedly dipped into the RecA-RFP protein solution for the incubation times indicated. The trapped bead position is indicated by an arrow; the bead is fluorescent due to the nonspecific binding of the RecA-RFP.

7. Move the complex to the second channel to confirm by visual inspection that a single molecule of intact l DNA is attached to the bead. 8. Incubate the complex for 2–3 min in the second channel to dissociate the YOYO-1. In this step, the shutter for excitation light is closed to avoid the photocleavage of DNA. 9. Move the complex to the third channel containing the RecA-RFP. 10. Incubate the complex for 5–30 s in the third channel. In this step, the shutter for excitation light is closed to protect the CCD camera from the strong signal due to a high concentration of fluorescent protein. 11. Move the complex to the second channel to observe the RecA-RFP clusters. To avoid the photobleaching of RecA-RFP, the illumination time should be minimized (1–2 s). 12. Repeat steps 9 and 10.

6. Data Analysis Methods To improve the efficiency of the single-molecule studies and increase the reliability of the data analysis, we developed programs to analyze the fluorescence images. The software is available upon request.

288

Ichiro Amitani et al.

6.1. Two-dimensional Gaussian fitting To determine the center of a fluorescent intensity distribution (e.g., for Cy3-end-labeled DNA or FITC–Rad54/Tid1), the distribution is fitted to the following two-dimensional Gauss function (Hilario et al., 2009; Nimonkar et al., 2007): " # ðx  xc Þ2 ðy  yc Þ2 f ðx; yÞ ¼ A  exp  þB ð13:8Þ  s2x s2y Here, f(x, y) is the point spread function (PSF), A the maximum intensity of PSF, x, y the coordinates of image, xc, yc the center of PSF, sx, sy the width of the PSF, and B the background intensity. To perform nonlinear regression, Eq. (13.8) is expanded to a Taylor series, and the zeroth and first derivative terms are used. The fitting process is iterated until the relative change of fitting parameters falls within a predefined threshold (5%; empirically chosen as a threshold because the fluctuation of fluorescent spot position is greater than the fluorescent spot size). To reduce the calculation time, a small region surrounding the fluorescent spot is selected for fitting. The image in the region of interest is median-filtered and averaged before fitting. The initial parameters for fitting are determined automatically from the fluorescent intensity distribution in the region. For the DNA length measurement, the radius of bead is known and subtracted from the observed length. Figs. 13.10C, 13.11B, and D show the results of a two-dimensional Gaussian fitting.

6.2. Automatic DNA length measurement To measure the length of DNA that is either fluorescently labeled with YOYO-1 or decorated with fluorescent proteins, we developed a plug-in for ImageJ to automate the analysis. Fig. 13.9B shows an example of the analysis. 1. To improve the data quality, fluorescence images are first averaged every 5–10 frames. 2. The position of the trapped bead is initially determined manually, taking advantage of the nonspecific binding of dye to the bead. 3. For each frame, a radial line scan originating from the bead center is determined. The orientation of the DNA is determined as the direction that has the maximum mean gray value. 4. The profile along the orientation of the DNA is calculated. 5. The derivative of the profile is calculated. The position of the maximum (excluding the bead) in the derivative is the position of the DNA end. The length of the DNA is calculated as the distance from the DNA end to the bead center, minus the radius of the bead.

Watching Individual Proteins Acting on Single Molecules of DNA

289

ACKNOWLEDGMENTS We wish to thank Jason Bell, Aura Carreira, Petr Cejka, Anthony Forget, Joe Hilario, Taeho Kim, Hsu-Yang Lee, Katsumi Morimatsu, Amitabh Nimonkar, Behzad Rad, and Lisa Vancelette for their comments on this manuscript, and members of the Kowalczykowski lab for their contribution to this research. The research in our lab has been funded by grants from National Institutes of Health T32 CA-108459 to C. C. D., and GM-41347, GM-62653, and GM-64745 to S. C. K.

REFERENCES Alexeev, A., Mazin, A., and Kowalczykowski, S. C. (2003). Rad54 protein possesses chromatin-remodeling activity stimulated by the Rad51-ssDNA nucleoprotein filament. Nat. Struct. Biol. 10, 182–186. Amitani, I., Baskin, R. J., and Kowalczykowski, S. C. (2006). Visualization of Rad54, a chromatin remodeling protein, translocating on single DNA molecules. Mol. Cell 23, 143–148. Anderson, D. G., and Kowalczykowski, S. C. (1997a). The recombination hot spot w is a regulatory element that switches the polarity of DNA degradation by the RecBCD enzyme. Genes Dev. 11, 571–581. Anderson, D. G., and Kowalczykowski, S. C. (1997b). The translocating RecBCD enzyme stimulates recombination by directing RecA protein onto ssDNA in a w-regulated manner. Cell 90, 77–86. Benson, F. E., Stasiak, A., and West, S. C. (1994). Purification and characterization of the human Rad51 protein, an analogue of E. coli RecA. EMBO J. 13, 5764–5771. Bianco, P. R., Tracy, R. B., and Kowalczykowski, S. C. (1998). DNA strand exchange proteins: A biochemical and physical comparison. Front. Biosci. 3, D570–D603. Bianco, P. R., Brewer, L. R., Corzett, M., Balhorn, R., Yeh, Y., Kowalczykowski, S. C., and Baskin, R. J. (2001). Processive translocation and DNA unwinding by individual RecBCD enzyme molecules. Nature 409, 374–378. Brewer, L. R., and Bianco, P. R. (2008). Laminar flow cells for single-molecule studies of DNA-protein interactions. Nat. Methods 5, 517–525. Campbell, M. J., and Davis, R. W. (1999). Toxic mutations in the recA gene of E. coli prevent proper chromosome segregation. J. Mol. Biol. 286, 417–435. Celliers, P. M., and Conia, J. (2000). Measurement of localized heating in the focus of an optical trap. Appl. Opt. 39, 3396–3407. Chen, Z., Yang, H., and Pavletich, N. P. (2008). Mechanism of homologous recombination from the RecA-ssDNA/dsDNA structures. Nature 453, 489–494. Conway, A. B., Lynch, T. W., Zhang, Y., Fortin, G. S., Fung, C. W., Symington, L. S., and Rice, P. A. (2004). Crystal structure of a Rad51 filament. Nat. Struct. Mol. Biol. 11, 791–796. Dillingham, M. S., and Kowalczykowski, S. C. (2008). RecBCD enzyme and the repair of double-stranded DNA breaks. Microbiol. Mol. Biol. Rev. 72, 642–671. Dixon, D. A., and Kowalczykowski, S. C. (1993). The recombination hotspot w is a regulatory sequence that acts by attenuating the nuclease activity of the E. coli RecBCD enzyme. Cell 73, 87–96. Galletto, R., Amitani, I., Baskin, R. J., and Kowalczykowski, S. C. (2006). Direct observation of individual RecA filaments assembling on single DNA molecules. Nature 443, 875–878.

290

Ichiro Amitani et al.

Handa, N., Bianco, P. R., Baskin, R. J., and Kowalczykowski, S. C. (2005). Direct visualization of RecBCD movement reveals cotranslocation of the RecD motor after w recognition. Mol. Cell 17, 745–750. Handa, N., Amitani, I., Gumlaw, N., Sandler, S. J., and Kowalczykowski, S. C. (2009). Single molecule analysis of a red fluorescent RecA protein reveals a defect in nucleoprotein filament nucleation that relates to its reduced biological functions. J. Biol. Chem. 284, 18664–18673. Hilario, J., Amitani, I., Baskin, R. J., and Kowalczykowski, S. C. (2009). Direct imaging of human Rad51 nucleoprotein dynamics on individual DNA molecules. Proc. Natl. Acad. Sci. USA 106, 361–368. Holzen, T. M., Shah, P. P., Olivares, H. A., and Bishop, D. K. (2006). Tid1/Rdh54 promotes dissociation of Dmc1 from nonrecombinogenic sites on meiotic chromatin. Genes Dev. 20, 2593–2604. Karstens, T., and Kobs, K. (1980). Rhodamine-B and rhodamine-101 as reference substances for fluorescence quantum yield measurements. J. Phys. Chem. 84, 1871–1872. Kato, H., Nishizaka, T., Iga, T., Kinosita, K. Jr., and Ishiwata, S. (1999). Imaging of thermal activation of actomyosin motors. Proc. Natl. Acad. Sci. USA 96, 9602–9606. Klein, H. L. (1997). RDH54, a RAD54 homologue in Saccharomyces cerevisiae, is required for mitotic diploid-specific recombination and repair and for meiosis. Genetics 147, 1533–1543. Kowalczykowski, S. C. (2000). Initiation of genetic recombination and recombinationdependent replication. Trends Biochem. Sci. 25, 156–165. Liu, Y., Cheng, D. K., Sonek, G. J., Berns, M. W., Chapman, C. F., and Tromberg, B. J. (1995). Evidence for localized cell heating induced by infrared optical tweezers. Biophys. J. 68, 2137–2144. Mao, H., Arias-Gonzalez, J. R., Smith, S. B., Tinoco, I. Jr., and Bustamante, C. (2005). Temperature control methods in a laser tweezers system. Biophys. J. 89, 1308–1316. Mazin, A. V., Bornarth, C. J., Solinger, J. A., Heyer, W. D., and Kowalczykowski, S. C. (2000). Rad54 protein is targeted to pairing loci by the Rad51 nucleoprotein filament. Mol. Cell 6, 583–592. Mazin, A. V., Alexeev, A. A., and Kowalczykowski, S. C. (2003). A novel function of Rad54 protein: Stabilization of the Rad51 nucleoprotein filament. J. Biol. Chem. 278, 14029–14036. Modesti, M., Ristic, D., van der Heijden, T., Dekker, C., van Mameren, J., Peterman, E. J., Wuite, G. J., Kanaar, R., and Wyman, C. (2007). Fluorescent human RAD51 reveals multiple nucleation sites and filament segments tightly associated along a single DNA molecule. Structure 15, 599–609. Neuman, K. C., and Block, S. M. (2004). Optical trapping. Rev. Sci. Instrum. 75, 2787–2809. Nimonkar, A. V., Amitani, I., Baskin, R. J., and Kowalczykowski, S. C. (2007). Single molecule imaging of Tid1/Rdh54, a Rad54 homolog that translocates on duplex DNA and can disrupt joint molecules. J. Biol. Chem. 282, 30776–30784. Ogawa, T., Yu, X., Shinohara, A., and Egelman, E. H. (1993). Similarity of the yeast RAD51 filament to the bacterial RecA filament. Science 259, 1896–1899. Peterman, E. J. G., Gittes, F., and Schmidt, C. F. (2003). Laser-induced heating in optical traps. Biophys. J. 84, 1308–1316. Petukhova, G., Stratton, S., and Sung, P. (1998). Catalysis of homologous DNA pairing by yeast Rad51 and Rad54 proteins. Nature 393, 91–94. Prasad, T. K., Yeykal, C. C., and Greene, E. C. (2006). Visualizing the assembly of human Rad51 filaments on double-stranded DNA. J. Mol. Biol. 363, 713–728.

Watching Individual Proteins Acting on Single Molecules of DNA

291

Prasad, T. K., Robertson, R. B., Visnapuu, M. L., Chi, P., Sung, P., and Greene, E. C. (2007). A DNA-translocating Snf2 molecular motor: Saccharomyces cerevisiae Rdh54 displays processive translocation and extrudes DNA loops. J. Mol. Biol. 369, 940–953. Robertson, R. B., Moses, D. N., Kwon, Y., Chan, P., Chi, P., Klein, H., Sung, P., and Greene, E. C. (2009). Structural transitions within human Rad51 nucleoprotein filaments. Proc. Natl. Acad. Sci. USA 106, 12688–12693. Roman, L. J., and Kowalczykowski, S. C. (1989). Characterization of the helicase activity of the Escherichia coli RecBCD enzyme using a novel helicase assay. Biochemistry 28, 2863–2873. Romano, V., Zweig, A. D., Frenz, M., and Weber, H. P. (1989). Time-resolved thermal microscopy with fluorescent films. Appl. Phys. B Photophys. Laser Chem. 49, 527–533. Schatz, P. J. (1993). Use of peptide libraries to map the substrate specificity of a peptidemodifying enzyme: A 13 residue consensus peptide specifies biotinylation in Escherichia coli. Biotechnology (N. Y.) 11, 1138–1143. Shinohara, M., Shita-Yamaguchi, E., Buerstedde, J. M., Shinagawa, H., Ogawa, H., and Shinohara, A. (1997). Characterization of the roles of the Saccharomyces cerevisiae RAD54 gene and a homologue of RAD54, RDH54/TID1, in mitosis and meiosis. Genetics 147, 1545–1556. Shinohara, M., Gasior, S. L., Bishop, D. K., and Shinohara, A. (2000). Tid1/Rdh54 promotes colocalization of rad51 and dmc1 during meiotic recombination. Proc. Natl. Acad. Sci. USA 97, 10814–10819. Solinger, J. A., and Heyer, W. D. (2001). Rad54 protein stimulates the postsynaptic phase of Rad51 protein-mediated DNA strand exchange. Proc. Natl. Acad. Sci. USA 98, 8447–8453. Solinger, J. A., Lutz, G., Sugiyama, T., Kowalczykowski, S. C., and Heyer, W. D. (2001). Rad54 protein stimulates heteroduplex DNA formation in the synaptic phase of DNA strand exchange via specific interactions with the presynaptic Rad51 nucleoprotein filament. J. Mol. Biol. 307, 1207–1221. Solinger, J. A., Kiianitsa, K., and Heyer, W. D. (2002). Rad54, a Swi2/Snf2-like recombinational repair protein, disassembles Rad51:dsDNA filaments. Mol. Cell 10, 1175–1188. Spies, M., and Kowalczykowski, S. C. (2005). Homologous recombination by RecBCD and RecF pathways. In ‘‘The Bacterial Chromosome,’’ (N. P. Higgins, ed.), pp. 389–403. ASM Press, Washington, DC. Spies, M., Bianco, P. R., Dillingham, M. S., Handa, N., Baskin, R. J., and Kowalczykowski, S. C. (2003). A molecular throttle: The recombination hotspot w controls DNA translocation by the RecBCD helicase. Cell 114, 647–654. Spies, M., Amitani, I., Baskin, R. J., and Kowalczykowski, S. C. (2007). RecBCD enzyme switches lead motor subunits in response to w recognition. Cell 131, 694–705. Stasiak, A., Di Capua, E., and Koller, T. (1981). Elongation of duplex DNA by recA protein. J. Mol. Biol. 151, 557–564. Sung, P., and Robberson, D. L. (1995). DNA strand exchange mediated by a RAD51ssDNA nucleoprotein filament with polarity opposite to that of RecA. Cell 82, 453–461. Taylor, A. F., and Smith, G. R. (1995). Strand specificity of nicking of DNA at Chi sites by RecBCD enzyme: Modulation by ATP and magnesium levels. J. Biol. Chem. 270, 24459–24467. van der Heijden, T., Seidel, R., Modesti, M., Kanaar, R., Wyman, C., and Dekker, C. (2007). Real-time assembly and disassembly of human RAD51 filaments on individual DNA molecules. Nucleic Acids Res. 35, 5646–5657. Zaitsev, E. N., and Kowalczykowski, S. C. (2000). A novel pairing process promoted by Escherichia coli RecA protein: Inverse DNA and RNA strand exchange. Genes Dev. 14, 740–749.

C H A P T E R

F O U R T E E N

DNA Curtains for High-Throughput Single-Molecule Optical Imaging Eric C. Greene,*,† Shalom Wind,§ Teresa Fazio,§ Jason Gorman,‡ and Mari-Liis Visnapuu† Contents 1. Introduction 2. Total Internal Reflection Fluorescence Microscopy 2.1. General description of TIRFM 2.2. Building a simple prism-type TIRFM 2.3. Flowcells and injection system 3. DNA Curtains 3.1. DNA curtains as a method for aligning thousands of DNA molecules 3.2. Manually etched diffusion barriers 3.3. Nanofabricated linear diffusion barriers 3.4. More complex barrier patterns 3.5. Trouble-shooting 4. Visualizing Protein–DNA Interactions 4.1. Quantum dots as a general fluorescent labeling strategy 4.2. Visualizing ATP-dependent DNA translocation 4.3. Using DNA curtains to image nucleosomes 4.4. Diffusion of MMR along DNA 5. Conclusions and Future Directions Acknowledgments References

294 294 295 295 295 297 297 298 298 303 306 307 307 308 310 310 314 314 314

Abstract Single-molecule approaches provide a valuable tool in the arsenal of the modern biologist, and new discoveries continue to be made possible through the use of these state-of-the-art technologies. However, it can be inherently difficult to obtain statistically relevant data from experimental approaches specifically designed to * The Howard Hughes Medical Institute, Columbia University, New York, USA Department of Biochemistry and Molecular Biophysics, Columbia University, New York, USA Department of Biological Sciences, Columbia University, New York, USA } Department of Applied Physics and Applied Mathematics, Center for Electron Transport in Molecular Nanostructures, NanoMedicine Center for Mechanical Biology, Columbia University, New York, USA { {

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72006-1

#

2010 Elsevier Inc. All rights reserved.

293

294

Eric C. Greene et al.

probe individual reactions. This problem is compounded with more complex biochemical reactions, heterogeneous systems, and/or reactions requiring the use of long DNA substrates. Here we give an overview of a technology developed in our laboratory, which relies upon simple micro- or nanofabricated structures in combination with ‘‘bio-friendly’’ lipid bilayers, to align thousands of long DNA molecules into defined patterns on the surface of a microfluidic sample chamber. We call these ‘‘DNA curtains,’’ and we have developed several different versions varying in complexity and DNA substrate configuration, which are designed to meet different experimental needs. This novel approach to single-molecule imaging provides a powerful experimental platform that offers the potential for concurrent observation of hundreds or even thousands of protein–DNA interactions in real time.

1. Introduction Single-molecule techniques have grown into important experimental tools for scientists interested in understanding mechanisms involving biological macromolecules. However, while single-molecule approaches can be powerful, they also suffer limitations. For example, it is often challenging to acquire statistically meaningful data, and this problem is compounded with biological systems that are heterogeneous and/or contain rare or transient reaction intermediates. In addition, single-molecule techniques often require that one or more of the reactants under investigation be anchored to a solid support (Visnapuu et al., 2008a,b). Nonspecific interactions with the solid support can render a biological system experimentally inaccessible. Part of our research efforts have been devoted to minimizing these experimental difficulties by developing new methodologies making it possible to organize thousands of individual DNA molecules into defined patterns on optical surfaces coated with ‘‘bio-friendly’’ lipid bilayers that mimic cell membranes (Fazio et al., 2008; Gorman et al., 2010; Grane´li et al., 2006; Visnapuu et al., 2008a,b). We call these methodologies ‘‘DNA curtains,’’ and they enable us to image hundreds or even thousands of individual molecules in real time by fluorescence microscopy. This report provides detailed information on these experimental platforms such that they can be replicated by anyone with experience in general laboratory techniques and optical instrumentation.

2. Total Internal Reflection Fluorescence Microscopy Total internal reflection fluorescence microscopy (TIRFM) uses spatially selective laser excitation to limit fluorescence background (Axelrod, 1989), and a TIRF microscope is our instrument of choice for the types of wide-field fluorescent imaging studies that are described below.

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

295

2.1. General description of TIRFM For TIRFM, a laser beam is directed through a microscope slide and reflected off the interface between the slide and an aqueous buffer (Axelrod, 1989). This spatially selective illumination geometry makes use of the fact that when light is reflected at an interface between a transparent slide and an aqueous buffer of differing refractive indexes, the incident energy is not abruptly reflected, but rather penetrates a few hundred nanometers into the buffer. The practical consequence of this type of illumination is that it yields a very small excitation volume—typically on the order of a few femtoliters. This minimizes excitation of contaminants and molecules in bulk solution, thereby reducing the background signal by several orders of magnitude relative to conventional wide-field illumination techniques. For additional discussion of more detailed aspects of TIRFM, we refer the reader to several excellent reviews (Axelrod, 1989; Forkey et al., 2000; Ha, 2001).

2.2. Building a simple prism-type TIRFM We use a Nikon TE2000U microscope with a simple through-prism illumination configuration. The excitation source is provided by a diodepumped solid-state laser (DPSSL; 488 nm, 200 mW, Sapphire, Coherent, Inc.), which is focused through the face of a fused silica prism onto a microfluidic flowcell to generate an evanescent wave within the sample chamber. Alignment is controlled by a remotely operated motorized mirror (New Focus, Inc.) that guides the beam to the prism. Photons are collected with a microscope objective (100
, 1.4 NA, oil immersion Plan Apo, Nikon or 60
, 1.2 NA, water immersion Plan Apo, Nikon), passed through a notch filter (Semrock) to block scattered laser light, and detected using a back-illuminated electron-multiplying CCD (EMCCD; Cascade II, Photometrics). When used for multicolor operation, the photons are passed through an image-splitter (Roper Bioscience) containing a dichroic mirror that separates the optical paths. The entire TIRFM system is mounted on an optical table (any standard optical table will suffice for most applications) to minimize vibrations and facilitate mounting of optical components.

2.3. Flowcells and injection system Our TIRFM experiments are performed within microfluidic flowcells that are machined and assembled in-house (Fig. 14.1). Each flowcell is made from a fused silica glass slide (G. Finkenbeiner, Inc.) and the surface of the slide is prepared with micro- or nanofabricated barrier patterns (see below). Prior to assembly, inlet and outlet holes are bored into each slide, using a diamond-coated 1.4 mm drill bit (Shor International) mounted on a

296

Eric C. Greene et al.

A

B

C

D

E

F

G

H

I

J

K

Figure 14.1 Flowcells: (A) shows the drill used to bore through the fused silica slide glass, and during drilling the slide is immersed in a flowing water bath to remove glass particulates. The slide (B) is then thoroughly cleaned, a narrow strip of paper (red) is used to protect the surface and also serves as a template for the sample chamber, and the surface is covered with double-sided tape (C). A narrow channel is excised around the paper template, and a coverslip is placed over the sample chamber (D). The flowcell is clamped (E) and placed in a vacuum oven to seal the chamber and nanoports are attached (F). Reagents necessary for deposition of the bilayer are injected into the sample chamber using syringes as shown (G). When a flowcell is ready for use, it is positioned on the microscope stage (H) and held in place by heating units as shown in (I), (J) and (K ).

precision drill press (Servo Products Company). The slide is submerged under continuously flowing water while being drilled, which cools the bit and removes fused silica dust. The slides are cleaned by sequential submersion in 2% HELLMANEXTM, 1 M NaOH (for 30 min each), and rinsed under running MilliQTM water between each step. This is followed by a rinse in absolute methanol and drying at 120  C under vacuum for at least 1 h. The sample chamber is prepared from double-sided tape (3M) placed over the slide, which is overlaid with a borosilicate glass coverslip (Fisher Scientific). The flowcell is clamped between two additional slides to evenly distribute pressure and placed in a 120  C vacuum oven for 2 h. Inlet and outlet ports (Upchurch Scientific) are glued over the drilled holes.

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

297

The completed assembly is mounted on a remotely operated stage (Ludl Electronic Products Ltd.), a syringe pump (KD Scientific) is connected to control the rate of buffer flow through the chamber, and a six-way injection valve (SCIVEX) controls sample delivery.

3. DNA Curtains Our studies utilize supported lipid bilayers as a means for passivating the sample chamber surface. The advantages of bilayers over other types of surfaces is that they mimic the cellular environment, can be modified through the incorporation of lipids with alternative head groups, and are easy to deposit on fused silica. Moreover, the use of fluid bilayers, in combination with barriers to lipid diffusion (Cremer and Boxer, 1999; Groves and Boxer, 2002; Groves et al., 1997), have allowed the development of ‘‘DNA curtains,’’ in which hundreds or thousands of DNA molecules can be aligned and imaged (Fazio et al., 2008; Gorman et al., 2010; Grane´li et al., 2006; Visnapuu et al., 2008a,b).

3.1. DNA curtains as a method for aligning thousands of DNA molecules The idea for DNA curtains came from the understanding that DNA molecules anchored to a fluid bilayer would move in the direction of an applied hydrodynamic force. A barrier placed across the path of the moving DNA and oriented perpendicular to the flow force could be used to halt the forward progression of the DNA. This would also cause the DNA molecules to align with one another in the same orientation, which would make it possible to visualize numerous DNA molecules in a single field-of-view. The general procedure for making DNA curtains is the same regardless of the barrier type (Fazio et al., 2008; Gorman et al., 2010; Grane´li et al., 2006; Visnapuu et al., 2008a,b). All lipids are purchased from Avanti Polar Lipids and stored in chloroform at  20  C. The chloroform is evaporated prior to liposome preparation using a stream of nitrogen and dried further under vacuum onto the glass wall of a test tube for 2–12 h. Lipids are resuspended in buffer A, which contains 10 mM Tris (pH 8.0) and 100 mM NaCl, at a concentration of 10 mg/ml, and sonicated to form liposomes, which are stored at 4  C and used within one week of preparation. Liposomes can also be prepared by extrusion through a polycarbonate filter with 100-nm pores (Avanti Polar Lipids, Alabaster, AL); however, we have had more uniform results, using liposomes prepared by sonication. The liposomes are comprised of a mixture of DOPC (1,2-dioleoyl-sn-glycero-phosphocholine), 0.5% (w/v) biotinylated-DPPE (1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(cap biotinyl)), and

298

Eric C. Greene et al.

8% (w/v) mPEG 550-DOPE (1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-550]). The mPEG is necessary only when using quantum dots as fluorescent tags (see below). For bilayer deposition, liposomes are injected into the sample chamber and incubated for 30 min. Excess liposomes are flushed away with buffer A and incubated for another 30 minutes. The flowcell is then rinsed with buffer B (40 mM Tris–HCl (pH 7.8), 1 mM DTT, 1 mM MgCl2, and 0.2 mg/ml BSA) and incubated for an additional 15 min. Streptavidin (0.02 mg/ml) in buffer A is injected into the sample chamber and incubated for 10 min. Streptavidin binds the biotinylated lipids, providing an attachment point for biotinylated DNA. After rinsing thoroughly with additional buffer B, biotinylated DNA (10 pM) pre-stained with 1–2 nM YOYO1 is injected into the sample chamber, incubated for 10 min, and unbound DNA is removed by flushing with buffer. Our work relies primarily upon the 48.5-kb genome of bacteriophage l. This DNA is commercially available and contains natural 12-nucleotide overhangs, which can be tagged with complementary oligonucleotides. Application of buffer flow causes the lipid-tethered DNA molecules to align along the diffusion barriers. If the DNA molecules are not evenly aligned along the full length of the barrier edges, then the buffer flow can be paused briefly, allowing the DNA molecules to diffuse away from areas of high density. The flow is then resumed, and the flow on–off cycle is repeated roughly 3–5 times until DNA curtains of even density form at the barriers.

3.2. Manually etched diffusion barriers Diffusion barriers can be made either by manual etching or by nanofabrication (Fig. 14.2, and described below). Manually etched barriers do not require specialized fabrication instrumentation (we routinely use diamond-tipped drill bits; Grane´li et al., 2006). The slide is etched prior to flowcell assembly by lightly dragging the scribe across its surface, and this is done repeatedly to ensure the presence of numerous barriers. Once the etched slide has been cleaned, it can be assembled into a flowcell, as described above. The disadvantages of manually etched barriers are that it is impossible to control barrier width, depth, or location, and the procedure itself can yield barriers of drastically different quality (Fig. 14.3D–I). In practice, the slide is visually scanned after assembly of the DNA curtains to identify areas of sufficient quality for data acquisition.

3.3. Nanofabricated linear diffusion barriers Nanofabricated barrier patterns can be made by either electron-beam (ebeam) or nanoimprint lithography, and yield uniform DNA curtains of high quality (Fig. 14.2; Fazio et al., 2008; Gorman et al., 2010; Visnapuu

299

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

A

Slide

Manually etch

Scratch

C

Master

PMMA 35K

B

Slide

Aquasave

Imprint with master

Raster scan with electron-beam

PMMA 495K PMMA 25K Slide

Chromium (Cr)

Nanobarriers

Wash and develop

Cr deposition

Lift off

Remove master and descum

Chromium (Cr)

Nanobarriers

Cr deposition

Liftoff

Figure 14.2 Barriers: (A–C) show simplified diagrams for making barriers by manual etching (A), ebeam lithography (B), or -nanoimprint lithography (C). For manual etching (A), the slide is scratched with a diamond-tipped drill bit. For ebeam lithography (B), the slide is coated with two layers of PMMA, and a layer of Aquasave, and an electron beam is used to etch through these layers. Chromium is deposited on the surface, and the remaining PMMA is removed, leaving behind the nanofabricated barriers. With nanoimprint lithography (C), a master in pressed into the PMMA at high temperature under vacuum. The master is removed, residual PMMA is removed, metal is deposited on the entire surface, and the remaining PMMA is removed, yielding nanofabricated metallic barriers.

et al., 2008a,b). Both ebeam and nanoimprint lithography offer sub-10-nm pattern precision. However, ebeam lithography is inherently low-throughput because the beam must raster through each pattern individually. With our current setup, it takes 30–45 min to pump down the slide in a vacuum chamber, focus the electron beam, and to raster the beam through an array of preset patterns on a single slide. Development takes approximately 5 min, bringing the total prep time before metal evaporation to 35–50 min. Furthermore, personnel must learn scanning electron microscopy, which typically takes weeks of practice before they are self-sufficient enough to focus and set the beam properly. Ebeam lithography is, however, ideal for prototyping patterns prior to settling on a specific design. Nanoimprint lithography is faster; it takes a total time of approximately 15 min to stamp an array of nanopatterns into resist on a slide and descum any residual resist before metal evaporation. Personnel can learn and practice the entire process in less than a day. Nanoimprint lithography, however, is not without its challenges.

300

Eric C. Greene et al.

A

nm

C16

B

nm 16 12 8 4

12

10

4

200 mm

5 mm 0

D

E

F

G

H

I

mm

7.5

8

5 2.5 0

0

2.5

5 mm

7.5

10

Figure 14.3 DNA curtains made with manually etched barriers. (A) and (B) show SEM images of typical etched barriers. (C) shows an AFM image of an etched barrier. A good DNA curtain made with YOYO1-stained DNA at a manually etched barrier is shown with and without buffer flow in (D) and (E), respectively. (F) and (G) show a DNA curtain at a barrier that is too wide, with and without buffer flow, respectively. (H) and (I) show curtains at etched barriers that are not straight. In (D) through (I), each field is shown with and without buffer flow; this is a standard control that used to verify that the DNA is only anchored by the biotinylated ends.

The master (designed and patterned by ebeam lithography) must be free of defects. Furthermore, the descum step must be optimized to minimize pattern distortion. Lastly, particle adhesion between the master and polymer resist is always a challenge, even in a cleanroom. Once these obstacles are overcome, nanoimprint lithography offers the potential for large-scale slide manufacture. 3.3.1. Barriers made by ebeam lithography Slides are cleaned in NanoStrip solution (CyanTek Corp, Fremont, CA) for 20 min, rinsed with acetone and isopropanol, and dried with N2. The slides are then spin-coated with a layer of 3% (w/v) polymethylmethacrylate (PMMA), molecular weight 25,000 Da, in anisole, followed by a layer of 1.5% (w/v) PMMA (495,000 Da), in anisole (MicroChem, Newton, MA),

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

301

then followed by a final layer of Aquasave conducting polymer (Mitsubishi Rayon). Each layer is spun at 4000 rpm for 45 s using a ramp rate of 300 rpm/s. Barrier patterns are written by ebeam lithography using an FEI Sirion scanning electron microscope equipped with a pattern generator and lithography control system ( J. C. Nabity, Inc., Bozeman, MT). The Aquasave is rinsed off with deionized water, and the resist is developed using a 3:1 mixture of isopropanol to methyl isobutyl ketone (MIBK) for 1 min with ultrasonic agitation at 5  C. The slide is then rinsed in isopropanol and dried with N2. A 15–20-nm layer of gold (Au) atop a 3–5-nm adhesion layer of either chromium (Cr) or titanium (Ti) is deposited using a Semicore electron-beam evaporator. The remaining PMMA layers are removed, in a process called liftoff (Fig. 14.2), by incubating the slide at 80  C in a 9:1 ratio of methylene chloride to acetone, which exposes the underlying fused silica surface that now harbors the metallic patterns. Alternatively, barriers can be made with a 15–20-nm layer of Cr, and to remove the PMMA, the coated substrate is submerged in a 65-C acetone bath for 30 min, and then gently sonicated. Following liftoff, samples are rinsed with acetone to remove stray metallic flakes and dried with N2. Barrier quality can be assessed with a scanning electron microscope, atomic force microscopy, and/or optical microscopy (Figs. 14.4–14.6). 3.3.2. Barriers made by nanoimprint lithography Nanoimprint masters (see Fig. 14.2) are fabricated using ebeam lithography, liftoff, and inductively coupled plasma etching. First, a double layer of PMMA (25,000 and 495,000 Da) is spun onto a silicon wafer with a thin coating of silicon dioxide. Patterns are written by an FEI Sirion SEM outfitted with a Nabity Nanopattern Generation System, and then developed in a mixture of isopropanol:methyl isobutyl ketone (3:1) at 5  C in a bath sonicator. Samples are then rinsed with isopropanol and dried with N2. A Semicore ebeam evaporator is used to vapor deposit 20 nm of Cr onto the masters. Liftoff is performed in acetone at 65  C. The patterned masters are then plasma-etched to a depth of 100 nm in a mixture of C4F8:O2 (9:1) for 90 s at a power of 300 W using an Oxford ICP etch tool and coated with a fluorinated self-assembled monolayer (Nanonex, Princeton, NJ) to prevent adhesion between the master and PMMA resist. To make nanoimprinted barriers, PMMA of 35,000 Da (Microresist Technologies, Germany) is spin-coated on a clean fused silica microscope slide and baked on a hotplate for 5 min at 180  C. Nanoimprinting is performed in two stages: first, a 2-min preimprint phase at a pressure of 120 psi and temperature of 120  C, followed by a 5-min imprint phase with a pressure of 480 psi and temperature of 190  C. The second step heats the PMMA above its glass transition temperature and allows it to conform to the mold. After imprinting, any residual PMMA left within the impression made by the mold must be removed by a cleaning step referred to as

302

Eric C. Greene et al.

A

Upstream

Downstream

100 mm

B

C

nm 60 40 20 10

nm

60

20

mm

7.5 40

5

2.5

0 0

2.5

5

mm

7.5

10 1.00 mm

D

E

F

G

Figure 14.4 DNA curtains made with nanofabricated linear barrier patterns. (A) shows an optical image of chromium (Cr) barriers. (B) shows an AFM image of Cr barrier. An SEM image of a typical Cr barrier is shown in (C). (D) and (E) show examples of YOYO1-stained DNA curtains assembled at linear barriers and show the presence and absence of buffer flow, respectively. These barriers were made by ebeam lithography. (F) and (G) show DNA curtains aligned at nanoimprinted barriers. Adapted with permission from Fazio et al. (2008).

303

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

A

nm

B

25 20 15 10 5

0

2.5

mm

5

7.5

100 µm

0

2.5 5

C

D

200 nm

F

mm

7.5

10

E

200 nm

200 nm

G

16 µm

Figure 14.5 Nanowell barriers: (A) shows an optical image of two barrier patterns with nanowells. (B) shows an AFM image with a representative nanowell pattern that repeats at 1-mm intervals. SEM images of nanowell barriers viewed from above are shown in (C–E). The peak-to-peak distances of these nanowells are 1900, 1000, 750 nm for (C–E), respectively. (F) and (G) show YOYO1-stained DNA curtains assembled at barriers with 1900 nm spacing, in the presence of buffer flow and immediately after pausing flow, respectively. Adapted with permission from Visnapuu et al. (2008a,b).

descum. This descum process is done in an inductively coupled plasma under CHF3:O2 (1:1) and a power of 200 W for 40 s total (two iterations of 20 s). After descum, 15–20 nm of Cr is vapor deposited onto the samples and liftoff is performed in acetone at 65  C for several hours, followed by bath sonication to remove stray metal flakes. Finally, the nanoimprinted slides are rinsed in acetone and dried with N2.

3.4. More complex barrier patterns Nanofabrication techniques offer the possibility of making more complex patterns that can be used for organizing the DNA. Below we describe barrier patterns that can control the lateral distribution of individual DNA molecules

304

Eric C. Greene et al.

nm 60 400 2

B

A

nm

1 µm

40

mm

100 nm

15

20

60

10

1.9 µm

500 nm

0

5

20

0

0

5

10

Linear barrier Pentagons

Height (nm)

mm

12 mm

15

20

25

30 20 10 0

0

5

15 10 Length (mm)

20

25

C

Channel 10 mm

D

Minus buffer flow

5 mm

E

Minus buffer flow

5 mm

Figure 14.6 Nanofabricated rack patterns for making ‘‘double-tethered’’ DNA curtains. (A) shows an optical image of a single barrier set and relevant pattern dimensions are indicated. An AFM image of a rack pattern is shown in (B) highlighting the height of the linear barriers and the pentagons as well as the distance between the barrier elements. An SEM image of a rack pattern is shown in (C). Examples of a ‘‘double-tethered’’ DNA curtain stained with YOYO1 are shown in (D) and (E). The ‘‘double-tethered’’ curtain shown in (D) was made by ebeam lithography, and the curtain shown in (E) was made by nanoimprint lithography. Adapted with permission from Gorman et al. (2010).

within the curtains (Visnapuu et al., 2008a,b), and ‘‘rack’’ patterns that can be used to make ‘‘double-tethered’’ substrates in which both ends of the DNA curtain are anchored to the slide surface (Gorman et al., 2010). 3.4.1. Geometric barrier patterns Linear barriers give no control over the lateral distribution of the DNA molecules within the curtains. The DNA molecules can overlap with one another or they can slip along the barrier edge. However, barriers

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

305

comprised of a repetitive triangular wave eliminate slippage, and also define the distribution of molecules within the curtain. We refer to these triangular features as nanowells (Fig. 14.5), and the peak-to-peak distance between the nanowells dictates the minimal lateral separation of the DNA molecules that make up the curtain. For example, nanowells that repeat at 500-nm intervals yield DNA molecules separated from one another by no less than 500 nm, provided that sufficiently low DNA concentrations are used. As with our other barrier patterns, the number of DNA molecules that make up the curtains can be varied by modulating several different parameters, the simplest of which is the amount of DNA injected into the sample chamber. At high concentrations, multiple molecules of DNA can accumulate within each nanowell. To avoid this problem, these experiments can be conducted with a relatively small amount of DNA (determined empirically, and 100 ml of a 30 pM solution of biotinylated lambda DNA is a good starting point), such that less than one DNA molecule is expected per nanowell. This ensures that some of the nanowells will remain unoccupied, many of the wells will have a single DNA molecule, and some of the wells will have multiple DNA molecules. This can be confirmed by measuring the fluorescence intensity of the DNA in each well. For example, nanowells harboring two molecules are twice as bright as those harboring just one, therefore allowing easy discrimination. 3.4.2. ‘‘Rack’’ patterns for anchoring both DNA ends The DNA curtains described above use hydrodynamic force to stretch the DNA, and if flow is turned off, the molecules quickly disappear from view as they drift outside of the detection volume defined by the penetration depth of the evanescent field. This ‘‘single-tethered’’ configuration is fine for many applications. However, in certain cases it is advantageous to be able to view the DNA in the absence of buffer flow, such as when measuring one-dimensional diffusion of proteins along the DNA or when reagents are limiting. Therefore, we designed ‘‘double-tethered’’ DNA curtains where both ends of the DNA are linked to the surface (Fig. 14.6). Double-tethered curtains utilize two pattern elements: linear barriers to lipid diffusion and pentagons that serve as solid anchor points for attachment of the second end of the DNA. One end of the DNA is first anchored via a biotin–streptavidin interaction to a supported lipid bilayer coating the surface of the sample chamber. Application of flow pushes the DNA into the linear barrier (Fig. 14.6); the linear barriers halt the movement of the lipid-tethered DNA molecules, causing them to accumulate at the leading edge of the barriers where they then extend parallel to the surface. The pentagons are positioned behind the linear barriers and separated from one another by small channels, which help prevent DNA from accumulating at the leading edge of the pentagons. The distance between the linear barriers and the pentagons is optimized for the length of the DNA

306

Eric C. Greene et al.

to be used for the experiments. The pentagons themselves are coated with antibodies directed against a small hapten, such as digoxigenin (DIG) or fluorescein isothiocyanate (FITC), which is covalently linked to the ends of the DNA opposite the ends bearing the biotin tag. When the haptencoupled DNA ends encounter the antibody-coated pentagons, they become immobilized, and the DNA molecules remain stretched parallel to the surface even when no buffer is being pushed through the sample chamber (Fig. 14.6). The DNA rack relies upon the selective, but nonspecific adsorption of antibodies to the large exposed surface of the metallic pentagons. Antibodies can also potentially adsorb to the linear barriers, but the larger surface area of the pentagons ensures that they are coated with more antibodies. As with the biotinylated end, the DIG or FITC tags are covalently attached to synthetic oligonucleotides that are complementary to the 12-nt overhang at the end of the lambda DNA. These oligonucleotides are annealed and ligated to lambda using T4 DNA ligase, and then the free oligonucleotide is removed by gel filtration (Gorman et al., 2007; Prasad et al., 2007). To assemble the ‘‘double-tethered’’ DNA curtains, the surface of the flowcell is first coated with a lipid bilayer, as described for our other curtain designs, with the exception that BSA must be omitted from all buffers used prior to deposition of the antibody. The omission of BSA is necessary to avoid blocking the exposed metallic surfaces. Once the bilayer is assembled, antibodies (0.025 mg/ml) directed against the small molecule hapten linked to the free end of the DNA are injected into the sample chamber where they are allowed to adhere nonspecifically to the exposed metal barriers. Following a brief incubation, the free antibody is rinsed from the flowcell and replaced with buffer containing 0.2-mg/ml BSA, which serves as a nonspecific-blocking agent to passivate any remaining exposed surfaces. DNA labeled at one end with biotin and at the other end with either DIG or FITC and also stained with the fluorescent intercalating dye YOYO1 is then injected into the sample chamber, incubated briefly without buffer flow, and then buffer flow is applied to push the anchored molecules into the linear barriers. Illumination of YOYO1 causes extensive DNA damage in the presence of molecular oxygen, but this can be suppressed by inclusion on an oxygen-scavenging system comprised of glucose oxidase, catalyse, and glucose (Gorman et al., 2007). Once aligned and stretched at the linear barriers, the hapten-tagged end of the DNA can bind to the antibody-coated pentagons. Flow is then terminated and the anchored DNA molecules are imaged by TIRFM.

3.5. Trouble-shooting Successful assembly of DNA curtains is reliant upon the integrity of the lipid bilayer, and failure can typically be traced to a problem with the bilayer. If the bilayer is not deposited on the surface, or if it is not mobile, then the

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

307

DNA curtains cannot be assembled. Prior to attempting to make a DNA curtain, it is absolutely essential to demonstrate the deposition of a bilayer and also to show that the bilayer is fluid. This is a relatively simple task and can be accomplished by spiking the lipid preparation (described above) with a small fraction of rhodamine-tagged lipids (e.g., [1,2-dimyristoyl-sn-glycero-3phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) (ammonium salt)]; Avanti Polar Lipids). The rhodamine provides a convenient signal that can be used to see whether a bilayer has been deposited on the surface of the flowcell, and it should appear as a uniform bright field image of fluorescence when viewed by TIRFM and it should not wash away when the flow chamber is flushed with buffer lacking lipids. In addition, a region of the surface should be photobleached. If the bilayer is fluid, the fluorescence in this photobleached region will recover over time. This FRAP (fluorescence recovery after photobleaching) control can be used to calculate the diffusion coefficient of the mobile lipids as well as the fraction of lipids that are mobile, and it can directly confirm whether or not the bilayers are of sufficient quality to support assembly of DNA curtains (Grane´li et al., 2006).

4. Visualizing Protein–DNA Interactions The primary motivation for developing DNA curtains is for use in single-molecule imaging of protein–DNA interactions. Below we present an overview of our general strategy for fluorescently labeling proteins with quantum dots, and we provide a very brief description of different examples of protein–DNA interactions that we have begun exploring using our DNA curtain approach. For more specific details regarding these experiments or data analysis, we refer the reader to the original publications (Gorman et al., 2007, 2010; Prasad et al., 2007; Visnapuu and Greene, 2009).

4.1. Quantum dots as a general fluorescent labeling strategy We use fluorescent semiconducting nanocrystals, also called quantum dots (QDs) or Qdots, as the fluorophore of choice rather than other types of fluorescent dyes. QDs are commercially available, they are relatively small (typically 10–20 nm in diameter), and they are extremely bright and photostable. A single laser source can be used to excite different colored QDs, eliminating the need for multiple illumination sources during applications requiring multicolor imaging. The primary disadvantage of QDs is that they are not as small as organic fluorophores, and it is critical to assess the effect of the QD for any protein under investigation through the use of standard in vitro ensemble assays. When making measurements of 1D diffusion (see below), the large QDs are expected to slow down the overall

308

Eric C. Greene et al.

motion (i.e., reduce the value of the diffusion coefficient by a few fold) of the proteins under investigation in proportion with the increased radius of the QD–protein complex compared to the unlabeled protein alone, but should have no impact on the actual mechanism of movement (Gorman et al., 2007; Kochaniak et al., 2009). We utilize a labeling strategy in which recombinant proteins are expressed as fusions with an epitope tag (e.g., HA, FLAG, thioredoxin, etc.) and these tags are used as handles for conjugating the protein of interest to a QD that is covalently coupled to the corresponding antibody. We use amine reactive QDs provided with an antibody conjugation kit from Invitrogen. The amines can be coupled to any antibody using the heterobifunctional crosslinking reagent SMCC (succinimidyl 4-[N-maleimidomethyl]cyclohexane-1-carboxylate). According to the manufacturer, this procedure yields on the order of 1–3 antibodies per QD, although reports in the literature suggest values closer to 0.01–0.1 antibodies per QD (Pathak et al., 2007). The resulting QD antibody conjugates can be purified by gel filtration to remove unreacted antibodies and stored in phosphate-buffered saline (PBS, pH 7.4) at 4  C for at least a few weeks without a noticeable decline in quality. The antibody-labeled QDs are then mixed with the epitope-tagged recombinant protein of interest. This labeling strategy can be applied to virtually any protein that has an epitope tag and is unaffected by the attachment of the Qdot.

4.2. Visualizing ATP-dependent DNA translocation Using DNA curtains aligned along manually etched barriers and QD tagged proteins, we have demonstrated that Rdh54 translocates along doublestranded DNA (Fig. 14.7) (Prasad et al., 2007). Rdh54 is an Snf2-family member and is involved in meiotic and mitotic homologous DNA recombination in Saccharomyces cerevisiae. Rdh54 is also capable of remodeling nucleosomes, and this activity may be related to its involvement in DNA recombination. Our studies have demonstrated that the translocation activity of Rdh54 is ATP-dependent, and translocation appears to coincide with the generation of looped DNA structures, which is consistent with the known biochemical properties of the related protein Rad54. The protein traveled at a mean velocity of 80 bp/s, but individual proteins could change direction and/or velocity, or even pause as they moved along the DNA. Similar behaviors have been reported in separate single-molecule studies from Kowalczykowski and colleagues (Amitani et al., 2006; Nimonkar et al., 2007). These findings suggest that this family of proteins may have multiple motor domains within a single multimeric protein complex, and the pauses, velocity, and direction changes may reflect the protein complex engaging different motor domains with the DNA.

309

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

A

0 4

Position (µm)

8 12 0 4 8 12

B

0 4 8 12 0

Position (µm)

4 8 12 6

106 bp/s

7 8

98 bp/s

9 42 bp/s

10 11 12 0

50

100

150

200

250

Time (s)

C

20 80 bp/s

30

n = 273

13 Kbp

15 Counts

Counts

40

20

n = 64

10 5

10

0

0 0

100

200 300 400 500 Translocation rate (bp/s)

600

0

10

20 30 Distance (Kbp)

40

Figure 14.7 Visualizing ATP-dependent translocation of Rdh54 using DNA curtains aligned at manually etched barriers. The upper panel in (A) shows Rdh54 bound to DNA. The lower panel in (A) shows the same field after transiently pausing buffer flow; (B) shows the kymogram of a single translocating complex of Rdh54 (upper panel), along with superimposed particle-tracking data (middle panel), or shown independently as a graph of the movement (lower panel). This data was collected using an algorithm that located and tracked the centroid position of each fluorescent particle within the DNA curtain. Linear fits to the data are also indicated along with the corresponding translocation rates. Histograms generated from the analysis of translocating Rdh54 complexes showing the distribution of translocation rates and total distance traveled during the 250-s observation windows are shown in (C). Adapted with permission from Prasad et al. (2007).

310

Eric C. Greene et al.

4.3. Using DNA curtains to image nucleosomes Eukaryotic DNA is compacted into chromatin, the fundamental unit of which is the nucleosome. Nucleosomes are comprised of a histone octamer wrapped by 147 bp of DNA. We have established a system for looking at QD tagged nucleosomes using DNA curtains (Fig. 14.8) (Visnapuu and Greene, 2009). In our initial study, we used DNA curtains assembled at nanofabricated linear barriers to demonstrate that the recent theoretical models of Widom, Segal, and colleagues (Field et al., 2008; Kaplan et al., 2009) can predict landscapes for nucleosome deposition on both l-DNA and on a 23-kb PCR fragment derived from the human b-globin locus. We have also confirmed that poly(dA-dT) tracts exclude nucleosomes, and the effects of these exclusionary sequences dominate the intrinsic binding landscape. We have shown that the deposition pattern for the human b-globin locus suggests an organizational mechanism consistent with a small number of strongly positioned nucleosomes near promotor and regulatory regions, and statistical packing of most other nucleosomes throughout the locus. We have also shown that octameric nucleosomes harboring a centromeric specific variant of histone H3 (CenH3) display intrinsic deposition patterns nearly identical to canonical nucleosomes. In contrast, hexameric nucleosomes harboring both CenH3 and Scm3, a centromer-specific nonhistone protein, overcome the exclusionary affects of poly(dA-dT), allowing them to be deposited in regions that disfavor normal nucleosome octamers.

4.4. Diffusion of MMR along DNA Post-replicative mismatch repair (MMR) is necessary to correct errors made during DNA synthesis. In eukaryotes, Msh2-Msh6 is responsible for locating and initiating repair of mispaired bases, and works in concert with Mlh1-Pms1, which coordinates downstream steps in the repair pathway. We have begun dissecting MMR by visualizing proteins as they interact with DNA curtains assembled at etched barriers, as well as isolated molecules of DNA that were not assembled into curtains (Gorman et al., 2007). Using anti-HA QDs, and HA-tagged Msh2-Msh6, we have demonstrated that Msh2-Msh6 can travel along DNA via 1D diffusion, and it is likely that the protein tracks the phosphate backbone as it slides along the DNA. Msh2-Msh6 also reversibly enters a nondiffusive state and we believe that this occurs when the protein stops to interrogate a site. The addition of ATP caused the nondiffusive Msh2-Msh6 to reenter a diffusive state and continue sliding along the DNA (Fig. 14.9). We hypothesize that reentry into a diffusive state mimics what occurs after lesion recognition, and represents the functional consequence of a conformational change that is triggered by ATP binding. We have also initiated new studies of the protein complex Mlh1–Pms1 and Mlh1 alone (Fig. 14.10; Gorman et al., 2010; Gorman

311

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

A

DNA

Lipid bilayer

Nucleosomes

Flow direction Nanobarrier

B

T1

Flow on

Flow off

Flow on

T2 T3 T4

Length (mm)

C

0 Flow

6 12 0

5

Count

Probability

D

Time (s)

15

10

1.0 0.8 0.6 0.4 0.2 0.0

Field et al. prediction

60

Nucleosome distribution (n = 2458)

40 20 0

10

30 20 Position (Kbp)

40

Figure 14.8 Visualizing fluorescently tagged nucleosomes. (A) depicts the experimental design. YOYO1-stained DNA curtains (green) bound by nucleosomes (magenta) that were tagged with anti-FLAG QDs are shown in (B). The tethered end of each curtain is indicated as T1-T4, and arrows indicate the direction of flow. A kymogram illustrating five nucleosomes on one DNA molecule is shown in (C). The nucleosomes disappear when flow is temporarily interrupted (blue arrowheads) and reappear when flow is resumed (green arrowheads), verifying that they are bound to the DNA and do not interact with the lipid bilayer. The top panel in (D) shows the theoretical distribution of nucleosomes on l-DNA as predicted by Field et al. (2008). The lower panel shows the observed distribution of nucleosomes obtained from DNA curtain experiments. Adapted with permission from (Visnapuu and Greene, 2009).

312

Eric C. Greene et al.

A

T

No buffer flow immobile Diffusing

Evanescent field

T

Buffer plus ADP T

F

F T

ATP chase T

F

F

B

Flow ON/OFF control

Inject ATP T

F 0

0.5

1.0 Time (min)

1.5

2.0

C

Number of Msh2Msh6 complexes

210

N = 510 total Msh2-Msh6 complexes

50

Average sliding distance = 2.7 mm (~9,40010,000 bp)

40 30

181 Dissociate without sliding (35.5%) 237 Dissociate after sliding (46.5%) (88 dissociate from DNA ends) (149 dissociate from internal location) 42 Slide without dissociating (8%) 50 Remain stationary (10%)

20 10 0

0

2

4 10 8 6 Distance traveled before dissociation (mm)

12

Figure 14.9 ATP-triggered release of Msh2-Msh6 from DNA aligned at a manually etched barrier. (A) Assay used to probe the effects of ATP on the immobile population of Msh2-Msh6. The proteins are red and the DNA is green. The DNA is tethered to the surface by only one end (‘‘T’’) and the free end (‘‘F’’) is only observed when flow is applied. Flow is from top to bottom in each panel and the distance between T and F is  13 mm. The upper panel shows the field after transiently pausing buffer flow (Flow ON/OFF control) and the cartoon at the left depicts the behavior of the molecules the absence of flow. The middle panel shows the same field after resuming flow. Each red spot in the image corresponds to at least one Msh2-Msh6 complex, and there are 274 identifiable red spots in this experiment and 124 DNA molecules. The lower panel shows the same DNA molecules after the injection of 2 mM ATP. (B) Kymogram showing one DNA molecule and the bound Msh2-Msh6. Arrowheads highlight the

313

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

A

1

2 3

a

b

c

B a

b

c 0

10

20

30

40

50

30

40

50

Time (s) C 1

2

3 0

10

20 Time (s)

Figure 14.10 One-dimensional diffusion of Mlh1 on ‘‘double-tethered’’ curtains of DNA. (A) shows an example of a double-tethered DNA curtain bound by QD labeled Mlh1. The DNA is shown in green and the proteins are magenta. This image represents a single 100-ms image taken from a 1-min video (not shown). (B) shows three representative kymograms (designated a, b, and c) made from individual DNA molecules from within (A). (C) shows examples of DNA molecules taken from (A) that broke during the course of DNA collection (numbered 1, 2, and 3), demonstrating that both the DNA and the bound proteins diffuse rapidly away from the surface and out of the evanescent field, this confirming that the QD tagged proteins are not adsorbed to the bilayer. Reproduced with permission from Gorman et al. (2010).

dissociation of Msh2-Msh6. (C) Summary of the behavior of 510 total Msh2-Msh6 complexes after the injection of ATP into the sample chamber. ‘‘Distance traveled before dissociation’’ corresponds to the distance that Msh2-Msh6 moved prior to falling off the DNA. Dissociation events that occurred after sliding from internal positions on the DNA are colored red (N ¼ 149) and those that occurred at the end of the DNA molecules are colored blue (N ¼ 88). Reproduced with permission from Gorman et al. (2007).

314

Eric C. Greene et al.

et al., in preparation). These experiments are the first to take advantage of these rack patterns and double-tethered DNA curtains, and highlight the unique advantages of these designs.

5. Conclusions and Future Directions DNA curtains are amenable to many different experimental systems. Manually etched barriers can be easily implemented in any laboratory with experience in single-molecule detection. Nanofabricated barriers require access to clean room facilities, but offer greater precision and design flexibility. As illustrated above, we have established experimental systems for studying DNA recombination and molecular motor proteins, nucleosomes, and chromatin, and postreplication MMR. This initial work involved single protein participants that are part of more complex, multicomponent biochemical systems. Moving forward, a great challenge will be to study these proteins in combination with different factors and engineered DNA molecules containing specific binding sites or lesions, with a goal of recapitulating complete biochemical pathways that can be viewed at the singlemolecule level.

ACKNOWLEDGMENTS This research was funded by the Initiatives in Science and Engineering grant (ISE: awarded to E. C. G. and S. W.) program through Columbia University, and by an NIH grant GM074739 and an NSF PECASE Award to E. C. G. T. A. F. was supported in part by an NSF Graduate Research Fellowship. J. G. was supported by an NIH training grant for Cellular and Molecular Foundations of Biomedical Sciences (T32GM00879807). This work was partially supported by the Nanoscale Science and Engineering Initiative of the NSF under Award Number CHE-0641523 and by the New York State Office of Science, Technology, and Academic Research (NYSTAR).

REFERENCES Amitani, I., Baskin, R., and Kowalczykowski, S. (2006). Visualization of Rad54, a chromatin remodeling protein, translocating on single DNA molecules. Mol. Cell 23, 143–148. Axelrod, D. (1989). Total internal reflection fluorescence microscopy. Methods Cell Biol. 30, 245–270. Cremer, P. S., and Boxer, S. G. (1999). Formation and spreading of lipid bilayers on planar glass supports. J. Phys. Chem. B 103, 2554–2559. Fazio, T., Visnapuu, M. L., Wind, S., and Greene, E. C. (2008). DNA curtains and nanoscale curtain rods: High-throughput tools for single molecule imaging. Langmuir 24, 10524–10531.

DNA Curtains for High-Throughput Single-Molecule Optical Imaging

315

Field, Y., Kaplan, N., Fondufe-Mittendorf, Y., Moore, I., Sharon, E., Lubling, Y., Widom, J., and Segal, E. (2008). Distinct modes of regulation by chromatin encoded through nucleosome positioning signals. PLoS Comput. Biol. 4, e1000216. Forkey, J. N., Quinlan, M., and Goldman, Y. (2000). Protein structural dynamics by singlemolecule fluorescence polarization. Prog. Biophys. Mol. Biol. 74, 1–35. Gorman, J., Chowdhury, A., Surtees, J. A., Shimada, J., Reichman, D. R., Alani, E., and Greene, E. C. (2007). Dynamic basis for one-dimensional DNA scanning by the mismatch repair complex Msh2-Msh6. Mol. Cell 28, 359–370. Gorman, J., Fazio, T., Wang, F., Wind, S., and Greene, E. C. (2010). Nanofabricated racks of aligned and anchored DNA substrates for single molecule imaging. Langmuir 26, 1372–1379. Gorman, J., Plys, A., Visnapuu, M. L., Alani, E., and Greene, E. C., manuscript in preparation. Grane´li, A., Yeykal, C., Prasad, T. K., and Greene, E. C. (2006). Organized arrays of individual DNA molecules tethered to supported lipid bilayers. Langmuir 22, 292–299. Groves, J. T., and Boxer, S. G. (2002). Micropattern formation in supported lipid membranes. Acc. Chem. Res. 35, 149–157. Groves, J. T., Ulman, N., and Boxer, S. G. (1997). Micropatterning fluid lipid bilayers on solid supports. Science 275, 651–653. Ha, T. (2001). Single-molecule fluorescence resonance energy transfer. Methods 25, 78–86. Kaplan, N., Moore, I., Fondufe-Mittendorf, Y., Gossett, A., Tillo, D., Field, Y., LeProust, E., Hughes, T., Lieb, J., Widom, J., and Segal, E. (2009). The DNA-encoded nucleosome organization of a eukaryotic genome. Nature 458, 362–366. Kochaniak, A., Habuchi, S., Loparo, J., Chang, D., Cimprich, K., Walter, J., and van Oijen, A. (2009). Proliferating cell nuclear antigen uses two distinct modes to move along DNA. J. Biol. Chem. 284, 17700–17710. Nimonkar, A. V., Amitani, I., Baskin, R., and Kowalczykowski, S. (2007). Single molecule imaging of Tid1/Rdh54, a Rad54 homolog that translocates on duplex DNA and can disrupt joint molecules. J. Biol. Chem. 282, 30776–30784. Pathak, S., Davidson, M. C., and Silva, G. A. (2007). Characterization of the functional binding properties of antibody conjugated quantum dots. Nano Lett. 7, 1839–1845. Prasad, T. K., Robertson, R. B., Visnapuu, M. L., Chi, P., Sung, P., and Greene, E. C. (2007). A DNA-translocating Snf2 molecular motor: Saccharomyces cerevisiae Rdh54 displays processive translocation and extrudes DNA loops. J. Mol. Biol. 369, 940–953. Visnapuu, M. L., and Greene, E. C. (2009). Single-molecule imaging of DNA curtains reveals intrinsic nucleosome landscapes. Nat. Struct. Mol. Biol. 16, 1056–1062. Visnapuu, M. L., Duzdevich, D., and Greene, E. C. (2008a). The importance of surfaces in single-molecule bioscience. Mol. Biosyst. 4, 394–403. Visnapuu, M. L., Fazio, T., Wind, S., and Greene, E. C. (2008b). Parallel arrays of geometric nanowells for assembling curtains of DNA with controlled lateral dispersion. Langmuir 24, 11293–11299.

C H A P T E R

F I F T E E N

Scanning FCS for the Characterization of Protein Dynamics in Live Cells Zdeneˇk Petra´sˇek,*,1 Jonas Ries,*,1 and Petra Schwille* Contents 318 320 322 326 327 327 330 332 334 334 335

1. Introduction 2. Implementation 2.1. Scan paths 2.2. Calibration of the scan path 3. Data Analysis 3.1. Calculation of correlation curves 3.2. Corrections 3.3. Data fitting 4. Applications 4.1. sFCS in Caenorhabditis elegans embryo 4.2. Small-circle sFCS 4.3. sFCS with a perpendicular scan path to measure in unstable membranes 4.4. Dual-focus sFCS 4.5. Dual-color sFCS 5. Conclusion References

337 339 340 341 342

Abstract Scanning fluorescence correlation spectroscopy (sFCS) is the generic term for a group of fluorescence correlation techniques where the measurement volume is moved across the sample in a defined way. The introduction of scanning is motivated by its ability to alleviate or remove several distinct problems often encountered in standard FCS, and thus, to extend the range of applicability of fluorescence correlation methods in biological systems. These problems include poor statistical accuracy in measurements with slowly moving molecules, photobleaching, optical distortions affecting the calibration of the measurement * Biotec, TU Dresden, Dresden, Germany Both authors contributed equally

1

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72005-X

#

2010 Elsevier Inc. All rights reserved.

317

318

Zdeneˇk Petra´sˇek et al.

volume, membrane instabilities, etc. Here, we present an overview of sFCS methods, explaining their benefits, implementation details, requirements, and limitations, as well as relations to each other. Further, we give examples of different sFCS implementations as applied to cellular systems, namely largecircle sFCS to measure protein dynamics in embryo cortex and line sFCS to measure protein diffusion and interactions in unstable membranes.

1. Introduction Fluorescence correlation spectroscopy (FCS) is a powerful tool to measure local concentrations, molecular weights, translational and rotational diffusion coefficients, chemical rate constants, association and dissociation constants, and photodynamics in vitro as well as in vivo (Bacia and Schwille, 2003; Bacia et al., 2006; Kim et al., 2007; Rigler and Elson, 2001). It is based on the statistical analysis of intensity fluctuations caused by fluorophores diffusing through a small (
fL) detection volume (for details see Fig. 15.1) and requires only a low concentration (
0.1–100 nM) of labeled molecules, minimizing the interference of the labeling with the system. The introduction of commercial confocal FCS systems has promoted the use of FCS so that it can now be considered a well-established technique. However, standard FCS suffers from several limitations of applicability, especially in complex biological systems. Optical artifacts (Enderlein et al., 2005) such as A

B G(t) =

C 〈dF(t) dF(t + t)〉 2

〈F 〉

G(t) 1 N

Laser

Pinhole APD

t

t+t

0

tD

t

Figure 15.1 Principle of FCS. (A) The sample is illuminated by focusing the laser beam through an objective. The emitted photons are then spectrally filtered and detected with an avalanche photodiode. A pinhole provides axial confinement to result in a tiny (sub-femtoliter) detection volume. (B) Fluorophores diffusing through the detection volume give rise to a fluctuating intensity trace F(t) from which the autocorrelation curve G(t), which measures the self-similarity of the signal, can be calculated. (C) Parameters of interest, for example, the diffusion time tD or number of particles in the detection volume N, are obtained by fitting a mathematical model to the experimental correlation curve.

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

319

varying cover slide thickness, refractive index mismatch, optical saturation, or aberrations change the size of the detection volume, which precludes the precise calibration necessary for quantitative concentration or diffusion measurements. Slow diffusion in biological samples demands long measurement times—at least 105 times the slowest timescale found in the system (Tcherniak et al., 2009)—and the long residence times in the detection volume promote photobleaching. Finally, measurements on biological membranes are especially challenging, since even tiny membrane movements or instabilities lead to severe artifacts. To overcome these limitations, modifications of standard FCS have been developed (Dertinger et al., 2007; Digman et al., 2005; Ruan et al., 2004), one of the most successful being sFCS (Petra´ˇsek and Schwille, 2008c). sFCS employs a moving detection volume instead of a static detection volume, which has several advantages (Box 1): sampling of a larger volume increases Box 1 When to use sFCS

Decide whether to use standard FCS or sFCS. In the following situations sFCS may be superior to standard FCS: 

Slow motion. The molecules diffuse (or move) very slowly, with diffusion times in the range of tens (hundreds) of milliseconds or longer. There are not enough fluctuation events during the permissible measurement time and the calculated autocorrelation is too noisy, especially at long lag times. By scanning the sample, the measurement is performed at more locations, thus improving the statistics.  Optical distortions. The measurement is performed within an optically inhomogeneous medium, such as cells or tissues, where the focused laser spot can be deformed, changing its size and shape. This invalidates the calibration of the volume size by an independent measurement, necessary in standard FCS to determine diffusion coefficients and particle numbers. Choose one of the types of sFCS where the volume calibration is replaced by an exact knowledge of the scan path. Smallcircle sFCS is useful when the measurement has to be limited to a small part of the sample, while large-circle or line sFCS simultaneously improves the signal-to-noise ratio in case of slowly diffusing molecules by sampling a larger area.  Photobleaching. Due to slow motion, the molecules stay too long in the laser focus and are more likely to become photobleached, resulting in distortions of the correlation curve. Choose a type of sFCS with a larger scan path, whereby more molecules are probed for a shorter time, thus reducing the risk of photobleaching and at the same time improving the statistics. (continued)

320

Zdeneˇk Petra´sˇek et al.

Box 1 (continued ) 

Membrane motion. When measuring motion on surfaces, for example, in biomembranes or on cell cortex, the surface itself may be moving in an uncontrolled way. Scanning perpendicularly to the surface makes it possible to eliminate these motions in the data analysis.  Complex motion. The molecular motion is more complex, possibly a combination of diffusion and flow with static features, or multicomponent or anomalous diffusion. Choose a scan path in combination with spatiotemporal correlation analysis that can reveal complex motion patterns.  Parallel measurement. Need to measure simultaneously at more locations? It may be possible to choose a scan path that passes through all these locations, and the subsequent data analysis produces separate results for each position.

the statistical accuracy for slowly moving molecules and leads to shorter measurement times; short residence times in the detection volume reduce the effect of photobleaching; and the scan speed can be determined with high accuracy, eliminating the need for calibrating the detection volume. The application of sFCS is especially beneficial in measurements on biological membranes where diffusion is usually very slow. Here, an alternative implementation of a moving detection volume has proven very useful: choosing a scan path that is perpendicular to the membrane plane eliminates the effect of membrane movements and instabilities (Ries and Schwille, 2006). In the following, we describe experimental steps to perform sFCS measurements (see also Box 2) and discuss applications and limitations. We start with a general description of the common features of sFCS. Then we will discuss specific implementations (circular sFCS and line sFCS with a perpendicular scan path) in detail and present their applications.

2. Implementation sFCS is typically performed on a confocal laser scanning microscope (CLSM) where the sample can be imaged before the scan path is chosen, and which makes it possible to scan the measurement volume during the actual scanning fluorescence correlation spectroscopy (sFCS) measurement across the sample in a user-defined way. The CLSM determines what scan paths can be used. While linear paths of any length and angle are typically possible

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

321

Box 2 How to proceed

1. Choose the scan path based on the motivation for using sFCS, and the scan parameters depending on the speed of molecular motion, sample size, etc. 2. Perform the measurement. Ideally, the full raw data (photon arrival times) are saved, and the correlation is calculated by the analysis software. 3. Calculate the correlation. The algorithm in general depends on the scan path, and some preprocessing may be necessary, for example, due to a moving membrane when measuring diffusion within unsupported membranes. 4. Apply corrections for photobleaching and other irregularities, if necessary. 5. Fit the data with a model dependent on the scan path, and obtain the desired parameters.

with available commercial instruments, circular paths can usually be realized only with homebuilt setups. Apart from the possibility to scan the measurement volume, the technical requirements for sFCS are identical with those of standard FCS. The main points making a CLSM capable of FCS measurements are briefly summarized in the following paragraphs, while details are extensively described in the literature (Petrov and Schwille, 2008). The critical parameter for a successful FCS measurement is the size and shape of the measurement volume. This is determined by the focusing of the excitation laser beam, and by the detection geometry of the emission signal. The water immersion objective used should have high numerical aperture to produce a small focus and collect a maximum signal, and a correction collar that allows adjustment for different coverslip thicknesses. Exact positioning of the correction collar yields an undistorted focus, a crucial requirement for any FCS measurement. On the detection side, the confocal pinhole should be adjustable, in both its size and its lateral (ideally also axial) position. The pinhole is not necessary if two-photon excitation is employed. Another important element is the detector. The standard is an avalanche photodiode, a photon counting detector with high quantum efficiency. Other important parameters of the detector are low dark count (few hundreds of counts per second or less) and weak afterpulsing (ideally of low amplitude and short decay, microseconds or less). The same detector can be used for photon counting imaging, which is becoming increasingly common nowadays in lowlight, single molecule, and quantitative imaging applications (Becker et al., 2004).

322

Zdeneˇk Petra´sˇek et al.

The signal from the detector consisting of pulses for each detected photon is processed by the computer. It is desirable to store the whole raw data, that is, the arrival times of all photons, with sufficiently high temporal resolution. This gives the user maximal freedom in the subsequent data analysis. Hardware correlators are practical for alignment and monitoring purposes because they provide the autocorrelation in real time, but should be used in sFCS only together with other means for obtaining the raw fluorescence signal. Commercial laser scanning microscope systems that provide FCS capability include Confocor 3 with LSM-510 or LSM-710 microscopes (Zeiss, Jena, Germany), TCS SMD FCS (Leica, Wetzlar, Germany), DCS-120 Confocal FLIM system (Becker & Hickl, Berlin, Germany), and Micro Time 200 (Picoquant, Berlin, Germany). The latter two companies also provide all hardware necessary for an implementation of FCS and sFCS in an existing, either commercial or homebuilt, confocal scanning microscope. LSM-710 (Zeiss) is available with an implemented raster image correlation spectroscopy (RICS, see below) option, a type of sFCS.

2.1. Scan paths The choice of the scan path depends on several factors, mainly on the type of molecular motion motivating the use of sFCS, the sample size and shape, and the options provided by the instrument. All the scan paths considered here lie within the focal plane (do not move to different z positions) and are realized by scanning the laser beam, not the sample stage, which would not be possible to realize at sufficient speed. The scan paths can be divided into two groups: linear and circular. This division is to some extent due to practical reasons. Line scans can be easily performed with commercial CLSMs. Although the commercial scanning microscopes equipped with galvanometer scanners are also capable of circular scans, this option is usually not software-implemented. The user is limited to homebuilt instruments where the scan path can be controlled with full flexibility. The circular path with large radius is more or less equivalent to a linear path, and what is said about one is mostly valid for the other. The only difference is in the calculation of the correlation. With the linear path, the termination of the line and possible changes of direction in case of bidirectional scanning have to be taken into account. With the circular path, all points along the circle are equivalent; therefore, no presorting of the data stream is necessary, simplifying the calculation of the autocorrelation. Additionally, circular scans can be faster than linear scans and feature more constant scanning velocities since they require no rapid acceleration. The most relevant differences between the line and circle scans can be found in the double-line scan, which would be difficult to realize in a circular configuration, and in a small-circle scan, which does not have a line equivalent.

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

323

In the following, we briefly describe each scan path and its main area of application (see also Table 15.1). 2.1.1. Single-line and large-circle scans Large-range scans are typically applied in the following situations: when the molecules move slowly and averaging over a larger population is desired, when spatiotemporal correlation of the molecular motion is of interest, or when we wish to perform correlation analysis at many locations along a line simultaneously. By large range we mean scan path lengths many times larger than the linear dimension of the measurement volume. When the molecules move slowly, the FCS measurement at one position suffers from poor statistics, because insufficient numbers of molecules pass through the measurement volume during the measurement time (Tcherniak et al., 2009). This results in low accuracy of the fluorescence correlation especially at long correlation times, affecting the values of recovered parameters (diffusion time and particle number). By scanning the focused excitation beam, the measurement is effectively performed at many locations, their number approximately given by the ratio between the length of the scan path and the size of the volume, thus improving the signal-to-noise ratio. The fluorescence autocorrelation in standard FCS describes the temporal dynamics of molecules leaving the measurement volume. Spatiotemporal correlation curves additionally contain information about the spatial spreading of the molecules from their original location. This additional information is needed to characterize more complex motion than simple diffusion, such as Table 15.1 Scan paths and their properties Scan type

Main benefit/application

No scanning

Rigler and Elson (2001) Fast diffusion, simple implementation, inhomogeneous sample Slow motion, photobleaching, Petra´ˇsek et al. (2008b) and Ries et al. (2009a) robustness, spatiotemporal correlation Robustness, precision, Skinner et al. (2005) and small area Petra´ˇsek and Schwille (2008b) Robustness, precision, Ries and Schwille (2006) membrane motion Slow motion, Digman et al. (2005) spatiotemporal correlation Membrane motion Ries and Schwille (2006)

Single-line, large-circle Small-circle

Double-line Raster Perpendicular to membrane

References

324

Zdeneˇk Petra´sˇek et al.

multicomponent diffusion, combination with flow, reaction kinetics, or anomalous diffusion. This has been exploited in spatiotemporal image correlation spectroscopy (STICS), a type of correlation analysis applied to a sequence of images (Hebert et al., 2005). sFCS also gives us access to spatiotemporal correlation, as described below. Although sFCS features only one spatial dimension (along the scan path), it has higher temporal resolution than STICS, enabling studies of faster processes. Additionally, the knowledge of spatiotemporal correlation makes it possible to determine from one experiment the diffusion coefficient D and the size of the measurement volume w0. In standard FCS, these two parameters are combined in the diffusion time tD ¼ w02 =4D, and to calculate D, the volume size has to be determined in an independent measurement with a molecule of a known diffusion coefficient. The effect of possible optical distortions, which can invalidate this calibration in standard FCS, is avoided in sFCS, because the volume size is determined independently from the fit. The spatial calibration of the scan path is determined by the instrument only, can be performed relatively easily, and is generally not affected by the optical properties of the sample. Another useful feature of long scan paths is the reduction of the probability that any observed molecule will be photobleached. Due to the motion of the focused laser beam, the total light dose is distributed over a larger area, and the residence time of individual molecules within the focus is reduced. This lowers the chance of photobleaching compared to situations where the slow-moving molecule diffuses through a stationary focus (Petra´ˇsek and Schwille, 2008a; Ries et al., 2009a; Satsoura et al., 2007). When we are interested in correlation analysis at more positions but cannot perform the measurements sequentially, for example, because the sample changes in time, sFCS can be used to perform the measurements in a semi-parallel way. The scan path is chosen so that the beam passes through all the points of interest, the raw data (fluorescence signal) are then sorted depending on the position, and the data from each position are correlated, producing autocorrelation curves at each location. The sorting of the raw data requires that the relation between the fluorescence signal and the position at which it was recorded is known. The size and shape of the scan path is largely governed by the sample. With the exception of the application described in the last paragraph, the sample should be homogeneous along the scan path, since the obtained results represent an average over the measured area. Generally, larger scan paths are preferable due to better averaging. 2.1.2. Small-circle scan Circular paths with a radius comparable to the size of the measurement volume are suitable for robust and accurate measurements of diffusion coefficients in situations when the size of the probed volume is not

325

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

known or can be affected by the sample (Petra´ˇsek and Schwille, 2008b). The principle is the same as with other correlation methods where spatiotemporal correlation is measured: by introducing another spatial measure (scan radius (Petra´ˇsek and Schwille, 2008b), distance covered per unit time (Ries et al., 2009a), distance between two foci (Dertinger et al., 2007)), the diffusion coefficient and the volume size in the model function become decoupled and can be determined independently from the fit, resulting in robustness to optical distortions. Small-circle scan is optimal for this purpose: since the radius is small, the correlation values at all lag times are nonzero, and therefore carry useful information, unlike the large-circle scan, where the correlation values outside the narrow peaks are zero (Fig. 15.2). Another important feature is that the scan with a small radius covers a minimal area, meaning that the measurements can be performed also on samples without large homogeneous areas, as is typical in cells.

A

p

a

j /rad.

C

−0.4 −0.3 −0.2 −0.1 0

0.1 0.2 0.3 0.4

0.07

2

0.06

1.5

0.05

1

0.04

0.5

0.04

0.06

0.03

logt / s

g(t )

B

0.02 0.01 0

0.05

0

0.03

−0.5

0.02

−1

0.01

−1.5

−0.01

0

−2

−0.02 10−6 10−5 10−4 10−3 10−2 10−1

t /s

−0.01 100

101

102

−1

−0.5

0

0.5

1

x/mm

Figure 15.2 Large-circle sFCS in a C. elegans embryo. (A) The laser beam is scanned in a circle on the flat part of the posterior (p) cortex of the embryo where GFP::PAR-2 localizes; the anterior half (a) is without PAR-2. (B) The autocorrelation of the fluorescence signal consists of peaks spaced by the scan period 1/f with their amplitude decreasing as in standard FCS. (C) The spatiotemporal correlation of the same signal shows additionally the lateral spreading of the GFP-labeled molecules.

326

Zdeneˇk Petra´sˇek et al.

2.1.3. Double-line scan An alternative way to reach the same goal as described in the previous paragraph is to use double-line scan, whereby two parallel lines separated by a known distance comparable to the measurement volume size are scanned sequentially. Spatial cross-correlation of the signal between the two lines is then used to determine the diffusion coefficient without the need to know the volume size (Chiantia et al., 2006; Ries and Schwille, 2006). 2.1.4. Raster scan Combination of line scans offset in a perpendicular direction results in a raster scan, commonly used for imaging in laser scanning microscopes. Correlation of raster scan data is the basis of raster image correlation microscopy (RICS) (Digman et al., 2005). In RICS, the correlation of the signal is performed along the scan line and in the direction perpendicular to the scan line, giving access to two different temporal scales. Raster correlation is closely related to image correlation spectroscopy, which is described elsewhere (Wiseman et al., 2000). 2.1.5. sFCS with membrane motion When molecules diffuse within a membrane, or another two-dimensional structure, and the membrane moves, due to undulations, cell growth or dynamics, or other reasons, the standard FCS faces the problem that the autocorrelation function contains the contributions of both, molecular and membrane motion, which cannot be easily separated. The solution by means of sFCS is to scan the laser focus in a line perpendicular to the fluctuating membrane surface and to correlate the intensities from subsequent passages of the beam through the membrane. Only the signal from around the maximum in each scan line is correlated, regardless of where this maximum occurred, thus eliminating any effects of membrane motion (Fig. 15.4). The temporal resolution in this case is limited to the scan period; this is, however, often sufficient due to relatively slow motion of molecules in membranes. This approach can be combined with a double-line scan described above, thus gaining the advantage of robustness against optical distortions (Fig. 15.5).

2.2. Calibration of the scan path Many of the implementations of sFCS rely on the exact calibration of the scan path. In circular scan methods, the scan radius must be known; in double-line and double-focus approaches the knowledge of the distance between the two scanned lines is required; and with a single-line scan, the scan speed is important. These dimensions can be determined in the slow

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

327

scan limit by imaging a calibration standard, for example, a Ronchi ruling, a stripe pattern with well-defined spacing. This calibration need not be valid at high scan speeds, since the actual response of the scanning device to the driving commands may vary. Then a ‘‘dynamic calibration’’ is required, whereby the same scan parameters as during the measurement are used. In the small-circle scan method and the line-scan method, the dynamic calibration can be performed by analyzing the signal reflected from the Ronchi ruling at high scan frequencies (Petra´ˇsek and Schwille, 2008b; Ries et al., 2009a). Another possibility is to use the sensor signal from the scanners (if available) to determine the exact scanner position. The distance between two lines can be determined by scanning over a film of immobilized fluorophores. The bleached traces are then analyzed in a high-resolution LSM image by fitting the profiles with a double-Gaussian (Ries and Schwille, 2006).

3. Data Analysis Data analysis in sFCS can be roughly divided into three steps: calculation of the correlation from the raw data, application of corrections, and fitting of the data to extract the parameters of interest. These steps are described in more detail in the following sections.

3.1. Calculation of correlation curves The correlation g12(t) of the fluorescence signals F1(t) and F2(t) is defined in the following way: g12 ðtÞ ¼

hdF1 ðtÞdF2 ðt þ tÞi ; hF1 ðtÞihF2 ðtÞi

ð15:1Þ

where hF1 ðtÞi and hF2 ðtÞi are the averages of F1(t) and F2(t), and dF1 ðtÞ ¼ F1 ðtÞ  hF1 ðtÞi, dF2 ðtÞ ¼ F2 ðtÞ  hF2 ðtÞi. The signals F1(t)and F2(t) can be equal, in which case Eq. (15.1) gives the autocorrelation gðtÞ of FðtÞ ¼ F1 ðtÞ ¼ F2 ðtÞ. Alternatively, the two signals can originate from two different spatial channels, giving spatial cross-correlation, or from two different spectral channels, as in dual-color versions of FCS, yielding spectral cross-correlation (see Box 3). Due to the variability of scan paths and goals in sFCS, several different ways of calculating the correlation from Eq. (15.1) are employed, in contrast to standard FCS, where the autocorrelation is either directly provided by the hardware correlator, or calculated by a multiple-tau method (Scha¨tzel, 1990).

328

Zdeneˇk Petra´sˇek et al.

Box 3 Dual-color cross-correlation spectroscopy

To measure binding between two distinctly labeled molecules, dual-color fluorescence cross-correlation spectroscopy (FCCS) (Schwille et al., 1997) can be employed. Here, two spectral channels are used to calculate the autocorrelation curves and also the spectral cross-correlation curve. Only if the two binding partners diffuse as an entity, do they give rise to a significant cross-correlation amplitude that can then be used to study the degree of binding. Dual-color crosscorrelation can be used in general with all implementations of sFCS discussed here if two spectral channels are used. Under the assumption that the red and green detection areas are equal in size and completely overlapping, we can calculate the relative binding from the amplitudes of the autocorrelation curves Gr(0) and Gg(0) and the cross-correlation amplitude Grg(0): Crg Grg ð0Þ ¼ ; Gr ð0Þ Crg þ Cg

Crg Grg ð0Þ ¼ ; Gg ð0Þ Crg þ Cr

ð15:13Þ

where Cg, Cr and Crg are the concentration of green-labelled, redlabelled and double-labelled molecules, respectively. Overlapping emission spectra and spectral cross talk, usually from the green fluorophore into the red channel, introduces additional similarities between the fluorescence fluctuations in the two channels and can therefore result in a false-positive cross-correlation. Spectral cross talk can be avoided by using alternating excitation (Muller et al., 2005; Thews et al., 2005) or has to be taken into account during data analysis (Ricˇka and Binkert, 1989). In the case of scanning FCCS spatial heterogeneities can also induce a false-positive cross-correlation. Correction schemes for heterogeneities must be explicitly tested on negative controls to avoid this artifact.

The multiple-tau method takes advantage of the fact that autocorrelations typically encountered in FCS decay progressively slower at longer lag times t, that is, the longer the lag time, the smaller the change in the correlation value. The correlation is calculated at time intervals and bin widths increasing with the lag time, rather than at fixed intervals with fixed bin width. Since typical correlations span time periods of many orders of magnitude, the number of lag times at which the correlation is evaluated is greatly reduced without losing much information. In commonly used implementations of the multiple-tau method the lag channel width and interval are doubled every m channels, where m is usually 8, 16, or more. The algorithms for the

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

329

calculation of g(t) from the sequence of the photon arrival times are described in detail elsewhere (Magatti and Ferri, 2003; Scha¨tzel, 1990). In sFCS with a circular path, the periodic motion introduces modulations to the correlation of the fluorescence signal. When the molecular motion is slow, the periodic modulation persists even at long correlation times, and may be averaged out by the wide bins in the multiple-tau method. Sometimes, it is sufficient to increase the number of constant-width channels m to 32, 64, or more in the multiple-tau method to refine the sampling (Petra´ˇsek and Schwille, 2008b), in other cases linear correlation with constant channel width has to be used (Petra´ˇsek et al., 2008b). The term ‘‘linear correlation’’ refers to the correlation calculated at lag times t spaced equally in time and with equal bin width. The direct calculation from the definition in Eq. (15.1) is computationally too intensive, due to the high number of correlation channels ti and real-time channels ti over which the averages in Eq. (15.1) are calculated. The standard way is therefore to exploit the relationship between the autocorrelation and the Fourier transform of F(t) together with the efficient algorithms for Fourier transform calculation. A possible problem may be the size of high-resolution data sets compared to the available computer memory. In such cases, a solution with an intermediate disk storage for large data files has been described (Petra´ˇsek et al., 2008a). In sFCS methods with the linear scan path, one typically sorts the raw data first, depending on the position from which the signal originates. To do this, the relationship between the signal and the measurement volume position has to be known at all times. The fluorescence is then a function of both the position r and the time t: F(r, t), and a spatiotemporal correlation g(x, t) can be defined as gðx; tÞ ¼

hdFðr; tÞdFðr þ x; t þ tÞi : hFðtÞi2

ð15:2Þ

A spatiotemporal cross-correlation for two different signals can be defined in a way analogous to Eq. (15.1). The spatiotemporal correlation is calculated either by using the linear correlation method for both time and space or by combining the multiple-tau correlation for time and the linear correlation for space. In applications where sFCS is used to perform measurements at multiple locations simultaneously only temporal correlation at desired positions ri is calculated from the appropriate parts of the raw data F(ri, t). In the application of sFCS to molecules in fluctuating membranes, the spatial resolution of F(r, t) is used to select only the part of the signal originating from the membrane and only temporal correlation is performed. This principle allows the elimination of membrane motion from the temporal correlation, and is described in more detail in Section 4.

330

Zdeneˇk Petra´sˇek et al.

3.2. Corrections The correlation calculated from the raw data often does not only represent the studied molecules but may be influenced by other factors, such as fluorescence background, gradual fluorophore depletion due to photobleaching, unspecific long-term fluctuations, etc. Here we describe how these effects can be corrected for before the correlation is fitted with a theoretical model. 3.2.1. Fluorescent background The signal background is often caused by the autofluorescence of native sample structures, but can also originate from optical components of the instrument, or poor blocking of Raman scattering. The background may be constant and uncorrelated in time, or exhibit an autocorrelation if it comes from diffusing molecules. It can be either uniform in space or vary as the scanning beam passes through different structures. The effect of a constant background is rescaling of the amplitude of the autocorrelation g(t) and can be corrected for if the background value B can be determined independently:
2 hF i gc ðtÞ ¼ gðtÞ: ð15:3Þ hF i  B If the background B(t) is correlated in time and can be measured independently from, for example, a control sample without fluorescence labeling, then the correlation of B(t) can be incorporated into the analysis as a fixed second component in a two-component fit (Rigler and Elson, 2001). 3.2.2. Photobleaching Photobleaching—the irreversible photochemical destruction of a fluorophore—leads to two kinds of artifacts on different timescales. (1)The sudden disappearance of a fluorophore results in shorter residence times in the detection volume and apparent faster diffusion. (2) Measurements within a closed small volume, for example, a cell or a vesicle, can lead to a gradual fluorescence depletion due to photobleaching. The same effect is encountered in two-dimensional systems, such as membranes, where the fluorophore depletion is not sufficiently compensated by influx of fresh molecules through diffusion. Under these conditions, the system is no longer in a steady state and correlation curves are seriously distorted. To avoid the first artifact—photobleaching in the detection volume— the excitation laser power has to be reduced until only a negligible fraction of fluorophores is bleached. It is advisable to determine the onset of photobleaching with an intensity series. The second artifact—depletion of fluorophores around the detection area due to photobleaching—is more difficult to avoid, since it can play a

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

331

role even for minimal excitation laser powers, especially if high concentrations are used. To correct for depletion (Ries et al., 2009a), the slow decay of the intensity trace F(t) is first approximated by an analytical function b(t). For a restricted reservoir, the decay can be assumed to be exponential: bðtÞ ¼ b0 et=tb . If depletion is due to the limited geometry in membrane measurements or the reservoir is connected weakly to a larger reservoir, a multiexponential is a good choice, and often two exponentials are sufficient. With the knowledge of b(t), the intensity trace can be corrected in the following way prior to calculating correlation curves:  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi FðtÞ Fc ðtÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ bð0Þ 1  bðtÞ=bð0Þ : ð15:4Þ bðtÞ=bð0Þ This transformation leads to a constant mean value and constant fluctuations with time, not distinguishable from a system in a steady state. The correlation curve calculated from the corrected intensity trace Fc(t) is no longer distorted by the decaying intensity and the concentration inferred by fitting is the initial concentration. Note that this approach does not correct for the apparent reduction of the concentration and diffusion time due to bleaching in the detection volume as described in the previous paragraph. 3.2.3. Spatial heterogeneities In sFCS, the beam passes through different parts of the sample, potentially crossing brighter or darker static structures or regions with varying background. This can lead to periodic oscillations in the correlation curve, or to longtime distortions. Similar distortions may be present at long correlation times, due to poor statistics. Although these effects cannot be removed by an independent measurement on a control sample, as with constant background described above, it is still desirable to eliminate these distortions to be able to better fit the data. One way to filter out these unwanted fluctuations from the correlation curve is to assume that the measured fluorescence intensity F(t) can be written as FðtÞ ¼ f ðtÞhðtÞ;

ð15:5Þ

where f(t) is the ‘‘pure’’ signal and h(t) is a modulation function reflecting either the periodic signal modulation along the scan path or long-term fluctuations in F(t). In case of no distortions, h(t) ¼ 1. The modulation function h(t) can be determined either by smoothing the raw data with a user-defined bin width or by fitting the raw data with an analytical function, as described above for the depletion case. It follows from the definition of the correlation function that the desired autocorrelation gc(t) of f(t) is related to the autocorrelation g(t) of F(t) and the autocorrelation gh(t) of h(t) in the following way:

332

Zdeneˇk Petra´sˇek et al.

gc ðtÞ ¼ ðgðtÞ þ 1Þ=ðgh ðtÞ þ 1Þ  1:

ð15:6Þ

Care should be taken with the choice of the bin width when calculating h(t), since this correction filters out all fluctuations on the longer timescale, without discriminating between the ‘‘true’’ signal and the interfering background. The correction of Eq. (15.6) remains approximately valid if the modulation by h(t) is not multiplicative as in Eq. (15.5), but additive, and if h(t) is small compared to f (t). The filtering using Eq. (15.6) affects only the temporal correlation profile, but does not correct the amplitude. This is because the value of the amplitude depends on the nature of the fluctuations (their variance). If the nature of the fluctuations is known throughout the measurement, the amplitude can also be corrected. For example, in the case of photodepletion as described above, the temporal profile can be corrected by Eq. (15.6) with hðtÞ ¼ bðtÞ=bð0Þ. The variance of the particle number changes as the concentration decreases, and it can be shown that the correct amplitude can be obtained by dividing gc(t) by hbð0Þ=bðtÞi. It is also possible to use the effects of the static structures on the autocorrelation to identify immobilized particles, rather than to filter this information out, as implemented in one form of small-circle sFCS (Skinner et al., 2005). It should be stressed that all corrections described here are not specific to sFCS but are applicable to standard FCS as well.

3.3. Data fitting 3.3.1. Fitting models Scanning affects the fluctuations of the measured signal, therefore it is only natural that it becomes visually apparent in the calculated autocorrelation, and also has to be included in the fitting models. Since the parameters of the scan are known, scanning does not increase the number of fitting parameters that might lead to less stable or undetermined fits. To the contrary, sFCS correlations contain more information, making some of the fitting parameters less correlated and the fits more stable. In the absence of scanning the model function G0(t) for free diffusion in three dimensions is G0 ðtÞ ¼




 1 1 t 1=2 1 t 1=2 1 t 1=2 pffiffiffi pffiffiffi pffiffiffi 1þ 1þ 1þ ; c pw0 tD tD tD S 2 pw0 pw0 S |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} x

y

z

ð15:7Þ

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

333

where c is the concentration, w0 the lateral size of the measurement volume, w0S its axial size, and tD ¼ w02 =ð4DÞ the diffusion time. The equation has been purposefully divided into three terms, each corresponding to diffusion along one dimension. If the molecular motion is restricted to two dimensions, for example, to a vertical yz or horizontal xy plane, only the corresponding terms remain. Here we use the capital symbol G(t) for models while the lowercase g(t) is reserved for correlations calculated from the data. Many organic dyes and fluorescent proteins exhibit fast, approximately exponential dynamics at sub-ms timescales. The origins of these fluctuations are singlet–triplet transitions, protonation reactions, etc. They are usually taken into account by introducing a multiplicative term into Eq. (15.7): 0

G 0 ðtÞ ¼ G0 ðtÞ

1  T þ T et=tT ; 1T

ð15:8Þ

where T is the fraction of the molecules in the dark (triplet, protonated) state, and tT is the characteristic relaxation time. The phenomena investigated with sFCS often occur on considerably longer timescales, therefore the fast dynamics term can usually be omitted. To see how the beam motion enters the models it is advantageous to start with the model for a circular scan:
 4R2 sin2 ðot=2Þ GðtÞ ¼ G0 ðtÞ exp  2 : w0 ð1 þ t=tD Þ

ð15:9Þ

The autocorrelation function with a fixed measurement volume is multiplied by a scan factor that depends on the scan frequency o and the radius R, both of which are known. Additionally, the scan factor depends on the diffusion coefficient D via tD and beam size w0 in a way that decouples these two parameters, making it possible to determine each of them independently from one fit. This decoupling is the basis of the precision and robustness of both circle- and line-scan methods. The autocorrelation for a linear scan with velocity v follows from Eq. (15.9) by assuming the limit of large radius, R!1, while maintaining the constant velocity v ¼ Ro:
GðtÞ ¼ G0 ðtÞ exp 

 v2 t2 : w02 ð1 þ t=tD Þ

ð15:10Þ

The spatiotemporal correlation G(x, t) for a linear scan follows directly from Eq. (15.10) by substituting x ¼ n t:
Gðx; tÞ ¼ G0 ðtÞ exp 

 x2 : w02 ð1 þ t=tD Þ

ð15:11Þ

334

Zdeneˇk Petra´sˇek et al.

This equation at the same time represents a cross-correlation between two volumes separated by the distance d  x. In practice, the sFCS measurement does not provide us with the spatiotemporal correlation at all values of x and t, but only at a subset of these. With a linear scan, the correlation is obtained at points where t ¼ x=v þ nT , where T is the scan period. In case of the circular scan, it is obtained at points where x ¼ 2R sinðot=2Þ. 3.3.2. Fitting The calculated and corrected correlation is fitted to the model using standard nonlinear, least squares fitting procedures as implemented in common data analysis software (Matlab, Origin, etc.). The results of the fit are the optimal parameter values that minimize the difference between the model and the data. We recommend to use weighting of the data points to achieve fits unbiased toward noisier data points. The weights wi are usually estimated from the correlations by comparing the variations between n ¼ 2k þ 1 data points surrounding the data point gi : gik ; . . . ; giþk . The weight wi is then taken as the inverse value of the standard deviation of these n points. A problem may arise when the deviation between the neighboring data points is not only due to random noise but due to abrupt variations of the correlation gi, as can happen with correlations modulated by the periodic scanning motion. The weights in these regions are then underestimated. In this case, the weights can be estimated from residuals of a preliminary unweighted fit, or the weights can be approximated by a smooth function, thus eliminating the bias at points of large variation.

4. Applications In this section, we describe several applications of sFCS, explaining the motivation behind using sFCS instead of standard (fixed-focus) FCS, the choice of the particular scan path, the details of implementation and analysis not mentioned already in the general part above, and the results obtained, demonstrating the advantages of the various sFCS approaches.

4.1. sFCS in Caenorhabditis elegans embryo We have applied circular sFCS with large-circle diameter to the investigation of PAR-2 protein dynamics on the cortex of living Caenorhabditis elegans embryo in the first division stage (Petra´ˇsek et al., 2008a,b). Several problems prevented us from using standard FCS in this case: the cortex does not stay in the same position for a sufficiently long time, the molecules move very slowly making it impossible to obtain good statistics within the

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

335

measurement time window limited by embryo development, and the slow motion causes photobleaching artifacts to be overwhelming. To circumvent these problems, we focused onto the bottom part of the embryo where a large part of the cortex is flattened and steady due to contact with the coverslip. Then we chose the largest circular scan path that still lies within the posterior half of the embryo and along which the fluorescence intensity is uniform (Fig. 15.2A). It would also be possible to use a linear scan; however, the length of the circle perimeter is larger than any line that would fit within the fluorescent part of the embryo, resulting in better averaging. Additionally, the data processing of the circle-scan data is more straightforward and no time is wasted due to the laser fly back and turning points as in the line scan. Scanning the beam practically eliminates the problems with photobleaching, because the total light dose is distributed along the whole scan path. Figure 15.2B shows the measured autocorrelation. The maxima of the uniformly spaced peaks correspond to a common autocorrelation as known from standard FCS. The peak spacing is determined by the scan frequency, 300 Hz in this case, a value close to the maximum technically possible with the setup used. A larger scan frequency would increase the temporal resolution at short correlation times. The better averaging achieved by probing many locations along the scan path results in meaningful correlation values up to the range of 1–10 s, which is not achievable with standard FCS within the measurement time of 100 s. More information than available from the amplitude of the correlation peaks shown in Fig. 15.2B can be obtained by replotting the data as a spatiotemporal correlation (Fig. 15.2C). Spatial information is contained in the shape of the peaks and is obtained by converting the correlation time into the spatial correlation coordinatex : x ¼ 2R sinðpf tÞ. This spatial correlation contains information about lateral spreading, and, in principle, makes it possible to distinguish lateral motion from detachment of proteins from the cortex. An analysis of the spatiotemporal correlation data shows that the best description requires a model with at least two diffusion components. These experiments have been performed using two-photon excitation, which has the benefit of reduced photobleaching in the embryo as a whole, since the molecules outside the focal volume are not excited. Additionally, no signs of phototoxicity are observed. Taken together, these two features of two-photon excitation enable long observation times, permitting monitoring of embryo development.

4.2. Small-circle sFCS sFCS with a small circular path is a convenient way to measure spatiotemporal correlations within a minimum sample volume. In its original application (Skinner et al., 2005), it was implemented as position-sensitive

336

Zdeneˇk Petra´sˇek et al.

correlation, which allows determination of not only diffusion, but also direction and magnitude of flow, immobilization, and any combination of these. We have applied a simpler analysis, without the position sensitivity, and used the spatial information encoded in the correlation to determine the diffusion coefficients in a precise and robust way, even in the presence of optical distortions and photobleaching (Petra´ˇsek and Schwille, 2008b). When the beam is scanned in a small circle (Fig. 15.3A), the autocorrelation is modulated as a result of scanning (Fig. 15.3B), but contrary to the large-circle scan, the width of the peaks at multiples of the scan period is comparable to their spacing. In both cases, the model autocorrelation is expressed by Eq. (15.9). This modulation makes it possible to determine A f R

B

0.1 eGFP f R D a c2r

0.08

g(t)

0.06 0.04

1 kHz 0.300 mm 92.8 mm2 s−1 0.148 mm 1.03

Residuals

0.02 0 2 0 −2

10−2

10−1

100 t(ms)

101

102

Figure 15.3 Small-circle sFCS for robust and precise determination of diffusion coefficients. (A) The laser beam is scanned in a circle with frequency f and radius R which is comparable to the measurement volume size w0. (B) The autocorrelation of the fluorescence signal is modulated as a result of scanning. The upper and lower envelopes correspond to the autocorrelation at a fixed position and the spatial crosscorrelation at a distance equal to the diameter 2R of the scanned circle, respectively. From a fit of the data with Eq. (15.9) the diffusion coefficient D and the volume size w0 ¼ 2a are determined. The sample is eGFP (enhanced Green Fluorescent Protein) in buffer solution.

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

337

both the diffusion coefficient D and the volume size w0 independently. The upper envelope of the oscillating autocorrelation is the curve that would be measured with standard FCS. The lower envelope corresponds to a spatial cross-correlation between two extreme positions of the moving laser focus, that is, at a distance 2R, as shown in Fig. 15.3A. The choice of the scan frequency f will typically be limited by the instrument and often has such a value that the scan period 1/f is longer than the diffusion time tD of the investigated molecules. The optimal scan pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi radius R was shown to lie near the value of 0:5w0 2 þ 1=ðf tD Þ, which is comparable to the measurement volume size w0. The optimal scan radius approximately corresponds to the situation where the first maximum of the autocorrelation at the time 1/f coincides with the maximum of the crosscorrelation forming the lower envelope of the scanning autocorrelation (Fig. 15.3B). This implementation of sFCS, apart from yielding the diffusion coefficient without any knowledge of the volume size w0, was shown to be more robust than standard FCS against photobleaching and optical distortions that change the volume size. Its potential has been demonstrated both in solutions and in living cells.

4.3. sFCS with a perpendicular scan path to measure in unstable membranes For measurements in biological membranes, membrane movements or instabilities can be a major problem. Considering the lateral and axial sizes of the focal spot of
250 and
700 nm, respectively, shifts as small as 100 nm during a measurement time of several minutes are already devastating and lead to strong distortions of the correlation curves. The choice of a scan path that is perpendicular to the membrane plane avoids this problem as the detection volume crosses the membrane in a reproducible way independently of membrane movements. This method was first introduced with a circular scan path to measure membrane dynamics in presence of a strong background in solution (Ruan et al., 2004). Here we concentrate on the implementation with a linear scan path (Ries and Schwille, 2006), since it can be readily used with a commercial laser scanning microscope. In addition, it can be easily extended to dual-focus FCS for exact diffusion measurements without the need for calibrating the detection volume, and to dual-color FCS to measure interactions in membranes. In this implementation of sFCS, the scan path is chosen in a way that it crosses the vertical membrane perpendicularly (see Fig. 15.4). The intersection of the scan path with the membrane defines the detection area. During data analysis, the intensity fluctuations in this detection area have to be extracted to calculate correlation curves. The necessary steps are as follows:

338

Time

A

B

Align Side view

Position

C

G(t)

Zdeneˇk Petra´sˇek et al.

D

E

Correlate

Sum F

t

Figure 15.4 Principle of sFCS with the scan path perpendicular to a membrane. (A) In sFCS the detection volume is repeatedly scanned perpendicularly through a vertical membrane. The individual line scans can be arranged as a pseudo-image, (B) where the vertical axis denotes the time. In this pseudo-image the membrane is clearly visible. Due to membrane movements, the position of the membrane is not constant in time. These instabilities can be corrected for by shifting each line scan in such a way that the membrane becomes a straight line. (C) For each scan, membrane contributions are added up to give one point in the intensity trace (D) which can then be used to calculate the autocorrelation curve G(t) (E).

1. Construction of pseudo-image. In case a commercial laser scanning microscope, such as Zeiss LSM-510, is used in the line-scan mode, the raw data will already be in the form of a pseudo-image, where the individual line scans are arranged next to each other. Then the horizontal axis denotes the position in the sample and the vertical axis the time. If the raw data consist of photon arrival times, the reconstruction of the pseudo-image relies on knowledge of the repetition rate. The use of beginning-of-line marker signals, automatically inserted into the data file by the software controlling the scanners, is most accurate. If those are not available, the repetition rate can be determined from the periodicity of the signal (Ries and Schwille, 2006). 2. Identification of the membrane. To reduce the background and improve the signal-to-noise ratio, only the signal in a small window around the membrane is used to construct the intensity trace. The position of the membrane at a given time ti can be determined by evaluating the position of the intensity maximum in an average of several hundred scans around the ti. 3. Construction of the fluctuating intensity trace. Photons collected within a time window Dt around the intensity maximum are summed up to result in the intensity Fi. The time window should be large enough that no signal from the membrane is lost, but small enough to reject most of the background noise. A good choice is to include all photons within 2s  2s of the Gaussian approximation of the apparent membrane profile.

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

339

4. Calculation of the correlation curve. The correlation curve can be calculated using a standard multiple-tau correlation algorithm (Magatti and Ferri, 2003). 5. Data fitting. The effective detection area is described by a two-dimensional elliptical Gaussian with an aspect ratio of S ¼ wz/wx. The correlation function describing the experimental correlation curve is given by Eq. (15.7) using only the y and z parts (Ries and Schwille, 2006). Note that due to the limited time resolution of approximately 1 ms, triplet dynamics need not be included in the model. A prerequisite for the use of this implementation of sFCS is a vertical membrane. On the length scale of the detection area (1 mm), the curvature should be small. This need for a vertical membrane seems to preclude its application to adherent cells. A solution is to grow cells on glass beads with a diameter of 10–50 mm. Plasma membranes of cells at the equator of the beads are predominantly vertical and are well suited for sFCS. However, inhomogeneities on the micrometer scale due to protrusions or interactions with the cytoskeleton still limit the applicability of FCS. Plasma membrane spheres (Lingwood et al., 2008) allow the investigation of processes in plasma membranes in a much simpler, spherical geometry which is optimally suited for sFCS. By measuring diffusion coefficients with dual-focus sFCS, we could show that the cross-linking of the lipid GM1 by the protein CTxB leads to a lipid-driven raft-like phase separation in plasma membrane spheres (Lingwood et al., 2008).

4.4. Dual-focus sFCS To overcome problems caused by optical aberrations, saturation, and other experimental uncertainties that render calibration measurements to determine the size of the detection area inaccurate, dual-focus FCS (Dertinger et al., 2007) can be employed. It employs two spatial detection areas with a well-defined and known distance. Here, the calibration of the detection area is replaced by the determination of this distance, which can be measured more precisely, and, most importantly, is constant for all measurements. In two-focus sFCS (Fig. 15.5), two parallel lines, spatially offset by a distance d, are scanned through the membrane in an alternating fashion. The intersections with the membrane give rise to two intensity traces F1(t) and F2(t), from which the autocorrelation curves g1(t) and g2(t) and also the spatial cross-correlation curves g12(t) and g21(t) can be calculated using Eq. (15.1). The model function for the experimental cross-correlation curves in case of free diffusion is given by Eq. (15.7) (y and z parts) and Eq. (15.11) with x ¼ d (Ries and Schwille, 2006):

340

Zdeneˇk Petra´sˇek et al.

C

B

D Autocorrelation Cross-correlation

G(t)

A

d

Top view

d

t

t+t

t

Figure 15.5 Dual-focus sFCS. (A) Two parallel lines are alternatingly scanned in a perpendicular path through the membrane. (B) The intersections with the membrane define two detection areas. From the intensities in the two detection areas (C) the autoand spatial cross-correlation curves can be calculated. (D) If the distance d between the detection areas is known, a fit of the correlation curves allows direct determination of concentrations and diffusion coefficients without the need for calibrating the size of the detection area.

G12 ðtÞ ¼ G21 ðtÞ ¼



 1 4Dt 1=2 4Dt 1=2 1 þ 1 þ cpSw02 w02 w02 S2
 d2  exp  2 : w0 þ 4Dt

ð15:12Þ

Once the distance d is known, the diffusion coefficient D, the concentration c and the waist of the laser focus w0 can be determined directly by fitting the data with Eq. (15.12) without any additional calibration measurement (Dertinger et al., 2007). The autocorrelation function follows from Eq. (15.12) for d ¼ 0. The photons in the two foci are not collected within the same time window, but with a delay td, which is usually given by the scan period. Therefore, the cross-correlation curves are shifted with respect to the autocorrelation curves by this delay time, and this needs to be taken into account during data analysis (Ries and Schwille, 2006). Two lines can be scanned with commercial laser scanning microscopes by using the frame mode with N  2 pixels. However, for the maximum repetition rate, the lines are not necessarily parallel. The following settings result in parallel lines in the Zeiss LSM-510: bidirectional multitrack mode, 32  2 pixels per frame, illumination only during the second track.

4.5. Dual-color sFCS In membranes, binding can be measured with dual-color sFCS using two spectral channels (see Box 3). By scanning every other line with a different color and detecting the photons only in the corresponding channel,

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

341

contributions from two different fluorophores can be separated completely and spectral cross talk is avoided (Muller et al., 2005; Ries and Schwille, 2006; Thews et al., 2005). From the two intensity traces, red and green autocorrelation curves and the cross-correlation curve can be calculated. By fitting the correlation curves with Eq. (15.7), their amplitudes can be obtained. The relative binding can then be calculated with Eq. (15.13) below. We applied dual-color sFCS to study the binding affinities of the apoptotic proteins tBid and BclXL in giant unilamellar vesicles, micrometer sized liposomes (Garcı´a-Sa´ez et al., 2009). We could show that their interaction is drastically enhanced in the membrane compared to free solution. The application of sFCS is not limited to single cells, but can be applied to multicellular organisms. In tissue, most cells exhibit large vertical membranes, well suited for sFCS. We combined dual-focus sFCS and dual-color sFCS with conventional static-volume FCS to determine the mobilities and binding affinities of fibroblast growth factor receptors (Fgfr1, Fgfr4) to their extracellular ligand Fgf8 directly in living gastrulating zebra fish embryos. Here, dual-focus sFCS was used to measure the diffusion coefficients of the receptors in the plasma membranes, and the size of the detection area. This internal calibration is very useful due to optical distortions in tissue. Dual-color sFCS with alternating excitation was used to determine the concentrations of the receptors and receptor–ligand complexes in the membrane, whereas static-volume FCS measured the free-ligand concentration in the extracellular space (Ries et al., 2009b). Using global data analysis we could infer the dissociation constants between Fgfr1 or Fgfr4 and Fgf8 and estimate the endogenous receptor concentration.

5. Conclusion We aimed to present the family of sFCS methods with an eye on the problems they address and the measurement geometries used to implement them (i.e., the scan path). Successful application of sFCS requires understanding of the motivation for using scanning in any given implementation, as well as a knowledge of its limitations. The general aspects of sFCS, including implementation and data analysis, were supplemented by several examples of applications of different sFCS approaches. Altogether, this article should guide the user not only to reproduce an existing sFCS method, but also to modify it or to design a new implementation particularly suited to the problem investigated.

342

Zdeneˇk Petra´sˇek et al.

REFERENCES Bacia, K., and Schwille, P. (2003). A dynamic view of cellular processes by in vivo fluorescence auto-and cross-correlation spectroscopy. Methods 29, 74–85. Bacia, K., Kim, S., and Schwille, P. (2006). Fluorescence cross-correlation spectroscopy in living cells. Nat. Methods 3, 83–89. Becker, W., Bergmann, A., Hink, M. A., Konig, K., Benndorf, K., and Biskup, C. (2004). Fluorescence lifetime imaging by time-correlated single-photon counting. Microsc. Res. Tech. 63, 58–66. Chiantia, S., Ries, J., Kahya, N., and Schwille, P. (2006). Combined AFM and two-focus SFCS study of raft-exhibiting model membranes. ChemPhysChem 7, 2409–2418. Dertinger, T., Pacheco, V., von der Hocht, I., Hartmann, R., Gregor, I., and Enderlein, J. (2007). Two-focus fluorescence correlation spectroscopy: a new tool for accurate and absolute diffusion measurements. ChemPhysChem 8, 433–443. Digman, M. A., Brown, C. M., Sengupta, P., Wiseman, P. W., Horwitz, A. R., and Gratton, E. (2005). Measuring fast dynamics in solutions and cells with a laser scanning microscope. Biophys. J. 89, 1317–1327. Enderlein, J., Gregor, I., Patra, D., Dertinger, T., and Kaupp, U. B. (2005). Performance of fluorescence correlation spectroscopy for measuring diffusion and concentration. ChemPhysChem 6, 2324–2336. Garcı´a-Sa´ez, A. J., Ries, J., Orza´ez, M., Pe´rez-Paya`, E., and Schwille, P. (2009). Membrane promotes tBid interaction with Bcl-xL. Nat. Struct. Mol. Biol. 16, 1178–1185. Hebert, B., Costantino, S., and Wiseman, P. W. (2005). Spatiotemporal image correlation spectroscopy (STICS) theory, verification, and application to protein velocity mapping in living CHO cells. Biophys. J. 88, 3601–3614. Kim, S. A., Heinze, K. G., and Schwille, P. (2007). Fluorescence correlation spectroscopy in living cells. Nat. Methods 4, 963–973. Lingwood, D., Ries, J., Schwille, P., and Simons, K. (2008). Plasma membranes are poised for activation of raft phase coalescence at physiological temperature. Proc. Natl. Acad. Sci. USA 105, 10005–10010. Magatti, D., and Ferri, F. (2003). 25 ns software correlator for photon and fluorescence correlation spectroscopy. Rev. Sci. Instrum. 74, 1135–1144. Muller, B., Zaychikov, E., Brauchle, C., and Lamb, D. (2005). Pulsed interleaved excitation. Biophys. J. 89, 3508–3522. Petra´ˇsek, Z., and Schwille, P. (2008a). Photobleaching in two-photon scanning fluorescence correlation spectroscopy. ChemPhysChem 9, 147–158. Petra´ˇsek, Z., and Schwille, P. (2008b). Precise measurement of diffusion coefficients using scanning fluorescence correlation spectroscopy. Biophys. J. 94, 1437–1448. Petra´ˇsek, Z., and Schwille, P. (2008c). Scanning fluorescence correlation spectroscopy. In ‘‘Single Molecules and Nanotechnology’’, Springer, Berlin, Vol. 12 of Springer Series in Biophysics, pp. 83–105, chapter 4. Petra´ˇsek, Z., Hoege, C., Hyman, A. A., and Schwille, P. (2008a). Two-photon fluorescence imaging and correlation analysis applied to protein dynamics in C. elegans embryo. Proc. SPIE 6860, 68601L. Petra´ˇsek, Z., Hoege, C., Mashaghi, A., Ohrt, T., Hyman, A. A., and Schwille, P. (2008b). Characterization of protein dynamics in asymmetric cell division by scanning fluorescence correlation spectroscopy. Biophys. J. 95, 5476–5486. Petrov, E. P., and Schwille, P. (2008). State of the art and novel trends in fluorescence correlation spectroscopy. In ‘‘Standardization and Quality Assurance in Fluorescence Measurements II: Bioanalytical and Biomedical Applications’’, Springer, Berlin, Heidelberg, New York, Vol. 6 of Springer Series on Fluorescence, pp. 145–197.

Scanning FCS for the Characterization of Protein Dynamics in Live Cells

343

Ricˇka, J., and Binkert, T. (1989). Direct measurement of a distinct correlation-function by fluorescence cross-correlation. Phys. Rev. A: At. Mol. Opt. Phys. 39, 2646–2652. Ries, J., and Schwille, P. (2006). Studying slow membrane dynamics with continuous wave scanning fluorescence correlation spectroscopy. Biophys. J. 91, 1915–1924. Ries, J., Chiantia, S., and Schwille, P. (2009a). Accurate determination of membrane dynamics with line-scan FCS. Biophys. J. 96, 1999–2008. Ries, J., Yu, S. R., Burkhardt, M., Brand, M., and Schwille, P. (2009b). Modular scanning FCS quantifies receptor-ligand interactions in living multicellular organisms. Nat. Methods 6, 643–645. Rigler, R., and Elson, E. (2001). Fluorescence Correlation Spectroscopy: Theory and Applications Springer. Ruan, Q., Cheng, M., Levi, M., Gratton, E., and Mantulin, W. (2004). Spatial-temporal studies of membrane dynamics: scanning fluorescence correlation spectroscopy (SFCS). Biophys. J. 87, 1260–1267. Satsoura, D., Leber, B., Andrews, D. W., and Fradin, C. (2007). Circumvention of fluorophore photobleaching in fluorescence fluctuation experiments: a beam scanning approach. ChemPhysChem 8, 834–848. Scha¨tzel, K. (1990). Noise on photon correlation data. I. Autocorrelation functions. Quantum Opt. 2, 287–305. Schwille, P., Meyer-Almes, F., and Rigler, R. (1997). Dual-color fluorescence crosscorrelation spectroscopy for multicomponent diffusional analysis in solution. Biophys. J. 72, 1878–1886. Skinner, J. P., Chen, Y., and Muller, J. D. (2005). Position-sensitive scanning fluorescence correlation spectroscopy. Biophys. J. 89, 1288–1301. Tcherniak, A., Reznik, C., Link, S., and Landes, C. F. (2009). Fluorescence correlation spectroscopy: criteria for analysis in complex systems. Anal. Chem. 81, 746–754. Thews, E., Gerken, M., Eckert, R., Zapfel, J., Tietz, C., and Wrachtrup, J. (2005). Cross talk free fluorescence cross correlation spectroscopy in live cells. Biophys. J. 89, 2069–2076. Wiseman, P. W., Squier, J. A., Ellisman, M. H., and Wilson, K. R. (2000). Two-photon image correlation spectroscopy and image cross-correlation spectroscopy. J. Microsc. 200, 14–25.

C H A P T E R

S I X T E E N

Observing Protein Interactions and Their Stoichiometry in Living Cells by Brightness Analysis of Fluorescence Fluctuation Experiments Yan Chen,* Jolene Johnson,* Patrick Macdonald,* Bin Wu,*,† and Joachim D. Mueller* Contents 346 347 347 349 354 354 354 356 357 361 361

1. Introduction 2. Brightness Classification of Fluorescent Molecules 2.1. Single brightness state 2.2. Two brightness states 2.3. Fluorescent proteins 3. Brightness Measurements in Cells 3.1. Brightness titration 3.2. Control and calibration experiments 3.3. Cell selection and measurement Acknowledgments References

Abstract A single fluorescently labeled protein generates a short burst of light whenever it passes through a tiny observation volume created within a biological cell. The average amplitude of the burst is related to the stoichiometry of the fluorescently labeled protein complex. Fluorescence fluctuation spectroscopy quantifies the burst amplitude by introducing the brightness parameter. Brightness provides a spectroscopic marker for observing protein interactions and their stoichiometry directly inside cells. Not all fluorescent proteins are suitable for brightness experiments. Here we discuss how brightness properties of the fluorophore influence brightness measurements and how to identify a well-behaved fluorescent protein. Protein interactions and stoichiometry are determined from a brightness titration. * School of Physics and Astronomy, University of Minnesota, Minneapolis, USA Albert Einstein College of Medicine, Bronx, New York, USA

{

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72026-7

#

2010 Elsevier Inc. All rights reserved.

345

346

Yan Chen et al.

Experimental details of brightness titration measurements are described together with the necessary calibration and control experiments.

1. Introduction Protein interactions are at the heart of virtually every cellular process. These interactions lead to the assembly of complexes comprised of several protein molecules. These protein complexes interact with other protein assemblies as they carry out their biological function. Because protein assemblies and their interactions are central to biological function, it is highly desirable to have methods that detect and quantify protein complexes and their composition directly in the living cell. Brightness measurements with fluorescence fluctuation spectroscopy (FFS) provide a general framework for addressing this challenge. Each fluorescent object passing through a small optical observation volume gives rise to a fluctuation in the fluorescent signal. FFS exploits these fluctuations to gain information about the nature of these fluorescent objects (Magde et al., 1972; Thompson et al., 2002). Brightness is an FFS parameter that characterizes the average fluorescence intensity of a single molecule (Chen et al., 1999; Palmer and Thompson, 1989; Qian and Elson, 1990a). The fluorescent molecules of greatest interest for cellular FFS experiments are fluorescent proteins, because they offer an ideal tool for the labeling of cellular proteins. A fluorescently labeled protein passing through the observation volume is excited by the laser beam and produces a burst of fluorescence. Now consider two fluorescently labeled proteins that associate into a dimeric complex. The dimer carries two independently fluorescing labels, which are simultaneously excited whenever the dimer crosses the observation volume. We expect to observe twice the fluorescent signal from a dimer carrying two labels compared to a monomeric protein carrying a single label. In other words, the brightness of the dimer is twice the brightness of the monomer (Chen et al., 2003). This conceptual example illustrates that brightness encodes the stoichiometry of a protein complex (Fig. 16.1). Accurate FFS measurements in cells provide a noninvasive method for observing protein interactions through changes in the brightness. Brightness experiments in cells are quite finicky, and a thorough characterization of the experimental system is vital in order to arrive at a faithful interpretation of the experiment. There are a large number of potential pitfalls and while our lab has close to 10 years of experience with this subject, we continue to be surprised by new issues that surface as we advance the methodology. Given the space constraints of this article, we focus on a few fundamental issues that we have directly encountered. We hope this approach provides the most authoritative and useful account for the reader. Consequently, we limit

Observing Protein Interactions and Their Stoichiometry in Living Cells

Fluorophore

Brightness: l

Monomer

Dimer

Trimer

2l

3l

347

Protein of interest

l

Figure 16.1 Brightness and stoichiometry. A fluorophore with brightness l is attached to the protein of interest. The brightness of the monomeric protein equals that of the fluorophore alone. A dimer has a brightness of 2l, because it carries two fluorophores. Similarly, the brightness of a trimer is 3l.

our commentary to two-photon excitation (Berland et al., 1995; Denk et al., 1990), which the lab has used exclusively for brightness measurements in cells. We describe brightness experiments in the context of freely diffusing proteins in the cytoplasm or nucleoplasm of cells. Such systems are conveniently characterized by brightness experiments with a stationary excitation beam. While FFS experiments with a stationary observation volume are the most common, fluctuation experiments using a scanned beam or imaging methods are becoming more widely used (Berland et al., 1996; Digman et al., 2005; Wiseman and Petersen, 1999). We start by discussing the brightness properties of fluorescent labels, because quantitative interpretation of brightness measurements hinges upon a well-behaved fluorophore.

2. Brightness Classification of Fluorescent Molecules Not all fluorophores are suited for brightness experiments. Ideally, the brightness of an object should be directly proportional to the number of fluorophores carried (Fig. 16.1). In this case, the brightness of a dimer is twice the brightness of the monomer. We now present a brief classification of fluorophores with different brightness properties. This classification provides a succinct overview of the most crucial aspects to consider when evaluating fluorophores for FFS experiments.

2.1. Single brightness state Fluorescent molecules with a single brightness state are the most suitable for quantitative FFS experiments provided that the absence of quenching is established.

348

Yan Chen et al.

2.1.1. Quenching Local changes in the environment of a fluorophore can affect its spectroscopic properties. Changes in the environment typically results in quenching of the fluorescence intensity. We are specifically concerned with quenching occurring as a result of attaching the fluorescent label to a protein or when labeled proteins associate into a complex. Another source for quenching is the direct interaction between two fluorophores in close proximity, which is referred to as self-quenching. There are two types of quenching, static and dynamic (Lakowicz, 2006). Dynamic quenching is easily detected, because it reduces the florescence lifetime of the fluorophore. Static quenching, on the other hand, is much harder to diagnose. Because both static and dynamic quenching reduce the brightness of the fluorophore, FFS experiments are well suited to detect the presence of quenching. Here we use EGFP as an example to illustrate the typical experiments that are performed to evaluate a new fluorophore (see Fig. 16.2). First, cells expressing EGFP are measured to determine the intrinsic brightness of the fluorescent label. Second, a cellular protein X is chosen that is known to form no homooligomers, such as the ligand-binding domain of the retinoic acid receptor (RARLBD) (Egea et al., 2001). Protein X is labeled with EGFP by constructing the fusion protein

Fluorophores/ proteins A

Brightness unquenched

Brightness quenched

l

--

l

lQ < l

2l

lQ < 2l

2l

lQ < 2l

B

C

D

Figure 16.2 Quenching of brightness. (A) A fluorophore with brightness l. (B) Protein labeled with the fluorophore. (C) Two fluorophores are covalently linked together to create an artificial brightness dimer. (D) A fluorescently labeled protein that is known to form a dimer. In the absence of quenching the brightness is exactly l times the number of fluorophores carried by the molecular complex. Any changes in the molecular environment of the fluorophore that lead to quenching result in a lower than expected brightness.

Observing Protein Interactions and Their Stoichiometry in Living Cells

349

X-EGFP, which is expressed in cells. The experimentally measured brightness of X-EGFP is compared to the intrinsic brightness of EGFP. If both have the same brightness, then labeling with the fluorophore is free of quenching. Next, a fusion protein containing two EGFPs is constructed to create an artificial dimer of the fluorescent protein, which we refer to as EGFP2. The brightness of EGFP2 is measured and compared with EGFP. The brightness of EGFP2 is only double that of EGFP in the absence of quenching. Finally, we choose a protein Y that is known to form a dimer, such as the ligand-binding domain of retinoic X receptor (RXRLBD) in the presence of its activating ligand. The protein Y-EGFP is expressed in cells and forms a dimer (Y-EGFP)2 at sufficiently high concentrations. A measurement of (Y-EGFP)2 only provides twice the brightness of EGFP if quenching is absent. These test experiments (summarized in Fig. 16.2) constitute a fairly rigorous examination of the fluorophore. A fluorophore that passes all these tests is well behaved and suitable for brightness experiments. 2.1.2. HomoFRET Fo¨rster resonance energy transfer between identical fluorophores is known as homoFRET. The fluorescence emission spectrum and fluorescence lifetime are unaffected by homoFRET (Lakowicz, 2006). Each fluorophore donates and receives energy at the same rate. These two effects compensate each other and keep the fluorescence intensity emitted per molecule unchanged. This argument implies that the brightness, which is the detected fluorescence intensity per molecule, is unchanged by homoFRET. However, a subtle effect might nevertheless influence the brightness. Fluorescent proteins, such as EGFP, have a longer rotational correlation time than fluorescence lifetime, which leads to a polarized fluorescence signal. HomoFRET introduces a depolarization of the fluorescence. This depolarization effect is sometimes exploited to detect homointeractions using a polarization sensitive detection setup (Gautier et al., 2001). Brightness experiments are performed without polarizers. However, other optical elements, such as high numerical aperture objectives and dichroic beam splitters, may inadvertently introduce a polarization dependence. A quick method to rule out a depolarization-dependent influence on brightness of the experimental setup is to measure a dimeric fluorescent protein construct that is known to exhibit homoFRET. For example, a tandem construct that joins two EGFPs with a short linker sequence brings both fluorescent (Widengren et al., 1994) construct is twice that of EGFP, then depolarization effects caused by homoFRET are negligible.

2.2. Two brightness states Not every fluorescent protein has a single spectroscopic state (Volkmer et al., 2000). Conformational substates that alter the environment of a chromophore within a protein are a potential source of brightness states. In addition, the

350

Yan Chen et al.

presence of different brightness states may also be the result of photophysical processes, such as intersystem crossing, which populates the triplet state (Widengren et al., 1994). A protein that can switch between different spectroscopic states has a brightness that depends on the properties of each state. To simplify the discussion we consider the presence of two states, which is sufficient to capture the essence of brightness heterogeneity of a fluorophore. For clarity, we assume that quenching of the fluorescent label is negligible. We refer to the two states of the fluorophore A as A(1) and A(2). The corresponding brightness of each state is described by l(1) and l(2). The fluorophore interconverts between both states (A(1) $ A(2)) on a timescale that depends on the activation energy separating the states. We contrast this interconversion time with the diffusion time, which represents the average time for the fluorophore to pass through the observation volume. The term ‘‘long-lived states’’ indicates that the interconversion time is slower than the diffusion time of the fluorophore. Under this condition the vast majority of fluorophores cross the observation volume without changing their state. Because the contribution of the interconversion process on the fluctuations is negligible, a fluorophore with long-lived states is modeled for the purpose of FFS experiments as having noninterconverting states (Hillesheim et al., 2006). On the other hand, if the interconversion time is similar or faster than the diffusion time, the fluorophore is ‘‘flickering.’’ The likelihood that the fluorophore changes its brightness state when passing through the observation volume is high. The additional fluctuations contributed from the switching of states cannot be ignored and need to be accounted for in quantitative brightness experiments. The easiest way of detecting flickering is through the fluorescence correlation function (Schwille et al., 2000). Because autocorrelation functions are plagued by afterpulsing artifacts (Enderlein and Gregor, 2005), it is best to measure the crosscorrelation of the fluorophore by evenly distributing the fluorescence signal into two separate detection channels. While this procedure weakens the amount of light received by the detector, the signal quality is typically good enough for performing a crosscorrelation measurement. A correlation function that shows a fast process in addition to the diffusion process indicates the presence of flickering (Fig. 16.3A). A correlation function modeled by a single diffusion process indicates the absence of flickering (Fig. 16.3B). Thus, if the fluorophores exist in different brightness states, these states have to be long lived. However, the number of long-lived states (whether one or more) is not determined by the correlation function. The identification and characterization of long-lived brightness states of fluorescent proteins by FFS is a fairly time consuming and complex endeavor. We refer to the literature for further details (Hillesheim et al., 2006; Wu et al., 2009). It is particularly difficult to detect flickering if the interconversion time coincides with the diffusion time. Slowing down the diffusion process by increasing the viscosity of the solvent (by adding glycerol or sucrose, e.g.) or by attaching

Observing Protein Interactions and Their Stoichiometry in Living Cells

351

A

g(t)

Flickering

Diffusion

t

g(t)

B

Diffusion

t C 0.04

g(t)

0.03 0.02

Residual

0.01 0.00 0.01 0.00 -0.01 -6

10

-5

10

10-4 t (s)

10-3

10-2

Figure 16.3 Brightness flickering. (A) The fluorescence correlation function of a fluorescent protein with fast brightness flickering shows, in addition to the diffusion process, a second kinetic process at short correlation times. (B) In the absence of flickering only the diffusion process is present. (C) Crosscorrelation function of EGFP measured in PBS (pH 7.4) with 50% glycerol at 905 nm using a 50/50 beam splitter. The correlation function is fit to a simple diffusion model (solid line). The residuals of the fit are shown in the bottom panel.

the fluorescent label to a large molecule separates the two processes, which is very useful for identifying flickering under these circumstances. 2.2.1. Long-lived states The most straightforward example of two long-lived states is a fluorescent protein with a dark state (Fig. 16.4A). We refer to the bright and dark states as B and D, respectively. The protein is present as a mixture with fraction

352

Yan Chen et al.

Fluorescent protein with a bright and dark state B

D =Dark, non-fluorescent =Bright, fluorescent population fraction: (1-f ) population fraction: f

A

1-f

B

lB = l

f

D

lD = 0

Monomeric fluorophore

2

(1-f )

B Dimeric fluorophore

2f (1-f ) f

C

Monomer sample

l sample = l

B

B

B

D

D

D

lBB = 2l lBD = l

2

D

lDD = 0

Dimer sample

l < l sample < 2l

Figure 16.4 Fluorescent protein with a long-lived bright and dark state. (A) The fluorophore exists with a population fraction (1  f ) in the bright state B and with a population fraction f in the dark state D. (B) A dimeric fluorescent protein exists as three brightness species: BB with brightness 2l, BD with brightness l, and DD with brightness 0. (C) The brightness of the monomer sample is l, because the dark state population is invisible. (D) The dimer sample reveals the presence of the dark state population, because dimers that carry a bright and dark fluorophore only contribute half the brightness of the dimers with two bright fluorophores. The mixture of brightness populations leads to a brightness of the sample that is in between l and 2l.

(1  f ) existing in the bright state B and a fraction f existing in the dark state D. Because the dark state has a brightness of zero, it is invisible to FFS. An FFS experiment of such a fluorescent protein only detects the bright state and identifies its brightness l and concentration (Fig. 16.4C). Thus, the FFS experiment fails to recognize the presence of the dark state D and underestimates the total concentration of fluorophores present. Expression of a

Observing Protein Interactions and Their Stoichiometry in Living Cells

353

dimeric protein results in three brightness species (BB, BD, DD). The BB species with brightness 2l contains two fluorophores in the bright state, the BD species with brightness l contains a bright and a dark fluorophore, and the DD species with brightness 0 contains two dark fluorophores (Fig. 16.4B). FFS experiments under cellular conditions cannot resolve this brightness mixture and instead report an apparent brightness that reflects the average over the different brightness populations present. It is straightforward to derive that a fluorophore with a dark-state fraction f leads to a brightness of (2  f )l for the dimer (Hillesheim et al., 2006). In other words, the presence of a dark state leads to less than doubling of the brightness for the dimer (Fig. 16.4D). As the dark state fraction grows, the brightness of the dimer gets closer to that of the monomer. The reduction in the brightness contrast between dimer and monomer makes the use of fluorophores with dark states in brightness experiments problematic, and quantitative brightness experiments are only feasible if the dark-state population is sufficiently small. Instead of a bright and dark state, a fluorophore may have a bright and a dim state. While the details for calculating the apparent brightness values differ from the dark-state case, the general results are very similar, and the brightness of the dimer is less than double the brightness of the monomer (Wu et al., 2009). Thus a fluorescent label with different brightness states presents a serious complication for determining the stoichiometry of protein complexes. The use of a fluorescent protein with long-lived states in brightness experiments requires extensive characterization of the fluorescent protein and complex modeling (Wu et al., 2009).

2.2.2. Short-lived states and flickering Flickering introduces an additional source of intensity fluctuations that complicates brightness analysis. The presence of flickering has been reported for many fluorescent proteins in single-photon excitation experiments (Haupts et al., 1998; Schwille et al., 2000). So far we have not observed flickering in fluorescent proteins when measured under two-photon excitation. For example, Fig. 16.3C shows the correlation function of EGFP excited at 905 nm, which is well described by a simple diffusion model. While one- and twophoton excitation obey different selection rules, the reason for the absence of flickering in two-photon excitation is, as far as we are aware, unknown. This phenomenon, however, significantly simplifies brightness measurements with two-photon excitation, which is the focus of this report. Therefore, the influence of flickering on brightness will not be further discussed. We refer to the literature to provide a starting point for further information on this subject (Palo et al., 2006; Saffarian et al., 2007).

354

Yan Chen et al.

2.3. Fluorescent proteins We have characterized a number of proteins and briefly summarize our findings here. Both EGFP and EYFP are well described by a single brightness state (Chen et al., 2003, 2009). No quenching of fluorescence has been observed. The brightness of a dimer carrying two EGFPs or two EYFPs doubles with an accuracy of better than 10%. We have compiled measurements over several weeks and determined a brightness ratio between dimer and monomer of EGFP as 1.95  0.05. The measured brightness ratio between dimer and monomer for EYFP is 2.02  0.05. We further demonstrated that complexes containing one to four EGFPs are correctly identified by their normalized brightness (Duckworth et al., 2007). The red fluorescent protein mRFP1 contains a bright and a dark state (Hendrix et al., 2008; Hillesheim et al., 2006). This protein is inadequate for brightness experiments because a large population of the protein is in the dark state. Another red fluorescent protein, mCherry, also exists in more than a single brightness state (Wu et al., 2009). We successfully modeled FFS experiments of mCherry using a long-lived bright state and a dim state. Unlike mRFP1, mCherry is suitable for brightness experiments as long as the two states are taken into account in the analysis of the experiment. These examples clearly demonstrate that each new fluorescent protein has to be thoroughly characterized before using it as a brightness label.

3. Brightness Measurements in Cells This section describes the concept of brightness titration experiments in cells and highlights some potential pitfalls. To present the material in a simple manner, we assume throughout this section a well-behaved fluorescent protein with a single brightness state. Brightness values are occasionally quoted as normalized values. The normalized brightness b is defined as the brightness divided by the brightness of the label. Thus the normalized brightness of a monomeric protein is b ¼ 1 and that of a dimeric protein is b ¼ 2. The normalized brightness of an oligomer is identical to its stoichiometry.

3.1. Brightness titration Protein association is a concentration-dependent phenomenon that derives from kinetic and thermodynamic factors. A binding curve of the protein reaction can be constructed by systematically varying the protein concentration, while measuring the brightness of the sample (Chen et al., 2003). We exploit the variations in protein expression from cell to cell in order to probe the exogenously expressed protein over a wide concentration range.

355

Observing Protein Interactions and Their Stoichiometry in Living Cells

An FFS measurement of a single cell expressing the fluorescently labeled protein determines the concentration and brightness of the sample (Chen et al., 2002). After collecting data from many cells, the brightness is graphed versus the protein concentration. This brightness titration plot characterizes the binding curve of a cellular protein in its native environment. An example of a brightness titration curve of testicular orphan receptor 4 (TR4) labeled with EGFP is shown in Fig. 16.5. The normalized brightness of TR4-EGFP at low protein concentrations is close to 1, which indicates a monomeric protein state. The normalized brightness increases with concentration and finally reaches b ¼ 2, indicating the presence of dimeric TR4-EGFP. The normalized brightness values at intermediate concentrations fall in between 1 and 2. These values reflect a mixture of monomeric and dimeric TR4-EGFP. A direct resolution of both protein fractions by brightness is not feasible, because the signal-to-noise ratio of brightness

2.5

b

2.0 1.5 1.0 0.5 100

1000

10,000

cTR4-EGFP (nM)

Monomer bsample = 1

Mixture 1 < bsample < 2

Dimer bsample = 2

Figure 16.5 Brightness titration. The brightness of cells expressing TR4 labeled with EGFP is measured as a function of protein concentration. Each data point represents the normalized brightness b from a different cell. Adapted from Fig. 5 of Chen et al., 2003. The normalized brightness increases from 1 to 2 as a function of TR4-EGFP concentration. A normalized brightness of 2 indicates the presence of dimeric TR4-EGFP, while a brightness of 1 represents monomeric TR4-EGFP. A normalized brightness b between 1 and 2 is indicative of a mixture of monomeric and dimeric TR4-EGFP.

356

Yan Chen et al.

measurements in the cellular environment is too low. Nevertheless, this experimental brightness is very useful, because it represents a value averaged over all brightness species. A normalized brightness close to 1 corresponds to a sample with most proteins in the monomeric state, while a normalized brightness close to 2 indicates a sample where the majority of labeled proteins are dimers. In this sense, the normalized brightness provides an approximate measure of the average degree of oligomerization. The brightness and degree of oligomerization increases with concentration until it plateaus. This saturation of the brightness indicates that the reaction has been driven to completion. Thus, all proteins are assembled in the highest oligomeric state available to the labeled protein. The normalized brightness of the plateau region of the titration graph specifies the stoichiometry of the oligomer, which in the case of Fig. 16.5 corresponds to a dimer.

3.2. Control and calibration experiments Every brightness experiment of a cellular protein requires additional control experiments. We illustrate the procedure for EGFP as the fluorescent label. Cells are seeded into the wells of a tissue culture coverglass. Two of the wells are reserved for control and calibration experiments. The first reserved well is transfected with the fluorescent label EGFP, while the second reserved well is transfected with the dimeric construct EGFP2. The other wells are transfected according to the needs of the experiment. The cells are measured 24–36 h posttransfection. The cell culture coverglass is mounted on the microscope, the well containing the cells expressing EGFP is selected, and several cells are measured in order to determine the intrinsic brightness of the fluorescent label. This calibration experiment establishes the brightness of the fluorophore. The interpretation of all other brightness experiments relies on the underlying brightness of the fluorescent protein. Therefore, a sufficient number of cells need to be measured in order to arrive at an accurate estimation of its value. This calibration must be performed for every experiment, because any change in the experimental setup affects the brightness. The laser power, wavelength, and pulse width are usually not constant on a day-to-day basis. Other factors, such as changing the objective and adjustment of the beam expander, also influence the brightness. It is advisable to measure the brightness of the fluorophore not only at the beginning, but also at the end of the experiment. If the brightness calibrations differ, a drift in instrumental parameters must have occurred during the duration of the experiment. Cells expressing EGFP2 provide a crucial control system for cellular brightness studies. Measuring a normalized brightness for EGFP2 of less than 2 reveals a serious problem for the experiment. To illustrate this point consider the case of incomplete maturation or misfolding of the fluorescent protein. Both events lead to a nonfluorescent protein. A cell that contains a

Observing Protein Interactions and Their Stoichiometry in Living Cells

357

mixture of fluorescent and nonfluorescent proteins is similar to the situation where the fluorophore exists in a bright and dark state (Fig. 16.4). As noted earlier cells that express EGFP alone cannot detect the presence of these darkstate proteins (Fig. 16.4B). However, a dimer that carries a fluorescent and nonfluorescent label only contributes the brightness of a single EGFP molecule. Thus, the presence of such dark states in EGFP2 leads to a normalized brightness of less than 2. We observed in a few instances a lower than normal brightness for EGFP2. These cases also resulted in a reduced brightness of EGFP-fusion proteins that are known to dimerize, such as RXRLBD-EGFP. Data taken under these circumstances are unreliable and should be discarded. A successful control experiment requires that the brightness of EGFP2 agrees within experimental uncertainty with the nominal value. We noticed that COS-1 cells have been the most troublesome among the cell lines measured so far. However, while transiently transfected COS-1 cells have given us less than double brightness for EGFP2 on a number of occasions, stably transfected COS-1 cells greatly reduce this problem. Because the properties of each cell line are different, the suitability of each for brightness experiments has to be evaluated on an individual basis. The experiments rely on a labeled protein that is faithfully expressed by the cell. Measuring the brightness of EGFP2 provides an important assay for probing the fidelity of cellular expression. Another useful check of the expressed protein utilizes Western blotting of the cell lysate with an anti-GFP antibody. The labeled protein should appear as a single band in the gel at the position that corresponds to its expected molecular weight. Figure 16.6 shows a Western blot gel of EYFP2 and EYFP. Both proteins appear as single bands at positions that correspond to their molecular weight. Cells with expression levels ranging from low to high are selected for the brightness titration experiment of the EGFP-labeled protein. It is imperative to measure the control cells over the same concentration range. The brightness of EGFP and EGFP2 must be stable and constant over the whole concentration range (Fig. 16.7). Observation of a concentration-dependent brightness unmistakably signals the presence of a bias in the experiment. For example, the detector introduces deadtime and afterpulsing artifacts that, if not properly corrected, lead to a concentration-dependent bias of the brightness (Hillesheim and Muller, 2005, 2003). Furthermore, fluorescence background can artificially reduce the brightness at low expression levels (Chen et al., 2002). Such concentration-dependent biases can be very hard to spot in a brightness titration curve, which by itself already varies with concentration.

3.3. Cell selection and measurement The first step of a brightness experiment is the selection of a suitable cell. The microscope is switched to epifluorescence mode and the stage is moved until a successfully transfected cell is found. Cells that look stressed are discarded.

358

Yan Chen et al.

MW (kDa) 100 75 50 37

25

EYFP EYFP2

Figure 16.6 Western blot gel. EYFP and EYFP2 are individually expressed in COS-1 cells. Cells were collected and lysed in the presence of protease inhibitor cocktail (Sigma) 36 h after transfection. Proteins were denatured in the NuPAGE LDS Sample buffer at 70  C for 10 min and electrophoresed in NuPAGE 4–12% Bis Tris SDS–PAGE, transferred onto nitrocellulose membrane, immunoblotted with anti-GFP antibody (Upstate, Millipore), and detected with TMB using One-Step WesternTM Complete Kit (Genescript, Piscataway, NJ). The two lanes each show a single band at the expected position (EYFP: 27 kDa, EYFP2: 54 kDa).

While epifluorescence is excellent for identifying transfected cells, the morphology of a cell is more accurately assessed by bright-field microscopy. Thus, the microscope is switched to bright-field imaging to judge the morphology of the cell. Only cells with a morphology that matches that of untransfected, healthy cells are selected for brightness measurements. The wide-field image is recorded by a CCD camera and displayed on a monitor. The monitor screen contains a mark that specifies the position of the two-photon spot. The microscope stage is adjusted until the cell location to be measured is aligned with the mark. The microscope is switched to two-photon FFS mode and the focus is adjusted to the desired height. Most measurements are conducted with the focus of the excitation beam at mid-height of the cell. We typically collect data for 30–60 s, which is sufficient for a determination of the brightness of EGFP-labeled proteins. We do not attempt to cover data analysis of FFS experiments, because the diversity of methods and their subtle

Observing Protein Interactions and Their Stoichiometry in Living Cells

359

2.5 2.0

b

1.5 1.0 0.5

EGFP EGFP2

0.0 100

1000 c(EGFP) [nM]

Figure 16.7 Normalized brightness of EGFP (squares) and EGFP2 (circles) as a function of protein concentration. Each data point represents the brightness measured in a different cell expressing either EGFP or EGFP2. The average normalized brightness of EGFP is 1  0.058, while the average normalized brightness of EGFP2 is 1.97  0.098.

but important differences requires a separate treatment. Most available analysis methods either use photon count distributions (Chen et al., 1999; Kask et al., 1999), photon count moments (Mu¨ller, 2004; Qian and Elson, 1990b), or correlation functions (Palmer and Thompson, 1989) to extract brightness information. Another important consideration of data analysis is the sampling time and its influence on the photon count statistics (Palo et al., 2000; Wu and Mu¨ller, 2005). Care should be exercised to avoid photobleaching during epifluorescence imaging. The power of the excitation light should be reduced as far as feasible and the exposure of each cell should be kept to a minimum. In our experience, the exposure of a cell can be reduced to a few seconds. Prolonged exposure of a cell to the excitation light leads to photobleaching, which may severely bias brightness experiments. Bleaching creates a population of nonfluorescent molecules (Fig. 16.8). As previously indicated, a population of nonfluorescent fluorophores has the same effect on brightness as a fluorophore with bright and dark states. Cells expressing EGFP are not suitable for testing the presence of photobleaching effects on brightness, because bleaching only reduces the concentration of the fluorescent population without changing its brightness. Bleaching of EGFP2, on the other hand, leads to molecules where none, one, or both fluorophores are bleached. This situation is analogous to the dimer with dark states and results in a reduction of the brightness. Thus, cells expressing EGFP2 are a good test system for detecting unacceptable levels of photobleaching.

360

Yan Chen et al.

Bleaching l=l

l=0

Monomer

Initial sample

Bleached sample Bleaching

Light source

lS = l

Dimer

lS = l

lS = 2l

l < lS < 2l

Figure 16.8 Brightness and bleaching. Photobleaching leads to nonfluorescent molecules (l ¼ 0). The introduction of nonfluorescent molecules by photobleaching leads to effects that are analogous to the presence of long-lived dark states. Photobleaching reduces the observable protein population, but preserves the brightness of a monomeric protein sample. In contrast, photobleaching reduces the brightness of dimeric and higher order protein complexes.

Any instrument for fluorescence correlation spectroscopy is also suitable for brightness experiments. In general, the signal-to-noise ratio of cellular FFS measurements is quite low even after optimizing the optics of the instrument. Increasing the excitation power is the only remaining option for improving the signal. However, when the power is raised beyond a certain level, saturation and photobleaching of the fluorophore kicks in (Cianci et al., 2004). These undesirable effects need to be avoided. A power study, which measures the brightness as a function of excitation power, is performed in an EGFP-expressing cell. A graph of such a study shows a quadratic power-dependence of the brightness, but starts to be subquadratic as the power increases. The power where the crossover from the quadratic to subquadratic dependence occurs is chosen as the upper limit. All FFS experiments are performed at excitation powers that do not exceed this upper limit. Finally, we like to add a comment about the cell thickness. The majority of FFS experiments conducted by our group are performed in the cell nucleus. The thickness of the cell at the nucleus is 4 m for the cell lines measured. This thickness is sufficient to have the whole two-photon

Observing Protein Interactions and Their Stoichiometry in Living Cells

361

observation volume enclosed within the cell. This situation is described by the standard FFS models. However, when FFS measurements are performed in a thin cytoplasmic section, the observation volume extends beyond the thickness of the cell. Because the fluorescent proteins cannot access the entirety of the observation volume, analysis of the fluctuations by the standard models recovers an incorrect brightness. The erroneous brightness is always larger than the correct value. We have observed a brightness increase of almost a factor of two for very thin cytoplasmic sections. One would mistakenly conclude that the protein dimerized, if the thickness dependence of the brightness is not recognized. One approach to counter this problem is to perform all measurements, including the brightness calibrations, at the same cell thickness. There are many additional topics, such as the influence of endogenous protein on brightness experiments, which are relevant, but cannot be covered within this chapter. Dual-color brightness methods and protein heterointeractions are of great interest, but are omitted because these topics require a lengthy discussion (Chen and Mu¨ller, 2007). We instead opted to focus on a few fundamental issues regarding single-color brightness measurements of protein homointeractions in cells. Although there are many additional details to be considered in dual-color brightness experiments of protein heterointeractions, the principles and experimental considerations described in this manuscript remain equally valid. Thus, we hope this chapter serves as a valuable starting point for anyone interested in brightness experiments inside cells.

ACKNOWLEDGMENTS This work was supported by grants from the National Institutes of Health (GM64589), the National Science Foundation (PHY-0346782), and the American Heart Association (0655627Z).

REFERENCES Berland, K. M., So, P. T. C., and Gratton, E. (1995). Two-photon fluorescence correlation spectroscopy: Method and application to the intracellular environment. Biophys. J. 68, 694–701. Berland, K. M., So, P. T., Chen, Y., Mantulin, W. W., and Gratton, E. (1996). Scanning two-photon fluctuation correlation spectroscopy: Particle counting measurements for detection of molecular aggregation. Biophys. J. 71, 410–420. Chen, Y., and Mu¨ller, J. D. (2007). Determining the stoichiometry of protein heterocomplexes in living cells with fluorescence fluctuation spectroscopy. Proc. Natl. Acad. Sci. USA 104, 3147–3152. Chen, Y., Mu¨ller, J. D., So, P. T. C., and Gratton, E. (1999). The photon counting histogram in fluorescence fluctuation spectroscopy. Biophys. J. 77, 553–567.

362

Yan Chen et al.

Chen, Y., Mu¨ller, J. D., Ruan, Q., and Gratton, E. (2002). Molecular brightness characterization of EGFP in vivo by fluorescence fluctuation spectroscopy. Biophys. J. 82, 133–144. Chen, Y., Wei, L. N., and Mu¨ller, J. D. (2003). Probing protein oligomerization in living cells with fluorescence fluctuation spectroscopy. Proc. Natl. Acad. Sci. USA 100, 15492–15497. Chen, Y., Wu, B., Musier-Forsyth, K., Mansky, L. M., and Mueller, J. D. (2009). Fluorescence fluctuation spectroscopy on viral-like particles reveals variable gag stoichiometry. Biophys. J. 96, 1961–1969. Cianci, G. C., Wu, J., and Berland, K. M. (2004). Saturation modified point spread functions in two-photon microscopy. Microsc. Res. Tech. 64, 135–141. Denk, W., Strickler, J. H., and Webb, W. W. (1990). Two-photon laser scanning fluorescence microscopy. Science 248, 73–76. Digman, M. A., Brown, C. M., Sengupta, P., Wiseman, P. W., Horwitz, A. R., and Gratton, E. (2005). Measuring fast dynamics in solutions and cells with a laser scanning microscope. Biophys. J. 89, 1317–1327. Duckworth, B. P., Chen, Y., Sham, Y., Mueller, J. D., Taton, T. A., and Distefano, M. D. (2007). A universal method for the preparation of covalent protein-DNA conjugates for use in creating protein nanostructures. Angew. Chem. Int. Ed. 46, 8819–8822. Egea, P. F., Rochel, N., Birck, C., Vachette, P., Timmins, P. A., and Moras, D. (2001). Effects of ligand binding on the association properties and conformation in solution of retinoic acid receptors RXR and RAR. J. Mol. Biol. 307, 557–576. Enderlein, J., and Gregor, I. (2005). Using fluorescence lifetime for discriminating detector afterpulsing in fluorescence-correlation spectroscopy. Rev. Sci. Instrum. 76, 033102–033105. Gautier, I., Tramier, M., Durieux, C., Coppey, J., Pansu, R. B., Nicolas, J. C., Kemnitz, K., and Coppey-Moisan, M. (2001). Homo-FRET microscopy in living cells to measure monomer–dimer transition of GFP-tagged proteins. Biophys. J. 80, 3000–3008. Haupts, U., Maiti, S., Schwille, P., and Webb, W. W. (1998). Dynamics of fluorescence fluctuations in green fluorescent protein observed by fluorescence correlation spectroscopy. Proc. Natl. Acad. Sci. USA 95, 13573–13578. Hendrix, J., Flors, C., Dedecker, P., Hofkens, J., and Engelborghs, Y. (2008). Dark states in monomeric red fluorescent proteins studied by fluorescence correlation and single molecule spectroscopy. Biophys. J. 94, 4103–4113. Hillesheim, L. N., and Mu¨ller, J. D. (2003). The photon counting histogram in fluorescence fluctuation spectroscopy with non-ideal photodetectors. Biophys. J. 85, 1948–1958. Hillesheim, L. N., and Muller, J. D. (2005). The dual-color photon counting histogram with non-ideal photodetectors. Biophys. J. 89, 3491–3507. Hillesheim, L. N., Chen, Y., and Mueller, J. D. (2006). Dual-color photon counting histogram analysis of mRFP1 and EGFP in living cells. Biophys. J. 91, 4273–4284. Kask, P., Palo, K., Ullmann, D., and Gall, K. (1999). Fluorescence-intensity distribution analysis and its application in biomolecular detection technology. Proc. Natl. Acad. Sci. USA 96, 13756–13761. Lakowicz, J. R. (2006). Principles of Fluorescence Spectroscopy. Springer, New York, Berlin. Magde, D., Elson, E., and Webb, W. W. (1972). Thermodynamic fluctuations in a reacting system: Measurement by fluorescence correlation spectroscopy. Phys. Rev. Lett. 29, 705–708. Mu¨ller, J. D. (2004). Cumulant analysis in fluorescence fluctuation spectroscopy. Biophys. J. 86, 3981–3992. Palmer, A. G. D., and Thompson, N. L. (1989). High-order fluorescence fluctuation analysis of model protein clusters. Proc. Natl. Acad. Sci. USA 86, 6148–6152.

Observing Protein Interactions and Their Stoichiometry in Living Cells

363

Palo, K., Mets, U., Jager, S., Kask, P., and Gall, K. (2000). Fluorescence intensity multiple distributions analysis: Concurrent determination of diffusion times and molecular brightness. Biophys. J. 79, 2858–2866. Palo, K., Mets, U., Loorits, V., and Kask, P. (2006). Calculation of photon-count number distributions via master equations. Biophys. J. 90, 2179–2191. Qian, H., and Elson, E. L. (1990a). Distribution of molecular aggregation by analysis of fluctuation moments. Proc. Natl. Acad. Sci. USA 87, 5479–5483. Qian, H., and Elson, E. L. (1990b). On the analysis of high order moments of fluorescence fluctuations. Biophys. J. 57, 375–380. Saffarian, S., Li, Y., Elson, E. L., and Pike, L. J. (2007). Oligomerization of the EGF receptor investigated by live cell fluorescence intensity distribution analysis. Biophys. J. 93, 1021–1031. Schwille, P., Kummer, S., Heikal, A. A., Moerner, W. E., and Webb, W. W. (2000). Fluorescence correlation spectroscopy reveals fast optical excitation-driven intramolecular dynamics of yellow fluorescent proteins. Proc. Natl. Acad. Sci. USA 97, 151–156. Thompson, N. L., Lieto, A. M., and Allen, N. W. (2002). Recent advances in fluorescence correlation spectroscopy. Curr. Opin. Struct. Biol. 12, 634–641. Volkmer, A., Subramaniam, V., Birch, D. J., and Jovin, T. M. (2000). One- and twophoton excited fluorescence lifetimes and anisotropy decays of green fluorescent proteins. Biophys. J. 78, 1589–1598. ¨ . (1994). Triplet-state monitoring by fluorescence Widengren, J., Rigler, R., and Mets, U correlation spectroscopy. J. Fluoresc. 4, 255–258. Wiseman, P. W., and Petersen, N. O. (1999). Image correlation spectroscopy. II. Optimization for ultrasensitive detection of preexisting platelet-derived growth factor-beta receptor oligomers on intact cells. Biophys. J. 76, 963–977. Wu, B., and Mu¨ller, J. D. (2005). Time-integrated fluorescence cumulant analysis in fluorescence fluctuation spectroscopy. Biophys. J. 89, 2721–2735. Wu, B., Chen, Y., and Muller, J. D. (2009). Fluorescence fluctuation spectroscopy of mCherry in living cells. Biophys. J. 96, 2391–2404.

C H A P T E R

S E V E N T E E N

Detection of Individual Endogenous RNA Transcripts In Situ Using Multiple Singly Labeled Probes Arjun Raj* and Sanjay Tyagi† Contents 1. Introduction 2. Design and Synthesis of Fluorescent Oligonucleotide Probe Sets 2.1. Design 2.2. Synthesis and purification 3. Preparation of Samples for In Situ Hybridization 3.1. Fixation solutions 3.2. Fixation protocols 4. Hybridization 4.1. Hybridization solutions 4.2. Hybridization protocols 5. Imaging 5.1. Microscopy equipment 6. Image Analysis Acknowledgments References

366 369 369 369 373 373 375 377 378 379 381 381 384 386 386

Abstract Measurements of gene expression within single cells have revealed startling variability otherwise hidden in bulk measurements. Here, we present an in situ hybridization method capable of detecting individual mRNA molecules, thus permitting the accurate quantification and localization of mRNA within fixed sample. Our in situ protocol involves probing the target mRNA using a series of singly labeled oligonucleotide probes. This method is simple to implement and is applicable to a variety of biological samples. We also discuss some aspects of image processing required for analyzing the resulting data.

* Department of Bioengineering, University of Pennsylvania, Philadelphia, Pennsylvania, USA Public Health Research Institute, New Jersey Medical School-UMDNJ, Newark, New Jersey, USA

{

Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72004-8

#

2010 Elsevier Inc. All rights reserved.

365

366

Arjun Raj and Sanjay Tyagi

1. Introduction Single cell measurements have revealed that gene expression in individual cells can deviate significantly from the average behavior of cell populations, with significant biological consequences (Larson et al., 2009; Maheshri and O’Shea, 2007; Raj and Van Oudenaarden, 2008, 2009; ). These findings have created a need for accurate methods of quantifying expression in single cells, ideally even yielding intracellular spatial information about the localization of mRNAs. A natural candidate for such a method is in situ hybridization, in which labeled nucleotide probes find their specific targets through Watson–Crick base pairing (Levsky and Singer, 2003). Initially, researchers performed in situ hybridizations using radioactive probes (Gall, 1968). Early improvements involved linking the probes to enzymes that catalyze chromogenic or fluorogenic reactions (Raap et al., 1995; Tautz and Pfeifle, 1989). Unfortunately, these reactions generated molecules that diffused away from the probe itself, making it difficult to ascertain the precise spatial location of the target. Alternatively, one could use fluorescently labeled probes, thus sidestepping the issues of localization, but the sensitivity of such methods was relatively poor. Singer and colleagues then developed an in situ hybridization procedure that was both sensitive enough to permit the detection of single mRNA molecules, but also restricted the fluorescence to reside close to the target (Femino et al., 1998). Their method involved the use of five 50mer DNA oligonucleotides, each of which was conjugated to five fluorophore moieties. The authors convincingly demonstrated single molecule sensitivity, with spots corresponding to individual mRNA molecules, and their method has seen subsequent use (Maamar et al., 2007; Zenklusen et al., 2008). However, they estimated that over 30% of the transcripts hybridized to either zero (5%) or to just one (25%) of the oligonucleotide probes (Table 1 of supplementary information in Femino et al. (1998)). This lack of coupling efficiency is worrisome because the detection of just a single probe cannot discriminate between legitimate binding to the target and nonspecific binding. Another issue with this method is that the probes are generally difficult to generate: it is difficult to efficiently label DNA oligonucleotides with multiple fluors, and each probe must be synthesized and purified individually. Our method involves probing target mRNAs using a larger number (> 30) of shorter oligonucleotides (20 bases), each of which hybridize to a different portion of the target mRNA (Fig. 17.1). We label each of these oligonucleotides with a single fluorophore at its 30 end; thus, upon hybridization, a large number of fluors are all brought within close proximity of

Detection of Individual Endogenous RNA Transcripts

367

30-50 probes labeled on 3⬘ end Positions selected to optimize GC content

Open reading frame of target mRNA

Figure 17.1 Depiction of scheme for imaging individual mRNA molecules using singly labeled oligonucleotide probes. The placement of the probes is often nonuniform in order to maintain an optimal GC content for all probes, thus matching hybridization conditions between probes. See Raj et al. (2008) for all sequences used in this chapter.

the target. The presence of so many fluors in a single location results in enough fluorescence that the spot can be made out as a diffraction-limited spot in a widefield fluorescence microscope. Our method achieves its specificity and sensitivity owing to the large number of probes used. The rate of false negatives is low because even if the target RNA molecule has been partly degraded or is partly obscured by RNA binding proteins, at least some fraction of the probes will still bind to it, yielding a detectable signal. The rate of false positives is also low because one only detects a signal when a significant fraction of the probes are bound. Thus, off-target binding of individual probes will not yield much signal above background. Such false positives are particularly a concern in other methods consisting of the hybridization of a single probe followed by an enzymatic signal amplification; in such cases, it is impossible to distinguish a single nonspecific binding event from a legitimate interaction. Moreover, our method is straightforward and easy to implement, owing to the simplicity of the chemistries involved. Advances in oligonucleotide synthesis make it cost effective to purchase large numbers of singly functionalized probes, and the labeling procedure can be performed on a pooled set of these oligonucleotides, greatly reducing the effort required. Another feature of our method is that it can be combined with other methods like DNA FISH (Vargas et al., 2005) (Arjun Raj unpublished observations) and immunofluorescence (Raj et al., 2008). Furthermore, the fact that we see diffraction-limited spots allows us to precisely determine the location of the center of the spot beyond the optical diffraction limit (Yildiz et al., 2003), technically making our method a version of super-resolution imaging. One question that often arises is how we know that each fluorescent spot represents a single RNA molecule rather than a conglomeration of multiple target mRNAs. We have several pieces of evidence supporting the conclusion that each spot corresponds to a single RNA, mostly outlined in Vargas et al. (2005). In one experiment, we synthesized mRNAs prehybridized with two different fluorophores (effectively, we made ‘‘red’’ and ‘‘green’’ mRNAs). We then injected a mixture of these red

368

Arjun Raj and Sanjay Tyagi

and green mRNAs into single cells. If the mRNAs were in fact clumping, the spots would have contained both red and green mRNAs and would thus appear yellow. What we found, however, was that each spot was either red or green but never both, showing that each spot corresponded to individual red or green mRNA molecules. Moreover, the signal intensities from these synthetic mRNAs were roughly identical to those from endogenously transcribed mRNAs, indicating that endogenously transcribed mRNAs also do not form clumps. It is possible, of course, that certain other mRNAs form clumps, and a formal proof requires this procedure be followed for each particular mRNA under study, but at least this case is consistent with the null hypothesis that mRNAs do not form conglomerates. Another piece of evidence comes from examining the distribution of spot intensities (Raj et al., 2008; Vargas et al., 2005). If some of the spots were conglomerates of small numbers of mRNA, one would expect that some spots would consist of one mRNA, some of two, and so on. The distribution of spot intensities would then show multiple peaks, as observed with MS2 binding-site-tagged mRNAs, which are know to clump (Golding et al., 2005). Instead, we always see a single peak (Raj et al., 2008; Vargas et al., 2005), consistent with each spot representing single mRNA molecules. We also compared mRNA counts obtained with our method (specifically, average number of mRNA per cell) to those obtained by quantitative RT-PCR and found that the results compared favorably, coming within 30% of each other. These measurements not only bolster our claim to be detecting individual molecules, but also show that our fixation procedure does not result in the loss of a significant fraction of the mRNAs in the sample. Another potential issue could be the occlusion of the target RNA by various RNA-binding proteins; for example, cytoplasmic mRNAs being occluded by ribosomes. We doubt this factor is significant, though, partly because of the quantitative RT-PCR experiments described above. Also, we simultaneously labeled both the open-reading frame and the 30 UTR (upon which there should be no ribosomes) simultaneously but with differently colored probes, and we found a high degree of colocalization (80%), showing that at least ribosome binding is not a significant impediment to RNA detection. In this chapter, we describe the procedures involved in detecting individual RNA molecules in situ. These are (1) designing and synthesizing the fluorescently labeled oligonucleotides, (2) fixation of the biological specimen, (3) hybridization, (4) imaging on a fluorescence microscope, and (5) data analysis. None of these steps utilize any exotic chemicals, procedures or equipment, and we will indicate as needed any aspects of the application of our method that require any special attention.

Detection of Individual Endogenous RNA Transcripts

369

2. Design and Synthesis of Fluorescent Oligonucleotide Probe Sets 2.1. Design Our method involves the synthesis of a set of fluorescently labeled oligonucleotides (which we call the ‘‘probe set’’) that will hybridize along the length of target RNA molecule. There are a few general guidelines we typically follow when designing these oligonucleotides. Firstly, the probe sets typically consist of anywhere between 30 and 96 (typically 48) different 20mer DNA oligonucleotides, each complementary to a different region of the target RNA, with no less than two bases separating any two oligonucleotides (Fig. 17.1). We have found that one can sometimes obtain signals with less than 30 oligonucleotides, but the signals are often fainter. Forty-eight probes appear to be sufficient to generate a robust signal in most instances, and many synthesis companies sell parallel orders of oligonucleotides in batches of 48, which is why that is the default number of oligonucleotides we utilize in our probe sets. Another issue is that of the GC content of the individual oligonucleotides. Given that GC content can significantly alter the hybridization parameters, we consider it desirable to make the GC contents of the various oligonucleotides as uniform as possible, thus ensuring that as many probes as possible will bind at a given hybridization stringency. In order to design such probe sets, we have deployed a web-based program (http://www. singlemoleculefish.com) that, given a target RNA sequence, a desired number of probes and a target GC percentage, will generate a set of oligonucleotides whose GC contents are as uniform as possible. Of course, for shorter target RNAs, there is a tradeoff between the number of probes one can generate and the GC uniformity of those probes, but we have not systematically studied these effects. Anecdotally, we find that beyond around 35 probes, optimizing the GC content of the probes is probably more useful than squeezing more probes onto the target mRNA. As reference points, we note that we have seen decent signals using as few as 20 probes and excellent signals using just 30 probes, meaning that one can detect mRNAs as short as 500 bases.

2.2. Synthesis and purification Once the oligonucleotide sequences are generated, we order the oligonucleotides synthesized with a 30 amine group, which we use for coupling the fluorophore. The oligonucleotides we order are desalted and resuspended in water rather than TE, since Tris can interfere with subsequent aminecoupling reactions. Since the amount of oligonucleotide used for each hybridization is typically very small, one should have the oligonucleotides

370

Arjun Raj and Sanjay Tyagi

synthesized on the smallest scale possible. We order our oligonucleotides from Biosearch Technologies (Novato, CA) at a scale of 10 nmol per oligo, which are delivered to us in 100 ml of water. We have found that other means of labeling oligonucleotides (especially internal amino-dTs) have far lower coupling efficiencies. The next step is to couple the oligonucleotides to the desired fluorophore. We utilize succinimidyl ester derivatives to couple to the amine group at the 30 end of the oligonucleotides (we will discuss the choice of fluorophore shortly). Rather than coupling and purifying each oligonucleotide individually, we instead couple and purify the oligonucleotides en masse via reverse phase HPLC, significantly reducing the labor involved: 1. Combine the uncoupled oligonucleotides by pooling 1 nmole of each oligonucleotide (10 ml in our case) together. 2. Add enough volume of 1 M sodium bicarbonate (pH 8.0) so that the oligonucleotide pool contains 0.1 M sodium bicarbonate. 3. Meanwhile, dissolve the fluorophore in 50 ml of 0.1 M sodium bicarbonate (pH 8.0). Note that some fluorophores, such as tetramethylrhodamine (TMR), are more readily soluble in organic solvents; we first dissolve those fluorophores in around 5 ml of DMSO and then add 50 ml of 0.1 M sodium bicarbonate. 4. Add the fluorophore solution to the oligonucleotide solution. 5. Let the reaction sit in the dark overnight at room temperature. At this point, the tube will contain uncoupled fluorophore, uncoupled oligonucleotides, and coupled oligonucleotides. In order to remove the uncoupled fluorophores, we perform an ethanol (EtOH) precipitation: 6. Add 0.13 vol of 3 M sodium acetate (pH 5.2) and 2.5 vol EtOH to the reaction. 7. Store at 80  C for at least 1 h. 8. Spin in a 4- C microcentrifuge at maximum speed for 15 min. A colored pellet containing the coupled and uncoupled oligonucleotides should form at the bottom of the tube. 9. Carefully pipette off as much of the supernatant as possible. This supernatant contains the uncoupled fluorophore. In order to separate the uncoupled oligonucleotides from the coupled oligonucleotides, we purify the oligonucleotides by HPLC. The typically hydrophobic organic fluorophores cause a large increase in hydrophobicity of the coupled oligonucleotides as compared to the rather hydrophilic uncoupled oligonucleotides. The size of this increase is much larger than the variation in the hydrophobicity of the individual oligonucleotides, thus enabling us to purify the entire pool of oligonucleotides at once. This procedure requires an HPLC equipped with a C18 column (C18TP104) and a dual wavelength detector (or diode array detector) set to

Detection of Individual Endogenous RNA Transcripts

371

detect DNA absorption (260 nm) as well as the absorption of the coupled fluorophore (e.g., 555 nm for TMR). We have found it easiest to collect the desired fractions manually rather than automatically from the outflow. For our gradient, we use 0.1 M Triethyl ammonium acetate, pH 6.5 (Buffer A) and acetonitrile, pH 6.5 (Buffer B), ranging from 7% to 30% Buffer B over the course of 30 min at a flow rate of 1 mL per min (after this, be sure to run the column at 70% Buffer B for 10 min in order to clear the column of extraneous molecules and then equilibrate the column at 7% Buffer B for 10 min before running another sample). The specific gradient may depend on the exact nature of your HPLC setup, but should be at least similar to that we describe here. While the gradient is running, continuously monitor the absorption in the 260 and 555 nm channels. Initially, you will see a large set of peaks at 260 nm while the 555 nm absorption remains low. This peak contains the uncoupled oligonucleotide. After that peak passes, you will observe another set of peaks in which there is large absorption in both the 260 and 555 nm channels (Fig. 17.2). This peak contains the coupled oligonucleotides. Collect this fraction as it passes through the HPLC (the total volume collected will typically be between 2 and 6 mL). Be sure to collect the entire peak rather than just the ‘‘top.’’ This is important, because different parts of the peak will contain different oligonucleotides. It also bears mentioning that even labeled oligonucleotide can generate multiple peaks due to incomplete deprotection or dye-induced chemical variation. We do not think, though, that these issues lead to any serious problems in oligonucleotide purity. Once the fraction is collected, dry the samples in a lyophilizer or a speedvac rated for use with acetonitrile, then resuspend the fractions in 50–100 ml of Tris EDTA (TE), pH 8.0. This is the stock of your probe (concentration of roughly 0.1–1 mM) from which you can make working dilutions (1:10, 1:20, 1:50, 1:100) for your hybridizations. As for the choice of fluorophore, the ones we commonly use are TMR (Molecular Probes, Invitrogen), Alexa 594 (Molecular Probes, Invitrogen), and Cy5 (GE Amersham). Using appropriate filter sets (Table 17.1), we are able to independently image these three colors reliably in most samples we have examined with no bleedthrough between channels (Fig. 17.3). While it is possible that one can use fluorophores that absorb and emit at even shorter wavelengths (e.g., Alexa 488), we have found that background autofluorescence at these wavelengths is usually strong enough that it is difficult to make out the signals (although we have had success with Alexa 488 on occasion). Even the signals from TMR and to some extent Alexa 594 are sometimes marred by autofluorescent blobs that make the particles hard to distinguish. In such situations, one can get some idea of whether or not the signals are real by taking pictures of the sample using GFP/fluorescein filters—background autofluorescence typically has a broad emission spectrum and will often show up in multiple channels, whereas the organic dyes will not appear in the GFP channel. Reducing this background is often

372

Arjun Raj and Sanjay Tyagi

1200

260 555

TMR

800

400

0

1200

Alexa 594

Absorbance (arbitrary units)

15

20

25

30

35

40

45

25

30

35

40

45

30 35 Time (min)

40

45

260 594

800

400

0 15 1200

20

260 650

Cy5

800

400

0 15

20

25

Figure 17.2 Sample HPLC chromatographs showing absorbance at 260 nm (blue) and 555, 594, and 650 nm (red; TMR, Alexa 594, and Cy5, respectively). The first 260 nm peak is the uncoupled oligonucleotides. The next peak appears in both the 260 nm and fluorophore absorbance channels, indicating that this is the coupled oligonucleotides. Collect the entire fraction between the gray dotted lines.

373

Detection of Individual Endogenous RNA Transcripts

Table 17.1 Optical filters for multiplex mRNA detection

TMR Alexa 594 Cy5

Excitation

Dichroic

Emission

Supplier

546DF10 590DF10 HQ620/60x

555DRLP 610DRLP Q660LP

580DF30 630DF30 HQ700/75m

Omega Omega Chroma

Details of the excitation, dichroic, and emission filters used for multiplex detection with TMR, Alexa 594, and Cy5. The nomenclature used is specific to the suppliers listed on the right (Omega Optical, Chroma). The first and second numbers refer to the center and width of the bandpass region, respectively.

difficult and is very sample dependent, with some cell lines and tissues exhibiting high levels of background, often related to cellular stress. Another factor to consider is the fact that Alexa 594 and especially TMR are fairly photostable and thus require no special care to be taken about photobleaching when imaging. Cy5, on the other hand, is notorious for being rapidly photobleached. In order to combat this, we use a glucose-oxidase (glox)-based oxygen scavenging mounting medium (described later in this chapter; adapted from Yildiz et al., 2003), which reduces the photobleaching rate of Cy5 to levels comparable to that of TMR. Given the low autofluorescent background at these far red wavelengths, Cy5 is an excellent choice for samples in which reduction in cellular autofluorescence is impossible.

3. Preparation of Samples for In Situ Hybridization In this section, we outline the procedures for fixation and permeabilization of various biological samples for use in in situ hybridization. These protocols are based on the protocols developed in the lab of Robert Singer (Femino et al., 1998; http://www.singerlab.org/protocols). While the specifics may change slightly from organism to organism, the fundamental procedure is roughly the same in all cases: fix the sample in 3.7% (v/v) formaldehyde (i.e., 10% formalin) in 1 phosphate buffered saline (PBS), then permeabilize in 70% ethanol, at which point samples can be stored at 4  C for weeks (even months) without degradation. Note that all solutions used postfixation should be made with nuclease-free water.

3.1. Fixation solutions 3.1.1. Fixation solution (3.7% formaldehyde/10% formalin, 1 PBS) 40 mL RNase free H2O (Ambion) 5 mL 37% (v/v) formaldehyde (100% formalin) 5 mL 10 PBS (RNase free, Ambion)

374

Arjun Raj and Sanjay Tyagi

A

B

C FAM105A mRNA particle

Channel:

TMR

A594

Cy5

D PTGS2 mRNA particle

TMR

A594

Cy5

FKBP5 mRNA particle TMR

A594

Cy5

0 200 0 200 0 200

Increasing Z

Fluorescence (A.U.)

200

0 200 0

Figure 17.3 Demonstration of three color mRNA detection. (A) Expression of FKBP5 (blue), PTGS2 (purple), and FAM105A (yellow) mRNAs in human carcinoma cell line A549. Scale bar is 5 mm long. (B–D) Examination of fluorescent spot bleedthrough. (B) Images of an FAM105A mRNA spot labeled with TMR as seen through the TMR, Alexa 594, and Cy5 filter channels. Linescans of fluorescent intensity corresponding to the line through the image are given below, with the different linescans corresponding to measurements taken at increasing z (0.25 mm spacing). The green linescan corresponds to the z-slice shown in the image itself. A similar analysis was performed for a PTGS2 mRNA spot labeled with Alexa 594 (C) and an FKBP5 mRNA particle labeled with Cy5 (D). All linescan intensity measurements had the camera background subtracted but range between 0 and 200 arbitrary fluorescence units.

3.1.2. Buffer B (1.2 M sorbitol, 0.1 M Potassium phosphate) 218 g sorbitol 17.4 g Potassium phosphate (dibasic) RNase free water to 1 L

Detection of Individual Endogenous RNA Transcripts

375

3.1.3. Spheroplasting buffer 10 mL Buffer B 100 ml 200 mM vanadyl ribonucleoside complex (New England Biolabs) 3.1.4. M9 5.8 g Na2HPO4 3.0 g KH2PO4 0.5 g NaCl 1.0 g NH4Cl Double-distilled ddH2O (ddH2O) to 1 L

3.2. Fixation protocols 3.2.1. Fixation of yeast cells 1. Grow yeast to an optical density (OD, at 260 nm) of around 0.1–0.2 in a 45-mL volume of minimal media. 2. Add 5-mL of 37% (v/v) formaldehyde directly to growth media and let sit for 45 min. 3. Wash 2 twice with 10 mL ice-cold Buffer B. 4. Add 1 mL of spheroplasting buffer, transferring to a microcentrifuge tube. 5. Add 1 ml of zymolyase and incubate at 30  C for 15 min. 6. Wash 2 twice with 1 mL ice-cold Buffer B, spinning at low speed (2000 rpm). 7. Add 1 mL of 70% (v/v) EtOH and leave at least overnight at 4  C. 3.2.2. Fixation of adherent mammalian cells 1. Grow cells on #1 coverglasses set in six-well culture dishes or in LabTek chambered coverglass (with #1 coverglass on the bottom; we have had bad results with #1.5 coverglass). 2. Aspirate growth medium. 3. Wash with 1 PBS. 4. Add fixation solution and incubate at room temperature for 10 min. 5. Wash 2 twice with 1 PBS. 6. Add 70% (v/v) EtOH and store at 4  C at least overnight. 3.2.3. Fixation of Caenorhabditis elegans larvae (L1–L4) 1. Grow larvae in a plate seeded with OP50. 2. Add 5 mL M9 buffer and swirl in plate to release worms from surface, then move worms to a 15 mL conical centrifuge tube.

376

Arjun Raj and Sanjay Tyagi

We often use DI water instead of M9 in this step and in step 4 and get fine results. Spin down worms and aspirate. Wash with 5 mL M9 buffer. Spin down worms and aspirate. Add 1 mL fixation solution, transfer to microcentrifuge tube, and incubate for 45 min. 7. Wash 2 with 1 mL 1 PBS. 8. Resuspend in 1 mL of 70% EtOH and leave for at least overnight at 4  C. We have sporadic reports that longer incubations at 4  C in EtOH (i.e., 5 days) can reduce autofluorescence, but we do not think it really matters. 3. 4. 5. 6.

3.2.4. Fixation of C. elegans embryos 1. Add 5 mL M9 buffer to a plate of gravid hermaphrodites and swirl to release worms from surface. Move worms to a 15-mL conical centrifuge tube. We often use DI water instead of M9 in this and subsequent steps and get fine results. 2. Spin down and add bleaching solution (40 mL H2O, 7.2 mL 5 N NaOH, 4.5 mL 6% NaHOCl). 3. Vortex for roughly 4–8 min until worms disappear and only embryos remain. 4. Spin down and aspirate, then wash 2 twice in M9 buffer. 5. Resuspend in 1-mL fixation solution and incubate at room temperature for 15 min. 6. Vortex and then immediately submerge tube in liquid nitrogen for 1 min to freeze crack the embryos’ eggshells. 7. Thaw in water at room temperature. 8. Once thawed, vortex and place on ice for 20 min. 9. Wash twice with 1 mL 1 PBS. 10. Resuspend in 1 mL of 70% (v/v) EtOH and store at least overnight at 4  C. 3.2.5. Fixation of Drosophila melanogaster wing imaginal discs 1. Submerge 3rd instar larvae in 1 mL 1 PBS and dissect to release wing imaginal discs. 2. Place discs at the bottom of a chambered coverglass. They should stick readily.

Detection of Individual Endogenous RNA Transcripts

377

3. Fix wing discs by aspirating PBS and adding 1 mL fixation solution; incubate at room temperature for 45 min. 4. Wash 2 twice with 1 mL 1 PBS to remove fixative. 5. Add 1 mL of 70% EtOH and leave at least overnight at 4  C. 3.2.6. Fixation of tissue sections 1. Freeze tissue section in optimal cutting temperature compound (OCT). We have heard reports that using sucrose-based cryoprotectants can lead to high background and so should be avoided. 2. Slice the tissue section into 4–10 micron sections using a cryotome and affix the sections to #1 coverslips; the sections can then be stored for months at 80  C. Notes: Although it is a more standard procedure, do not affix the tissue sections to slides, as this greatly hinders the visualization of the fluorescent spots. Also, the use of poly-L-lysine or some similar surface treatment may enhance the degree to which your tissue section sticks to the coverglass. 3. Thaw the section and immediately fix in fixation solution, either in a coplin jar or a six-well plate. Note: we perform this procedure by affixing a perfusion chamber (Grace Biolabs) to a 24-mm  50-mm coverglass and adding all solutions, etc. to this chamber. Using the perfusion chamber greatly reduces the amount of fixing/washing reagents required. 4. Wash twice with 1 mL 1 PBS to remove fixative. 5. Add 1 mL of 70% (v/v) EtOH and leave for 1 h at room temperature. Note: some researchers have reported trouble with their sections floating off of the coverslip when stored for prolonged periods of time in 70% EtOH; thus, we recommend beginning the in situ hybridization less than 1 h following fixation.

4. Hybridization Hybridization consists of a brief prehybridization followed by an overnight hybridization with the oligonucleotide probes. In the morning, two washes in a washing buffer remove the nonhybridized probes and the samples are essentially ready for imaging. There are three basic parameters involved in the hybridization. One is the concentration of the probes used in the hybridization. We typically determine the appropriate concentration empirically, but we have found that (generally) a concentration in the vicinity of 5–50 nM works. Moreover, we have found that there is a fairly

378

Arjun Raj and Sanjay Tyagi

significant range (roughly a factor of 2 or 3 in either direction) of concentrations over which the signals are similar and readily quantifiable. When possible, it is best to optimize the probe concentration rather than changing other factors in the hybridization, since multiplex detection requires shared hybridization conditions, which are easy to match when the only variable is different concentrations of probes. The other two parameters of the hybridization are related and concern the stringency of the hybridization itself: one is the temperature at which one hybridizes the probes, and the other is the concentration of formamide used in the hybridization and washes. Regarding the former, higher temperature generally leads to higher stringency, as fewer of the probes will bind (nonspecifically or specifically) as the temperature increases. We typically use either 30  C or room temperature, but rarely change this variable, especially since changing the formamide concentration is largely equivalent and is easier to control in a fine-grained manner. To adjust the stringency via formamide, the main point is that the higher the concentration of formamide, the higher the stringency. We typically use 10% (v/v) for most of our hybridizations, but sometimes probes with GC contents of 55–60% require the use of 25% (v/v) formamide. It is important, however, to note that more stringent conditions can lead to a dramatic rate of false negatives: one can only see a few faint looking spots, when in reality, there are many more mRNAs present. Thus, it is generally better to begin with the less stringent 10% conditions and then work up from there. Also, we have found that adding some wet paper towels in the hybridization chamber is NOT helpful and often causes a spotlike background.

4.1. Hybridization solutions 4.1.1. Hybridization buffer (10 mL) Dextran sulfate (1 g) Escherichia coli tRNA (10 mg) Vanadyl ribonucleoside complex (NEB) (100 ml of 200 mM stock) BSA (RNase free) (Ambion) (40 ml of 5 mg/mL) 20 SSC (nuclease free, Ambion) (1 mL) Formamide (deionized, Ambion) (1 mL for 10% final concentration, can increase formamide to increase stringency) Nuclease free (NF) water (Ambion) (to 10 mL final volume) First, mix the dextran sulfate in about 4 mL of water with gentle agitation at room temperature until dissolved (can take min to h, depending on the batch). Then add the other components. We then keep the hybridization solution in 0.5 mL aliquots at 20  C.

Detection of Individual Endogenous RNA Transcripts

379

4.1.2. Wash buffer (50 mL): Wash/prehybridization buffer: 40 mL RNase free water (Ambion) 5 mL Formamide (deionized, Ambion) 5 mL 20 SSC (RNase free, Ambion) Note: one can increase the stringency by increasing the amount of formamide. 4.1.3. Antifade buffer and enzymes: 10% (w/v) glucose in nuclease free water 2 M Tris–HCl, pH 8.0 20 SSC (Ambion) Nuclease free water (Ambion) Glox (Sigma) (diluted to 3.7 mg/mL stock in 50 mM sodium acetate, pH  5) Catalase (Sigma) Mix together 0.85 mL of NF water and add 100 ml of 20 SSC, 40 ml of 10% (w/v) glucose and 5 ml of 2 M Tris–HCl. Vortex and then transfer 100 ml of this ‘‘glox’’ buffer to another tube, to which one should add 1 ml of glox stock and 1 ml of (nicely vortexed) catalase suspension. The remainder will be used as an equilibration buffer.

4.2. Hybridization protocols 4.2.1. Hybridization in solution 1. Prepare the hybridization solution: to 100 ml of hybridization buffer, add 1–3 ml of probe at the appropriate concentration, then vortex and centrifuge. a. Be sure to warm the hybridization solution to room temperature before opening it. b. For the initial test of a set of probes, it is best to start four separate hybridization reactions by adding 1 ml each of the 1:10, 1:20, 1:50, and 1:100 working dilutions of probes to see which one is optimal. 2. Centrifuge the fixed sample and aspirate away the ethanol. 3. Resuspend in 1 mL wash buffer that contains the same percentage formamide as the hybridization buffer you will be using. Let stand for 2–5 min. 4. Centrifuge sample and aspirate wash buffer, then add hybridization solution. Incubate in the dark overnight at 30  C.

380

Arjun Raj and Sanjay Tyagi

5. In the morning, add 1 mL of wash buffer to the sample, vortex, centrifuge, and aspirate, then resuspend in another 1 mL of wash buffer and incubate at 30  C for 30 min. 6. Vortex, centrifuge, and aspirate the wash buffer, then resuspend in another 1 mL of wash buffer containing 5 ng/mL DAPI for nuclear counterstaining. Incubate at 30  C for 30 min. 7. If you are imaging without using glox antifade solution (e.g., if you are using TMR), then just resuspend in an appropriate volume (>0.1 mL) of 2 SSC and proceed to imaging. 8. If you are imaging with the glox antifade solution, aspirate the buffer and resuspend in the glox buffer without enzymes for equilibration; incubate for 1–2 min. 9. Aspirate the buffer and resuspend in the 100 ml of glox buffer to which the enzymes (glox and catalase) have been added. Proceed to imaging. 4.2.2. Hybridization for samples adhered to coverglass 1. Prepare the hybridization solution: to 100 ml of hybridization buffer, add 1–3 ml of probe at the appropriate concentration, then vortex and centrifuge. a. Be sure to warm the hybridization solution to room temperature before opening it. b. For the initial test of a set of probes, it is best to start four separate hybridization reactions by adding 1 ml each of the 1:10, 1:20, 1:50, and 1:100 working dilutions of probes to see which one is optimal. 2. Aspirate the 70% ethanol off of the sample. 3. Add 1 mL wash buffer that contains the same percentage formamide as the hybridization buffer you will be using. Let stand for 2–5 min. 4. Aspirate wash buffer and then add hybridization solution. Place a carefully cleaned coverslip over the sample to prevent drying of the hybridization solution during the incubation. Incubate in the dark overnight at 30  C. Note: if using perfusion chambers on a coverslip containing a tissue section, one can remove the perfusion chamber before performing the hybridization. 5. In the morning, add 1 mL of wash buffer to the sample, remove the coverslip, then incubate at 30  C for 30 min. a. Be sure to remove the coverslip very carefully so as not to disturb the cells underneath very much. b. For tissue sections, add 100 ml wash buffer to the edges of the coverslip and gently remove the coverslip. Then reaffix a perfusion chamber and proceed as usual. 6. Aspirate the wash buffer, then resuspend in another 1 mL of wash buffer containing 5 ng/mL DAPI for nuclear counterstaining. Incubate at 30  C for 30 min.

Detection of Individual Endogenous RNA Transcripts

381

7. If you are imaging without using glox antifade solution (e.g., if you are using TMR), then just add 1 mL of 2 SSC and proceed to imaging. 8. If you are imaging with the glox antifade solution, aspirate the buffer and resuspend in 2 SSC. 9. Aspirate the SSC and add the glox buffer without enzymes for equilibration; incubate for 1–2 min. 10. Aspirate the buffer and resuspend in the 100 ml of glox buffer to which the enzymes (glox and catalase) have been added. 11. Place a carefully cleaned coverslip over the sample. This will spread the glox buffer over the entire sample and also slow evaporation. 12. Proceed to imaging.

5. Imaging At this point, the samples are essentially ready for imaging. The microscopy equipment required is fairly standard.

5.1. Microscopy equipment 1. Standard widefield fluorescence microscope (e.g., Nikon TE2000 or Ti, Zeiss Axiovert). 2. Strong light source, such as a mercury or metal-halide lamp (e.g., ExFo Excite, Prior Lumen 200). We have found that the metal-halide lamps are generally brighter, especially for the far red dyes such as Cy5. 3. Filter sets appropriate for the fluorophores chosen (see Table 17.1). 4. Standard cooled CCD camera, ideally optimized for low-light level imaging rather than speed (13 mm pixel size or less is ideal; for example, Pixis, Princeton Instruments, CoolSNAP HQ). We have found that EMCCDs do not provide any additional signal-to-noise benefits over nongain amplified cameras. 5. High NA (>1.3) 100 DIC objective (be sure to check transmission properties when using far red dyes such as Cy5 or Cy5.5). We have also seen spots using an oil-immersion 60 objective, but the reduced spatial resolution makes the spots somewhat more difficult to identify computationally. Generally speaking, the imaging of single mRNAs using a widefield fluorescence microscope is relatively straightforward; see Fig. 17.4 for some examples. The only difference between this and many more standard applications of fluorescence microscopy is that the signals are much weaker than, say, a DAPI stain, thus requiring exposure times on the order of 2–3 s. We have found that widefield microscopy works best due to the relatively

382

Arjun Raj and Sanjay Tyagi

A

B

Gut

Merge of focal planes

Single focal plane

C

D

E

A

dpp En P

F

G No salt

H 0.4 M NaCl

Figure 17.4 Imaging mRNAs in a variety of biological samples. (A) elt-2 mRNA molecules (red) in an early stage embryo (100 cell stage) from C. elegans; the nuclei have been counterstained with DAPI (blue). (B) elt-2 mRNA molecules in an L1 larva from C. elegans. Inside the blue box, a single focal plane is shown in which the intestinal track is visible. (C) A schematic depiction of dpp and engrailed expression in the imaginal wing discs of third instar larvae from D. melanogaster. (D) Image showing the locations of the computationally identified dpp mRNA molecules (light blue circles) and Engrailed expression detected by immunofluorescence (dark blue). (E) Image containing enhanced dpp mRNA molecule signals (light blue) and Engrailed protein expression detected by immunofluorescence (dark blue). (F) Image of FKBP5 mRNAs in human carcinoma cell line A549 induced with dexamethasone (nuclei in purple). (G–H) STL1 mRNA particles in both unperturbed cells (G) and cells subjected to a 10-mi 0.4 M NaCl salt shock (H), with nuclear DAPI counterstaining in purple. STL1 is one among a number of yeast genes whose expression is significantly upregulated by the addition of salt to the growth medium. All images except the boxed portion of (B) are maximum merges of a z-stack of fluorescent images, and all scale bars are 5 mm long.

Detection of Individual Endogenous RNA Transcripts

383

large amount of light gathered as compared to confocal imaging setups. That said, we have had some success using both spinning disk and laser scanning confocal microscopes, but it seems that one issue with their use is that the high intensity of the laser excitation rapidly bleaches samples. Since the total light collection is much lower than in a widefield microscope (per illumination), this bleaching limits the signal one can gather and is especially problematic when one takes multiple z sections to generate three-dimensional image stacks. However, the use of widefield microscopy places a tight limitation upon the thickness of the sample one can image, because thicker samples lead to far more out-of-focus light that can obscure the relatively faint mRNA signals. We have found that the single mRNA signals are most readily detectable when the sample is less than 7–8 microns thick. Some samples (notably C. elegans embryos and larvae) are considerably thicker than this limit, so we generally flatten them between two coverslips to reduce the z-extent of the sample considerably before imaging. For other samples, such as tissue sections and cell lines, the specimens are already sufficiently thin so as to obviate the requirement for flattening. Also, for imaging multiple slices, we recommend using at least a 0.2-mm z spacing between sections, and larger spacings such as 0.3 or even 0.4 mm are also probably fine. The main consideration is an empirical one: Aim for each RNA spot showing up in at least two adjacent optical sections. This gives confidence that the spots identified are legitimate. There are also some common microscopy practices that one should avoid when doing single molecule FISH. One of these is the use of commercial antifade mounting media. We have found that while these media do decrease the rate of photobleaching, they also lower the overall fluorescence of the sample and also introduce a strong background that interferes with the FISH signals (most likely from the glycerol included in many of these solutions). We recommend avoiding these entirely and just imaging with the antifade glox solution (or just 2 SSC if photobleaching is not a concern). Another thing to avoid is the use of nail polish to seal samples. This introduces a high background into the sample, again obscuring the FISH signals. We recommend sealing with silicone-based vacuum grease instead. Regarding the mounting of the samples, we use #1 coverglass to image all of our samples. We have found that our signals are better with #1 than with #1.5, even though our objectives (like most) are designed for use with #1.5 coverglass. Also, one should avoid having ones samples on a microscopy slide and then ‘‘covering’ them with coverglass. We have found that the subsequent layer of liquid between the top of the coverglass and the sample causes the signals to blur. If the target RNAs are stained properly, you will see clear diffractionlimited spots, such as those depicted in Fig. 17.4. The width of the spots is

384

Arjun Raj and Sanjay Tyagi

roughly 200–500 nm, depending on the dye and the imaging setup. While the intensities of the individual spots may vary, the spot size should be essentially identical. (An exception to this are active sites of transcription, at which many nascent mRNAs accumulate, resulting in a somewhat larger and significantly brighter spot (Femino et al., 1998; Raj et al., 2006; Vargas et al., 2005) Variability in the spot size is an indication that the spots are not actually target mRNAs but rather are some form of autofluorescent background. One way to check for this is to perform the hybridization without adding the probes to check if the allegedly nonspecific spots persist. Another way to see if the spots are merely autofluorescent background is to acquire images with different filter sets. Typically, the autofluorescent background will show up in multiple channels, owing to the rather broad spectral properties of cellular autofluorescence—with appropriate filters, there is essentially no bleedthrough between the different organic dyes used to label the oligonucleotides (Fig. 17.3) (Raj et al., 2008).

6. Image Analysis The analysis of images acquired using this method involves the computer-assisted identification of spots in a three-dimensional set of images (Fig. 17.5). Given that the spot-like signals are significantly brighter than the background, one might assume that a simple threshold would be sufficient. Unfortunately, due to out-of-focus light, the background itself often varies greatly throughout the image, making the simple application of a threshold impossible. To remove this (typically slowly varying) background, we employ Laplacian of Gaussian (LoG) filters (Fig. 17.5B). The LoG filter has essentially one parameter, which is the width of the filter. For any particular microscope/camera combination, we usually determine the optimal filter width by trial and error (theoretically, the width of the filter should be the same as the width of the spots one is trying to identify). We should note that we apply our filters in three-dimensions, thus using the three-dimensionality of the spots to further enhance the signals. Another option for removing the out-of-focus light is deconvolution software. We have found, though, that while the results from deconvolution are often nice, they seldom yield results that are better in terms of spot counting accuracy. Moreover, they are extremely expensive, both monetarily and computationally. For these reasons, we find our simple linear filtering approach to be more appropriate, especially for large data sets. After performing the filtering, one must select an appropriate threshold. We have found that this task is difficult to automate, since it is difficult to say a priori what the appropriate threshold is. Instead, we compute the number of spots detected for all possible thresholds. Upon graphing this relationship,

Detection of Individual Endogenous RNA Transcripts

B

C

D Number of spots found

A

385

1000

500

0 0

0.1 0.2 0.3 0.4 Threshold (normalized)

0.5

Figure 17.5 Computational identification of mRNA spots. (A) Raw image data (maximum intensity merge) obtained from imaging FKBP5 mRNA particles in A549 cells induced with dexamethasone. (B) Image (maximum merge) obtained by running raw data through Laplacian of a Gaussian filter designed to enhance spots of the correct size and shape while removing the slowly varying background. (C) The number of spots (i.e., connected components) found upon thresholding the filtered image from (B) is plotted as a function of the threshold value, ranging from 0 to the maximum intensity of the filtered image (normalized to 1). The presence of a plateau indicates that there is a region over which the number of particles detected is fairly insensitive to the particular threshold chosen. The gray line represents the threshold used (within the plateau) for determining the actual number of mRNA in the image. (D) Image showing the results of using the threshold represented by the gray line in C on the filtered image in (B), with each distinct spot assigned a random color. The spots detected correspond very well with those identified by eye. All scale bars are 5 mm long. Adapted with permission from Supplementary Fig. 1 of Raj et al. (2008).

we found that there was a plateau region in the graph, which means that there is a broad region of thresholds over which the spot count does not vary significantly (Fig. 17.5C). This is generally the correct threshold to choose, as spots identified at those thresholds correspond nicely to those identified by eye (Fig. 17.5D). Our image processing pipeline is thus to first preprocess the data via filtering and applying all possible thresholds, then manually picking thresholds based on the graph (with some visual feedback). We find that this facilitates rapid processing of many images, allowing one to threshold hundreds of images in a matter of hours. Software demonstrating these algorithms (implemented in MATLAB and including some sample data) is free for download at: http://rajlab.seas.upenn. edu/pdfs/raj_nat_meth_2008_software.zip.

386

Arjun Raj and Sanjay Tyagi

ACKNOWLEDGMENTS Sanjay Tyagi acknowledges support from National Institutes of Health grant NIMH079197. Arjun Raj acknowledges support from National Science Foundation postdoctoral fellowship DMS-0603392 and a Burroughs-Wellcome Fund Career Award at the Scientific Interface.

REFERENCES Femino, A. M., Fay, F. S., Fogarty, K., and Singer, R. H. (1998). Visualization of single RNA transcripts in situ. Science 280, 585–590. Gall, J. G. (1968). Differential synthesis of the genes for ribosomal RNA during amphibian oo¨genesis. Proc. Natl. Acad. Sci. USA 60, 553–560. Golding, I., Paulsson, J., Zawilski, S. M., and Cox, E. C. (2005). Real-time kinetics of gene activity in individual bacteria. Cell 123, 1025–1036. Larson, D. R., Singer, R. H., and Zenklusen, D. (2009). A single molecule view of gene expression. Trends Cell Biol. 19, 630–637. Levsky, J. M., and Singer, D. (2003). Fluorescence in situ hybridization: Past, present and future. J. Cell Sci. 116, 2833–2838. Maamar, H., Raj, A., and Dubnau, D. (2007). Noise in gene expression determines cell fate in Bacillus subtilis. Science 317, 526–529. Maheshri, N., and O’shea, E. K. (2007). Living with noisy genes: How cells function reliably with inherent variability in gene expression. Ann. Rev. Biophys. Biomol. Struct. 36, 413–434. Raap, A. K., Van de corput, M. P., Vervenne, R. A., Van gijlswijk, R. P., Tanke, H. J., and Wiegant, J. (1995). Ultra-sensitive FISH using peroxidase-mediated deposition of biotinor fluorochrome tyramides. Hum. Mol. Genet. 4, 529–534. Raj, A., and Van oudenaarden, A. (2008). Nature, nurture, or chance: Stochastic gene expression and its consequences. Cell 135, 216–226. Raj, A., and Van oudenaarden, A. (2009). Single-molecule approaches to stochastic gene expression. Ann. Rev. Biophys. 38, 255–270. Raj, A., Peskin, C. S., Tranchina, D., Vargas, D. Y., and Tyagi, S. (2006). Stochastic mRNA synthesis in mammalian cells. PLoS Biol. 4, e309. Raj, A., Van den bogaard, P., Rifkin, S. A., Van oudenaarden, A., and Tyagi, S. (2008). Imaging individual mRNA molecules using multiple singly labeled probes. Nat. Methods 5, 877–879. Tautz, D., and Pfeifle, C. (1989). A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback. Chromosoma 98, 81–85. Vargas, D. Y., Raj, A., Marras, S. A., Kramer, F. R., and Tyagi, S. (2005). Mechanism of mRNA transport in the nucleus. Proc. Natl. Acad. Sci. USA 102, 17008–17013. Yildiz, A., Forkey, J. N., Mckinney, S. A., Ha, T., Goldman, Y. E., and Selvin, P. R. (2003). Myosin V walks hand-over-hand: Single fluorophore imaging with 1.5-nm localization. Science 300, 2061–2065. Zenklusen, D., Larson, D. R., and Singer, R. H. (2008). Single-RNA counting reveals alternative modes of gene expression in yeast. Nat. Struct. Mol. Biol. 15, 1263–1271.

C H A P T E R

E I G H T E E N

Single mRNA Tracking in Live Cells Hye Yoon Park, Adina R. Buxbaum, and Robert H. Singer Contents 388 389 391 391 392 393 394 394 395 396 396 397 398 399 401 402 403 403

1. Introduction 2. Significance of Tracking mRNA 3. Labeling mRNA in Living Cells 3.1. Selection of probes for SPT 3.2. The MS2-GFP system 3.3. Minimizing photobleaching and phototoxicity 4. Imaging mRNA Movements 4.1. Experimental considerations 4.2. Instrumentation 4.3. 3D tracking 5. Analyzing mRNA Motions 5.1. Localization algorithms 5.2. Tracking algorithms 5.3. Categories of single particle motion 5.4. Interpretation of mRNA tracking data 6. Conclusions Acknowledgments References

Abstract Asymmetric distribution of mRNA is a prevalent phenomenon observed in diverse cell types. The posttranscriptional movement and localization of mRNA provides an important mechanism to target certain proteins to specific cytoplasmic regions of their function. Recent technical advances have enabled real-time visualization of single mRNA molecules in living cells. Studies analyzing the motion of individual mRNAs have shed light on the complex RNA transport system. This chapter presents an overview of general approaches for single particle tracking and some methodologies that are used for single mRNA detection. Anatomy and Structural Biology and Gruss-Lipper Biophotonics Center, Albert Einstein College of Medicine, New York, USA Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72003-6

#

2010 Elsevier Inc. All rights reserved.

387

388

Hye Yoon Park et al.

1. Introduction Localization of mRNA is an important mechanism to generate cell polarity crucial in diverse cellular functions from motility to differentiation (for reviews, see Condeelis and Singer, 2005; Martin and Ephrussi, 2009; Shav-Tal and Singer, 2005). The asymmetrical distribution of mRNA provides a means for a cell to regulate the protein synthesis at high spatial and temporal resolution. Localized mRNAs can be translated repeatedly to produce high concentrations of proteins in specific subcellular compartments in response to local stimuli. To date, thousands of mRNAs are found to exhibit spatially distinct patterns in many different cell types, including budding yeast, fruit fly oocyte, fibroblasts, and neurons (Martin and Ephrussi, 2009). Technical developments in intracellular RNA imaging have been indispensable to increase our knowledge about the mechanisms of mRNA localization. When the localization of b-actin mRNA was first observed in the lamellipodia of fibroblasts (Lawrence and Singer, 1986), the mRNAs were hybridized with radioactive DNA probes and visualized by autoradiography, which required exposure times in the range of weeks. Now, it is possible to observe the movement of single mRNA molecules in living cells in real time (Bertrand et al., 1998; Fusco et al., 2003; Shav-Tal et al., 2004). Single particle tracking (SPT) is used in many different research fields to investigate the dynamics of individual objects by regarding them as punctate points while ignoring the internal conformations. By following the trajectories of particles, we can characterize the types of motion and measure the velocity or diffusion coefficient. Jean Perrin, probably in the first SPT akin to modern methods, observed the movements of gamboges with submicron precision (Perrin, 1913). His quantitative analysis of the trajectories supported Einstein’s microscopic theory of Brownian motion (Einstein, 1905). In cell biology, the use of SPT was pioneered by Barak and Webb (1982). They observed the motion of fluorescently labeled low-density lipoprotein (LDL) receptors on plasma membrane. De Brabander et al. (1985) microinjected colloidal gold particles of 20–40 nm in living cells, and visualized their motion using transmitted light To date, SPT has been extensively used to study complex cellular dynamics, including ligand–receptor interactions, membrane organization, secretory granules, locomotion of motor proteins, and transport within nuclei (reviewed in Kusumi et al., 2005; Levi and Gratton, 2007; Saxton and Jacobson, 1997; Wieser and Schutz, 2008). There are other optical techniques for measuring the lateral mobility. In the technique of fluorescence recovery after photobleaching, or FRAP (Axelrod et al., 1976), a region of interest is irreversibly photobleached by

Single mRNA Tracking

389

intense laser irradiation and then, recovering fluorescence in the area is recorded in time. From the recovery curve, one can derive the fraction and the diffusion coefficient D of mobile fluorescent molecules. Caution is required, however, in the presence of multiple species with distinct characteristics of mobility: FRAP data are an ensemble average of the total population, and the specific dynamics of a subpopulation may be hidden. SPT overcomes this limitation of FRAP by directly observing individual particles. Furthermore, the spatial resolution of SPT exceeds that of FRAP by more than an order of magnitude. SPT considers only the center of particles which can be determined with a precision of one to tens of nanometers, whereas the diffraction-limited focal volume dictates the minimum area in FRAP or fluorescence correlation spectroscopy (FCS). Consequently, SPT is suitable for high-resolution studies, far below the diffraction limit, of nanometer-scale displacements and structures, such as motor proteins and membrane microdomains. Here, we describe SPT techniques that have been applied to the studies of mRNA trafficking in living cells. Methods to label, visualize, and track single mRNA molecules are reviewed. The ‘‘MS2 system’’ (Beach et al., 1999; Bertrand et al., 1998) for labeling mRNA is emphasized, which has been established in our laboratory. Various analysis techniques are reviewed and the information obtained by combining SPT with the MS2 system is discussed toward the end of the chapter.

2. Significance of Tracking mRNA Many aspects of mRNA transport and localization have been discovered by single mRNA imaging and tracking. Whereas in situ hybridization shows the distribution of mRNA fixed at different stages, tracking of single mRNAs can reveal the in vivo dynamics that occur in the native environment. Tracking single mRNA particles in the cytoplasm of COS cells revealed that the movement of a reporter mRNA in the cytoplasm could be diffusive, static, corralled, and directed, with diffusive motion dominating (Fusco et al., 2003). The authors of this study were also able to show for the first time the movement of mRNA along cytoskeletal fibers. Interestingly, the addition of the b-actin 30 UTR to the construct, which contains a localization sequence necessary for the localization of b-actin mRNA, increased in the relative amount of directed movements and their average length. In a related study, single molecule tracking allowed the measurement of the diffusion coefficient of b-actin mRNA in different regions of the COS cells. b-actin mRNA was found to diffuse freely in the leading edge of the cell, however, in the perinuclear region, mRNA diffusion was restricted. Disruption of the actin cytoskeleton delocalized mRNA

390

Hye Yoon Park et al.

and increased the diffusion coefficient of mRNA in the perinuclear region, indicating that cytoskeletal barriers may play a role in the localization of b-actin mRNA (Yamagishi et al., 2009). In an additional study where SPT was critical to probing a mechanism of mRNA localization, Bertrand et al. (1998) employed single mRNA tracking to address the question of how ASH1 mRNA travels to the bud tip in yeast. It was known that SHE1/MYO4, a type V myosin, as well as an intact actin cytoskeleton were necessary for ASH1 mRNA localization, however, it was not clear whether ASH1 mRNA was actively transported to the bud tip or if myosin was transporting another protein necessary for ASH1 mRNA anchoring at the bud tip. Real-time imaging and particle tracking indicated that ASH1 was transported from mother to daughter yeast cell with a velocity consistent with motor-based transport and that mRNA particles colocalized with myosin. Single mRNA tracking in the nucleus was used to address the controversial question of how mRNA travels in the nucleus, revealing movements indicative of corralled diffusion (Shav-Tal et al., 2004). In this study, it was shown that mRNAs are not actively transported in the nucleus but passively diffuse. Zimyanin et al. (2008) also used live cell visualization and tracking of mRNA to address a controversy in the field of oskar mRNA localization in Drosophila oocytes. Prior to their study, it had been known that kinesin was necessary for posterior oskar mRNA localization, so seemingly oskar mRNA localization depended on kinesin-based transport; however, paradoxically, the microtubule network in the Drosophila oocyte lacks uniform polarity. Other theories postulated that cytoplasmic flow or exclusion from specific regions are responsible for oskar mRNA localization, with kinesin playing an indirect role. By direct observation of the mRNA, the authors showed that the mRNA moves along microtubules in many directions with a 14% bias toward the posterior region. Over time, this is sufficient to localize the mRNA to the correct region in the appropriate time frame. An interesting cellular model for active transport of mRNA is the study of mRNA localization in neuronal processes, as diffusion alone is insufficient to transport mRNA into long dendritic processes, thus, active transport is a necessity for mRNA to reach the distal regions of neurons. Live imaging of calcium/calmodulin kinase II alpha reporter mRNA revealed a kinesin and microtubule-dependent oscillatory movement of the mRNA in the dendrites. Following stimulation, there is an increase of mRNA movement in the anterograde direction, bringing mRNA granules into dendrites and increasing the probability of arriving at activated synapses (Rook et al., 2000). These representative examples emphasize the significant discoveries in the field of mRNA trafficking where live mRNA imaging and tracking played a pivotal role in understanding the mechanisms of mRNA localization.

Single mRNA Tracking

391

3. Labeling mRNA in Living Cells 3.1. Selection of probes for SPT A wide variety of probes have been used in SPT, including gold particles, quantum dots, small organic dyes, and fluorescent proteins. Colloidal gold particles of 20–40 nm in diameter have been used with bright field microscopy (De Brabander et al., 1985) or differential interference contrast (DIC) microscopy (Sheetz et al., 1989). A small number of ligands or Fab fragments of the antibody IgG for target molecules are conjugated to the gold particles. Labeling by gold is advantageous for longer duration of tracking because there is no photobleaching and little saturation. Also, it allows the manipulation of single particles by using an optical trap (Edidin et al., 1991; Kusumi et al., 1998). However, gold probes have artifacts such as nonspecific charge interactions and crosslinking (Kusumi et al., 2005) and have not been yet applied successfully to mRNA labeling in living cells. Fluorescent probes are more amenable to specific labeling. Simultaneous tracking of different species is readily achieved by multicolor imaging with diverse fluorescent tags. When using fluorescent probes, photostability and brightness are the primary figures of merit for SPT. Quantum dots have been widely used for SPT since they are 10- to 100-times brighter and 100to 1000-times more photostable than organic dyes (Smith et al., 2008). Another advantage of semiconductor nanocrystals is that the emission wavelength can be tuned by the size; larger quantum dots emit redder fluorescence. However, quantum dots exhibit intermittent emission, or ‘‘blinking’’ (Nirmal et al., 1996), which can complicate the analysis of SPT data. Using quantum dots, Ishihama and Funatsu observed the movement of single mRNAs for over 60 s with a time resolution of 30 ms (Ishihama and Funatsu, 2009). Organic dyes and fluorescent proteins have been predominantly used for labeling mRNAs (for reviews, see Querido and Chartrand, 2008; Rodriguez et al., 2007; Tyagi, 2009). To image total mRNA in live cells, nonspecific nucleic acid stains such as SYTO 14 can be used (Knowles et al., 1996). Visualization of specific mRNA has been typically achieved through the microinjection of fluorescently labeled RNAs (Ainger et al., 1993; Shan et al., 2003; for a review of fluorescent RNA cytochemistry, see Pederson, 2001). An alternate technique that allows the labeling of endogenous mRNA is a variation of FISH performed on live cells. Santangelo et al. (2009) describe a technique where the membranes of live cells are reversibly permeabilized with the Streptolysin O, which delivers fluorescently labeled oligonucleotides into cells. Finally, molecular beacons have also been used to visualize endogenous mRNAs in live cells (Bratu et al., 2003), where delivery also typically involves microinjection.

392

Hye Yoon Park et al.

3.2. The MS2-GFP system To label native mRNA with GFP in living cells, the MS2-labeling technique has been devised (Bertrand et al., 1998). High autofluorescence in the cytoplasm can significantly confound tracking single molecules in a live cell. In order to enhance the signal-to-background ratio, the system expresses mRNAs that contain multiple MS2 stem loops, to each of which a dimer of fluorescent protein-fused MS2 coat proteins (FP-MCP) specifically binds. We have empirically determined that 24 copies of the MS2 binding sites (MBS) binding up to 48 FP-MCPs are sufficient to visualize single mRNA molecules (Fusco et al., 2003; Shav-Tal et al., 2004). Plasmids containing multiple MBS cassettes and FP-MCP are available upon request at http:// singerlab.aecom.yu.edu/requests/. The benefit of using genetically encoded fluorescent proteins to label mRNAs is that the mRNA is transcribed and labeled in the nucleus, which should ensure proper binding of mRNA binding proteins, necessary for proper export, transport, and translation (Farina and Singer, 2002). Additionally, the MS2 system involves minimal perturbation to the cellular structure as opposed to other methods of delivery of exogenous mRNA such as microinjection of fluorescently labeled mRNAs or delivery through the perturbation of the plasma membrane. Many previous chapters have addressed the methodology using the MS2 system to fluorescently label mRNAs (Chao et al., 2008a; Grunwald et al., 2008b). This chapter will focus on technical considerations as opposed to specific instructions. An MS2-GFP labeling system should be designed properly with several considerations. It is important that the MS2-GFP construct includes appropriate untranslated regions (UTR), which play an essential role in the mRNA localization by regulating the mRNA’s interaction with the cytoskeleton or RNA binding proteins. Moreover, the MS2 repeats must be inserted in a carefully selected location. It is highly recommended to verify proper trafficking of mRNA using FISH in order to avoid potential problems. Other unknown elements not included in the reporter construct may be important for correct localization. Or the MS2 repeats may interfere with trafficking or induce nonsense-mediated degradation of the mRNA. Secondly, appropriate levels of expression are crucial. If both the reporter mRNA and the FP-MCP are overly expressed, they may form fluorescent aggregates in the cytoplasm of the cell. Overexpression of mRNA may also lead to abnormal localization, because RNA-binding proteins and transport machinery may exist in limiting amounts. Therefore, it is most desirable that the reporter constructs are expressed under their own promoters. Retrovirus or lentivirus infection is widely used for creating stably expressing cells. Because each cell will only contain a few copies of the transgene, this method not only eliminates the concern of overexpression but also reduces the cell-to-cell variations in expression.

Single mRNA Tracking

393

When performing live cell imaging using the MS2 system, multiple controls are essential to perform to verify the correct trafficking of the mRNA. MS2-tagged mRNAs should be visualized in combination with FISH to measure the relative abundance of labeled mRNAs, as in Fusco et al. (2003). Furthermore, FISH should be performed on cells that express the stem-loop-tagged mRNA in the absence and presence of the FP-MCP to ensure proper targeting of the mRNA with the stem loops and while bound to multiple fluorescent proteins. An additional necessary control is to express the FP-MCP in cells that do not contain the stemloop-tagged mRNA for the purpose of verifying that the expression of the coat protein does not lead to artifactual aggregation of fluorescent protein in the cells. In our laboratory, a transgenic mouse line with 24 MS2 repeats inserted into the 30 UTR of the b-actin gene has been created recently (manuscript in preparation). This system will allow the visualization and tracking of endogenous b-actin mRNA in various cell types, and moreover in vivo, which has not been achieved before. An orthogonal system for RNA labeling has also been developed using PP7 bacteriophage coat protein (Chao et al., 2008b), which will enable the tracking of multiple mRNA species.

3.3. Minimizing photobleaching and phototoxicity Ultimately, long time-lapse imaging experiments are limited by photobleaching and phototoxicity. The average number of photons emitted by a dye molecule before photobleaching is approximately 10,000–100,000. Photobleaching occurs by several complex mechanisms and strongly depends on the environmental conditions such as solvent polarity and temperature (Eggeling et al., 2005). The most notable mechanism for photobleaching is photooxidation. Fluorophores in triplet excited state react with ground-state triplet oxygen and generate singlet oxygen (1O2). The highly reactive singlet oxygen causes both photobleaching and phototoxicity. Several reagents such as ascorbic acid and enzymatic deoxygenation systems have been used to reduce the detrimental effects. However, the removal of oxygen can enhance or reduce the photobleaching effect depending on the experimental condition. This is because photooxidation processes cause both the ground-state recovery of the dyes and the formation of irreversible photoproducts. Therefore, the concentration of oxygen scavengers needs to be optimized for sufficiently long tracking experiments. Addition of triplet quenchers such as Trolox (a water-soluble analog of vitamin E) and mercaptoethylamine can also improve the photostability (Rasnik et al., 2006; Widengren et al., 2007).

394

Hye Yoon Park et al.

4. Imaging mRNA Movements 4.1. Experimental considerations In order to track single mRNA movement in real time, it is important to achieve high sensitivity for single molecule detection and fast image acquisition. In fluorescence microscopy, photobleaching phenomena inherently limit the number of photons available from the probe. Therefore, one needs to find a good balance in the image-acquisition protocol. First, the camera exposure time needs to be optimized to detect single molecules in motion. The precision to locate the center of a particle is proportional to the total number of collected photons (Bobroff, 1986). Once the imaging system is optimized for the highest signal-to-noise ratio, the exposure time needs to be long enough to locate particles with a desirable precision. On the other hand, the camera exposure has to be short enough to capture an image of a highly mobile object. If the particle travels a significant distance during the exposure time, it will show up as a streak or a blurred object, which impairs the detection of the object. Secondly, a high frame rate is desired to follow the trajectory of a diffusing particle. In order to identify the same particle in two successive image frames, it is ideal to meet the Nyquist criterion in temporal sampling, that is, the displacement during the time interval should be less than half the spatial resolution. The previously measured diffusion coefficients of messenger ribonucleoprotein particles (mRNPs) in living cells are 0.1–0.8
m2/s (Fusco et al., 2003; Shav-Tal et al., 2004), thus, the sampling time interval needs to be 5–40 ms. This requirement can be relaxed when the particle density in the image is sufficiently low. If the average distance between particles is much larger than the average particle displacement between frames, two successive images of an object can be linked to each other unambiguously. However, with increasing particle density, it becomes more difficult to solve the motion correspondence problem. Lastly, a sufficient tracking range is crucial to identify the type of motion. The total number of frames in the image sequence determines the statistical accuracy of the analysis (Qian et al., 1991; Saxton, 1997). Monte Carlo simulations can be performed without limitation on the tracking period to quantitatively estimate the deviations from Brownian motion. For instance, Saxton examined the statistical variation of the diffusion coefficient D by simulations (Saxton, 1997). In addition, longer observation enables the detection of motion-type transitions. Since it is difficult to meet all of these requirements with a limited number of photons, one needs to find a good compromise between the high acquisition rate and the total duration of the experiment.

Single mRNA Tracking

395

4.2. Instrumentation SPT may be performed in various forms of light microscopy, including widefield, confocal, and total internal reflection microscopy (TIRFM). A standard wide-field epi-illumination microscope has been successfully used to visualize single mRNA molecules labeled by the MS2 system. The microscope system should be optimized to observe single molecule dynamics in living cells. For a sensitive detection of weak fluorescence, the photon collection efficiency needs to be maximized while the background noise is minimized. Microscope objectives with higher NA are desirable to obtain a higher photon collection efficiency and tighter point-spread function. Large magnification may be beneficial to minimize the pixelation noise, as long as the particle under study does not travel beyond the field of view. When using multiple fluorophores with different emission colors, chromatic aberration must be appropriately corrected by using achromat or apochromat objectives. For colocalization of single molecules labeled with different fluorophores, multichannel image registration is also necessary (Churchman and Spudich, 2007). The most common light source for wide-field microscopy is either a mercury or xenon arc lamp. If the power of the lamp at the excitation wavelength is not sufficient to detect single molecules, a laser light source can be employed. Laser illumination provides not only higher power but also narrower excitation bandwidth in the subnanometer range, which reduces the excitation background (Grunwald et al., 2008b). Since the viability of the cell also needs to be ensured, the illumination power must be balanced to protect the specimen against photodamage and photobleaching. For single-molecule detection, there are many different types of cameras and spot detectors. The most commonly used detector for SPT is the electron-multiplying charge-coupled device (EMCCD). The electronmultiplying shift register increases the gain while keeping the noise level low. In order to achieve shot noise-limited detection, it is desirable to obtain maximum quantum efficiency and minimum camera noise. For higher quantum efficiency, back-illuminated type CCD chips are preferable. The dark noise of the CCD is due to the thermal fluctuation in the amount of charge in total darkness and can be reduced by cooling the chip down to 80  C. The readout noise increases approximately with the square root of the readout speed (Rasnik et al., 2007). Therefore, there is a tradeoff between the acquisition rate and the noise level. A frame-transfer feature alleviates this constraint and is highly desirable for sufficiently frequent acquisition. Finally, cells must be kept in physiological conditions to ensure that the dynamics observed is an appropriate representation of the behavior in vivo. For mammalian cells, the temperature should be maintained at 37  C.

396

Hye Yoon Park et al.

There are several commercial systems to keep the sample warm during imaging sessions. A simple economical method is to use heating elements for the specimen and the objective lens. However, there can be a thermal drift due to the cycles of heating and cooling. A more reliable method is to build an incubator around the microscope body. A custom-designed incubator that can enclose the majority part of the microscope can keep the system at a stable temperature and prevent thermal drift. Also, incubators are desirable to control the CO2 level and humidity for extended periods of time.

4.3. 3D tracking SPT has been mainly employed in two-dimensional systems such as cell surface or immobilized cytoskeletons in vitro. It is highly desirable to extend the technology into three-dimensional imaging since most biological processes occur within the 3D space of the cell. Kao and Verkman (1994) introduced a weak cylindrical lens in the detection optics of an epifluorescence microscope, which caused astigmatism in a particle image. Images of fluorescent beads are circular in the focus but become ellipsoidal when out of the focus. The major axis of the ellipsoid is rotated by 90 above and below the focus. They retrieved the x, y, and z positions by analyzing the shape, orientation, and position of the particle’s image. Simple defocusing methods have also been used for 3D tracking. Speidel et al. (2003) calibrated the radii of the ring patterns in the defocused image of a particle as a function of the axial position of the object. They found a linear dependence of the ring radii on the z-offset within an axial range of 3 mm. Toprak et al. (2007) employed a similar method but with simultaneous imaging of the focused and defocused planes, and improved the localization accuracy in 3D. 3D tracking is also demonstrated using a two-photon microscope by tracing the laser beam in four circular orbits surrounding the object (Levi et al., 2005). The position of the particle is calculated on the fly, and those coordinates are used to set the next scanning position.

5. Analyzing mRNA Motions A number of ideas and techniques for tracking objects in a sequence of images can be found in the context of fluid mechanics, computer vision, and radar surveillance. In cell biological applications, two types of tracking algorithms have typically been used. The first category detects the changes in particle positions by crosscorrelating consecutive frames. The second category generally consists of two steps: find the center of each particle in time-lapse images, and connect the positions to reconstruct the trajectory. Cheezum et al. (2001) compared the tracking algorithms quantitatively by

397

Single mRNA Tracking

simulations. They concluded that the crosscorrelation method performs better for particles larger than the wavelength. Conversely, for particles that are smaller than the emission wavelength, it is more accurate and precise to perform particle detection followed by trajectory linking. Since the size of the mRNPs is smaller than the visible light wavelength, we will consider the second category of the tracking method here.

5.1. Localization algorithms In light microscopy, an object that is smaller than the dimension of wavelength appears as a diffraction-limited spot. Because of the limited resolution, the details of the object cannot be discerned. However, the center of the object can be determined with a much better precision when a sufficient number of photons are collected from the particle. There are two major categories of algorithms to identify the location of single particles: searching for the intensity-weighted center of mass (centroid) or fitting image intensities by point-spread function. In a centroid method, the image is filtered to remove high-frequency noise, a binary mask is applied to exclude the background below threshold intensity, and the weighted center of mass of contiguous pixels is calculated (Ghosh and Webb, 1994). Gelles et al. (1988) demonstrated a localization precision of 1–2 nm by usingDIC images of plastic beads. They crosscorrelated the sequence of images with a kernel segment of a single bead image and computed the centroid of each particle. The centroid method is computationally efficient and valid for asymmetric particles. However, the precision and accuracy of the particle position found by centroid methods are highly dependent on the background threshold level (Cheezum et al., 2001). Alternatively, the fluorescent intensity distribution of a single particle can be fit with a 2D Gaussian function. It provides a higher localization precision and an accurate measure of the intensity (Anderson et al., 1992). Cheezum et al. (2001) compared the efficacy of the centroid and Gaussian fitting routines and concluded that a direct Gaussian fit to the intensity profile is superior in terms of both accuracy and precision. Thompson et al. (2002) derived an approximate equation for the localization precision: hðDxÞ2 i ¼

s2 þ a2 =12 8ps4 b2 þ 2 2 aN N

ð18:1Þ

where s is the standard deviation of the point-spread function, a is the pixel size, b is the background noise, and N is the photon number. In the shot noise limit (the first term in Eq. (18.1)), the localization error is inversely

398

Hye Yoon Park et al.

proportional to the square root of the number of photons. When the background noise dominates (the second term in Eq. (18.1)), the uncertainty scales as the inverse of the number of photons. They also introduced a simplified fitting algorithm called ‘‘Gaussian mask,’’ which is equivalent to a nonlinear least-squares fit to a Gaussian distribution ignoring the shot noise. In this method, the centroid is calculated in convolution with a Gaussian distribution around the candidate position, and iterated until the centroid position converges. When the number of photons originating from the molecule of interest is small, the Gaussian mask algorithm can be more robust than the full least-squares fit.

5.2. Tracking algorithms After the particles are located in a sequence of frames, the next step is to link a position in each frame with a corresponding position in the next frame. In general, the particles are not distinguishable from one another. With increasing particle density, it becomes more difficult to determine the next position of a given particle. Therefore, an important parameter to gauge the difficulty of particle tracking is the spacing-displacement ratio, which is the average distance between particles divided by the average particle displacement between two successive frames (Malik et al., 1993). If the spacing-displacement ratio is much larger than one, tracking can be reliably done by simple nearest-neighbor approaches (Anderson et al., 1992; Ghosh and Webb, 1994). However, it becomes more difficult to connect the trajectories as the spacing-displacement ratio becomes smaller. Because there are many possible pairs of particles between two images, it is necessary to find the most probable set of connection. Various algorithms have been developed to seek a unique solution to the motion correspondence problem, and they can be divided into two broad categories: deterministic and statistical methods (Yilmaz et al., 2006). Deterministic methods are also called combinatorial optimization techniques. They define a cost function of associating each spot in the previous frame to a single spot in the next frame. By minimizing the cost function, an optimal assignment can be obtained. For example, Crocker and Grier (1996) described a simple cost function to track noninteracting Brownian particles. If we consider an ensemble of indistinguishable noninteracting M particles, most probable set of linkages between two frames is obtained P the ! ! 2 when M ð j¼1 r j ði þ 1Þ  r j ðiÞÞ is minimized. If the particles can be distinguished by additional information such as size, color, and intensity, these data can be treated as another dimension of the particles in the cost function. The algorithm is available at http://www.physics.emory.edu/ weeks/idl.

Single mRNA Tracking

399

If the scope of the tracking is extended to more than two image frames, it becomes a multidimensional optimal assignment problem (Sbalzarini and Koumoutsakos, 2005; for the associated ImageJ plugin, see http://www. mosaic.ethz.ch/Downloads/ParticleTracker). Most multiframe tracking algorithms are based on heuristic assumptions such as smoothness of the particle trajectories motion (Sage et al., 2005; for the associated ImageJ plugin, see http://bigwww.epfl.ch/sage/soft/spottracker/; Vallotton et al., 2003). By tracking objects across multiple frames, the history of the particle movement is considered. Therefore, these methods can resolve problems arising from crossing trajectories, temporary occlusion, blinking, and detection failure. However, multiframe tracking is computationally expensive and becomes difficult to solve as the frame number increases. Therefore, greedy search techniques and heuristic approaches are used to obtain approximate solutions (reviewed in Ja¨hne et al., 2007; Yilmaz et al., 2006). Statistical data association methods take the uncertainty of the position measurements into account and assign a probability density function for a particle state. The probability distribution propagates over time and is updated by the measurements in each frame. The simplest statistical tracking method is the Kalman filter. In a Kalman filter, the initial particle state and noise have a Gaussian distribution. The next position of a particle is predicted by a linear model of motion, and the actual observation in the predicted search region is used to adjust the particle state. Kalman filtering can also be extended to multiple frame processing. The multiple hypotheses tracking (MHT) algorithm defers the correspondence decision until several frames are examined (Reid, 1979). Probabilities for multiple hypotheses are calculated, and the most likely set of track is chosen. The MHT algorithm seeks the globally optimal solution by considering all particle positions at all time frames. However, it is computationally intense both in time and memory. Thus, various approximate solutions to MHT were developed and applied to SPT in living cells ( Jaqaman et al., 2008; Serge et al., 2008).

5.3. Categories of single particle motion It is not known whether molecular motion in biology is finite, but the effort to categorize it is well worthwhile for SPT. Random walk is one of a few simple and universal models in physics, which analytically describes unobstructed, or ‘‘normal,’’ diffusion. Therefore, it is natural in SPT that normal diffusion serves as a reference while complicated motions are treated as deviations from this null model. A molecule may exhibit one of the five modes of motion, depending on the nature of interactions; stationary, normal diffusion, anomalous subdiffusion, corralled diffusion, and directional movement by active transport. Moreover, it is also possible that an

400

Hye Yoon Park et al.

mRNA molecule makes transitions between the modes (Fusco et al., 2003). It was only after SPT was applied that researchers began to recognize the significance of nonBrownian microscopic motions in biology (Feder et al., 1996; Kusumi et al., 1993). A measurable parameter most commonly employed in SPT analysis is the mean-squared displacement (MSD) as a function of time. If we consider a trajectory ! r ðtÞ recorded every dt for N time steps, the MSD for a given time lag ndt is calculated by: hr 2 ðnÞi ¼

n X 1 N ! 2 ½! r ði þ nÞ  r ðiÞ N  n i¼1

The MSD curves for the different types of motion are shown in Fig. 18.1 and the analytical forms can be expressed as follows: hr 2 i < ðDxÞ2 hr 2 i ¼ 2dDt hr 2 i ¼ Gt a   hr 2 i ¼ R2 1  A1 exp  AR22t hr 2 i ¼ 2dDt þ v2 t2

stationary (Dx: localization precision) normal diffusion (d: spatial dimension) anomalous subdiffusion (a < 1) corralled diffusion (R: radius of the corral) directed motion with diffusion (v: speed)

Mean-squared displacement

Normal diffusion Directed motion Anomalous subdiffusion Corralled diffusion

Time

Figure 18.1 Mean-squared displacement as a function of time for various modes of motion.

401

Single mRNA Tracking

When fitting experimental MSD data with these analytical functions, one needs to add a constant to the fit-function because the localization precision Dx leads to an offset in the MSD plot (Wieser and Schutz, 2008). Another approach to analyze SPT trajectories is obtained by looking at the statistics of displacements, rather than the average. The probability distribution (Anderson et al., 1992), or equivalently, the jump-distance distribution (Grunwald et al., 2008a; Siebrasse et al., 2008), permits different perspectives from MSD. While ensemble MSD analysis measures an average of a population, jump-distance histogram detects different mobility populations. Jump distance analysis measures the probability P to find a particle ! recorded within a distance of ! r ði þ 1Þ from the initial position r ðtÞ after time t according to the following equation: ! ! 2 1 ð! r ði þ 1Þ  r ðiÞÞ ! ! Pð r ði þ 1Þj r ðiÞ; tÞdV ¼ dV ; exp  4Dt ð4pDtÞd=2 normal diffusion The probability distribution is suited to distinguish subpopulations with multiple diffusion coefficients, which can be nontrivial to identify in MSD plots.

5.4. Interpretation of mRNA tracking data Upon successful labeling, imaging, and tracking of mRNAs, SPT data provide a rich source of information. Linking the quantitative analysis of mRNA movement to a biological process or function is another big challenge. SPT often yields observational information about the nature of mRNA movement. Observational reports about the travels of mRNA in the nucleus have utilized mean-squared displacement (Politz et al., 2006; Shav-Tal et al., 2004) as well as jump-distance histograms (Grunwald et al., 2008a; Siebrasse et al., 2008) to describe the nuclear environment that the mRNA encounters. In these reports, MSD measurements yielded an average diffusion coefficient, while mean jump distance was used to calculate the mean diffusion coefficient of discernable populations in unique compartments. While nuclear SPT of mRNA usually yields diffusion coefficients of mRNA in the various nuclear compartments or the entire nucleus, utilization of SPT of mRNA in the cytoplasm needs to distinguish between diffusing mRNAs and ones that are transported along cytoskeletal elements. The use of MSD to analyze SPT data is capable of comparing the distribution of distinct motility populations of mRNA in the cytoplasm of cells. For example, MSD has been used to compare the relative population of

402

Hye Yoon Park et al.

diffusing mRNAs compared to transported or static mRNAs of reporter constructs with and without the b-actin 30 UTR (Fusco et al., 2003). Alternatively, a specific aspect of active transport may be measured, such as the average velocity, maximum velocity, or the average length of transport path. Rook et al. (2000) measured a variety of aspects of active transport of CamKII alpha mRNA in the neuronal dendrites pre- and post-potassium chloride (KCl) stimulation. They measured the percent motile mRNA granules, distance traveled, average rate, and the maximum rate of active transport. The comparison of the motility of mRNA prior and following a treatment or knockdown of RNA-binding proteins is a direct way to measure cellular elements responsible for mRNA localization or means in which mRNA localization can be induced. In this study, it was shown that following KCl stimulation, there was a shift of movement from oscillatory to anterograde. In a more recent study of CamKII mRNA in dendrites, MSD was used to compare the relative abundance of actively transported and nonmotile mRNAs in wild type and FMRP knockout neurons. SPT of mRNA also allowed the measurement of the maximal and mean granule velocity in both the anterograde and retrograde directions in dendrites (Dictenberg et al., 2008). The measurement of mRNA velocity along the cytoskeleton is an important stepping stone toward understanding more about the nature of active transport in different cell types and situations. Because cytoskeletal filaments are required for active transport, studying the contribution of cytoskeletal elements and molecular motors on mRNA localization is often accomplished by chemical disruption of the cytoskeleton or overexpression of the dominant negative motors (Mingle et al., 2005; Sundell and Singer, 1991; Zhang et al., 1999). Conversely, live measurements of mRNA being actively transported can provide an insight into how cells actively facilitate the localization of mRNA to discrete locations.

6. Conclusions SPT is a useful tool for monitoring the behavior of individual molecules in living cells, providing new information about dynamic heterogeneity. Current technological advances in SPT used in conjunction with the MS2-labeling system have allowed more accurate and extended tracking of mRNAs in cells. Single mRNA tracking studies are now elucidating the mechanisms of mRNA transport and localization in various cell types. Despite the remarkable recent progress, many important questions remain to be answered. A clear picture of the cause-and-effect relationship between mRNA localization and cell physiology will likely emerge as the in vivo dynamics of mRNA is revealed. Furthermore, multicolor imaging of single

Single mRNA Tracking

403

mRNAs interacting with their diverse binding partners will provide a more comprehensive picture of the molecular pathways in live cells.

ACKNOWLEDGMENTS This work was supported by National Institutes of Health grant EB2060. H. Y. P. was also supported by National Research Service Awards 5T32 HL007675 and 1F32 GM087122.

REFERENCES Ainger, K., et al. (1993). Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes. J. Cell. Biol. 123, 431–441. Anderson, C. M., et al. (1992). Tracking of cell surface receptors by fluorescence digital imaging microscopy using a charge-coupled device camera. Low-density lipoprotein and influenza virus receptor mobility at 4 degrees C. J. Cell. Sci. 101(Pt 2), 415–425. Axelrod, D., et al. (1976). Mobility measurement by analysis of fluorescence photobleaching recovery kinetics. Biophys. J. 16, 1055–1069. Barak, L. S., and Webb, W. W. (1982). Diffusion of low density lipoprotein–receptor complex on human fibroblasts. J. Cell. Biol. 95, 846–852. Beach, D. L., et al. (1999). Localization and anchoring of mRNA in budding yeast. Curr. Biol. 9, 569–578. Bertrand, E., et al. (1998). Localization of ASH1 mRNA particles in living yeast. Mol. Cell. 2, 437–445. Bobroff, N. (1986). Position measurement with a resolution and noise-limited instrument. Rev. Sci. Instrum. 57, 1152. Bratu, D. P., et al. (2003). Visualizing the distribution and transport of mRNAs in living cells. Proc. Natl. Acad. Sci. USA 100, 13308–13313. Chao, J. A., et al. (2008a). Using the bacteriophage MS2 coat protein-RNA binding interaction to visualise RNA in living cells. In ‘‘Probes and Tags to Study Biomolecular Function,’’ (L. W. Miller, ed.), Wiley-VCH, Weinheim. Chao, J. A., et al. (2008b). Structural basis for the coevolution of a viral RNA–protein complex. Nat. Struct. Mol. Biol. 15, 103–105. Cheezum, M. K., et al. (2001). Quantitative comparison of algorithms for tracking single fluorescent particles. Biophys. J. 81, 2378–2388. Churchman, L. S., and Spudich, J. A. (2007). Colocalization of fluorescent probes: Accurate and precise registration with nanometer resolution. In ‘‘Single-Molecule Techniques: A Laboratory Manual,’’ (P. Selvin and T. Ha, eds.), pp. 73–84. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. Condeelis, J., and Singer, R. H. (2005). How and why does beta-actin mRNA target? Biol. Cell 97, 97–110. Crocker, J. C., and Grier, D. G. (1996). Methods of digital video microscopy for colloidal studies. J. Colloid Interface Sci. 179, 298–310. De Brabander, M., et al. (1985). Probing microtubule-dependent intracellular motility with nanometer particle video ultramicroscopy (nanovid ultramicroscopy). Cytobios 43, 273–283. Dictenberg, J. B., et al. (2008). A direct role for FMRP in activity-dependent dendritic mRNA transport links filopodial-spine morphogenesis to fragile X syndrome. Dev. Cell 14, 926–939.

404

Hye Yoon Park et al.

Edidin, M., et al. (1991). Lateral movements of membrane-glycoproteins restricted by dynamic cytoplasmic barriers. Science 254, 1379–1382. Eggeling, C., et al. (2005). Molecular photobleaching kinetics of Rhodamine 6G by oneand two-photon induced confocal fluorescence microscopy. Chemphyschem 6, 791–804. Einstein, A. (1905). Investigations on the Theory of the Brownian Movement. Dover Publications, Inc., New York. Farina, K. L., and Singer, R. H. (2002). The nuclear connection in RNA transport and localization. Trends Cell Biol. 12, 466–472. Feder, T. J., et al. (1996). Constrained diffusion or immobile fraction on cell surfaces: A new interpretation. Biophys. J. 70, 2767–2773. Fusco, D., et al. (2003). Single mRNA molecules demonstrate probabilistic movement in living mammalian cells. Curr. Biol. 13, 161–167. Gelles, J., et al. (1988). Tracking kinesin-driven movements with nanometre-scale precision. Nature 331, 450–453. Ghosh, R. N., and Webb, W. W. (1994). Automated detection and tracking of individual and clustered cell-surface low-density-lipoprotein receptor molecules. Biophys. J. 66, 1301–1318. Grunwald, D., et al. (2008a). Probing intranuclear environments at the single-molecule level. Biophys. J. 94, 2847–2858. Grunwald, D., et al. (2008b). Cell biology of mRNA decay. Methods Enzymol. 448, 553–577. Ishihama, Y., and Funatsu, T. (2009). Single molecule tracking of quantum dot-labeled mRNAs in a cell nucleus. Biochem. Biophys.l Res. Commun. 381, 33–38. Ja¨hne, B., et al. (2007). Data analysis. In ‘‘Springer Handbook of Experimental Fluid Mechanics,’’ (C. Tropea, et al., eds.), pp. 1474–1481. Springer, Berlin/Heidelberg. Jaqaman, K., et al. (2008). Robust single-particle tracking in live-cell time-lapse sequences. Nat. Methods 5, 695–702. Kao, H. P., and Verkman, A. S. (1994). Tracking of single fluorescent particles in three dimensions: Use of cylindrical optics to encode particle position. Biophys. J. 67, 1291–1300. Knowles, R. B., et al. (1996). Translocation of RNA granules in living neurons. J. Neurosci. 16, 7812–7820. Kusumi, A., et al. (1993). Confined lateral diffusion of membrane-receptors as studied by single-particle tracking (nanovid microscopy)—Effects of calcium-induced differentiation in cultured epithelial-cells. Biophys. J. 65, 2021–2040. Kusumi, A., et al. (1998). Application of laser tweezers to studies of the fences and tethers of the membrane skeleton that regulate the movements of plasma membrane proteins. Methods Cell Biol. 55, 173–194. Kusumi, A., et al. (2005). Paradigm shift of the plasma membrane concept from the twodimensional continuum fluid to the partitioned fluid: High-speed single-molecule tracking of membrane molecules. Ann. Rev. Biophys. Biomol. Struct. 34, 351–378. Lawrence, J. B., and Singer, R. H. (1986). Intracellular-localization of messenger-rnas for cytoskeletal proteins. Cell 45, 407–415. Levi, V., and Gratton, E. (2007). Exploring dynamics in living cells by tracking single particles. Cell Biochem. Biophys. 48, 1–15. Levi, V., et al. (2005). 3-D particle tracking in a two-photon microscope: Application to the study of molecular dynamics in cells. Biophys. J. 88, 2919–2928. Malik, N. A., et al. (1993). Particle tracking velocimetry in 3-dimensional flows. 2. Particle tracking. Exp. Fluids 15, 279–294. Martin, K. C., and Ephrussi, A. (2009). mRNA localization: Gene expression in the spatial dimension. Cell 136, 719–730.

Single mRNA Tracking

405

Mingle, L. A., et al. (2005). Localization of all seven messenger RNAs for the actinpolymerization nucleator Arp2/3 complex in the protrusions of fibroblasts. J. Cell Sci. 118, 2425–2433. Nirmal, M., et al. (1996). Fluorescence intermittency in single cadmium selenide nanocrystals. Nature 383, 802–804. Pederson, T. (2001). Fluorescent RNA cytochemistry: Tracking gene transcripts in living cells. Nucleic Acids Res. 29, 1013–1016. Perrin, J. (1913). Atoms. D. Van Nostrand Company, New York. Politz, J. C. R., et al. (2006). Rapid, diffusional shuttling of Poly(A) RNA between nuclear speckles and the nucleoplasm. Mol. Biol. Cell 17, 1239–1249. Qian, H., et al. (1991). Single particle tracking. Analysis of diffusion and flow in twodimensional systems. Biophys. J. 60, 910–921. Querido, E., and Chartrand, P. (2008). Using fluorescent proteins to study mRNA trafficking in living cells. In ‘‘ Fluorescent Proteins,’’ 2nd ed., Vol. 85, 273 pp. Elsevier Academic Press Inc., San Diego. Rasnik, I., et al. (2006). Nonblinking and long-lasting single-molecule fluorescence imaging. Nat. Methods 3, 891–893. Rasnik, I., et al. (2007). Electronic cameras for low-light microscopy. In ‘‘Digital Microscopy,’’ (G. Sluder and D. E. Wolf, eds.), p. 228. Elsevier Academic Press, Amsterdam. Reid, D. B. (1979). An algorithm for tracking multiple targets. IEEE Trans. Autom. Control AC-24. Rodriguez, A. J., et al. (2007). Imaging mRNA movement from transcription sites to translation sites. Semin. Cell Dev. Biol. 18, 202–208. Rook, M. S., et al. (2000). CaMKII alpha 30 untranslated region-directed mRNA translocation in living neurons: Visualization by GFP linkage. J. Neurosci. 20, 6385–6393. Sage, D., et al. (2005). Automatic tracking of individual fluorescence particles: Application to the study of chromosome dynamics. IEEE Trans. Image Process. 14, 1372–1383. Santangelo, P. J., et al. (2009). Single molecule-sensitive probes for imaging RNA in live cells. Nat. Methods 6, 347–349. Saxton, M. J. (1997). Single-particle tracking: The distribution of diffusion coefficients. Biophys. J. 72, 1744–1753. Saxton, M. J., and Jacobson, K. (1997). Single-particle tracking: Applications to membrane dynamics. Ann. Rev. Biophys. Biomol. Struct. 26, 373–399. Sbalzarini, I. F., and Koumoutsakos, P. (2005). Feature point tracking and trajectory analysis for video imaging in cell biology. J. Struct. Biol. 151, 182–195. Serge, A., et al. (2008). Dynamic multiple-target tracing to probe spatiotemporal cartography of cell membranes. Nat. Methods 5, 687–694. Shan, J. G., et al. (2003). A molecular mechanism for mRNA trafficking in neuronal dendrites. J. Neurosci. 23, 8859–8866. Shav-Tal, Y., and Singer, R. H. (2005). RNA localization. J. Cell Sci. 118, 4077–4081. Shav-Tal, Y., et al. (2004). Dynamics of single mRNPs in nuclei of living cells. Science 304, 1797–1800. Sheetz, M. P., et al. (1989). Nanometre-level analysis demonstrates that lipid flow does not drive membrane glycoprotein movements. Nature 340, 284–288. Siebrasse, J. P., et al. (2008). Discontinuous movement of mRNP particles in nucleoplasmic regions devoid of chromatin. Proc. Natl. Acad. Sci. USA 105, 20291–20296. Smith, A. M., et al. (2008). Bioconjugated quantum dots for in vivo molecular and cellular imaging. Advan. Drug Deliv. Rev. 60, 1226–1240. Speidel, M., et al. (2003). Three-dimensional tracking of fluorescent nanoparticles with subnanometer precision by use of off-focus imaging. Opt. Lett. 28, 69–71. Sundell, C. L., and Singer, R. H. (1991). Requirement of microfilaments in sorting of actin messenger-RNA. Science 253, 1275–1277.

406

Hye Yoon Park et al.

Thompson, R. E., et al. (2002). Precise nanometer localization analysis for individual fluorescent probes. Biophys. J. 82, 2775–2783. Toprak, E., et al. (2007). Three-dimensional particle tracking via bifocal imaging. Nano Lett. 7, 2043–2045. Tyagi, S. (2009). Imaging intracellular RNA distribution and dynamics in living cells. Nat. Methods 6, 331–338. Vallotton, P., et al. (2003). Recovery, visualization, and analysis of actin and tubulin polymer flow in live cells: A fluorescent speckle microscopy study. Biophys. J. 85, 1289–1306. Widengren, J., et al. (2007). Strategies to improve photostabilities in ultrasensitive fluorescence spectroscopy. J. Phys. Chem. A. 111, 429–440. Wieser, S., and Schutz, G. J. (2008). Tracking single molecules in the live cell plasma membrane—Do’s and Don’t’s. Methods 46, 131–140. Yamagishi, M., et al. (2009). Single-molecule imaging of beta-actin mRNAs in the cytoplasm of a living cell. Exp. Cell Res. 315, 1142–1147. Yilmaz, A., et al. (2006). Object tracking: A survey. ACM Comput. Surv. 38, 1–45. Zhang, H. L., et al. (1999). Neurotrophin regulation of beta-actin mRNA and protein localization within growth cones. J. Cell Biol. 147, 59–70. Zimyanin, V. L., et al. (2008). In vivo Imaging of oskar mRNA transport reveals the mechanism of posterior localization. Cell 134, 843–853.

C H A P T E R

N I N E T E E N

Single-Molecule Sequencing: Sequence Methods to Enable Accurate Quantitation Christopher Hart, Doron Lipson, Fatih Ozsolak, Tal Raz, Kathleen Steinmann, John Thompson, and Patrice M. Milos Contents 1. Introduction 2. Basic Principles of Single-Molecule Sequencing 3. Preparation of Genomic DNA for Single-Molecule Sequencing 3.1. DNA fragmentation and quantitation 3.2. Poly-A tailing 3.3. 30 end blocking 4. Bacterial Genome Sequencing 4.1. Preparation and sequencing of bacterial DNA 4.2. Assessment of coverage and lack of bias 5. Human Genome Sequencing and Quantitation 5.1. Copy number variation 6. Chromatin Immunoprecipitation Studies 6.1. Preparation of ChIP DNA 6.2. ChIP DNA poly-A tailing 6.3. ChIP DNA 30 blocking 7. Digital Gene Expression for Transcriptome Quantitation 7.1. Methodology for single-molecule sequencing digital gene expression 7.2. Demonstration of DGE counting reproducibility 8. Summary Acknowledgments References

408 409 410 411 415 415 416 416 417 418 420 421 422 423 423 423 424 428 428 428 430

Abstract HelicosÒ Single-Molecule Sequencing provides a unique view of genome biology through direct sequencing of cellular and extracellular nucleic acids in an unbiased manner, providing both quantitation and sequence information. Using Helicos BioSciences Corporation, One Kendall Square, Cambridge, Massachusetts, USA Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72002-4

#

2010 Elsevier Inc. All rights reserved.

407

408

Christopher Hart et al.

a simple sample preparation, involving no ligation or amplification, genomic DNA is sheared, tailed with poly-A and hybridized to the flow-cell surface containing oligo-dT for initiating sequencing-by-synthesis. RNA measurements involving direct RNA hybridization to the flow cell allows for the direct sequencing and quantitation of RNA molecules. From these methods, a diverse array of applications has now been successfully demonstrated with the HelicosÒ Genetic Analysis System, including human genome sequencing for accurate variant detection, ChIP Seq studies involving picogram quantities of DNA obtained from small cell numbers, copy number variation studies from both fresh tumor tissue and formalin-fixed paraffin-embedded tissue and archival tissue samples, small RNA studies leading to the identification of new classes of RNAs, and the direct capture and sequencing of nucleic acids from cell quantities as few as 400 cells with our end goal of single cell measurements. Helicos methods provide an important opportunity to researchers, including genomic scientists, translational researchers, and diagnostic experts, to benefit from biological measurements at the single-molecule level. This chapter will describe the various methods available to researchers.

1. Introduction The revolution in genomic sequencing that is currently occurring in the scientific community is heralding an exciting era of biology where experiments can be performed at a scale that fully elucidates the genome, its corresponding architecture, and the resulting transcriptome (all RNA molecules transcribed from the genome), revealing amazing new findings (Kahvejian et al., 2008). This revolution is continuing as we move into the era of single-molecule sequencing where, for the first time, we are sequencing and measuring the actual molecules present in cells and tissues. This new era offers the promise of a better understanding of the fundamental basis of health and disease. Helicos single-molecule sequencing offers the opportunity to examine billions of DNA or RNA molecules in a highly parallel fashion, scalable to sequencing of an entire human genome (Harris et al., 2008; Ozsolak et al., 2009; Pushkarev et al., 2009). While other technologies may offer similar approaches, the simplicity and the scalability of single-molecule sequencing sets it distinctly apart from next-generation sequencing technologies. Further, these same principles contribute directly to the absolute quantitative nature of the technology. By eliminating cumbersome sample preparation steps, including complex ligations and polymerase chain reactions for amplification, single-molecule sequencing offers both sequence information and reliable quantitation for many different applications. Often referred to as ‘‘third-generation’’ sequencing (Hayden, 2009), the methods involved in single-molecule sequencing demonstrate these unique principles.

Single Molecule Sequencing Methods

409

This chapter describes the methodological details for a variety of genomic applications used by the research and translational biology communities, including preparation of genomic DNA for complete genomic sequencing, copy number variation detection and chromatin immunoprecipitation (ChIP) studies. Quantitative aspects of single-molecule measurements for RNA are also described for methods associated with digital gene expression.

2. Basic Principles of Single-Molecule Sequencing Helicos single-molecule sequencing utilizes sequencing-by-synthesis methodology, involving individual nucleic acid molecules that are initially fragmented in the case of genomic DNA, melted into single strands of DNA, and poly-A tailed. These DNA molecules are then captured as individual strands of DNA through deposition onto a glass HelicosÒ Flow Cell (Fig. 19.1B) surface coated with oligo-dT-50 oligonucleotides, which are then filled with dTTP and polymerase for the purpose of filling in any remaining nucleotides complementary to the poly-A tail. Following the fill, nucleic acid templates are locked in place by the addition of fluorescently labeled dCTP, dGTP, and dATP Virtual TerminatorTM nucleotides, which incorporate as a single complementary nucleotide and prohibit subsequent extension prior to terminator cleavage. This ‘‘fill and lock’’ step ensures that each template become available for the sequencing-by-synthesis reaction (Bowers et al., 2009; Harris et al., 2008). Following the fill and lock step, sequencing-by-synthesis is initiated through the addition of fluorescently labeled Virtual TerminatorTM nucleotides added one at a time. Nucleotide incorporation occurs at the complementary position in the individual growing strands of DNA, using a DNA polymerase. After incorporation, unincorporated nucleotides are rinsed through the flow cell. The flow-cell surface is then illuminated with a laser and incorporation is detected by the fluorescent emission of light. The HeliScope Sequencer captures the images via a CCD camera and records which strands have incorporated a nucleotide and records positional information as well as cycle information to ensure conversion of the image to the individual DNA molecules as well as the A, C, G, or T nucleotide sequence information. After the images are captured, the terminator moiety is cleaved from the incorporated nucleotide, allowing subsequent addition of the next complementary nucleotide. Once the thousands of images, which correspond to all the channels of the flow cell, have been recorded, the fluorescent label is cleaved from the nucleotide, allowing the instrument to continue incorporation of the next nucleotide in the addition cycle. In a standard run, the HeliScope Sequencer

410

Christopher Hart et al.

A

B

25 channels - 2 flow cells per run

Figure 19.1 (A) The HeliScopeÒ Single-Molecule Sequencer. A powerful genetic analyzer that performs single-molecule sequencing chemistry and captures images of single-molecule incorporation of fluorescently labeled nucleotides, producing accurate sequences of billions of templates at a time. (B) The HelicosÒ Flow Cell. Specifically designed for sequencing chemistry used with the Sequencer, two flow cells – each with 25 channels enable a multitude of applications all benefiting from Helicos proprietary chemistry.

completes 120 cycles of individual nucleotide additions. A representative visual image taken from the HeliScope Sequencer is shown in Fig. 19.2. At the end of the run, real-time image processing has converted all the images into a complete sequence file, recording both the DNA strand position and nucleotide string addition; scientists are then able to download the sequence file and begin the alignment to appropriate reference genomic or transcriptomic sequences. To date, numerous genomes have been sequenced using the HeliScope Sequencer, including genomes from M13 virus (Harris et al., 2008), bacterial species, yeast, and Caenorhabditis elegans (Bowers et al., 2009), culminating in the world’s first sequencing of a human genome using single-molecule sequencing (Pushkarev et al., 2009). The following will describe the basic methodologies one requires in order to prepare genomic templates for single-molecule DNA and cDNA sequencing.

3. Preparation of Genomic DNA for SingleMolecule Sequencing The basic principles involved in the preparation of genomic DNA for subsequent sequencing-by-synthesis involve DNA fragmentation and quantitation, poly-A tailing, and 30 end blocking to ensure that sequence obtained

411

Single Molecule Sequencing Methods

3 2 1

Position Cycle Cycle Cycle Cycle Cycle Cycle

1 2 3 4 5

1 C A C

2 C A -

3 G A T -

X

G

G

-

Figure 19.2 Helicos Flow Cell Image and Virtual TerminatorÒ Nucleotide Incorporation. A true image derived from a section of the sequencing flow cell showing a closeup image of single molecules of DNA which have incorporated fluorescent Virtual TerminatorÒ nucleotides. The right insert shows a close-up view of the single molecules and the definition of the nucleotide incorporated at the positions 1, 2, or 3 during the cyclic addition of nucleotides.

from the 30 end of surface-bound oligoT is not contaminated with sequence from the 30 end of the hybridized DNA strands. Figure 19.3 outlines the process described in detail below.

3.1. DNA fragmentation and quantitation When quantities are not limiting, 1–3 mg of genomic DNA is typically used for single-molecule DNA sequencing of whole genomes, although much smaller quantities are also possible (see subsequent ChIP DNA Sequencing section). When the amount of DNA is low, care should be taken to use low-loss tubes and pipette tips. Addition of any type of carrier nucleic acid should be done cautiously as it could become a significant contaminant in sequencing. 3.1.1. DNA shearing 1. Prepare 1–3 mg of genomic DNA in a final volume of 120 ml 10 mM Tris 1 mM EDTA (1  TE). 2. Any method of DNA shearing can be used; however, if complete coverage is desired, the method chosen should cleave the DNA randomly and provide a 30 hydroxyl end for subsequent tailing. In the current protocol, ultrasonic shearing of the DNA is achieved using the Covaris

412

Christopher Hart et al.

Shearing

3′

3′ 5′

5′

Denature, terminal transferase + dATP AAAAAAAAAA3⬘

5′

5′ AAAAAAAAAA3⬘ 5′

AAAAAAAAAA3⬘

AAAAAAAAAA3⬘ 5′ Terminal transferase + ddATP AAAAAAAAAAddT3⬘

5′

5′ 5′

AAAAAAAAAAddT3⬘ 5′ AAAAAAAAAAddT3⬘ AAAAAAAAAAddT3⬘

True single molecule sequencing (tSMS)TM analysis

Figure 19.3 Depiction of DNA sequencing methodology. Illustration of the basic sample preparation steps of genomic DNA for single-molecule sequencing.

S2 instrument, resulting in fragmentation suitable for sequencing of the entire genomic sample. Conditions have been optimized by Covaris to allow for the use of genomic DNA ranging in length, at present, from 100 to 3000 base pairs (bp) so that the researcher can select the desired fragment size (Covaris, Woburn, MA; http://www.covarisinc.com). For typical genomic DNA sequencing using single reads, DNA is fragmented to an average size of 200–300 bp. For paired read sequencing in

Single Molecule Sequencing Methods

413

which two or more regions of the same DNA fragment are sequenced, fragmentation of 1500 bp is optimal to provide spacer gaps ranging from 100 to 700 nucleotides in length. 3. Conditions vary for the various fragmentation sizes: For example, to shear DNA to 200 bp, the DNA is sheared in Covaris microTubes using 3 cycles of 60 s, 10% duty cycle, intensity 5, and 200 cycles per burst. 4. Transfer the DNA to a clean 1.5-ml microtube. At this point, the DNA sample can be stored at 20  C. 3.1.2. DNA size selection The DNA sample is subsequently cleaned using Agencourt AMPureÒ beads to remove small nucleic acids, nucleotides, and salts that may be present in the sheared sample. 1. Adjust the DNA volume to 100 ml. 2. Warm AMPure Bead solution to room temperature (RT). Vortex to resuspend. 3. Transfer DNA sample to 1.5-ml tube and add water to bring each sample to 100 ml. Vortex the beads again and add 300 ml AMPure Bead slurry. 4. Incubate at RT for 30 min. Shake the tube every 10 min. 5. Briefly centrifuge at low speed, capture beads on DynalÒ magnetic stand for 5 min and carefully aspirate supernatant. 6. Wash beads twice with 700 ml freshly prepared 70% (v/v) ethanol. 7. Briefly centrifuge, place on magnet, remove ethanol, and dry pellet completely at RT for 5–7 min. Cracks will form when the pellet is dried sufficiently. 8. To elute the sheared DNA from the AMPure beads add 20 ml of water, pipette the beads and water up and down 20 times and place the tube back on the Dynal magnet. 9. Collect the 20 ml volume and transfer to a new 1.5-ml tube. 10. Repeat this process again to remove any remaining DNA on the AMPure beads. DNA will now be in the 40 ml volume. 3.1.3. Concentration estimation of 30 ends for subsequent poly-A tailing 1. In order to effectively tail the 30 ends of the genomic DNA, one must determine the approximate concentration of 30 ends, which requires a determination of the average fragment size of the sheared DNA obtained by running a 2-ml DNA aliquot on a 4–20% gradient Tris Borate EDTA (TBE) polyacrylamide gel. 2. DNA standards of 1000 and 25 bp ladders are included for size comparison.

414

Christopher Hart et al.

3. To estimate the size of the sheared product, compare the middle of the DNA smear to the size standards. An example gel is shown in Fig. 19.4. 4. Determine the double-stranded DNA concentration using a NanoDrop 1000 or 8000 spectrophotometer. Calculate the pmoles of the ends in the sample using the following formula. pmol 30 termini=ml ¼ XXng DNA=ml  ð103 pg=ngÞ  ðpmole=660 pgÞ  ð1=average fragment size as determined from gelÞ  2ð30 termini=dsDNA moleculeÞ

1

2

3

1: 1 kb DNA ladder 2: Sheared sample 3: 25 bp DNA ladder

Middle of smear (~200 bp)

Figure 19.4 Example of gel sizing for sheared DNA. A 4–20% TBE gradient gel is used to assess the successful fragmentation of genomic DNA for subsequent poly-A tailing. Size standards of 1 kilobase ladder and 25 bp ladder are used to estimate average fragmentation length. Compare the average size in the middle of the sample smear to the size standards.

Single Molecule Sequencing Methods

415

3.2. Poly-A tailing The DNA fragments must be modified at their 30 ends with a poly-A tail to allow for efficient hybridization onto the oligonucleotide-coated Helicos Flow Cell. Conditions are optimized to allow the addition of 90–200 polyAs to the single-stranded DNA molecules. 1. Prepare a sample DNA Tailing Mix assuming a 3.0-pmole sample reaction. For one reaction—4 ml 10 Terminal Transferase buffer, 4 ml 2.5 mM CoCl2, 2 ml Terminal Transferase Enzyme (20 U/ml), 3.9 ml Helicos supplied Poly-A Tailing dATP and 1.1 ml deionized water (dH2O). Please note: For the Tailing Control Tube, adjust the Poly-A Tailing dATP to 1.3 ml and the dH2O to 3.7 ml. 2. Place the 3.0 pmole sheared DNA sample into a 200-ml PCR tube. 3. At the same time, prepare a separate 200-ml PCR tube with tailing control sample which consists of 0.8 pmoles of your DNA sample and 0.2 pmoles of tailing oligo control supplied by Helicos to monitor efficiency of tailing. 4. Denature the DNA by placing the sheared DNA and tailing oligo control tubes in a PCR Thermocycler, at 95  C for 5 min. Snap cool by placing tubes in an aluminum block prechilled on an ice slurry for 2 min to prevent reannealing of the denatured single-stranded DNA. 5. Add 15 ml of Sample Tailing Mix or Control Tailing Mix to each DNA tube. Mix by pipetting up and down 10 times. Collect liquid contents by centrifuging briefly. 6. Place the tubes in the thermocycler using the following conditions: 37  C for 60 min, 70  C for 10 min, maintain at 4  C until ready to proceed to next step. 7. Success of tailing is determined by monitoring the oligo control tailing. Twenty microliters of the oligo control is run on a 4–20% polyacrylamide gel in TBE alongside 100 and 25 bp ladders. An example of successfully sheared and tailed DNA is shown in Fig. 19.4. 8. Control-tailed oligos should migrate anywhere between 250 and 600 bp, indicating the sample is properly poly-A tailed with a desired tail length of between 90 and 200 dA.

3.3. 30 end blocking During flow-cell hybridization, the poly-A tail on the DNA sequencing templates may align imperfectly to the oligo-dT surface on the Helicos Flow Cell surface. This may result in the generation of a recessed 30 end that can serve as a substrate for the sequencing-by-synthesis reaction. To prevent the incorporation of fluorescent Virtual Terminator nucleotides at that end

416

Christopher Hart et al.

of the duplex, the 30 ends of sheared DNA molecules are modified with a dideoxy terminator, using the following protocol. 1. Following the poly-A tailing, heat denature the DNA at 95  C for 5 min in the thermocycler. Immediately remove and snap cool for a minimum of 2 min by placing in the ice-cooled aluminum block. 2. Add 0.3 ml of 500 mM Biotin ddATP to each tube. 3. Add 2 ml Terminal Transferase (20 U/ml) to each tube. Mix thoroughly by pipetting up and down 10 times. 4. Collect contents by brief centrifugation. 5. Return to the thermocycler and run the following conditions: 37  C for 60 min, 70  C for 10 min, maintain at 4  C until ready to proceed to next step. Samples are now ready for hybridization to the Helicos Flow Cell for subsequent sequencing-by-synthesis. DNA concentrations in the range of 150–300 pM are utilized for each Helicos Flow Cell Channel typically in a 20-ml loading volume.

4. Bacterial Genome Sequencing Helicos BioSciences has applied the above DNA sample preparation methodology to the sequencing of three bacterial genomes to demonstrate the principles of single-molecule sequencing—the simplicity of the sample preparation, the lack of amplification requirement, and the corresponding lack of G þ C biases (Dohm et al., 2008), as well as the evenness of coverage across a broad range of bacterial genomes, including Escherichia coli K12 MG1655, Staphylococcus aureus USA 3000, and Rhodobacter sphaeroides 2.4.1. The percentage of guanine and cytosine nucleotides (%GC) content of the genomes of these organisms represents the entire range of %GC content of kilobase-sized windows found in the human genome (Table 19.1). They have therefore been employed as reference genomes to test the ability of sequencing platforms to sequence the human genome. Achieving accurate and even coverage across these bacterial genomes demonstrates an absence of sequence content bias, which thus provides both sequence information as well as quantitative accuracy.

4.1. Preparation and sequencing of bacterial DNA 1. Shear and prepare 1 mg of bacterial DNA obtained from each species to 250–300 bp as described in Section 3. 2. Following sample preparation, load 150–300 pM of each bacterial DNA into individual flow-cell channels in a volume of 20 ml and

417

Single Molecule Sequencing Methods

Table 19.1 Reference bacterial genomes containing diverse genome sequence content

E. coli R. sphaeroides S. aureus

Genome size (Mb)

GC content (%)

4.6 4.3 2.8

50.8 68.8 32.7

Genomic size and G þ C content of bacterial genomes used for demonstration of high-throughput sequencing methodologies due to diverse genomic content.

sequence-by-synthesis for 120 nucleotide cycle additions via the HeliScope Sequencer in an 8-day run in which both flow cells are utilized. 3. Align the single-molecule sequence reads obtained at run completion to the corresponding bacterial reference genomes using the Helicos IndexDP Genomic aligner (available at Helicos HeliSphere Technology Center http://open.helicosbio.com/mwiki/index.php/Main_Page). 4. The resulting throughput yields 12–20 million aligned reads per flowcell channel or, given the two flow cells totaling 50 channels per run, 0.6–1 billion alignable reads per run. 5. A single channel provides upward of 80–120 coverage for these bacterial genomes, depending on the genome size, and represents some 3–4 more coverage than is required for accurate consensus calling. Figure 19.5 shows the alignment view of reads and coverage within a selected region of the E. coli genome, which allows one to compare the sequence reads mapped to the region of a 5-kilobase pairs (kbp) window against the background of varying GC content in this same region. Coverage of sequence reads remains evenly distributed. Figure 19.5 also shows the read alignment, demonstrating the accuracy of the sequence information obtained.

4.2. Assessment of coverage and lack of bias The ability to achieve consistent coverage across these genomes with special emphasis on regions of highly varying GC content is a hallmark of singlemolecule sequencing. To demonstrate consistent coverage across the genomic regions of the three bacteria, we have plotted in Fig. 19.6 the coverage depth of single-molecule sequence reads binned across the bacterial genomic sequence and similarly plotted their known GC content in the same windows alongside the reference genomes. The HeliScope Sequencer produces even coverage across the entire span of genomic sequence content within a genome, even in the case of very G þ C rich (R. sphaeroides) and highly A þ T rich (S. aureus) genomes.

418

Christopher Hart et al.

1.0

130 Coverage GC content

120

0.8

Coverage

110 100

0.6

90 0.4

80 70

0.2

60 50 1,161,000

1,162,000

1,164,000 1,163,000 Genome position

1,165,000

0.0

CGTGAAATACTGGCGCATCCGCTGGTGGGAGGGCTGATTCTCTTTACGCGTAACTATCATGATCCTGCCCAGTTACGTGAACTGGTGCGCCAGATCCGCG

Figure 19.5 Detailed view of reads and coverage within an arbitrary position within the E. coli genome. Top panel shows sequence read coverage across each 5 kb region of the E. coli genome along (red smooth line) with the regional GC coverage (green jagged line). In each case positional statistics are derived from a sampling of the 500 bp upstream and 500 bp downstream regions. Bottom panel shows sequence reads as they aligned to the genome within the region demarcated by the vertical black lines in the top panel.

5. Human Genome Sequencing and Quantitation Whole genome sequencing has been successfully achieved by scientists at Stanford University using Helicos single-molecule sequencing methods and the HeliScope Sequencer. Pushkarev et al. (2009) utilized 200 pM of poly-A tailed human genomic DNA per Helicos Flow Cell channel and loaded some 170 channels with the genomic DNA. The researchers obtained 148 Gigabases of raw sequence of, on average, 33-nt read length to achieve, on average, a 28 coverage of a human genome. Some 90% of the human genome was covered using this initial genome sequencing methodology. Sequence variants were identified as described in Pushkarev et al. (2009), which included data on copy number variation found within the human genome sequence.

7

90

6

80

5

70 60

4

50

3

40

2

30

1

20

0 0.0

0.2

0.4

0.6

0.8

Mean coverage

Number of windows

A

10 1.0 140

5

120

4

100

3

80

2

60

1

40

0

20

−1 0.0

0.2

Number of windows

C

0.4 0.6 Fraction GC

0.8

0 1.0

7

60

6

50

5

40

4

30

3

20

2

10

1 0 0.0

Mean coverage

Number of windows

6

Mean coverage

Fraction GC

B

0.2

0.4

0.6

0.8

0 1.0

Fraction GC

Figure 19.6 Comparison of genomic sequence coverage across differing G þ C content within bacterial genomes. (A) E. coli (51% G þ C). (B) S. aureus (32% G þ C). (C) R. sphaeroides (70% G þ C). Single molecule DNA sequencing provides minimal sequence bias across diverse genomic content. Sequence reads were mapped to each genome and the number of reads which map in discrete bins of the genome are plotted (red line) versus the expected bins if the mapping was perfect. Obtaining a signal as nearly identical to each other demonstrates the unique ability to sequence across diverse GC and AT rich regions. The analysis utilized a 200-bp sliding window, the local GC content and observed mean sequencing coverage were tabulated. Windows were then aggregated into GC-content bins ranging from 0 to 1 with a step size of 0.1. Plotted is the mean coverage (RED; Right Y axis) for each window within each of the aggregated GC content bins (BLACK; Left Y axis). A distinguishing feature of the Helicos SMS approach appears to be the minimal shifts in coverage across the vast majorities of sequence contexts.

420

Christopher Hart et al.

5.1. Copy number variation Copy number variation studies provide a methodology for detecting amplification and deletion of genomic regions across the human genome and often represent critically important knowledge of mutational events occurring in cancer genomes. Given the demonstration of evenness of sequence coverage in the bacterial strains representing the diversity of sequence found in the human genome, the use of single-molecule sequencing with the HeliScope Sequencer for an assessment of copy number variation represents an important, cost-effective method. 1. When available, prepare 1–2 mg of genomic DNA as described in Section 3. Less material may be utilized if sample is limited. This material may be obtained as genomic DNA prepared from tissue, blood, and formalin-fixed paraffin-embedded (FFPE) genomic DNA. 2. In the case of FFPE DNA, visualize the isolated DNA on a 1% agarose gel to determine the size of the genomic DNA. It is possible that, depending on the fixation of the tissue from which the DNA was obtained, the DNA still consists of high-molecular weight DNA and, if above 2–3 kbp, will require additional shearing as described in Section 3.1.1. 3. If the FFPE DNA falls below the size range of 2–3 kbp, proceed directly to Section 3.1.2 (DNA size selection) to ensure removal of small molecular weight DNA that can interfere with DNA sequence yields. 4. Following preparation of poly-dA tailed genomic DNA, load 150– 300 pM of genomic DNA on each Helicos Flow Cell channel for the HeliScope Sequencer. 5. Depending on the desired level of resolution required for localization of the regions of amplification and duplication, a decision will be required regarding the depth of sequence coverage desired. At present performance, one channel of the HeliScope Sequencer provides 0.2–0.3 coverage of the human genome. This allows you to group sequence reads by using ‘‘bins’’ which can be between 10 and 50-kilobase-sized segments of the human genome. This resolution allows sufficient coverage for detection of amplification and duplication events, including loss of heterozygosity and two- to threefold amplification across the entire human genome. Figure 19.7 summarizes the copy number variation data obtained from a human cancer cell line in which approximately 100 Mio sequence reads were mapped to the genome at a read bin size of 1 kbp intervals. These data are compared to existing comparative genomic hybridization data using an array technology. Peaks of amplification are easily detected, and the peak intensities reflect the extent of amplification. We refer also to the copy number variation data obtained from the first single-molecule human genome sequence (Pushkarev et al., 2009). To further demonstrate the power of single-molecule sequencing technology, data used for the comprehensive

421

Single Molecule Sequencing Methods

140 120 Coverage

100 80 60 40 20 0 Position

~30 Mb

Figure 19.7 Detection of genomic amplification comparing Helicos single-molecule sequencing to comparative genomic hybridization. Genomic DNA from a breast cancer cell line was isolated, sheared, tailed, and sequenced according to the methods described in Section 5.1. Following sequencing, sequence reads were aligned to the human genome, binned into genomic bin sizes of 10 kb and bin sizes are plotted along the a 30 Megabase region chromosomal 20 (top panel). Regions of amplification are clearly detected in well described regions of Chr 20 previously identified using CGH arrays (bottom panel). (CGH Data: Courtesy of Genome Institute of Singapore).

view shown in Fig. 19.7 are replotted as individual channels of HeliScope Sequencer data and displayed in a 14-kbp region of the genome with 1-kbp smoothing of the read peaks (Fig. 19.8). These data reflect the consistency as well as the resolution achieved in single channels, allowing one to detect a region of five- to sevenfold amplification in this region.

6. Chromatin Immunoprecipitation Studies Helicos single-molecule sequencing technology is ideally suited also for another area of genomic science where accurate quantitation is key, ChIP studies (Goren et al., 2010). This method requires no ligation, amplification,

422

Christopher Hart et al.

2.0

1.5

1.0

0.5

0.0

2000

4000

6000

8000

10,000

12,000

14,000

Figure 19.8 Visualization of genomic amplification obtained from single HeliScope Sequencer Channels. Genomic reads used in Fig. 19.7 shown aligned to a segment of Chr 20 defining the boundary of amplification. Ten independent HeliScope Sequencer channels demonstrate the ability of a single channel of HeliScope Sequencer reads to clearly define the boundary of genomic amplification.

or complicated cleanup steps—all of which have the potential to induce sample loss and bias. The Helicos ChIP Seq methodology consists of a 1-h 30 poly-A tailing step followed by a 1-h 30 dideoxy-blocking step. Recommended starting material consists of 6–9 ng ChIP DNA (average fragment size 400–500 bp), although as little as 1–3 ng DNA prepared using this same method can be successfully employed. Typical yields obtained with the recommended 6–9 ng ChIP DNA from mouse or human studies allow one to load 3–6 Helicos Flow Cell Channels with a yield of 7–12 Million aligned sequence reads per channel.

6.1. Preparation of ChIP DNA 1. The quantity of ChIP DNA should first be determined with the Quant-iTTM PicoGreen dsDNA Reagent Kit (Invitrogen). 2. Samples should be free of RNA contamination and the use of the Qiagen Reaction Cleanup Kit (Qiagen) is recommended. 3. The micrococcal nuclease treatment used for fragmentation in some selected ChIP methods will generate phosphate groups on the 30 ends and thus will require end repair prior to initiating the ChIP tailing protocol. 4. One must also consider the alternative types of shearing used for fragmentation to ensure the 30 ends are amenable to direct poly-A tailing. Recommendations vary with shearing devices and one must check with the manufacturer on their advice for subsequent end repair.

Single Molecule Sequencing Methods

423

6.2. ChIP DNA poly-A tailing 1. Prepare a mix of 2 ml 10 Terminal Transferase buffer (NEB), 2 ml 2.5 mM CoCl2, ChIP DNA and Nuclease-free water 10.8 ml in a 200 ml PCR tube. 2. Place mixture in a thermocylcer and heat to 95  C for 5 min to denature the DNA. 3. Remove tube from the thermocylcer and quickly chill in an aluminum block held in an icy slurry to prevent renaturation. 4. Prepare a mix of 1 ml Terminal Transferase (1:4 diluted, 5 U/ml; NEB), 4 ml 50 mM dATP, and 0.2 ml BSA (NEB). 5. Add 5.2 ml mix to the denatured DNA on ice to bring total volume to 20 ml. 6. Place tube in the thermocycler and run the following program: 37  C for 1 h, 70  C for 10 min, maintain at 4  C until ready to proceed to next step.

6.3. ChIP DNA 30 blocking 1. Denature the 20 ml poly-A tailed ChIP DNA at 95  C for 5 min in the thermocycler, followed by immediate transfer to a prechilled aluminum block kept in an ice and water slurry. 2. Prepare a 10 ml mixture of 1 ml 10 Terminal Transferase buffer (NEB), 1 ml 2.5 mM CoCl2, 1 ml Terminal Transferase (1:4 diluted, 5 U/ml), 0.5 ml 200 mM Biotin-ddATP and 6.5 ml Nuclease-free water. 3. Add the 10 ml mixture to the denatured, poly-adenylated ChIP DNA mixture for a final volume of 30 ml. 4. Place the tube in a thermocycler and run the following program: 37  C for 1 h, 70  C for 20 min, followed by 4  C until ready to proceed to next step. 5. Add 2 pmol of a 50–80 nucleotide carrier oligonucleotide to the above terminal transferase reaction to minimize ChIP DNA loss during the sample loading steps. Since it does not contain a poly-A tail, the oligonucleotide will not hybridize to the Helicos Flow Cell. 6. Hybridize ChIP DNA sample to Helicos Flow Cell and sequence.

7. Digital Gene Expression for Transcriptome Quantitation Full transcriptome sequencing using high-throughput sequencing platforms (RNA Seq) has increased the sensitivity and accuracy of gene expression analysis. However, RNA Seq results in an inherent bias as a result of more reads from longer transcripts and thus has reduced the

424

Christopher Hart et al.

sensitivity for quantification of shorter transcripts (Oshlack and Wakefield, 2009). Further, assessing expression levels requires prior knowledge of transcript length for count normalization, which will not always be a reasonable demand, say in the case where there may be alternative splicing variants. Single-molecule sequencing digital gene expression (smsDGE) answers these difficulties and provides a route to quantitative analyses. smsDGE differs from RNA Seq in that only a single sequence read is generated per transcript molecule, regardless of its length. This permits short transcripts to be detected with the same sensitivity as long ones. Thus, whereas it would require 50 million RNA Seq reads to quantify 95% of the human transcriptome, with smsDGE 10 million reads will suffice (Lipson et al., 2009).

7.1. Methodology for single-molecule sequencing digital gene expression Sample preparation for smsDGE is minimal, requiring neither PCR amplification nor ligation. A summary of the method is shown in Fig. 19.9. Single-stranded cDNA is made directly from total RNA or poly-A þ RNA using poly-U primed reverse transcription. The RNA is then digested away using RNase, and a poly-A tail is added to the cDNA’s 30 end using terminal transferase. The sample can then be hybridized to the HeliScope flow-cell surface and sequenced (Lipson et al., 2009). 7.1.1. Single-stranded cDNA preparation 7.1.1.1. cDNA synthesis 1. Thaw RNA on ice (1–8 mg total RNA or 100–200 ng poly-Aþ RNA) preferably in 8 ml volume. 2. For sample: Prepare Master Mix A stock of 1 ml poly-U primer dTU25V (50 mM) and 1 ml dNTP nucleotide mix. Keep on ice. Prepare Master Mix B from Invitrogen SuperscriptIII kit as follows: 2 ml 10 Reverse Transcriptase buffer, 4 ml 25 mM MgCl2, 2 ml 0.1 mM DTT, 1 ml RNaseOUTTM and 1 ml Superscript III Reverse Transcriptase. 3. Aliquot 2 ml Master Mix A into a PCR tube. 4. Pipette 8 ml of RNA Sample into the PCR tube. Mix thoroughly by pipetting up and down. 5. Incubate the RNA at 65  C for 5 min. Snap cool by placing in aluminum block held in an ice water bath. 6. Add 10 ml Master Mix B Reverse Transcriptase enzyme and buffer to each tube. Mix well and spin down. 7. Place PCR tubes in the thermocycler. Incubate at 40  C for 5 min, 55  C for 50 min, 85  C for 5 min, maintain at 4  C until ready to proceed to next step.

425

Single Molecule Sequencing Methods

mRNA

5⬘

3⬘ AAAAAAAAAA Reverse transcriptase + poly(U) primer AAAAAAAAAA3⬘ UUUUUUUUUU5⬘

5⬘ 1st strand cDNA 3⬘

RNA/primer digestion

1st strand cDNA

3⬘

5⬘

Terminal transferase + dATP

1st strand AAAAAAAAAA cDNA 3⬘

5⬘ Terminal transferase + ddATP

1st strand 3⬘ ddTAAAAAAAAAA cDNA

5⬘

True single molecule sequencing (tSMS)TM analysis

Figure 19.9 Overview of method utilized for single molecule sequencing digital gene expression. Principles employed in the single-molecule sequencing digital gene expression methodology described in Section 7.

7.1.1.2. RNA digestion

1. Add 1 ml RNase H (2 U/ml) to the cDNA Synthesis reaction. Mix well and incubate at 37  C for 15 min, maintain at 4  C until ready to proceed to next step.

426

Christopher Hart et al.

2. Add 1 ml USERTM (1 U/ml) enzyme. Mix well and incubate at 37  C for 15 min, maintain at 4  C until ready to proceed to next step. 3. Add 1 ml RNase I (50 U/ml) enzyme. Mix well and incubate at 37  C for 15 min, maintain at 4  C until ready to proceed to next step. 7.1.1.3. cDNA sample cleanup

1. Warm AMPure Bead solution to RT. Vortex to resuspend. 2. Transfer cDNA sample to a 1.5-ml tube and add water to bring each sample to 50 ml. Vortex the beads again and add 65 ml AMPure Bead slurry. 3. Incubate at RT for 30 min. Shake tube every 10 min. 4. Briefly centrifuge at low speed, capture beads on Dynal magnetic stand for 5 min and carefully aspirate supernatant. 5. Wash beads twice with 200 ml freshly prepared 70% (v/v) ethanol. 6. Briefly centrifuge, place on magnet, remove ethanol, and dry pellet completely at RT for 5–7 min. 7. Elute the cDNA sample from the beads with 20 ml distilled water twice. 8. Repeat entire cDNA sample cleanup once more. Final product will be in 40 ml volume. 7.1.1.4. cDNA quantification

1. Determine the concentration and yield for each cDNA sample preparation using a small volume spectrophotometer. If the sample concentration is likely below 2 ng/ml use the Quant-iTTM OliGreenÒ ssDNA Reagent Kit and obtain spectrofluorometer reads accordingly. 2. Store samples at 20  C to continue sample preparation the next day if desired. 7.1.2. Poly-A tailing of the cDNA 7.1.2.1. Poly-A tailing reaction 1. Obtain control oligonucleotides from HelicosÒ Digital Gene Expression Assay Reagent Kit. 2. Place 20–60 ng of cDNA into a PCR tube. Add water to bring each to 28 ml. 3. Add 7.5 ml of HelicosÒ Control Oligonucleotide. Mix well and store on ice. 4. Incubate at 95  C for 5 min. Snap cool on ice. Briefly centrifuge. 5. Prepare poly-A Tailing mix of 5 ml 10 Terminal Transferase buffer, 5 ml CoCl2, 2.5 ml HelicosÒ poly-A Tailing dATP, and 1.5 ml Terminal Transferase. Mix well. 6. Add 14 ml of poly-A Tailing Mix to the cDNA and pipette up and down.

Single Molecule Sequencing Methods

427

7. Incubate at 42  C for 60 min, 70  C for 10 min, and maintain at 4  C until ready to proceed to next step. 7.1.2.2. Determining the success of the tailing reaction

1. Success of the tailing reaction is determined by monitoring the oligo control tailing. Run an aliquot of the control oligonucleotide without poly-A tail addition and the control poly-A tailed oligonucleotide alongside your cDNA tailing reaction on a 4–20% gradient polyacrylamide gel in 1 TBE, together with a 25-bp ladder. 2. Since the cDNA molecules will be of a very broad size range, assess the length of the tail added to the control oligonucleotide as a measure of the tail added to the cDNA molecules. 3. Control-tailed oligos should migrate anywhere between 225 and 450 bp of the 25-bp ladder to ensure a proper poly-A tail with a desired length between 90 and 140 dA. 7.1.3. cDNA blocking 7.1.3.1. cDNA blocking reaction 1. Incubate the cDNA sample at 95  C for 5 min. Snap cool on ice to denature. 2. Add 0.3 ml biotin-ddATP and 1.5 ml of Terminal Transferase enzyme. Mix well and spin down. 3. Incubate at 37  C for 30 min, then 70  C for 10 min, and maintain at 4  C until ready to proceed to next step. 7.1.3.2. Poly-A tailing control oligonucleotide digestion

1. Add 1 ml USER Enzyme (1 U/ml) to the cDNA sample. Mix well and spin down. 2. Incubate at 37  C for 30 min, maintain at 4  C until ready to proceed to next step. 7.1.3.3. Sample cleanup

1. Transfer cDNA from digestion step above to a 1.5-ml tube. Add water to bring volume to 60 ml. 2. Mix cDNA with 60 ml AMPure Bead slurry and incubate at RT for 30 min. Shake every 10 min. 3. Capture the beads on Dynal magnetic stand for 5 min and carefully aspirate supernatant. 4. Wash beads twice with 200 ml freshly prepared 70% (v/v) ethanol.

428

Christopher Hart et al.

5. Dry pellet completely at RT for 5–7 min. 6. Elute cDNA sample from beads with 20-ml distilled water twice. Hybridize 150–300 pmol smsDGE cDNA in 20 ml volume to Helicos Flow Cell and sequence.

7.2. Demonstration of DGE counting reproducibility To assess smsDGE reproducibility, we independently prepared three brain samples from the same RNA (poly-A RNA, Ambion, Austin TX) and sequenced each sample in a single HeliScope flow-cell channel. The three channels yielded 15, 14, and 12 million transcriptome-aligned reads. Transcript abundance ranged from 0 to 370,000 transcripts per million (tpm) with the highest seen for mitochondrial transcripts (chromosome M). Of the 28,800 transcripts included in our reference (UCSC genome database), 18,700 were present at a level higher than 1 tpm (>12 mapped reads). Transcript count reproducibility between samples was high (r ¼ 0.99) with coefficient of variation (%CV) ranging from 4% at 100 tpm to 20% at 1 tpm (Fig. 19.10).

8. Summary Methods for single-molecule sequence analysis of nucleic acids provide a diverse repertoire for quantitative and qualitative investigation of the genome and transcriptome. As such, we have attempted to describe many of the simple sample preparation methods offered to the research community. We will continue to optimize our sample preparation protocols to allow preparation and sequencing from picogram quantities of nucleic acid (Ozsolak et al., 2010)—all important for maximizing researchers abilities to perform important biological experiments with limiting biological sample amounts. These methods will serve as the starting point for the next edition of methods for single-molecule sequencing.

ACKNOWLEDGMENTS Special thanks to the many individuals who have contributed to the success of Helicos BioSciences technology—for their scientific excellence and passions to develop a remarkable new technology and all its broad applications.

429

Single Molecule Sequencing Methods

Brain sample 2 (tpm, single channel, 14.5 million reads)

A 1,000,000

100,000

10,000

1000

100

10

1 1

10

100

1000

10,000 100,000 1,000,000

Brain sample 1 (tpm, single channel, 15 million reads)

B

90 80

Coefficient of variation

70 60 50 40 30 20 10 0 0.1

1

10

100

1000

10,000

Average transcript abundance (tpm, 12–15 M reads/channel)

Figure 19.10 Demonstration of transcript counting reproducibility obtained using smsDGE methods with human brain RNA. smsDGE transcript quantification of independently prepared human brain samples. (A) Transcript count comparison between two samples run on a single flow-cell channel each. Each sample represents a single transcript (r ¼ 0.99). (B) Coefficient of variation (%CV) across transcript abundance levels between three samples at 12, 14, and 15 million transcriptome-aligned reads per channel.

430

Christopher Hart et al.

REFERENCES Bowers, J., Mitchell, M., Beer, E., Buzby, P. R., Causey, M., Efcavitch, J. W., Jarosz, M., Krzymanska-Olejnik, E., Kung, L., Lipson, D., Lowman, G. M., Marappan, S., et al. (2009). Virtual terminator nucleotides for next generation DNA sequencing. Nat. Methods 6, 593–595. Dohm, J. C., Lottaz, C., Borodina, T., and Himmelbauer, H. (2008). Substantial biases in ultra-short read data sets from high-throughput DNA sequencing. Nucleic Acids Res. 36 (16), e105. Goren, A., Ozsolak, F., Shoresh, N., Ku, M., Adli, M., Hart, C., Gymrek, M., Zuk, O., Regev, A., Milos, P. M., and Bernstein, B. E. (2010). Chromatin profiling by directly sequencing small quantities of immunoprecipitated DNA. Nat. Methods 7(1), 47–49. Harris, T. D., Buzby, P. R., Babcock, H., Beer, E., Bowers, J., Braslavsky, I., Causey, M., Colonell, J., Dimeo, J., Efcavitch, J. W., Giladi, E., Gill, J., et al. (2008). Single-molecule DNA sequencing of a viral genome. Science 320(5872), 106–109. Hayden, E. (2009). Genome sequencing: The third generation. Nature 457, 768–769. Kahvejian, A., Quackenbush, J., and Thompson, J. F. (2008). What would you do if you could sequence everything? Nat. Biotechnol. 26, 1125–1133. Lipson, D., Raz, T., Kieu, A., Jones, D. R., Giladi, E., Thayer, E., Thompson, J. F., Letovsky, S., Milos, P., and Causey, M. (2009). Quantification of the yeast transcriptome by single-molecule sequencing. Nat. Biotechnol. 27, 652–658. Oshlack, A., and Wakefield, M. J. (2009). Transcript length bias in RNA-seq data confounds systems biology. Biol. Direct 4, 14. Ozsolak, F., Platt, A., Jones, D., Reifenberger, J., Sass, L. E., McInerney, P., Thompson, J. F., Bowers, J., Jarosz, M., and Milos, P. (2009). Direct RNA sequencing. Nature 461, 814–818. Ozsolak, F., Goren, A., Gymrek, M. A., Guttman, M., Regev, A., Bernstein, B. E., and Milos, P. M. (2010). Digital transcriptome profiling from attomole-level RNA samples. Genome Res. [Epub ahead of print]. Pushkarev, D., Neff, N. F., and Quake, S. R. (2009). Single-molecule sequencing of an individual human genome. Nat. Biotechnol. 27, 847–850.

C H A P T E R

T W E N T Y

Real-Time DNA Sequencing from Single Polymerase Molecules Jonas Korlach, Keith P. Bjornson, Bidhan P. Chaudhuri, Ronald L. Cicero, Benjamin A. Flusberg, Jeremy J. Gray, David Holden, Ravi Saxena, Jeffrey Wegener, and Stephen W. Turner Contents 1. Introduction 2. Principle of Single-Molecule, Real-Time DNA Sequencing 3. Components of SMRT Sequencing 3.1. Zero-mode waveguides for observation volume confinement 3.2. ZMW surface derivatization for targeted enzyme immobilization 3.3. Phospholinked nucleotides for uninterrupted DNA polymerization 3.4. DNA polymerase—the sequencing ‘‘engine’’ 3.5. Instrument for highly parallel monitoring of sequencing reactions 3.6. DNA sequencing assay example 3.7. Data analysis 4. Single-Molecule DNA Polymerase Dynamics 4.1. Determination of single-molecule kinetic parameters 4.2. DNA polymerase pausing 5. Conclusions Acknowledgments References

432 433 435 435 435 437 440 441 443 444 446 446 447 451 452 452

Abstract Pacific Biosciences has developed a method for real-time sequencing of single DNA molecules (Eid et al., 2009), with intrinsic sequencing rates of several bases per second and read lengths into the kilobase range. Conceptually, this sequencing approach is based on eavesdropping on the activity of DNA polymerase carrying out template-directed DNA polymerization. Performed in a Pacific Biosciences, Menlo Park, California, USA Methods in Enzymology, Volume 472 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)72001-2

#

2010 Elsevier Inc. All rights reserved.

431

432

Jonas Korlach et al.

highly parallel operational mode, sequential base additions catalyzed by each polymerase are detected with terminal phosphate-linked, fluorescence-labeled nucleotides. This chapter will first outline the principle of this single-molecule, real-time (SMRTTM) DNA sequencing method, followed by descriptions of its underlying components and typical sequencing run conditions. Two examples are provided which illustrate that, in addition to the DNA sequence, the dynamics of DNA polymerization from each enzyme molecules is directly accessible: the determination of base-specific kinetic parameters from singlemolecule sequencing reads, and the characterization of DNA synthesis rate heterogeneities.

1. Introduction The ability to rapidly determine nucleic acid sequences has fundamentally transformed the biological sciences, both with respect to inquiries toward understanding biological processes and the approaches to manipulating them. Next-generation DNA sequencing methods have changed whole-genome sequencing projects into routine procedures (reviewed in Mardis, 2008) and have been adapted to other areas, such as transcriptome sequencing and epigenetics (Cloonan et al., 2008; Cokus et al., 2008; Fullwood et al., 2009; Maher et al., 2009; Yassour et al., 2009). However, despite their gains in sequencing throughput, these methods still fall short of providing the means to elicit fundamental changes in the fields of medical diagnostics, disease prevention, and treatment. Further improvements are required for higher quality and even more cost-effective sequencing of complete individual genomes and transcriptomes. DNA polymerases can be viewed as efficient DNA sequencers— engineered by nature—as they decode the sequence of a template strand by virtue of synthesizing its complementary strand. Over millions of years of molecular evolution, DNA polymerases have been optimized to rapidly and faithfully replicate genomes, and they have in turn developed many features attractive for artificial DNA sequencing methods. DNA polymerases can be very fast, with DNA synthesis rates reported in vitro as high as 750 bases per second (Tabor et al., 1987). Tens to hundreds of thousands of bases can be synthesized from a single polymerase binding event (Blanco et al., 1989). DNA polymerases can also be viewed as very frugal as only one nucleotide is consumed during each incorporation cycle. Error rates can be as low as one in 105 bases (Esteban et al., 1993) and even lower with associated proofreading activities. Finally, DNA polymerases are physically very small, enabling a high level of multiplexing on a small footprint.

Real-Time DNA Sequencing from Single Polymerase Molecules

433

Exploiting all of these characteristics directly by using polymerase as the actual sequencing engine had not been commercially feasible until recently. We have overcome the underlying technical challenges by innovations in the fields of nanofabrication, surface derivatizations, nucleotide and protein chemistries, and optics, to enable the direct, real-time interrogation of individual polymerase activities (Eid et al., 2009). Essentially, DNA is sequenced by watching with base-pair resolution what normally constitutes DNA replication occurring in dividing cells.

2. Principle of Single-Molecule, Real-Time DNA Sequencing The Single-Molecule, Real-Time (SMRTTM) DNA sequencing concept is illustrated in Fig. 20.1. The two principal technological components that facilitate SMRT sequencing are (i) zero-mode waveguide (ZMW) confinement that allows single-molecule detection at concentrations of labeled nucleotides relevant to the enzyme, and (ii) fluorescence-labeled, phospholinked nucleotides that permit observation of uninterrupted DNA polymerization. ZMW nanostructures (Fig. 20.1A) consist of dense arrays of holes,
100 nm in diameter, fabricated in a
100-nm metal film deposited on a transparent substrate (e.g., silicon dioxide) (Foquet et al., 2008; Levene et al., 2003). Each ZMW becomes a nanophotonic visualization chamber for recording an individual polymerization reaction, providing a detection volume of just
100 zeptoliters (10 21 L). This volume represents a
1000-fold improvement over diffraction-limited confocal microscopy, making it possible to observe single nucleotide incorporation events against the background created by diffusing fluorescence-labeled nucleotides. In addition to reducing the number of labeled nucleotides present inside the observation volume, the highly confined volume results in drastically shorter diffusional visitation times. This enables better temporal differentiation between events involving diffusion of labeled nucleotides through the observation volume (now typically lasting only a few microseconds) and enzymatic nucleotide incorporation events (typically lasting several milliseconds for polymerases). The second important component is phospholinked nucleotides for which the fluorescent label is attached to the terminal phosphate rather than the base, typically via a linker (Fig. 20.1B). A 100% replacement of unmodified nucleotides by phospholinked nucleotides is achieved because the enzyme cleaves away the fluorophore as part of the incorporation process, leaving behind a completely natural, double-stranded nucleic acid

434

Jonas Korlach et al.

B

A

= H N

Fluorophore

O

H N O

CH2SO3H

COOH CH SO H 2 3

O

NH

O O O O O O O O O O O O O P P P P P P OH OH OH OH OH OH

HN

N

O

O OH

Linker

Nucleotide

C Step 1

Step 2

Step 3

Step 4

Step 5

Primer

Intensity

Polymerase

Template

A

T

Time

Figure 20.1 Principle of single-molecule, real-time (SMRT) DNA sequencing. (A) Single DNA polymerase molecules with bound DNA template are immobilized to the bottom of zero-mode waveguide (ZMW) nanostructure arrays. Polymerization of the complementary DNA strand is observed in real time by detecting enzymatic processing of fluorescent phospholinked nucleotides. (B) Molecular structure of phospholinked nucleotides. Alexa Fluor 568-aminohexyltriphosphate-dTTP is shown by example (Eid et al., 2009). The arrow indicates the a-b phosphodiester bond cleavage mediated by the DNA polymerase. (C) Schematics of reactions steps involved in SMRT DNA sequencing (top), and corresponding fluorescence intensity time trace (bottom). Step 1: The DNA template/primer/polymerase complex is surrounded by diffusing phospholinked nucleotides which probe the active site. Step 2: A labeled nucleotide makes a cognate binding interaction with the template base in the DNA. During the time it is bound in the active site (typically lasting tens of milliseconds) fluorescence is emitted continuously, giving rise to a detectable pulse in the fluorescence intensity time trace. The identity of the fluorescent dye indicates which base is incorporated. Step 3: The polymerase incorporates the nucleotide into the growing nucleic acid chain by cleaving the a-b phosphodiester bond, thereby subsequently releasing the pyrophosphatelinker-fluorophore. Step 4: The polymerase translocates to the next template position. Step 5: The process repeats.

product. In SMRT sequencing, each of the four different nucleobases is labeled with a distinct fluorophore to discriminate base identities during incorporation events, thus providing sequence determination of the complementary DNA template (Fig. 20.1C). During incorporation, the enzyme holds the labeled nucleotide in the ZMWs detection volume for several milliseconds, orders of magnitude longer than the average diffusing nucleotide is present. Fluorescence is emitted continuously from the fluorophore

Real-Time DNA Sequencing from Single Polymerase Molecules

435

label during the incorporation process, causing a detectable pulse of increased fluorescence in the corresponding color channel. The pulse is terminated naturally by the polymerase releasing the pyrophosphatelinker-fluorophore group which diffuses out of the observation volume. The polymerase then translocates to the next base, and the process repeats.

3. Components of SMRT Sequencing 3.1. Zero-mode waveguides for observation volume confinement Fabrication of ZMWs with aluminum or gold as the metal cladding material was first described using a positive-tone, electron-beam lithography technique followed by reactive ion etching (Levene et al., 2003; Liu and Blair, 2003). Subsequently, other fabrication methods have been described, including negative-tone, electron-beam lithography followed by metallization and resist removal (Foquet et al., 2008; Miyake et al., 2008), focused ion beam milling (Rigneault et al., 2005), and photolithography (Foquet et al., 2008). Of these, photolithography is very attractive because of the lower cost of fabrication and its compatibility with high-volume manufacturing processes. We have recently improved the photolithographic process of ZMW fabrication, resulting in greater reproducibility, size uniformity, and ZMW shape control (Fig. 20.2). Detailed protocols for the different fabrication methods can be found in the references cited above.

3.2. ZMW surface derivatization for targeted enzyme immobilization The selective placement of an active polymerase molecule into a ZMW observation volume, immediately above the transparent ZMW floor, is an important prerequisite for efficient SMRT DNA sequencing. ZMWs put additional demands on the quality of surface preparations to achieve this because the functionalization target area is very small relative to the cladding surface area. At the same time, the surfaces should be well-passivated to prevent corrosion and nonspecific adsorption of phospholinked nucleotides which are used at much higher concentrations compared to conventional single-molecule assays. In addition, the employed surface coating reagents should exhibit low fluorescence levels, and should not interfere with enzymatic activities. For aluminum-clad ZMWs, we have developed protocols and reagent formulations meeting all of these criteria by exploiting an inherent feature of ZMW architecture in that the ZMW substrate and cladding are made of

436

Jonas Korlach et al.

B

A

Frequency

C 900 800 700 600 500 400 300 200 100 0 80 85 90 95 100 105 110 115 120 ZMW diameter (nm)

Figure 20.2 Zero-mode waveguide (ZMW) fabrication. (A) Scanning electron micrograph of the top side and (B) transmission electron micrograph of the center cross-section of a ZMW fabricated by the described photolithography protocol. The desired ZMW shape exhibits vertical bottom walls and a slightly tapered opening toward the top midway up the cladding. Scale bars ¼ 100 nm. (C) Distribution of ZMW floor diameters over an array of 3000 ZMWs with nominal diameter of 100 nm. (mean ¼ 98.4 nm; standard deviation ¼ 5.8 nm).

different materials. It combines selective passivation of the cladding surfaces using polyphosphonate chemistries (Korlach et al., 2008b) with selective functionalization of the ZMW substrate using biotin polyethylene glycol (PEG) silane (Eid et al., 2009). It results in ZMW arrays that present biotin for specific enzyme immobilization only at the ZMW glass floor, above a layer of PEG to preserve enzyme activities (Fig. 20.3A). Using this procedure, we have achieved high contrast ratios of biotin functionalization (in excess of 100:1, Fig. 20.3B), with undetectable levels of biotin PEG silane on the aluminum surface, as measured by X-ray photoelectron spectroscopy (XPS). The specificity of protein binding was also in excess of 100:1 (Fig. 20.3C). For polymerase immobilization, depending on the ZMW diameter chosen for sequencing, the concentrations of streptavidin and polymerase are adjusted to yield optimal loading which is governed by Poisson statistics (Korlach et al., 2008b) (Section 3.6).

437

Real-Time DNA Sequencing from Single Polymerase Molecules

B

A

SiO2

Al Selectivity (SiO2/Al) 110.7 ± 18.8

Polymerase/DNA complex

Aluminum cladding

C

-

Biotin block:

+

Specificity (SiO2/ SiO2biotin block) 102.4 ± 10.6

Glass substrate Polyphosphonates Biotin-PEG-silane

Streptavidin Native aluminum oxide

Figure 20.3 Surface derivatization for specific, selective immobilization of DNA polymerase in ZMWs. (A) The cladding surface is passivated with polyphosphonates, the ZMW substrate surface is functionalized with biotin PEG silane to mediate enzyme immobilization targeted to the ZMW floor. (B) Material selectivity of the surface derivatization, measured with 40 nm fluorescent neutravidin beads (Korlach et al., 2008b). The assay uses patterned substrates of 0.5 mm aluminum squares on fused silica. (C) Specificity of neutravidin binding to the glass surface. Specificity was determined by comparing neutravidin binding with a sample for which neutravidin was blocked with excess biotin before the immobilization step. Contrast means and standard deviations are for n ¼ 4 chips.

3.3. Phospholinked nucleotides for uninterrupted DNA polymerization Nucleotides with fluorescent labels attached to the terminal phosphate were first described as efficient substrates for Escherichia coli DNA-dependent RNA polymerase (Chatterji and Gopal, 1996; Schlageck et al., 1979; Yarbrough et al., 1979). Utilization of phospholinked dNTPs was subsequently demonstrated in conjunction with DNA polymerases and reverse transcriptases (Kumar et al., 2005; Mulder et al., 2005; Sood et al., 2005), but high concentrations had to be used because incorporation efficiencies were generally much lower than with unmodified dNTPs. It was observed that a linker containing one or more additional phosphate groups extending the triphosphate moiety proved beneficial for improving incorporation efficiencies (Kumar et al., 2005; Sood et al., 2005), presumably due to attenuation of steric hindrance effects of the bulky fluorophore proximal to the active site, and partial restoration of the negative charge lost by the linker conjugation at the g-phosphate. We have found that for f29 DNA polymerase, a linker that is extending the natural triphosphate by an additional two or three phosphates, followed by a short aminohexyl aliphatic

438

Jonas Korlach et al.

chain, yields nucleotides that are incorporated as efficiently as unmodified, natural dNTPs with respect to synthesis rates and processivity (Korlach et al., 2008a). Depending on the specific DNA polymerase under study, the exact nature of the linker and number of extra phosphates may vary. An exemplary synthesis protocol for the phospholinked dNTP Alexa Fluor 488 aminohexyl-dG5P (Korlach et al., 2008a) is given below. Based on carbonyldiimidazole (CDI) activation, it builds the additional diphosphate moiety on an aliphatic linker, allowing for flexibility of dye conjugation as the final step. The aliphatic linker also allows a larger spatial separation between nucleotide and fluorophore. During the third CDI activation for coupling the linker-triphosphate to the nucleotide, MgCl2 is included, which significantly improves the yield (Kadokura et al., 1997). Fmoc-6-aminohexylphosphate 1. Coevaporate 1 g (2.94 mM) of Fmoc-6-aminohexanol 2 with 20 ml anhydrous acetonitrile, then suspend in 10 ml anhydrous triethylphosphate. 2. Add two equivalents (550
l, 5.88 mM) of phosphorus oxychloride to the stirring suspension (Yoshikawa et al., 1967). After 2 h, HPLC shows disappearance of the Fmoc-6-aminohexanol. 3. Quench the reaction by the addition of 100 ml 0.1 M triethylamine bicarbonate (TEAB) (pH 6.8) and stir for 30 min. Purify by reverse phase HPLC on a Waters Xterra C18 RP 30  100 column using an acetonitrile gradient in 0.1 M TEAB. 4. Evaporate the fractions containing product, followed by coevaporation with methanol (2). 5. Triturate the residue twice with 100 ml diethylether and dry under vacuum to give a white powder. Yield: 1.24 g, 68% as bis-triethylamine salt. Purity (HPLC): 98%. Fmoc-6-aminohexyldiphosphate 1. Coevaporate 200 mg (320 mM) of Fmoc-6-aminohexylphosphate twice with anhydrous acetonitrile, then suspend in 2 ml anhydrous DMF. 2. Add four equivalents of 1,10 -carbonyldiimidazole (CDI; 207 mg, 1280 mM) and stir at ambient temperature for 4 h (Hoard and Ott, 1965). 3. Add six equivalents of methanol (77 ml, 1920 mM) and stir for 30 min. 4. Add to the reaction 10 equivalents of tributylamine–H2PO4 (3200 mM; prepared by mixing equimolar amounts of tributylamine and 85% phosphoric acid, followed by coevaporation three times with anhydrous acetonitrile, and dissolved in 4 ml anhydrous DMF). Stir the reaction mixture for 16 h. HPLC shows 3% Fmoc-aminohexylphosphate remaining. 5. Dilute the reaction mixture to 50 ml with 0.1 M TEAB, and purify by RP HPLC on a Waters Xterra C18 RP 30  100 column using an acetonitrile gradient in 0.1 M TEAB.

Real-Time DNA Sequencing from Single Polymerase Molecules

439

6. Evaporate the fractions containing product, followed by coevaporation with methanol (2). 7. Coevaporate the residue with anhydrous acetonitrile. Yield: 186 mg, 73% as Tris–TEA salt. Purity (HPLC): 96%. Aminohexyl-dG5P 1. Coevaporate 186 mg (233 mM) Fmoc-6-aminohexyldiphosphate twice with anhydrous acetonitrile, then suspend in 3 ml anhydrous DMF. 2. Add four equivalents of CDI (150 mg, 930 mM) and stir at ambient temperature for 4 h. 3. Add six equivalents of methanol (56 ml, 1400 mM) and stir for 30 min. 4. Coevaporate 1.5 equivalents dGTP (TEA salt, 350 mM) 3 with anhydrous acetonitrile, then suspend in 2 ml anhydrous DMF. 5. Add the Fmoc-aminohexyldiphosphoimidazolate reaction to the dGTP solution, followed by 10 equivalents of anhydrous MgCl2 (333 mg, 3500 mM) (Kadokura et al., 1997). Stir the reaction for 18 h. HPLC shows 28% of the Fmoc-aminohexyldiphosphate converted to Fmocaminohexyl-dG5P. 6. Dilute the reaction mixture to 125 ml with 0.1 M TEAB, and purify by RP HPLC on a Waters Xterra C18 RP 30  100 column using an acetonitrile gradient in 0.1 M TEAB. 7. Evaporate the fractions containing product, followed by coevaporation with methanol (2). 8. Take up the residue in 20 ml 10% TEA/water and stir for 16 h to remove the Fmoc protecting group from the amine on the linker. 9. Evaporate triethylamine, add water to 25 ml, and extract the solution three times with 25 ml diethyl ether. 10. Purify the product from the aqueous layer by anion exchange chromatography on Q sepharose FF using a TEAB gradient from 0.05 to 1 M. Yield: 42 mM, 18%. Purity (HPLC): 98%. Alexa Fluor 488-aminohexyl-dG5P 1. Dissolve 1 mM aminohexyl-dG5P in 200 ml of 50 mM NaHCO3, pH 8.7, and add to 1 mg Alexa Fluor 488-TFP ester (Invitrogen). Briefly sonicate the mixture. 2. After 4 h, HPLC showed no active ester remaining (the product is identified by characteristic PDA scan). Purify the compound by IEX on Q sepharose FF with a TEAB gradient from 0.05 to 1 M. Purify the product further by RP HPLC on a Waters Xterra RP C18 19  100 column using an acetonitrile gradient in 0.1 M TEAB. 3. Evaporate the fractions containing pure product, followed by coevaporation with methanol (2). 4. Dissolve the residue in water and quantitate by UV–Vis spectrophotometry. Yield: 370 nM, 37%. Purity (HPLC): 99%.

440

Jonas Korlach et al.

In contrast to base-linked nucleotide derivatizations, this synthesis scheme proceeds identically for all four nucleobases, and for labeling with different fluorophores using the appropriate dye NHS esters. For phospholinked nucleotides containing a triphosphate linker (Eid et al., 2009) instead of the diphosphate moiety described here, Fmoc-aminohexyl-triphosphate is substituted for Fmoc-aminohexyl-diphosphate in the condensation reaction with nucleoside triphosphate. Demands on the purity of phospholinked nucleotides for accurate SMRT DNA sequencing are high, as even small traces of unlabeled nucleotides, when incorporated by the polymerase, could lead to missed bases in the sequencing read. Exposure of materials to ambient light should be minimized during the dye conjugation step to avoid fluorophore bleaching. In addition, in contrast to base-linked nucleotides, it is straightforward to subject phospholinked nucleotides to an additional enzymatic purification. For example, one can take advantage of the specificity of alkaline phosphatases which rapidly degrade unmodified dNTPs to the corresponding nucleoside, but are completely inactive on dNTPs that contain moieties coupled to the terminal phosphate (Sood et al., 2005; Yarbrough, 1978). Such post-chemical-synthesis purification can be carried out before using phospholinked nucleotides in SMRT sequencing reactions, or conveniently, the phosphatase can be included in the sequencing reaction.

3.4. DNA polymerase—the sequencing ‘‘engine’’ Various DNA polymerases can be used in conjunction with SMRT DNA sequencing, and the sequencing performance will depend on their specific properties. We have applied wild-type and mutant DNA polymerases from bacteriophage f29 to our SMRT DNA sequencing method, taking advantage of several favorable characteristics. f29 DNA polymerase is extremely processive (tens of kilobases), relatively fast (
50–100 bases/s) and highly accurate (error rate of
10 5–10 6) (Baner et al., 1998; Blanco et al., 1989; Esteban et al., 1993). It is also very stable, maintaining constant enzymatic activities for up to several days (Dean et al., 2001; Nelson et al., 2002). The use of doublestranded DNA templates is possible by its efficient DNA strand displacement synthesis activity, thus simplifying sample preparation procedures. The following protocol outlines the expression and purification of f29N62D, a mutant with reduced 30 -50 exonuclease activity while maintaining essentially identical polymerizing properties (Blanco and Salas, 1996; de Vega et al., 1996). N-terminal Histidine (His) and GST tags were cloned into pET41 (Invitrogen, Carlsbad, CA) for ease of purification (Korlach et al., 2008a). 1. Overproduce polymerase in E. coli by addition of IPTG (1 mM) at midlog phase.

Real-Time DNA Sequencing from Single Polymerase Molecules

441

2. Disrupt the cells using lysozyme (chicken egg white, Sigma-Aldrich, St. Louis, MO) in a buffer containing 50 mM Tris–HCl, pH 7.5, 7 mM 2-mercaptoethanol, and 5% glycerol (buffer B), and additionally containing 0.2 M NaCl for 30 min. Sonicate for 2 min with a sonication probe. Degrade DNA with DNase I (bovine pancreas, Sigma-Aldrich) for 30 min at room temperature while shaking. Remove cell debris by centrifugation for 30 min at 15,000g. 3. Adjust the supernatant to 0.5 M NaCl for purification on a 1-ml HisTrap FF column (Ni-resin, GE Healthcare, Piscataway, NJ), equilibrated with buffer B containing 0.5 M NaCl. Wash His-tagged polymerase retained on the HisTrap column with at least 50 column volumes, first using buffer B containing 1 M NaCl, 0.2% Tween-20, and 20 mM imidazole, followed by buffer B containing 0.5 M NaCl and 50 mM imidazole. 4. Elute the polymerase with buffer B containing 300 mM imidazole. Pool this ‘‘HisTrap’’ fraction and adjust it to 0.2 M NaCl using buffer C (50 mM Tris–HCl, pH 7.5, 1 mM EDTA, 7 mM 2-mercaptoethanol, 5% (v/v) glycerol). 5. Load the sample onto a Heparin-Sepharose CL-6B column (10 ml, GE Healthcare), equilibrated in buffer C. Elute polymerase with buffer C and a gradient of 0–1 M NaCl. Pool these ‘‘Heparin’’ fractions and concentrate using Centricon YM-50 (Millipore, Billerica, MA). 6. Adjust to the final storage buffer (50 mM Tris–HCl, pH 7.5, 0.2 M NaCl, 1 mM EDTA, 7 mM 2-mercaptoethanol, 50% (v/v) % glycerol). 7. Analyze protein fractions by sodium dodecyl sulfate—polyacrylamide gel electrophoresis (SDS–PAGE, 10% or 15% polyacrylamide). The purified polymerase is at least 97% homogenous. Determine protein concentrations both by measuring the absorbance at 280 nm, and by the Bradford method using known amount of BSA (Bio-Rad). Various strategies exist to mediate specific enzyme immobilization in ZMWs. They will depend on the particular coupling chemistry chosen. For a surface derivatization providing selectively biotinylated ZMW floors (described in Section 3.2), they include nonspecific biotinylation of the enzyme (Hermanson, 1996), introduction of specific biotin tags (Beckett et al., 1999), or fusion proteins (Nilsson et al., 1997).

3.5. Instrument for highly parallel monitoring of sequencing reactions As for all single-molecule fluorescence recording systems, optimization of light collection efficiencies is paramount for maximal signal-to-noise detection. For SMRT DNA sequencing, additional demands exist for simultaneously illuminating and monitoring many ZMWs containing immobilized DNA polymerase molecules. We have described the development of a

442

Jonas Korlach et al.

highly parallel optical system that is capable of continuously analyzing thousands of concurrent sample locations (Lundquist et al., 2008) (Fig. 20.4). In this system, wavelength-specific holographic phase masks (Kress and Meyrueis, 2000) act as illumination multiplexers by dividing the laser beams into several thousand subbeams. Relay lens assemblies convert these beams into corresponding arrays of spots focused at a plane conjugate to the front focus of a microscope objective. After combining multiple wavelengths paths, all of the illumination light is transmitted through a common dichroic filter and brought to an array of diffraction-limited focal spots in the sample plane where it excites fluorescence in each ZMW observation volume. Multiple laser sources, coupled with demands for detecting several different fluorophores with single-molecule sensitivity, put stringent demands on the dichroic filter. We have found that better performance

Multiplex zero-mode waveguide array

Objective lens

Zero-mode waveguide array Monochrome detector Confocal pinhole array

Color separation

ZMW 1 ZMW 2 ZMW 3

Dichroic

ZMW 4 Prism

Illumination confocal plane

Laser 2

ZMW 5

Laser 1

Holographic phase masks

Figure 20.4 Schematics of an optical system for SMRT DNA Sequencing. This instrument provides simultaneous illumination of 3000 ZMWs with two different lasers, and wavelength-specific real-time detection of fluorescence from phospholinked nucleotides processed by DNA polymerase immobilized in the ZMWs. A photograph of an instrument (top cover removed) is shown on the bottom.

Real-Time DNA Sequencing from Single Polymerase Molecules

443

with respect to narrow excitation bandwidths and high fluorescence throughput can be achieved by reversing the conventional positions of the illumination and collection paths. Upon reflecting off the dichroic filter, fluorescence light is first imaged onto a congruent array of confocal pinholes to reject stray light not originating from ZMW locations. It is subsequently reimaged through a prism dispersive element onto a monochrome, electron-multiplication charge-coupled device (EM-CCD) array. The compound prism serves to linearly disperse the wavelength content of the emitted light, transforming the image from each confocal volume element from a diffraction-limited spot into a ‘‘rainbow’’ pattern. Emitted fluorescence from different phospholinked nucleotides falls onto different spatial locations on the detector, thereby enabling identification of the type of nucleotide incorporated by the polymerase at any given time. A single camera thus collects both spatial and spectral information for the entire ZMW array. The prism assembly we have chosen allows for hightransmission, continuous color separation. The resulting opportunity for oversampling the fluorescence spectra in wavelength improves the accuracy of classification in cases of overlapping dye emission spectra. For example, in the implementation described in Lundquist et al., (2008), a three-wedge compound prism was optimized to provide linear angular dispersion of 1.25 mrad between the wavelengths of 490 and 730 nm, with a zero-deviation angle at 550 nm. With this choice, the system can be applied to a variety of fluorophore combinations. If resolution of highly overlapping emission spectra were a limitation, dispersion could be increased in the critical regions to improve spectral performance.

3.6. DNA sequencing assay example A typical protocol for performing a SMRT DNA sequencing reaction is described below. An oxygen scavenging system, consisting of protocatechuate dioxygenase (PCD) and its substrate protocatechuic acid (PCA), is used to remove oxygen and thereby suppress photophysical effects that might be deleterious to the polymerase (Aitken et al., 2008; Eid et al., 2009). A nitrobenzyl-based triplet state quencher, nitrobenzoic acid (NBA), is included to shorten the time fluorophores can reside in the dark triplet state (Dave et al., 2009). 1. Incubate DNA polymerase carrying an N-terminal biotin-tag (Beckett et al., 1999) with 1.5 molar excess of primed DNA template at 4  C for 10 min in a buffer containing 50 mM MOPS, pH 7.5, 75 mM potassium acetate, 5 mM dithiothreitol and 0.05% (v/v) Tween-20. 2. Simultaneously, incubate streptavidin (Invitrogen) at a twofold stoichiometric excess over polymerase in the same buffer at 22  C on the ZMW array. Wash the array five times with buffer.

444

Jonas Korlach et al.

3. Immobilize polymerase/template complexes onto the arrays at 4  C for 15 min. Remove unbound complexes by washing five times with reaction buffer (50 mM ACES, pH 7.1, 75 mM potassium acetate, and 5 mM dithiothreitol). 4. Add the oxygen scavenging system (1 PCD, 4 mM PCA), triplet state quencher (6 mM NBA (all Sigma-Aldrich, St. Louis, MO), and all phospholinked dNTPs (250 nM final concentration each), except the one corresponding to the first base to be incorporated into the DNA template. 5. Initiate the sequencing reactions by simultaneous addition of the first phospholinked dNTP to be incorporated (250 nM) and manganese acetate (0.5 mM final concentration) (Korlach et al., 2008a; Kumar et al., 2005).

3.7. Data analysis Fluorescence pulse calling is performed by a threshold algorithm on the dye-weighted intensities using fluorescence emission calibration spectra for each of the phospholinked dNTPs (Horne, 1986). For base identification, light collected by the detector is summed over the duration of the pulse, allowing each fluorescent phospholinked nucleotide to be spectrally evaluated and classified. Automated classification is performed by least-squares fitting to the four known dye reference spectra. The spectrum that yields the minimum chi-squared difference when compared to the pulse spectrum identifies the pulse as a particular phospholinked dNTP. The degree of spectral cross talk between different fluorophore types depends on the specific dyes chosen and their intrinsic brightness. A typical set of dyes we have used in conjunction with SMRT DNA sequencing, with spectral separation between the emission maxima of the pairs of dyes excited with the same laser of 31 and 23 nm, respectively, showed misidentification rates of less than 1% using this method (Eid et al., 2009). A section of a typical sequencing read after such automated base calling is shown in Fig. 20.5A. DNA polymerase activity is marked by a train of pulses corresponding to phospholinked nucleotide incorporations. Incorporation signals from the four phospholinked nucleotides show fluorescence intensity ‘‘level setting’’ characteristics which are due to (i) the stationary location of the enzyme’s active site with respect to the illumination profile, and (ii) different excitation efficiencies and intrinsic fluorophore brightness of the four fluorescent dyes. High detector frame rates (e.g., 100 Hz; Eid et al., 2009) ensure that enzymatic turnovers are oversampled in time, allowing precise measurements

445

Real-Time DNA Sequencing from Single Polymerase Molecules

A

Intensity (a.u.)

600

A

Interpulse duration (IPD)

Pulse width (PW)

A

450

T

300

150

T C

C

C

C G

G

0 194.5

194.75

195

B IPD:

195.25

195.5

PPi

195.75 196 Time (s)

196.25

196.5

196.75

197

N E:D+1

E:D+1PPi

E:D

E:D:N

PW: E*:D+1PPi

E*:D:N

Figure 20.5 SMRT DNA sequencing example. (A) Four-color time trace of dyeweighted fluorescence intensities with automated base annotations, along with definitions of pulse width (PW) and interpulse duration (IPD). While significant spectral overlap can be present from two dyes excited by one laser, its magnitude is known from the dye reference spectra and thus it does not affect misidentification rates. (B) Generalized enzymatic reaction cycle of DNA polymerization, with the red rectangle differentiating the steps which constitute the ‘‘bright’’ (PW) state. E: polymerase enzyme; D: DNA template; N: nucleotide.

of ‘‘on’’ and ‘‘off’’ times. The pulses exhibit stochastic intensity fluctuations because of counting statistics and dye photophysics. Hallmarks of singlemolecule fluorescent events are characteristic: single-frame rise and fall times at the start and end of the pulse, respectively ( 10 ms), which facilitate pulse detection and base calling even when pulses are close together. Single-molecule events corresponding to phospholinked dNTP incorporations manifest as fluorescent pulses whose variable duration and spacing directly reflect the underlying enzyme kinetics. In the generalized enzymatic DNA polymerization cycle (Fig. 20.5B), kinetic steps starting with phospholinked nucleotide binding, then proceeding through the transition from the ‘‘open’’ to the ‘‘closed’’ polymerase conformation, catalysis, reverse conformational transition, and up to the moment of pyrophosphate-linkerfluorophore release, have the nucleotide bound in the active site and thereby

446

Jonas Korlach et al.

make up the duration of the fluorescence pulse (defined hereafter as pulse width (PW)). The time elapsing after release of the reaction product, translocation of the polymerase along the template by one base, and waiting time for the binding of the next incoming phospholinked nucleotide constitute the duration between successive pulses (interpulse duration (IPD)).

4. Single-Molecule DNA Polymerase Dynamics While the base sequence of the synthesized DNA strand constitutes the main output of the SMRT sequencing method, the real-time aspect of this approach generates unprecedented information about DNA polymerase kinetics. Because the system reports the kinetics of every base incorporation through PW and IPD, it can be used to investigate the dynamics of DNA polymerization with base-pair resolution, and to provide the distribution of kinetic parameters over many different sequence contexts in a single 5-min experiment. The method thereby allows direct assessments of static variation (differences in enzymatic activity between different molecules) and dynamic variation (fluctuation of catalytic rate constants over time for a single enzyme molecule).

4.1. Determination of single-molecule kinetic parameters From the multitude of nucleotide incorporations recorded for single DNA polymerases during a SMRT sequencing read, it is possible to determine effective kinetic parameters for each enzyme molecule. For certain polymerases and reaction conditions, a simplified model can successfully be used for which single steps limit the rates of transitions between the ‘‘bright’’ and ‘‘dark’’ states, as defined above (Fig. 20.5B). In this case, the kinetic cycle of DNA synthesis reduces to k1

k2

N þ E : D > E : D : N ! E : Dþ1 k1

with the ternary complex of enzyme, DNA, and nucleotide (E:D:N) now denoting all states for which the phospholinked nucleotide is bound in the active site. Here, the average fluorescence PW is determined by the summation of kinetic steps exiting the ‘‘bright’’ state PW ¼

1 1 ¼ kPW ðk2 þ k1 Þ

ð20:1Þ

Real-Time DNA Sequencing from Single Polymerase Molecules

447

The IPD is governed by the combination of DNA polymerase translocation along the template and time to binding of the next nucleotide. For enzymes with fast translocation kinetics relative to nucleotide binding (e.g., below nucleotide substrate saturation concentrations), the latter becomes rate-limiting, and the average IPD is then given by IPD ¼

1 1 ¼ kIPD k1 ½N 

ð20:2Þ

The two expressions are combined to yield a measurement of the Km for a single polymerase molecule from measured average PWs and IPDs: Km ¼

ðk2 þ k1 Þ IPD ¼ ½N  k1 PW

ð20:3Þ

Base-specific Km values determined in this way for a 10-min, 2.4-kb sequencing read are depicted in Fig. 20.6. The single-exponential fits to the PW and IPD histograms indicate that this simplified, single rate-limiting kinetic model is a good approximation for the DNA polymerase employed in this experiment. Km values for this molecule are in good agreement with bulk data measuring the transient kinetics of single nucleotide turnovers using a stopped-flow instrument (KinTek), except for dTTP which has a lower Km for the single polymerase by a factor of
2.

4.2. DNA polymerase pausing The above-mentioned analysis does not account for transient changes in DNA polymerization dynamics during the course of DNA synthesis for a single enzyme. While the validity of the Michaelis–Menten equation has been demonstrated theoretically and experimentally in the presence of such dynamic variation (English et al., 2006; Kou et al., 2005; Min et al., 2006; Velonia et al., 2005), the extracted parameters do not inform about it, so a more detailed analysis is needed. DNA polymerases present a more complex case compared to other, single-substrate turnover enzymes, as there are four different nucleotides competing for one active site, several different enzymatic activities are possible in one protein, and the sequence of the DNA template introduces a large number of different sequence contexts. A detailed description of these phenomena is beyond the scope of this chapter, and only a few examples illustrating the richness of these singlemolecule data are given here. We have previously described polymerase pausing caused by hairpin formation in single-stranded DNA templates (Eid et al., 2009). Here, we extend our analysis to double-stranded DNA templates several hundred

Phospholinked nucleotide

80

80

80

60 40

0.2 0.3 0.4 Pulse width (s)

40

60 40

20

20

0 0.5 0.0

0 0.5 0.0

0.1

0.2 0.3 0.4 Pulse width (s)

60 40 20

0.1

0.2 0.3 0.4 Pulse width (s)

0 0.5 0.0

80

80

70

70

70

60 50

60 50

60 50

50 40 30 20 10 0 0.0

40 30

0.33 ± 0.03 (0.28)

0.8

0 0.0

40 30 20 10

20 10 0.2 0.4 0.6 Interpulse duration (s)

Frequency

80

70 60

Frequency

80

Frequency

Frequency

0.1

60

Frequency

100

Frequency

100

Frequency

100

0 0.0

Km(mM) (bulk)

T

80

20

Interpulse duration

G

C

100

Frequency

Pulse width

A

0.2 0.4 0.6 Interpulse duration (s) 0.24 ± 0.03 (0.27)

0.8

0 0.0

0.1

0.2 0.3 0.4 Pulse width (s)

0.5

0.2 0.4 0.6 Interpulse duration (s)

0.8

40 30 20 10

0.2 0.4 0.6 Interpulse duration (s) 0.46 ± 0.05 (0.56)

0.8

0 0.0

0.46 ± 0.05 (0.96)

Figure 20.6 Determination of effective Km values from a SMRT sequencing read. Distributions from a 10-min, 2.4-kb read for pulse widths of, and interpulse duration before each of the four phospholinked nucleotides are fit to single exponential decay functions used to derive Km values (Eq. (20.3)). Values are compared to bulk measurements of single-turnover kinetics (gray).

Real-Time DNA Sequencing from Single Polymerase Molecules

449

bases in length, in conjunction with DNA strand displacement synthesis carried out by the polymerase. While double-stranded DNA is characterized by dramatically less secondary structure in comparison with singlestranded DNA, there still remains a degree of higher order structure due to different sequence contexts (Gimenes et al., 2008; Hagerman, 1990). In addition, the dynamics of strand displacement activity of the polymerase, separating the nontemplate strand from the template strand before its entry into the polymerase active site, is also likely influencing DNA polymerization rates (Kamtekar et al., 2004; Rodriguez et al., 2005). Single-molecule sequencing trajectories of template position over time allow a detailed view into polymerase-mediated DNA synthesis dynamics (Fig. 20.7). Variations in polymerization rates are apparent with respect to heterogeneity among different molecules for a given template position, allowing the creation of instantaneous rate distributions for every template position when analyzed over all molecules. Template sequence context effects over many time scales are apparent. For the specific example shown in Fig. 20.7A, the DNA synthesis rate roughly doubles after incorporation of
250 nucleotides. Analyzed over shorter time scales, a range of interpulse distances is observed, including occurrences of cessation of DNA synthesis activity on a time scale of many seconds, followed by resumption of DNA synthesis. Such polymerase pausing is specific to certain DNA template locations where it is common to several molecules (Fig. 20.7A, arrows). To rule out that such variations are caused by systematic effects of the instrumentation or reaction conditions, and instead represent dynamic changes of DNA polymerization caused by different DNA template sequence contexts, circular DNA templates can be employed (Eid et al., 2009). In this configuration, polymerases capable of strand displacement synthesis, such as f29 DNA polymerase (Blanco et al., 1989), will encounter the same DNA sequence on a template molecule multiple times. DNA synthesis trajectories encompassing multiple rounds of circular template sequencing (Fig. 20.7B) show that polymerases repeat the dynamic signatures upon consecutive encounters of the same DNA template sequence context. For example, the transition to the faster overall rate at around 250 nucleotides into the template (described above) can be discerned for each round of synthesis. Similarly, polymerase pausing occurs at the same DNA template site during each round of synthesis, albeit with variable IPDs (Fig. 20.7B, lower right panel, arrows). To within the resolution of the data in this experiment, the correlation between IPDs of the same sequence context in different molecules was the same as between different laps around the circular template carried out by the same molecule. IPD profiles from several hundred polymerases for multiple laps illustrate these phenomena further (Fig. 20.7C), showing the slower rate for the first half of the template due to overall increased IPDs, and highlighting template locations of increased pausing frequencies.

450

Jonas Korlach et al.

A

B Template position

700

Base incorporation

600 500

500

500

400

400

300

300

200

200

100

~2.5 bases/s

100

0

0 0

400 Template position

300 ~1.3 bases/s

200 100

100

200

300

400

500

600

500

500

400

400

300

300

200

200

0

50

100

150

200 250 Time (s)

300

350

C

400

200

0

100

200

300

400

500

600

300 400 Time (s)

500

600

0

0

0

100

100

100

0

0

100

200

300 400 Time (s)

500

600

D 1st lap:

IPD (s)

5

Template position 34

4 3 2 1 0 0

200 300 Template position

100

nd

2

lap:

IPD (s)

5

400

500

4 3 2 1 0 600

700

800 900 Template position

1000

Sequence

Consensus

GTCCTCAGCAA

76 90 152 180 194

ATTTAAAGATT AGATAAATGGA TGTTAAACAGT xxxxAAAxxxx TAGTAAACCCT TTACAAACACC

65 165 359

CTACAGGTTTG ATACGGGTAAA xxACxGxTxxx AGACTGTTTAA

Figure 20.7 SMRT DNA synthesis dynamics. (A) DNA synthesis trajectories of template position over time for 33 polymerase molecules. The two dashed lines illustrate a transition to higher polymerization rates occurring at
250 bases into the DNA template. Arrows indicate common sites in the DNA template at which pausing occurs for several enzyme molecules. (B) Polymerization trajectories for four DNA polymerase molecules on circular DNA templates, showing multiple laps of continuous, processive DNA synthesis. The white dotted line indicates the same template position as in (A) where the speed transition occurs. Arrows in the lower right panel show an example of template locations at which the polymerase molecule pauses during each round of synthesis. (C) 90th percentile of the IPD versus template position for several hundred polymerases sequencing two successive laps around a circular DNA template. The arrows indicate the pause sites for which the sequences and consensus contexts (D) are given.

Analysis of DNA sequences for these pauses indicate common, relatively short consensus sequence contexts (Fig. 20.7D) which we have confirmed using templates from different sources (data not shown). While the mechanism and biological significance of these pausing signatures remains to be elucidated, the ‘‘AAA’’ sequence is reminiscent of the initiation mechanism by f29 DNA polymerase (characterized by multiple dATP incorporations and sliding-back movements; Mendez et al., 1992). We speculate that perhaps the polymerase has an increased propensity for slippage in this template sequence context, creating the pause signature before DNA synthesis resumes.

451

Real-Time DNA Sequencing from Single Polymerase Molecules

5. Conclusions SMRT DNA sequencing harnesses the intrinsic power of DNA polymerases, allowing their speed, processivity, efficiency, and fidelity to be exploited directly. Rapid intrinsic DNA synthesis rates translate to short sequencing run cycle times. Long continuous sequence reads preserve the molecular integrity of the DNA template, simplifying the downstream bioinformatics for genome assembly and analysis in the context of structural variations and allelic polymorphism linkages. Many sequencing-by-synthesis techniques employ DNA polymerase as a bulk reagent consumable, synchronizing its activity with various termination approaches (reviewed in Mardis, 2008). Such gating allows for an increase in multiplex and overall sequence throughput, but comes at a cost of long singlebase cycle times and relatively short read lengths due to incomplete cycle efficiencies. These methods utilize uniform protocols for each incorporation cycle and are therefore insensitive to sequence context effects on polymerization efficiencies, leading to variable systematic errors (Kong, 2009) and even inaccessible genomic regions. In contrast, sequencing by observing uninterrupted DNA polymerization allows the enzyme to spend variable amounts of

120 100

Frequency

80 60 40 20 0 0

1 2 3 4 5 6 7 Instantaneous DNA synthesis rate (bases/s)

8

Figure 20.8 DNA polymerization rate variations over different sequence contexts. In this example, SMRT sequencing reads were recorded from 128 polymerase molecules. Instantaneous DNA synthesis rates for each position of the 1.1 kb DNA template were extracted for each molecule. Mean values from the rate distributions at each template position are used for the histogram. The coefficient of variation of DNA synthesis rate for this template is 70%.

452

Jonas Korlach et al.

time synthesizing different sequence contexts. Even when analyzed over many polymerase molecules, significant variation of DNA synthesis rates exists for different DNA template positions (Fig. 20.8). This dispersion of DNA synthesis rates as a function of sequence context suggests that in an ensemble sequencing-by-synthesis system, the ideal protocol for incorporation should vary by that same amount. This requirement is moot in a single-molecule, realtime method, as every incorporation is allowed precisely the correct amount of time, thus providing an intrinsic adjustment and optimal sequencing performance in every sequence context. Heterogeneities between enzymes (static variation) or in the catalytic rates for a single enzyme molecule (dynamic variation) have been observed for many enzymes (reviewed in Blank et al., 2009), and DNA polymerases are no exception. The information can be used advantageously to improve the quality of sequencing, and at the same time provide insights into the dynamics of DNA polymerization. Because polymerase kinetics is sensitive to biological perturbation, this information can be further developed for investigating DNA binding proteins, DNA polymerase inhibitors, and effects of base methylation.

ACKNOWLEDGMENTS We are indebted to the entire staff at Pacific Biosciences for their dedicated work that brings this technology to fruition. We also thank J. Puglisi, M. Hunkapiller, R. Kornberg, K. Johnson, D. Haussler, W. Webb, and H. Craighead for many helpful discussions. Aspects of this research were supported by National Human Genome Research Institute grant R01HG003710.

REFERENCES Aitken, C. E., Marshall, R. A., and Puglisi, J. D. (2008). An oxygen scavenging system for improvement of dye stability in single-molecule fluorescence experiments. Biophys. J. 94, 1826–1835. Baner, J., Nilsson, M., Mendel-Hartvig, M., and Landegren, U. (1998). Signal amplification of padlock probes by rolling circle replication. Nucleic Acids Res. 26, 5073–5078. Beckett, D., Kovaleva, E., and Schatz, P. J. (1999). A minimal peptide substrate in biotin holoenzyme synthetase-catalyzed biotinylation. Protein Sci. 8, 921–929. Blanco, L., Bernad, A., Lazaro, J. M., Martin, G., Garmendia, C., and Salas, M. (1989). Highly efficient DNA synthesis by the phage phi 29 DNA polymerase. Symmetrical mode of DNA replication. J. Biol. Chem. 264, 8935–8940. Blanco, L., and Salas, M. (1996). Relating structure to function in phi29 DNA polymerase. J. Biol. Chem. 271, 8509–8512. Blank, K., De Cremer, G., and Hofkens, J. (2009). Fluorescence-based analysis of enzymes at the single-molecule level. Biotechnol. J. 4, 465–479. Chatterji, D., and Gopal, V. (1996). Fluorescence spectroscopy analysis of active and regulatory sites of RNA polymerase. Methods Enzymol. 274, 456–478.

Real-Time DNA Sequencing from Single Polymerase Molecules

453

Cloonan, N., Forrest, A. R., Kolle, G., Gardiner, B. B., Faulkner, G. J., Brown, M. K., Taylor, D. F., Steptoe, A. L., Wani, S., Bethel, G., Robertson, A. J., Perkins, A. C., et al. (2008). Stem cell transcriptome profiling via massive-scale mRNA sequencing. Nat. Methods 5, 613–619. Cokus, S. J., Feng, S., Zhang, X., Chen, Z., Merriman, B., Haudenschild, C. D., Pradhan, S., Nelson, S. F., Pellegrini, M., and Jacobsen, S. E. (2008). Shotgun bisulphite sequencing of the Arabidopsis genome reveals DNA methylation patterning. Nature 452, 215–219. Dave, R., Terry, D. S., Munro, J. B., and Blanchard, S. C. (2009). Mitigating unwanted photophysical processes for improved single-molecule fluorescence imaging. Biophys. J. 96, 2371–2381. de Vega, M., Lazaro, J. M., Salas, M., and Blanco, L. (1996). Primer-terminus stabilization at the 30 -50 exonuclease active site of phi29 DNA polymerase. Involvement of two amino acid residues highly conserved in proofreading DNA polymerases. EMBO J. 15, 1182–1192. Dean, F. B., Nelson, J. R., Giesler, T. L., and Lasken, R. S. (2001). Rapid amplification of plasmid and phage DNA using Phi 29 DNA polymerase and multiply-primed rolling circle amplification. Genome Res. 11, 1095–1099. Eid, J., Fehr, A., Gray, J., Luong, K., Lyle, J., Otto, G., Peluso, P., Rank, D., Baybayan, P., Bettman, B., Bibillo, A., Bjornson, K., et al. (2009). Real-time DNA sequencing from single polymerase molecules. Science 323, 133–138. English, B. P., Min, W., van Oijen, A. M., Lee, K. T., Luo, G., Sun, H., Cherayil, B. J., Kou, S. C., and Xie, X. S. (2006). Ever-fluctuating single enzyme molecules: Michaelis– Menten equation revisited. Nat. Chem. Biol. 2, 87–94. Esteban, J. A., Salas, M., and Blanco, L. (1993). Fidelity of phi 29 DNA polymerase. Comparison between protein-primed initiation and DNA polymerization. J. Biol. Chem. 268, 2719–2726. Foquet, M., Samiee, K. T., Kong, X. X., Chauduri, B. P., Lundquist, P. M., Turner, S. W., Freudenthal, J., and Roitman, D. B. (2008). Improved fabrication of zero-mode waveguides for single-molecule detection. J. Appl. Phys. 103, 034301. Fullwood, M. J., Wei, C. L., Liu, E. T., and Ruan, Y. (2009). Next-generation DNA sequencing of paired-end tags (PET) for transcriptome and genome analyses. Genome Res. 19, 521–532. Gimenes, F., Takeda, K. I., Fiorini, A., Gouveia, F. S., and Fernandez, M. A. (2008). Intrinsically bent DNA in replication origins and gene promoters. Genet. Mol. Res. 7, 549–558. Hagerman, P. J. (1990). Sequence-directed curvature of DNA. Annu. Rev. Biochem. 59, 755–781. Hermanson, G. T. (1996). Bioconjugate Techniques. Academic Press, San Diego. Hoard, D. E., and Ott, D. G. (1965). Conversion of mono- and oligodeoxyribonucleotides to 50 -triphosphates. J. Am. Chem. Soc. 87, 1785–1788. Horne, K. (1986). An optimal extraction algorithm for CCD spectroscopy. Publ. Astron. Soc. Pac. 98, 609–617. Kadokura, M., Wada, T., Urashima, C., and Sekine, M. (1997). Efficient synthesis of gamma-methyl-capped guanosine 50 -triphosphate as a 50 -terminal unique structure of U6 RNA via a new triphosphate bond formation involving activation of methyl phosphorimidazolidate using ZnCl2 as a catalyst in DMF under anhydrous conditions. Tetrahedron Lett. 38, 8359–8362. Kamtekar, S., Berman, A. J., Wang, J., Lazaro, J. M., de Vega, M., Blanco, L., Salas, M., and Steitz, T. A. (2004). Insights into strand displacement and processivity from the crystal structure of the protein-primed DNA polymerase of bacteriophage phi29. Mol. Cell 16, 609–618.

454

Jonas Korlach et al.

Kong, Y. (2009). Statistical distributions of sequencing by synthesis with probabilistic nucleotide incorporation. J. Comput. Biol. 16, 817–827. Korlach, J., Bibillo, A., Wegener, J., Peluso, P., Pham, T. T., Park, I., Clark, S., Otto, G. A., and Turner, S. W. (2008a). Long, processive enzymatic DNA synthesis using 100% dyelabeled terminal phosphate-linked nucleotides. Nucleosides Nucleotides Nucleic Acids 27, 1072–1083. Korlach, J., Marks, P. J., Cicero, R. L., Gray, J. J., Murphy, D. L., Roitman, D. B., Pham, T. T., Otto, G. A., Foquet, M., and Turner, S. W. (2008b). Selective aluminum passivation for targeted immobilization of single DNA polymerase molecules in zeromode waveguide nanostructures. Proc. Natl. Acad. Sci. USA 105, 1176–1181. Kou, S. C., Cherayil, B. J., Min, W., English, B. P., and Xie, X. S. (2005). Single-molecule Michaelis–Menten equations. J. Phys. Chem. B 109, 19068–19081. Kress, B., and Meyrueis, P. (2000). Digital diffractive optics: An introduction to planar diffractive optics and related technology. John Wiley, Chichester; New York. Kumar, S., Sood, A., Wegener, J., Finn, P. J., Nampalli, S., Nelson, J. R., Sekher, A., Mitsis, P., Macklin, J., and Fuller, C. W. (2005). Terminal phosphate labeled nucleotides: Synthesis, applications, and linker effect on incorporation by DNA polymerases. Nucleosides Nucleotides Nucleic Acids 24, 401–408. Levene, M. J., Korlach, J., Turner, S. W., Foquet, M., Craighead, H. G., and Webb, W. W. (2003). Zero-mode waveguides for single-molecule analysis at high concentrations. Science 299, 682–686. Liu, Y., and Blair, S. (2003). Fluorescence enhancement from an array of subwavelength metal apertures. Opt. Lett. 28, 507–509. Lundquist, P. M., Zhong, C. F., Zhao, P. Q., Tomaney, A. B., Peluso, P. S., Dixon, J., Bettman, B., Lacroix, Y., Kwo, D. P., McCullough, E., Maxham, M., Hester, K., et al. (2008). Parallel confocal detection of single molecules in real time. Opt. Lett. 33, 1026–1028. Maher, C. A., Kumar-Sinha, C., Cao, X., Kalyana-Sundaram, S., Han, B., Jing, X., Sam, L., Barrette, T., Palanisamy, N., and Chinnaiyan, A. M. (2009). Transcriptome sequencing to detect gene fusions in cancer. Nature 458, 97–101. Mardis, E. R. (2008). Next-generation DNA sequencing methods. Ann. Rev. Genomics Hum. Genet. 9, 387–402. Mendez, J., Blanco, L., Esteban, J. A., Bernad, A., and Salas, M. (1992). Initiation of phi 29 DNA replication occurs at the second 30 nucleotide of the linear template: A sliding-back mechanism for protein-primed DNA replication. Proc. Natl. Acad. Sci. USA 89, 9579–9583. Min, W., Gopich, I. V., English, B. P., Kou, S. C., Xie, X. S., and Szabo, A. (2006). When does the Michaelis–Menten equation hold for fluctuating enzymes? J. Phys. Chem. B 110, 20093–20097. Miyake, T., Tanii, T., Sonobe, H., Akahori, R., Shimamoto, N., Ueno, T., Funatsu, T., and Ohdomari, I. (2008). Real-time imaging of single-molecule fluorescence with a zero-mode waveguide for the analysis of protein–protein interaction. Anal. Chem. 80, 6018–6022. Mulder, B. A., Anaya, S., Yu, P., Lee, K. W., Nguyen, A., Murphy, J., Willson, R., Briggs, J. M., Gao, X., and Hardin, S. H. (2005). Nucleotide modification at the gamma-phosphate leads to the improved fidelity of HIV-1 reverse transcriptase. Nucleic Acids Res. 33, 4865–4873. Nelson, J. R., Cai, Y. C., Giesler, T. L., Farchaus, J. W., Sundaram, S. T., Ortiz-Rivera, M., Hosta, L. P., Hewitt, P. L., Mamone, J. A., Palaniappan, C., and Fuller, C. W. (2002). TempliPhi, phi29 DNA polymerase based rolling circle amplification of templates for DNA sequencing. Biotechniques (Suppl.), 44–47.

Real-Time DNA Sequencing from Single Polymerase Molecules

455

Nilsson, J., Stahl, S., Lundeberg, J., Uhlen, M., and Nygren, P. A. (1997). Affinity fusion strategies for detection, purification, and immobilization of recombinant proteins. Protein Expr. Purif. 11, 1–16. Rigneault, H., Capoulade, J., Dintinger, J., Wenger, J., Bonod, N., Popov, E., Ebbesen, T. W., and Lenne, P.-F. (2005). Enhancement of single-molecule fluorescence detection in subwavelength apertures. Phys. Rev. Lett. 95, 117401. Rodriguez, I., Lazaro, J. M., Blanco, L., Kamtekar, S., Berman, A. J., Wang, J., Steitz, T. A., Salas, M., and de Vega, M. (2005). A specific subdomain in phi29 DNA polymerase confers both processivity and strand-displacement capacity. Proc. Natl. Acad. Sci. USA 102, 6407–6412. Schlageck, J. G., Baughman, M., and Yarbrough, L. R. (1979). Spectroscopic techniques for study of phosphodiester bond formation by Escherichia coli RNA polymerase. J Biol. Chem. 254, 12074–12077. Sood, A., Kumar, S., Nampalli, S., Nelson, J. R., Macklin, J., and Fuller, C. W. (2005). Terminal phosphate-labeled nucleotides with improved substrate properties for homogeneous nucleic acid assays. J. Am. Chem. Soc. 127, 2394–2395. Tabor, S., Huber, H. E., and Richardson, C. C. (1987). Escherichia coli thioredoxin confers processivity on the DNA polymerase activity of the gene 5 protein of bacteriophage T7. J. Biol. Chem. 262, 16212–16223. Velonia, K., Flomenbom, O., Loos, D., Masuo, S., Cotlet, M., Engelborghs, Y., Hofkens, J., Rowan, A. E., Klafter, J., Nolte, R. J., and de Schryver, F. C. (2005). Single-enzyme kinetics of CALB-catalyzed hydrolysis. Angew. Chem. Int. Ed. Engl. 44, 560–564. Yarbrough, L. R. (1978). Synthesis and properties of a new fluorescent analog of ATP: Adenosine-50 -triphosphoro-gamma-1-(5-sulfonic acid) napthylamidate. Biochem. Biophys. Res. Commun. 81, 35–41. Yarbrough, L. R., Schlageck, J. G., and Baughman, M. (1979). Synthesis and properties of fluorescent nucleotide substrates for DNA-dependent RNA polymerases. J. Biol. Chem. 254, 12069–12073. Yassour, M., Kaplan, T., Fraser, H. B., Levin, J. Z., Pfiffner, J., Adiconis, X., Schroth, G., Luo, S., Khrebtukova, I., Gnirke, A., Nusbaum, C., Thompson, D. A., et al. (2009). Ab initio construction of a eukaryotic transcriptome by massively parallel mRNA sequencing. Proc. Natl. Acad. Sci. USA 106, 3264–3269. Yoshikawa, M., Kato, T., and Takenishi, T. (1967). A novel method for phosphorylation of nucleosides to 50 -nucleotides. Tetrahedron Lett. 50, 5065–5068.

Author Index

A Abelson, J., 31–40, 155, 156, 160, 162, 163, 170, 172–174 Aguzzi, A., 191 Ahn, K., 72, 73 Ainger, K., 391 Aitken, C. E., 78, 155, 443 Akiyama, B. M., 214 Alami, M., 91 Albright, L. M., 233 Alexeev, A., 264 Altschul, S. F., 243 Ames, B. N., 236 Amir, D., 188 Amirgoulova, E. V., 14 Amitani, I., 261–288, 308 Anderson, C. M., 397–398 Anderson, D. G., 264 Anderson, J. R., 121 Anna, S. L., 68 Antoun, A., 231 Aoki, H., 254 Arai, Y., 69 Ashkin, A., 69, 70 Atkinson, D., 91 Atkins, W. M., 89–112 Axelrod, D., 95, 139, 294–295, 388 B Baas, B., 97 Baba, T., 244 Bacia, K., 135, 186, 318 Bai, Y., 209 Baker, S., 92 Banci, L., 42, 51 Baner, J., 440 Barak, L. S., 388 Barrios-Rodiles, M., 134 Bartley, L. E., 216 Baruah, H., 20 Baskin, R. J., 261–288 Bastiaens, P. I., 242 Bateman, A., 160 Bauch, A., 134 Baum, D. A., 214 Bayburt, T. H., 90, 94 Beach, D. L., 389

Becker, W., 321–322 Beckett, D., 441, 443 Beer, N. R., 64 Benitez, J. J., 41–59 Benson, F. E., 264 Berland, K. M., 347 Bernard, A., 140 Bernath, K., 63 Bertozzi, C. R., 20, 21 Bertrand, E., 388–390, 392 Bianco, P. R., 262–265, 271, 281 Binkert, T., 328 Blair, S., 435 Blanchard, S., 91 Blanchard, S. C., 224, 226, 250, 252, 253 Blanco, L., 432, 440, 449 Blanco, M., 39, 153–176 Blank, K., 452 Block, S. M., 69, 70, 274, 278 Bobroff, N., 394 Bochner, B. R., 236 Bokinsky, G., 160, 161 Boldog, T., 91, 94 Bolen, D. W., 194 Bopp, M. A., 62 Borbat, P., 110 Bornholdt, S., 136 Boukobza, E., 43, 45, 47, 62, 214 Bowers, J., 409, 410 Boxer, S. G., 47, 297 Brameshuber, M., 133–148 Brandi, L., 236 Bratu, D. P., 391 Brau, R. R., 76 Breinbauer, R., 20 Brent, R., 229 Brewer, L. R., 262, 271 Brian, A. A., 47 Broder, Y. C., 135 Brody, J. P., 63 Bronson, J. E., 161, 163, 164, 167, 168 Brucale, M., 181, 200 Brundin, P., 110 Brunger, A., 96 Brustad, E. M., 188, 189 Bussell, R. Jr., 110, 195 Buxbaum, A. R., 387–403

457

458

Author Index C

Cai, L., 120 Cammack, K. A., 226 Campbell, M. J., 264 Caraculacu, A. A., 4 Carbon, J., 249, 250 Carignano, M. A., 3 Cavalcanti-Adam, E. A., 137 Celliers, P. M., 276 Chakrabarti, A. C., 58 Chakraborty, A., 19–28 Chambliss, G. H., 225 Chandra, S., 110 Chang, P. V., 20, 21 Chao, J. A., 392–393 Chartrand, P., 391 Chatterji, D., 437 Cheezum, M. K., 396–397 Chen, P., 41–59 Chen, Y., 345–361 Chen, Z., 264 Chiantia, S., 326 Chin, J. W., 21 Chiou, P. Y., 72 Chiu, D. T., 43 Cho, S. K., 72 Chugh, D., 72 Churchman, L. S., 395 Cianci, G. C., 360 Cisse, I., 43, 58 Clamme, J. P., 198 Clemens, M. J., 254 Cloonan, N., 432 Cobine, P. A., 42, 51 Cohen, A. E., 120 Cohen, H. M., 63 Cokus, S. J., 432 Condeelis, J., 388 Conia, J., 276 Conway, A. B., 264 Cookson, M., 109 Cooper, M., 21 Cormack, B. P., 81 Coseri, S., 4 Courtois, F., 64 Cremer, P. S., 297 Crick, S. L., 186 Crocker, J. C., 398 Cross, A. J., 80 Cruz, J. A., 206 Cui, B. X., 120 Curtis, J. E., 71 D Das, R., 206, 209 Datsenko, K. A., 244 Dave, R., 443

David, H., 249, 250 Davidich, M. I., 136 Davidson, W. S., 110, 195 Davis, R. W., 264 Daviter, T., 224 Davydov, D., 94, 97, 98 Dawson, P. E., 188, 190 Deamer, D. W., 58 Dean, F. B., 440 De Brabander, M., 388, 391 Deiters, A., 21 DeLano, W., 243 Denisov, I., 92, 94, 97, 98, 110 Deniz, A. A., 179–200 Denk, W., 347 Dertinger, T., 319, 325, 339–340 Desai, K. V., 83 de Vega, M., 440 Dickson, R. M., 62 Dictenberg, J. B., 402 Diez, M., 96 Digman, M. A., 135, 319, 323, 326, 347 Dillingham, M. S., 264 Dincbas-Renqvist, V., 232 Dirksen, A., 190 Dittrich, P. S., 64 Ditzler, M. A., 156, 160, 161, 217 Dixon, D. A., 264 Dohm, J. C., 416 Doi, N., 63 Dombrowski, C. C., 261–288 Dorsch, S., 136 Dorywalska, M., 214, 244 Douglas, J. F., 3 Draper, D. E., 206, 209 Drescher, M., 110 Dressman, D., 64 Dube, D. H., 20 Dubnoff, J. S., 227 Duckworth, B. P., 354 Dunker, A. K., 180 Dyson, H. J., 180 E Eaton, W. A., 180 Eberhard, M., 73 Ebright, R. H., 19–28 Ebright, Y. W., 19–28 Eddy, S. R., 160 Edidin, M., 391 Egea, P. F., 348 Eggeling, C., 393 Ehrenberg, M., 225 Eid, J., 433, 436, 440, 444, 447, 449 Eidne, K. A., 135 Einstein, A., 388 Elad, M., 102 Elbing, K., 229

459

Author Index

Elenko, M. P., 217 Eliezer, D., 110, 195 Elson, E. L., 318, 323, 330, 346, 359 Elvekrog, M. M., 221–254 Enderlein, J., 318, 350 Englander, M. T., 221–254 Englander, S. W., 78 English, B. P., 447 Ephrussi, A., 388 Ermolenko, D. N., 249 Erne, P., 73 Esteban, J. A., 432, 440 F Falke, J., 96 Farina, K. L., 392 Fazio, T., 293–314 Feder, T. J., 400 Fei, J., 221–254 Femino, A. M., 366, 373, 384 Fernando, H., 97 Ferreon, A. C., 96, 110, 179–200 Ferri, F., 329 Fields, S., 135 Field, Y., 310–311 Fiore, J. L., 156, 161 Fleming, G. R., 82 Fleury, S., 143 Fogarty, K., 120 Foquet, M., 433, 435 Forchhammer, K., 254 Forkey, J. N., 295 Fo¨rster, T., 11 Fourmy, D., 227 Frank, J., 222, 224, 253 Fredrick, K., 237 Freistroffer, D. V., 239 Fullwood, M. J., 432 Funatsu, T., 391 Fusco, D., 388–389, 392–394, 400, 402 G Gahagan, K. T., 71 Gallagher, S. R., 225, 230, 231 Galletto, R., 262, 264, 269, 286 Gall, J. G., 366 Gambin, Y., 179–200 Gao, F., 102, 105 Garcı´a-Sa´ez, A. J., 96, 341 Garstecki, P., 68 Gasteier, P., 2, 3, 6 Gauchet, C., 20 Gautier, I., 349 Gell, C., 95 Gelles, J., 397 George, J., 110 Georgieva, E. R., 110, 198

Ghadessy, F. J., 64–66 Ghosh, I., 135 Ghosh, R. N., 397–398 Gill, S. C., 27 Gimenes, F., 449 Goedert, M., 109 Goetz, H., 4 Goldberg, R. B., 244 Golding, I., 368 Goldner, L. S., 61–83 Gonzalez, R. L., 221–254 Gopal, V., 437 Gorman, J., 293–314 Gragerov, A., 42 Grakoui, A., 148 Grane´li, A., 294, 297–298, 307 Gratton, E., 388 Greene, E. C., 293–314 Greenfeld, M., 205–218 Gregor, I., 350 Grier, D. G., 398 Griffiths, A. D., 63, 64 Grilley, D., 209 Grinkova, Y., 94 Groll, J., 1–16 Groves, J. T., 297 Grunwald, D., 392, 395, 401 Guengerich, F., 97, 98 Guthrie, C., 31–40 Guttenberg, Z., 72 H Haas, E., 188 Hadjivassiliou, H., 31–40 Hagerman, P. J., 449 Hall, A., 42 Halperin, A., 2 Halpert, J., 97, 98 Hamadani, K. M., 198 Handa, N., 262, 264, 268, 281, 286 Hangauer, M. J., 20, 21 Hang, H. C., 20 Hapke, B., 226 Haran, G., 43, 165 Haran, G., 193 Harder, P., 3 Harriman, P. D., 244 Harris, J. M., 2, 3 Harris, L. J., 130 Harris, T. D., 408–410 Hart, C., 407–429 Hartz, D., 233, 235–237 Haselgru¨bler, T., 133–148 Hase, M., 70 Ha, T., 21, 62, 154, 182, 206, 215, 222, 295 Haugland, R. P., 11, 182 Haupts, U., 353 Hayden, E., 408

460

Author Index

Hebda, J., 106 Hebert, B., 324 Heid, C. A., 64 Heikal, A. A., 82 Heise, B., 133–148 Helenius, A., 195 He, M. Y., 68 Hendler, R., 102 Hendrix, J., 354 Henzler-Wildman, K., 91 Hermanson, G. T., 441 Herschlag, D., 205–218 Hesch, C., 133–148 Hesse, J., 139, 148 Heurgue-Hamard, V., 232 Heyer, W. D., 264 Heyes, C. D., 5, 15, 16 Hilario, J., 262, 264, 266, 269, 270, 284–286, 288 Hillesheim, L. N., 350, 353–354, 357 Hinnebusch, A. G., 254 Hinz, H. J., 195 Hirokawa, G., 241 Hoard, D. E., 438 Ho, C. K., 36 Hodak, J. H., 214 Hoffmann, A., 229 Hogrefe, H. H., 15 Hohng, S., 198, 242 Holzen, T. M., 264 Hope, M. J., 46, 47 Horne, K., 444 Huang, B., 125 Huang, Z., 217 Hu, C. D., 135 Hu, D., 82 Huebner, A., 65, 83 Huffman, D. L., 42, 51 Humenik, M., 20 I Irvine, D. J., 3 Isacoff, E., 102 Ishihama, Y., 391 Ishii, Y., 155 Isin, E., 98 J Jacobson, K., 388 Jager, M., 188 Ja¨hne, B., 399 Jakes, R., 193 Jao, C., 110 Jaqaman, K., 399 Jares-Erijman, E. A., 135 Jelenc, P. C., 225 Jencks, W. P., 207 Jensen, P. H., 195

Jia, Y. W., 62 Jo, E., 110, 195 Jofre, A. M., 61–83 Johnson, J. M., 46, 345–361 Jones, E., 104 Jones, T. B., 72 Joo, C., 171, 180, 210, 215 Jorgenson, J. W., 124 Joseph, S., 237, 238 Jovin, T. M., 135, 242 K Kadokura, M., 438, 439 Kahvejian, A., 408 Kaler, K., 72 Kaltenbrunner, M., 133–148 Kamtekar, S., 449 Kanaya, S., 15 Kane, R. S., 139 Kao, H. P., 396 Kapanidis, A. N., 21, 155, 185 Kaplan, N., 310 Kapp, L. D., 254 Karstens, T., 276 Kask, P., 359 Kato, H., 278, 279 Katsura, S., 70 Keller, A. M., 41–59 Keller, R. A., 120 Kelly, B. T., 63, 64 Kent, S. B., 188 Kiick, K. L., 20, 21 Kijac, A., 94 Kim, B.-E., 42, 51 Kim, P. W., 147 Kim, S. A., 119–131, 318 Kingshott, P., 3 Kiss, M. M., 64 Klein, H. L., 264 Knight, J., 96, 106, 108, 109 Knowles, R. B., 391 Kobs, K., 276 Kochaniak, A., 308 Kohn, M., 20 Kong, Y., 451 Koo, P., 89–112 Koopmans, W. J. A., 9, 10, 12 Korennykh, A. V., 217 Korlach, J., 431–452 Korostelev, A., 242 Korsmeyer, T., 72 Koshland, D., 97 Kotani, N., 135 Koumoutsakos, P., 399 Kou, S. C., 447 Kowalczykowski, S. C., 261–288 Kozuka, J., 96 Kress, B., 442

461

Author Index

Krieg, U. C., 21 Krishnan, R., 191 Kruger, R., 193 Kumaresan, P., 64 Kumar, S., 437, 444 Kusumi, A., 388, 391, 400 Kuszak, A., 91 Kuzmenkina, E. V., 14 L Lakowicz, J. R., 95, 243, 348–349 Lampe, J., 98 Lang, P., 148 Larson, D. R., 366 Laughlin, S. T., 20 Laurence, T. A., 43 Lawrence, J. B., 388 Lawrence, R., 161 Leamon, J. H., 63 64 Lee, A. I., 63 Lee, J. Y., 43 Lee, N. K., 185, 198 Lee, T. H., 160, 224 Legler, D. F., 139 Lehner, M., 133–148 Leitz, A., 94 Lemieux, G. A., 21 Lemke, E. A., 198 Leung, W.-Y., 21 Levene, M. J., 43, 433, 435 Levitus, M., 96, 181 Levi, V., 388, 396 Levsky, J. M., 366 Li, G., 12 Li, H. T., 191 Li, L., 161, 163 Liljas, A., 222 Lindquist, S. L., 191 Lingwood, D., 339 Ling, Y., 106 Link, A. J., 20 Link, D. R., 68 Lin, M. Z., 139 Lin, R. J., 32 Li, P., 206 Lipman, E. A., 120, 198 Lipson, D., 407–429 Li, Q. J., 136, 146 Liu, B., 261–288 Liu, C., 236 Liu, S., 160 Liu, Y., 276, 435 Li, W., 224 Lorenz, R. M., 69, 71 Lorsch, J. R., 254 Lotharius, J., 110 Lu, H. P., 62, 82, 91 Lukacs, K. D., 124

Lundquist, P. M., 442, 443 Luo, J. K., 71, 72 Luo, Y., 129 Lutsenko, S., 42, 51 M Maamar, H., 366 Maar, D., 236 Macdonald, P., 345–361 MacDonald, R. C., 46, 47 MacDougall, D. D., 221–254 Magatti, D., 329, 339 Magde, D., 95, 184, 346 Maher, C. A., 432 Maheshri, N., 366 Maheshwari, G., 4 Malik, N. A., 398 Malmsten, M., 3 Mammen, M., 208 Mardis, E. R., 432, 451 Margulies, M., 64 Marshall, R. A., 222, 224, 244, 253 Martin, K. C., 388 Mason, T. G., 65 Maurel, D., 135 Mazin, A. V., 264 MCnnell, H. M., 47 McIntosh, T. J., 137 McKenna, S. A., 225, 226 McKinney, S. A., 160, 161, 163, 171, 216 McPherson, T., 3 Melin, J., 122 Mendez, J., 450 Meyrueis, P., 442 Michalet, X., 154, 180, 181, 183 Milligan, J. F., 225, 226 Milos, P. M., 407–429 Mingle, L. A., 402 Min, W., 156, 447 Miranker, A. D., 89–112 Mishra, R., 106 Miyake, T., 435 Modesti, M., 264 Moeller, M., 1–16 Moerner, W. E., 120, 180 Mohr, D., 237, 254 Monnard, P.-A., 58 Monod, J., 97 Moore, D. D., 234 Moore, S., 244 Mora, L., 232 Moran, C. R., 179–200 Morrissey, J., 91, 94 Mossman, K. D., 137 Mueller, J. D. 345–361 Muir, T. W., 190 Mukhopadhyay, S., 181, 183, 185, 186, 191–193, 197

462

Author Index

Mulder, B. A., 437 Muller, B., 328, 341 Mu¨ller, J. D., 357, 359, 361 Munro, J. B., 157, 163, 175, 253 Murakoshi, H., 96 Muralidharan, V., 190 Musyanovych, A., 64–66 N Nakano, M., 64 Nanga, R., 106 Nath, A., 89–112 Neher, E., 163 Nelson, J. R., 440 Neuman, K. C., 69, 70, 274, 278 Ngo, J. T., 20 Nguyen, D. P., 21 Nilsson, J., 441 Nimonkar, A. V., 262, 264, 268, 283, 288, 308 Nirmal, M., 391 Noji, H., 62 Noller, H. F., 206, 226, 237, 238 Noll, H., 226 Nooren, I. M.A., 42 O Odom, O. W., 249 Ogawa, T., 264 Ogle, J. M., 224 O’Halloran, T. V., 42, 51 O’Hare, H. M., 139 Ohashi, R., 91 Ohno, S., 21 Okamoto, K., 158 Okumus, B., 43, 45, 48, 49, 62, 214 Orrit, M., 180 Orte, A., 185 Orth, R. N., 137 O’Shea, E. K., 366 Oshlack, A., 424 Ott, D. G., 438 Ozdemir, P., 64 Ozsolak, F., 407–429 P Paar, C., 133–148 Pace, C. N., 194 Palmer, A. G. D., 346, 359 Palo, K., 353, 359 Panchunk-Voloshina, N., 21 Papin, J. A., 134 Park, H. Y., 91, 387–403 Park, S.-Y., 72 Paster, W., 133–148 Pathak, S., 308 Patino, M. M., 191

Pavlov, M. Y., 225 Pederson, T., 391 Pereira, M. J. B., 154, 163, 165, 171, 214 Perrin, J., 388 Perroud, T. D., 128 Peske, F., 241 Peterman, E., 91 Peterman, E. J. G., 276 Petersen, N. O., 347 Petra´ˇsek, Z., 317–341 Petrov, E. P., 321 Petruschke, R., 98 Petukhova, G., 264 Pfeifle, C., 366 Pfleger, K. D., 135 Pleiss, J. A., 33, 34 Plumbridge, J. A., 250 Politz, J. C. R., 401 Pollack, M. G., 72 Polymenidou, M., 191 Polymeropoulos, M. H., 193 Poritz, A. B., 160 Powers, T., 226 Prasad, T. K., 264, 306–309 Prentice, P. A., 71 Prescher, J. A., 20 Priest, C., 73 Protter, M., 102 Puig, O., 134 Pulukkunat, D. K., 221–254 Pushkarev, D., 408, 410, 418, 420 Q Qian, H., 346, 359, 394 Qin, F., 161, 163 Qin, Y., 254 Quake, S. R., 122 Querido, E., 391 R Raap, A. K., 366 Rabiner, A., 161 Raj, A., 365–385 Raleigh, E. A., 229 Ramakrishnan, V., 224 Ramm, P., 148 Rasnik, I., 48, 50, 78, 393, 395 Raz, T., 407–429 Recht, M. I., 226 Reid, D. B., 399 Reiner, J. E., 62, 65, 70, 71, 73, 79, 80 Reynolds, J. A., 195 Rhoades, E., 43, 89–112, 181 Ries, J., 317–341 Rigler, R., 318, 323, 330 Rigneault, H., 435 Ricˇka, J., 328

463

Author Index

Ritchie, T., 90, 92, 93 Robberson, D. L., 264 Roberts, A., 98 Robertson, J. M., 226 Robertson, R. B., 264 Rodriguez, A. J., 391 Rodriguez, I., 449 Roeder, R. G., 229 Roman, L. J., 264 Romano, V., 276, 279 Rook, M. S., 390, 402 Rosato, A., 42, 51 Rosenzweig, A. C., 42, 51 Rosgen, J., 195 Rotman, B., 63 Roy, R., 21, 95, 155–157, 181, 183, 185, 213, 215, 222 Ruan, Q., 319, 337 Rueda, D., 160, 161 S Saffarian, S., 353 Sage, D., 399 Sakmann, B., 163 Salas, M., 440 Sanchez, S., 92 Sandal, M., 181, 200 Santangelo, P. J., 391 Sasaki, K., 71 Sasse, J., 230 Satsoura, D., 324 Sattin, B., 206, 210, 214–218 Sauer, R. T., 244 Saxon, E., 20, 21 Saxton, M. J., 388, 394 Sbalzarini, I. F., 399 Schaffer, J., 82 Scha¨tzel, K., 327, 329 Schatz, P. J., 268 Schlageck, J. G., 437 Schlapak, R., 141 Schmitt, E., 227, 228 Schneider, R. J., 254 Schuette, J. C., 224 Schuler, B., 180, 181, 183, 193 Schultz, P. G., 188 Schuster–Bo¨ckler, B., 160 Schu¨tz, G. J., 133–148, 388, 401 Schuwirth, B. S., 242 Schwartz, J. A., 72 Schwarzenbacher, M., 133–148 Schwille, P., 96, 135, 186, 317–341, 350, 353 Segrest, J., 91, 93 Selmer, M., 223 Serge, A., 399 Serio, T. R., 191 Seshadri, A., 241 Shan, J. G., 391

Sharon, R., 195 Shastry, M. C.R., 73 Shav-Tal, Y., 388, 390, 392, 394, 401 Shaw, A., 94 Sheetz, M. P., 391 Sherman, E., 193 Shimizu, Y., 227 Shinohara, M., 264 Shorter, J., 191 Shrager, R., 102 Shuman, S., 36 Siebrasse, J. P., 401 Silverman, S. K., 214 Silvius, J. R., 45 Simonian, M. H., 230 Simon, S. A., 137 Singer, D., 366 Singer, R. H., 387–403 Skinner, J. P., 323, 332, 335 Slatko, B. E., 233 Sletten, E. M., 20 Sligar, S., 90 Small, D., 91 Smith, A. M., 391 Smith, G. J., 214 Smith, G. R., 264 Smith, J. A., 230 Soderberg, O., 136 Sofia, S. J., 3 Solans, C., 65 Solinger, J. A., 264 Solomatin, S. V., 214, 217 Sonenberg, N., 254 Song, H., 63, 64, 68 Song, L., 58 Song, O., 135 Sonnleitner, A., 133–148 Sood, A., 437, 440 Spahn, C. M., 254 Speidel, M., 396 Spies, M., 262, 264, 281 Spillantini, M. G., 193 Spudich, J. A., 395 Stagljar, I., 135 Stark, H., 224 Stark, M. R., 36 Stasiak, A., 264 St. Claire, R. L., 124 Steinmann, K., 407–429 Sternberg, S. H., 221–254 Stockinger, H., 133–148 Stone, M. D., 214 Strausak, D., 42, 51 Striker, G., 82 Stryer, L., 11, 154, 182 Subramanian, M., 97 Suchanek, M., 135 Sugiura, S., 69 Sundell, C. L., 402

464

Author Index

Sung, P., 264 Sunzenauer, S., 133–148 Superti-Furga, G., 134 Suzuki, K. G., 148 Svoboda, K., 69, 70 Swaminathan, R., 82 Swartzlander, G. A., 71 Szleifer, I., 3 T Tabor, S., 432 Tada, H., 72 Takamoto, K., 212 Talaga, D. S., 62, 162 Taly, V., 63, 64 Tanaka, M., 137 Tan, E., 214 Tanford, C., 195 Tang, J., 61–83 Tautz, D., 366 Tawfik, D. S., 63, 64 Taylor, A. F., 264 Tcherniak, A., 319, 323 Terabe, S., 125 Terazima, M., 158 Thews, E., 328, 341 Thompson, J. D., 243, 407–429 Thompson, N. L., 346 Thompson, R. E., 397 Thornton, J. M., 42 Thorsen, T., 68 Thummel, K., 96 Tirrell, D. A., 20 Tompa, P., 180 Toprak, E., 396 Torres, T., 96, 181, 186, 197 Trexler, A. J., 89–112, 181, 183, 198 Tsalkova, T., 98 Tsao, M. L., 21 Tyagi, S., 365–385, 391 U Ueda, K., 109 Uetz, P., 135 Ulbrich, M., 102 Ulmer, T. S., 110, 196 Umbanhowar, P. B., 69 Unger, M. A., 121 Unsworth, L. D., 3 Uptain, S. M., 191 Uskova, M. A., 82 Utada, A. S., 69 Uversky, V. N., 180 V Valentin, G., 135 Valle, M., 224

Vallotton, P., 399 Vandelinder, V., 198 van der Heijden, T., 264 van Dijk, 76 Van Orden, A., 120 Van Oudenaarden, A., 366 Vargas, D. Y., 367, 368, 384 Varshney, U., 241 Veldhuis, G., 181, 183, 197 Velev, O. D., 72 Velonia, K., 447 Verkman, A. S., 396 Veronese, F. M., 139 Vijayraghavan, U., 38 Villa, E., 224 Visnapuu, M. L., 293–314 Vocadlo, D. J., 20 Vogelsang, J., 105 Volkmer, A., 82, 349 von Hippel, P. H., 27 W Wade, H. E., 226 Wagner, E. G., 225 Wakefield, M. J., 424 Wall, J. D., 244 Walter, N. G., 39, 153–176, 180, 183 Wang, C. C., 21 Wang, D., 19–28 Wang, J., 221–254 Wanner, B. L., 244 Wasserman, L., 167 Wazawa, T., 62 Webb, W. W., 186, 388, 397–398 Weghuber, J., 133–148 Weinger, J. S., 238 Weinreb, P. H., 110, 193 Weiss, S., 21, 43, 154, 155, 198 Wennmalm, S., 62 Westhof, E., 206 Whorton, M., 91 Wickner, R. B., 191 Widengren, J., 349–350, 393 Wieser, S., 388, 401 Wilkinson, G., 96 Williamson, J. R., 208 Williams, P., 98 Wilson, D. N., 222 Wilson, K., 228 Wind, S., 293–314 Wintermeyer, W., 226 Wiseman, P. W., 326, 347 Wright, P. E., 180 Wu, B., 345–361 Wu, H., 120 Wu, M., 137 Wyatt, J. R., 225, 226

465

Author Index X Xie, J., 188 Xie, X. S., 54 Xu, W., 54 Y Yamada, R., 72 Yamagishi, M., 390 Yanagawa, H., 63 Yang, Z., 3 Yao, H., 71 Yarbrough, L. R., 437, 440 Yassour, M., 432 Yildiz, A., 367, 373 Yilmaz, A., 398–399 Yim, P. B., 80 Yoon, T. Y., 62

Yoshikawa, M., 438 Youngman, E. M., 239 Yu, J., 181, 200 Z Zaitsev, E. N., 264 Zalipsky, S., 2, 3 Zare, R. N., 119–131 Zavialov, A. V., 239, 241 Zeldin, R., 98 Zeng, J., 72 Zenklusen, D., 366 Zhang, H. L., 402 Zhang, Y. H., 64 Zhuang, X. W., 160, 161, 180, 183, 206, 214, 216, 227 Zimyanin, V. L., 390

Subject Index

A Alexa488-phosphine synthesis, 23–24 Alexa647-phosphine synthesis, 25–26 Amine labeling, 188 Aminosilanization, 7 Amylin. See Islet amyloid polypeptide (IAPP) Automatic DNA length measurement, 288 Azide-specific biomolecule labeling, 26–27. See also Staudinger–Bertozzi ligation, bioorthogonal labeling B Bait–prey interactions, 148. See also Proteinprotein interaction detection Biotinylated l DNA biotin and digoxigenin, 267 preparation, 265 Bovine serum albumin (BSA) separation, 125, 128 C Caenorhabditis elegans fixation, in situ hybridization, 375–376 gene sequencing, 410 sFCS, 334–335 Capillary electrophoresis (CE), 120. See also Microfluidics Carbonyldiimidazole (CDI), 438, 439 Chromatin immunoprecipitation (ChIP) DNA 30 blocking, 423 poly-A tailing, 423 preparation, 422 Coalescence, droplet, 62–63, 73 Complex barrier patterns, DNA curtains geometric patterns, 304–305 rack patterns double-tethered, 304, 305 linear barriers, 305 pentagons, 306 Complex single-molecule FRET time trajectories FRET state number selection, 167–168 hidden Markov analysis, 160–162 HMM software, 163–164 post-HMM processing and data visualization data condensation and visualization, 169–172 local correlation analysis, 168–170 yeast pre-mRNA splicing, 171, 173–174

preprocessing trajectories formatting, 165–166 outliers removal, 164–165 smoothing (noise reduction), 165 stitching trajectories, 166–167 QuB data preparation, 174 data visualization, 176 HMM analysis, 175 molecule selection, 174 postprocessing, 176 preprocessing, 175 simple trajectories analysis FRET state distribution, 158 kinetic analysis, 158–160 trajectories selection, 157–158 Confocal laser scanning microscope (CLSM), 320–322 Cy3B-phosphine synthesis, 23, 25 Cy3-labeled antibody, DNA substrate, 266–267 Cysteine labeling, 187–188 Cytochrome P450 3A4 (CYP3A4)–Nanodiscs, 96 allosterism, 97–98 CYP3A4 incorporation, 99–101 effectors, 98 Nile Red dye, 98–99 surface attachment, 101–102 D Dielectrophoresis, 72 1,2-Dimyristoyl-sn-glycero-3-phosphocholine (DMPC) lipid membranes, 58 Dipole force. See Gradient force DNA–bead complexes Cy3-labeled antibody binding, 267 DNA labeling, biotin and digoxigenin, 267 fluorescent secondary antibody, 266 preparation, 265–266 DNA curtains, single molecule imaging complex barrier patterns geometric patterns, 303–305 rack patterns, 305–306 DNA molecules aligment bilayer deposition, 298 buffer, 298 fluid bilayer, 297 procedure, 297–298

467

468

Subject Index

DNA curtains, single molecule imaging (cont.) manually etched diffusion barriers, 298, 299 nanofabricated barrier patterns ebeam lithography, 299–304 nanoimprint lithography, 299, 301, 303 TIRFM description, 295 flowcells and injection system, 295–297 simple prism-type, 295 trouble-shooting, 306–307 visualizing protein-DNA interaction ATP-dependent DNA translocation, 308–309 mismatch repair, 310, 312–314 quantum dots, 307–308 tagged nucleosomes, 310, 311 DNA fragmentation and quantitation 30 ends estimation, poly-A tailing, 413–414 shearing, 411–413 size selection, 413 DNA molecule aligment bilayer deposition, 298 buffer application, 298 fluid bilayer, 297 procedure, 297–298 DNA polymerization kinetic parameter determination, 446–448 polymerase pausing circular template sequencing, 449–450 consensus sequence contexts, 450 double-stranded DNA templates, 447, 449 IPD profiles, 449–450 sequencing engine, 440–441 DNA sequencing method. See Single-molecule, real-time (SMRT) DNA sequencing Drosophila melanogaster, 376–377 Dual-color fluorescence cross-correlation spectroscopy (FCCS), 328 Dual-color sFCS, 340–341 Dual-focus sFCS, 339–340 E Ebeam lithography, 299–304 Egg phosphatidylcholine lipid, 45, 47, 48, 57 Electrokinetic injection, 125–127 Electron-multiplying charge-coupled device (EMCCD), 295, 395 Electrowetting, 72 Emulsification, 65–66 Ensemble averaging, 156, 180, 217, 262 Escherichia coli-based in vitro translation system, smFRET elongation polypeptide synthesis assay, 238–239 primer-extension inhibition assay, 237–238 initiation assays GTP hydrolysis assay, 236–237

primer-extension inhibition assay, 233–236 mRNA preparation, 226–227 phylogenetic analysis/structural modeling, 243 ribosome labeling fluorescently labeled r-proteins, 247–248 mutant ribosomes preparation, 244–247 r-proteins reconstitution, 248–249 ribosome recycling, 241–242 ribosomes and ribosomal subunits, 226 termination elongation reaction, 239–240 polypeptide release assay, 239–241 release reactions, 240 RF1/2 activity determination, 240–241 translation factors cleavage reaction, 230 EF-G, 232 EF-Tu, 232 IF1, 230–231 IF3, 232 IF2g, 231–232 labeling, 252–253 Ni2þ-NTA affinity purification, 229–230 RF1 and RF2, 232 tris–polymix buffer, 225–226 tRNA labeling, 249–252 tRNA synthetase, 227–228 F FCS. See Fluorescence correlation spectroscopy (FCS) Figures of merit, protein interaction detection, 147–148 Flow cell design, 270–271 fabrication, 271–273 illustration, 262–263 Fluorescence correlation spectroscopy (FCS), 95. See also Scanning fluorescence correlation spectroscopy (sFCS) IDP, 186 microfluidics autocorrelation function, 127–128 immunocomplex formation, 130 multivalent antibody hypothesis, 130–131 parameters, 129, 130 TMR-labeled BSA and anti-BSA antibody, 128–129 rIAPP binding Nanodiscs, 107–109 Fluorescence fluctuation spectroscopy (FFS), 346 Fluorescence in situ hybridization (FISH), 383 Fluorescence quenching EGFP and protein X labeling, 348–349 environmental changes, 348 types, 348 Fluorescence recovery after photobleaching (FRAP), 388–389

469

Subject Index

Fluorescence resonance energy transfer (FRET). See also Single-molecule fluorescence resonance energy transfer (smFRET) measurement, 154–155 membrane proteins, 95 ratio, 154 a-synuclein (aS), Nanodiscs, 111–112 Fluorescent oligonucleotide probe design, 369 synthesis and purification Alexa 594, 373 HPLC samples, 370–372 oligonucleotide synthesis, 369–370 optical filters, 371, 373 three color mRNA detection, 371, 374 TMR, 370 Fluorescent pre-mRNA substrates, smFRET microscope slide, 38–39 pre-mRNA mutation, 37–39 spliceosome assembly, 32 splicing efficiency dyes, 37, 38 Ubc4 pre-mRNA, 33, 35 synthetic fluorescent Ubc4 pre-mRNA design, 35 oligonucleotide ligation, 36–37 precipitation, 35–36 purification, 33–34 yeast pre-mRNA in vitro splicing exon length requirement, 35 prp2–1 mutant, 33 relative splicing efficiency, 34 ribosomal protein RPS6A, 33 Fluorescent probes, 21. See also Staudinger–Bertozzi ligation, bioorthogonal labeling Fluorescent semiconducting nanocrystals. See Quantum dots Fluorescent spot, 367–368 Fo¨rster resonance energy transfer. See Fluorescence resonance energy transfer (FRET) FRET. See Fluorescence resonance energy transfer (FRET) G Gaussian mask algorithm, 398 Genomic DNA preparation DNA fragmentation and quantitation 30 end estimation, poly-A tailing, 413–414 shearing, 411–413 size selection, 413 30 end blocking, 415–416 poly-A tailing, 415 Gradient force, 69 GTP hydrolysis assay, 236–237

H Helicos single-molecule sequencer bacterial genome sequencing coverage assessment and lack of bias, 417, 419 preparation, 416–418 cDNA blocking, 427–428 poly-A tailing, 426–427 single-stranded, preparation, 424–426 chromatin immunoprecipitation (ChIP) DNA 30 blocking, 423 poly-A tailing, 423 preparation, 422 copy number variation assessment, 420 genomic amplification visualization, 421–422 vs. genomic hybridization, 421–422 DGE reproducibility, 428 genomic DNA preparation DNA fragmentation and quantitation, 411–414 30 end blocking, 415–416 poly-A tailing, 415 Helicos Flow Cell Image and Virtual TerminatorÒ , 410, 411 nucleotid incorporation, 409 principles, 409–410 real-time image processing, 410 sequencing-by-synthesis reaction, 410 a-Hemolysin self-assembly, 58 Hidden Markov model (HMM), FRET trajectories. See also QuB program, FRET trajectories data condensation and visualization complex trajectories, 172 POKIT, 171 TDP, 169–171 five-state system analysis, 161, 162 FRET state number selection, 167–168 local correlation analysis, 168–170 parameter sets, 160–161 preprocessing trajectories formatting trajectories, 165–166 outliers removal, 164–165 smoothing (noise reduction), 165 stitching trajectories, 166–167 software programs, 161, 162 features, 164 HaMMy, 163 QuB, 163 vb-FRET, 163–164 transition rate constant, 161–162 yeast pre-mRNA splicing, 171, 173–174 Homologous recombination (HR), 263

470

Subject Index I

L

IAPP. See Islet amyloid polypeptide (IAPP) IDP. See Intrinsically disordered proteins (IDPs) Imaging mRNA movement image-acquisition protocol, 394 microscopic system noise reduction, 395 physiological conditions, 395–396 three-dimensional tracking, 396 Injection, droplet loading protocol, 67 micropipettes, pulling protocol, 67–68 schematic representation, 66–67 In situ hybridization buffer B, 374 FISH, 383 fixation protocols adherent mammalian cell, 375 Caenorhabditis elegans, 375–376 Drosophila melanogaster wing imaginal discs, 376–377 freezed tissue sections, 377 yeast cells fixation, 375 fixation solution, 373–374 M9 salt, 375 spheroplasting buffer, 375 Intrinsically disordered proteins (IDPs) single-molecule fluorescence method dual-color single-molecule coincidence, 185–186 FCS, 186 smFRET, 182–185 site-specific labeling amine labeling, 188 cysteine labeling, 187–188 dual-labeling, FRET, 189–190 orthogonal chemical reactions, 189 protein ligation, 190 unnatural amino acid functional groups, 188–189 structure and dynamics a-synuclein, 193–199 yeast prion protein Sup35, 191–193 Islet amyloid polypeptide (IAPP) FCS measurement, 107–109 membrane interactions, 106–107 type II diabetes, 106 Isocyanate terminated star molecules (NCO–sP (EO-stat-PO)) biosensors, 11 protein structure and function immobilized biomolecule, 12–13 nucleosomes, 12 RNase H, 14–16 surface coating aqueous solution, 4–5 reactivity, 5–6

Lab-on-chip method, 71–72 Laplacian of Gaussian (LoG) filters, 384 Lipid nanovesicle preparation lipid film, 46 lipid selection, 45 unilamellar nanovesicles, 46–47 Living cell mRNA labeling MS2-GFP system, 392–393 photobleaching and phototoxicity limitation, 393 probe selection, SPT, 391 Low-density lipoprotein (LDL) receptors, 388 M Microbiochip, 138, 143–144, 146 Microcontact printing experimental design, 139–140 procedure, 141–142 Microfluidics detergent-assisted microchannel electrophoresis electrokinetic injection, 125–127 PDMS microchip, 124–125 separation buffer and sample preparation, 125 droplets, 64–65, 68–69 electrophoretic biomolecule separation, 120 FCS autocorrelation function, 127–128 immunocomplex formation, 130 multivalent antibody hypothesis, 130–131 parameters, 129, 130 TMR-labeled BSA and anti-BSA antibody, 128–129 laser-induced fluorescence detection, 123–124 laser trap microscope, proteins on DNA dual optical trap imaging system, 275–276 schematics, 274, 275 single optical trap imaging system, 273–275 microchip fabrication design drawing and photomask printing, 121–122 molding master fabrication, 122 PDMS chip fabrication, 122–123 single-molecule experiment, 120 Micropatterning technique advantages, 147–148 CD4-YFP protein interaction, 143 principle, 137–138 Mismatch repair (MMR), 310, 312–314 Mold master, 122 mRNA tracking, living cell. See also Quantum dots FRAP technique, 388–389 imaging mRNA movement image-acquisition protocol, 394

471

Subject Index

microscopic system, 395–396 three-dimensional tracking, 396 labeling MS2-GFP system, 392–393 photobleaching and phototoxicity limitation, 393 probe selection, SPT, 391 motion analysis data interpretation, 401–402 localization algorithms, 397–398 particle tracking, 398–399 single particle motion, 399–401 types, 396–397 significance distribution, COS cells, 389 localization, 390 single particle tracking (SPT), 388 technical development, 388 MS2-labeling technique, 392–393 Multiple hypotheses tracking (MHT) algorithm, 399 Multiple-tau method, 328–329 N Nanodiscs. See Phospholipid bilayer Nanodiscs Nanodroplet confinement cell biology, 83 compartmentalization, 63–64 droplet coalescence and mixing, 62–63, 73 droplet generation emulsification, 65–66 injection, 66–68 microfluidics, 68–69 lab-on-chip method dielectrophoresis, 72 electric field, 71–72 electrowetting, 72 surface acoustic wave, 72 microfluidics, 64–65 (see also Microfluidics) molecule confinement, 62–63 molecule reaction kinetics, 63 optical manipulation force, 69–70 optical tweezers, 70–71 optical vortices, 71 PCR, 64 single fluorophore detection acousto-optical modulators (AOMs), 76 apparatus alignment, 74, 76–77 droplet injection, 77–78 emulsion sample preparation, 77 filters, 75 polarizing beamsplitter (PC), 75 removable mirror (RM), 75 technique, 74–75 Ytterbium fiber laser, 73–74 single-molecule measurement FRET, RNA molecule, 80

photobleaching, dye molecule, 79–80 time-resolved fluorescence anisotropy, EGFP, 81–82 Nanofabricated barrier patterns, DNA curtains ebeam lithography, 299–304 nanoimprint lithography descum process, 301, 303 PMMA layers, 299, 301 preimprint phase, 301 Nanoimprint lithography, 299, 301, 303 Nanovesicle trapping, weak protein interactions advantages, 44 characteristics, 42 DMPC lipid membranes, 58 effective concentration vs. vesicle diameter, 45 Egg PC limitation, 57–58 a-hemolysin self-assembly, 58 lipid nanovesicle preparation lipid film preparation and hydration, 46 unilamellar nanovesicles, 46–47 lipid selection, 45 protein trapping, 46–47 single-molecule kinetic analysis conditional probability, 55 dissociation constant, 57 dwell time, 56–57 generic kinetic scheme, 54 kinetic rate equations, 54–55 probability density, 56 smFRET measurement acceptor-blinked/bleached states, 50–52 intracellular copper transporters, 51–53 lipid–protein interactions, 49 nanovesicle occupation, 49–50 nanovesicle surface immobilization, 47–49 schematics, 43–44 zero-mode waveguides, 43 Nile Red (NR) binded CYP3A4, 98–99, 103 O Optical distortions, 319 Optical trapping. See Optical tweezers Optical tweezers, 70–71 Optical vortices, 71 P PEO. See Poly(ethylene oxide) (PEO) surface passivation Phosphine Alexa488 synthesis, 22–23 Alexa647 synthesis, 24–26 biotin and fluorescent probes, 21 Cy3B synthesis, 23, 25 Phospholinked nucleotide principle fluorescence emission, 434–435 molecular structure, 434

472 Phospholinked nucleotide (cont.) SMRT sequence determination, 433–434 uninterrupted DNA polymerization Alexa Fluor 488-aminohexyl-dG5P, 439 aminohexyl-dG5P, 439 enzymatic purification, 440 Fmoc-6-aminohexyldiphosphate, 438–439 Fmoc-6-aminohexylphosphate, 438 linker, 437–438 unmodified dNTPs, 437–438 Phospholipid bilayer Nanodiscs application, 91 characteristics, 94 vs. conventional membrane, 91 cytochrome P450 3A4 (CYP3A4) allosterism, 97–98 effectors, 98 incorporation of, 99–101 Nile Red dye, 98–99 surface attachment, 100–102 HDL particles formation, 91–92 lipid–protein ratio, 93 POPC–MSP1D1 discs, 92–94 IAPP FCS measurement, 107–110 membrane interactions, 106–107 type II diabetes, 106 membrane protein study, 94 oligomerization, 94 self-assembly, 93 single-molecule techniques, 95–96 SVD based image filtering artifacts, photobleaching, 105 noise reduction, 102–104 pseudocode, 104–105 a-synuclein (aS) conformation, 110 smFRET measurement, 110–112 Photobleaching, 79–80, 319, 330–331 Photomask printing, 121–122 Photoresist, 122 Poly-A tailing cDNA, 426–427 ChIP DNA, 423 3’ ends estimation, 413–414 oligonucleotide digestion, 427 procedure, 415 Polymethylmethacrylate (PMMA), 300, 301 Poly(dimethylsiloxane) (PDMS) microfluidic chip fabrication, 122–123 microchannel electrophoresis, 124–125 Poly(ethylene oxide) (PEO) surface passivation end-functionalized groups, 3 grafting density, 3–4 linear vs. cross-linked PEO, 2 NCO-sP(EO-stat-PO) system protein structure and function, 11–16

Subject Index

surface coating, 4–6 PEO star size, 3–4 protein-repellant coatings, 2–3 quality, 10 sP(EO-stat-PO) substrate aminosilanization, 7 biotin and streptavidin formation, 9 layer precipitation procedure, 8–9 nonfouling properties, 9, 10 preparation, 6 schematics, 9–11 spin-coating, 7–8 POPC–MSP1D1 Nanodiscs characteristics, 94 oligomerization, 94 preparation, 92–94 Population-weighted and kinetically indexed transition density (POKIT), 171 Primer-extension inhibition assay EF-G(GTP), 238 IF1, 235–236 IF3, 236 IF2g activity, 235 initiation reaction, 234 Lys-tRNALys ternary complex, 238 Phe-tRNAPhe ternary complex, 237–238 primer labeling reaction, 233 Protein interactions and stoichiometry, living cells brightness classification fluorescent proteins, 354 single brightness state, 347–349 two brightness state, 349–353 brightness titration monomeric protein state, 355 normalized brightness values, 355–356 oligomeric state, 356 protein expression variability, 354 TR4-EGFP labeling, 355 cell selection bright-field imaging, 358 excitation powers, 360 morphology, 357–358 photobleaching, 359–360 signal-to-noise ratio, 360 thickness, 360–361 two-photon FFS mode, 358 control and calibration experiments fluorescent label EGFP, 356 normalized brightness measurement, EGFP2, 356–357, 359 Western blot gel, 357, 358 FFS parameter, 346 Protein ligation, 190 Protein-protein interaction detection applicability living cells, 136 plasma membrane proteins, 137 bait–prey interactions, 134–135, 148

473

Subject Index

capture ligand, 139 CD4-Lck interaction, 145–147 cellular expression system, 139 chip production, 139–140 dynamic range/sensitivity, 137 false negatives/false positives, 137 figures of merit, 147–148 high throughput capabilities, 137 instrumentation, 139 living cells, 135 micropatterning technique, 137–138 procedure cell incubation, micropatterned surface, 142–143 data analysis, 144–145 microcontact printing, 141–142 microscopy, 143–144 quantification, 136 resting state analysis, 148 signaling analysis, 148 weak interaction detection, 136 Protein repellant coating, 2–3 Proteins on DNA automatic DNA length measurement, 288 DNA substrate preparation biotinylated l DNA, 265 Cy3-labeled antibody, 266–267 DNA–bead complexes, 265–266 ensemble averaging, 262 flow cell design, 270–271 fabrication, 271–273 illustration, 262–263 fluorescent proteins preparation chemically modified fluorescent RecA/ Rad51 proteins, 269–270 Rad54/Tid1 labeling, 268–269 RecBCD labeling, 268 laser trap microscope and microfluidic system dual optical trap imaging system, 275–276 schematics, 274, 275 single optical trap imaging system, 273–275 RecA/Rad51, 264 RecBCD helicase/nuclease, 264 recombinational DNA repair, 263 single-molecule imaging DNA unwinding, 281 experimental steps, 280–281 Rad54/Tid1 translocation, 283–284 real-time Rad51 assembly, 284–285 real-time Rad51 disassembly, 285–286 RecAFAM/RecA-RFP/Rad51FAM filament formation, 286–287 RecBCD–nanoparticle translocation, 281–283 temperature determination and control, instruments schematics, 277

temperature control, 280 temperature determination, 276–279 thermal gradient, 279–280 two-dimensional Gaussian fitting, 288 Protein synthesis ribosome structure, 222, 223 smFRET studies (see Single-molecule fluorescence resonance energy transfer (smFRET)) Protein trapping, 46–47 Protocatechuate dioxygenase (PCD), 443 Protocatechuic acid (PCA), 443 Q Quantum dots, 391. See also mRNA tracking, living cell antibody conjugation, 308 disadvantage, 307 labeling strategy, 308 QuB program, FRET trajectories data preparation, distribution analysis and HMM, 174 data visualization, 176 HMM analysis, 175 molecule selection, 174 postprocessing, 176 preprocessing, 175 R Rad51 proteins, 264 assembly, 284–285 chemical modification, 269–270 disassembly, 285–286 Rad54/Tid1 proteins labeling, 268–269 translocation, 283–284 Raster image correlation microscopy (RICS), 326 RecA proteins, 264 chemical modification, 269–270 RecAFAM/RecA-RFP/Rad51FAM filament formation, 286–287 RecBCD enzyme DNA unwinding, 264, 281 labeling, 268 nanoparticle translocation, 281–283 RNase H, 14–16 RNA transcripts detection fluorescent oligonucleotide probe sets design, 369 synthesis and purification, 369–373 fluorescent spot, 367–368 hybridization protocol, 379–381 solution, 378–379 image analysis computational spot identification, 384–385 deconvolution software, 384

474

Subject Index

RNA transcripts detection (cont.) LoG filter, 384 microscopic imaging antifade mounting media, 383 single mRNA, various samples, 381–382 spot size variation, 384 staining method, 383–384 widefield microscopy, 381–383 oligonucleotide synthesis, 367 RNA binding protein, 368 in situ hybridization, 366 fixation protocols, 375–377 fixation solution, 373–375 target mRNAs probing, 366–367 S Scanning fluorescence correlation spectroscopy (sFCS) application Caenorhabditis elegans embryo, 325, 334–335 dual-color sFCS, 340–341 dual-focus sFCS, 339–340 perpendicular scan path, 337–339 small-circle sFCS, 335–337 correlation curves autocorrelation, 327 dual-color cross-correlation spectroscopy, 327–328 linear correlation, 329 multiple-tau method, 328–329 spatiotemporal cross-correlation, 329 data fitting autocorrelation function, 333 data points, 334 fitting parameters, 332 experimental steps, 320, 321 fluorescent background, 330 laser scanning microscope system, 322 measurement volume, 320–321 photobleaching, 330–331 photon counting detector, 321 principle, 318 scan paths calibration, 326–327 circular path, 324–325 double-line scan, 326 large-range scans, 323–324 membrane motion, 326 raster scan, 326 spatial heterogeneities, 331–332 uses, 319–320 Scan paths, sFCS calibration, 326–327 circular path, 324–325 double-line scan, 326 large-range scans correlation analysis, 324

spatiotemporal correlation, 323–324 membrane motion, 326, 338, 340 properties, 323 raster scan, 326 Scattering force, 69 Single brightness state, fluorescent molecules homoFRET, 349 quenching, 348–349 Single fluorophore detection, nanodroplet acousto-optical modulators (AOMs), 76 apparatus alignment, 74, 76–77 droplet injection, 77–78 emulsion sample preparation, 77 filters, 75 polarizing beamsplitter (PC), 75 removable mirror (RM), 75 technique, 74–75 Ytterbium fiber laser, 73–74 Single-molecule fluorescence resonance energy transfer (smFRET). See also Complex single-molecule FRET time trajectories elongation polypeptide synthesis assay, 238–239 primer-extension inhibition assay, 237–238 energetic coupling, tertiary RNA contacts cooperativity measurement, 217 design, single molecule construct, 212–215 identification, 210, 212 knocking out, 212 single molecule construct validation, 215–217 ensemble-averaging, 156 features, 156 fluorescent labeling, 224 fluorescently labeled translation components phylogenetic analysis/structural modeling, 243 ribosome labeling, 244–249 translation factor labeling, 252–253 tRNA labeling, 249–252 fluorescent pre-mRNA substrates microscope slide, 38–39 pre-mRNA mutation, 37–39 spliceosome assembly, 32 splicing efficiency, dyes, 37, 38 synthetic fluorescent Ubc4 pre-mRNA, 33–37 yeast pre-mRNA in vitro splicing, 33–35 fluorophores, 155–156 highly purified in vitro translation system, Escherichia coli mRNA preparation, 226–227 ribosomes and ribosomal subunits, 226 translation factors, 228–232 tris–polymix buffer, 225–226 tRNA synthetase, 227–228 IDP confocal format, 183, 185

Subject Index

protein dual-labeling, 189–190 protein ligation, 190 schematics, 184 a-synuclein, 193–199 TIRF format, 185 transfer efficiency, 182–183 yeast prion protein Sup35, 191–193 initiation assays GTP hydrolysis assay, 236–237 primer-extension inhibition assay, 233–236 instrumentation, 155 mRNA-encoded aa-tRNA, 224 protein structure and function FRET efficiency, 12–14 immobilized biomolecule, 12–13 nucleosomes, 12 RNase H, 14–16 ribosome labeling fluorescently labeled r-proteins, 247–248 mutant ribosomes preparation, 244–247 r-proteins reconstitution, 248–249 ribosome recycling, 241–242 termination, 239–241 weak protein interactions acceptor-blinked/bleached states, 50–52 advantages, 42 concentration limitation, 43 dynamic protein–protein interactions, 42 intracellular copper transporters, 51–53 lipid–protein interactions, 49 nanovesicle occupation, 49–50 nanovesicle trapping, 43–44 surface immobilization (see Surface immobilization, nanovesicle) Single-molecule fluorescence (SMF) spectroscopy bioorthogonal labeling (see Staudinger–Bertozzi ligation, bioorthogonal labeling) dual-color single-molecule coincidence, 185–186 fluorescence correlation spectroscopy (FCS), 186 microfluidics (see Microfluidics) Nanodiscs (see Phospholipid bilayer Nanodiscs) single-molecule fluorescence resonance energy transfer (smFRET), 182–185 (see also Single-molecule fluorescence resonance energy transfer (smFRET)) Single-molecule kinetic analysis, three-state protein interactions conditional probability, 55 dissociation constant, 57 dwell time, 56–57 generic kinetic scheme, 54 kinetic rate equations, 54–55 probability density, 56

475 Single-molecule, real-time (SMRT) DNA sequencing compound prism, 443 data analysis dye types, 444 enzymatic reaction cycle, 445 fluorescence pulse, 445–446 DNA polymerase, 440–441 DNA squencing assay, 443–444 dynamics DNA polymerase pausing, 447, 449–450 kinetic parameters determination, 446–448 optical system, DNA sequencing, 441–443 phospholinked dNTP Alexa Fluor 488-aminohexyl-dG5P, 439 aminohexyl-dG5P, 439 Fmoc-6-aminohexyldiphosphate, 438–439 Fmoc-6-aminohexylphosphate, 438 linker, 437–438 purity, 440 principle phospholinked nucleotide, 433–435 ZMW nanostructures, 433, 434 ZMW fabrication, 435, 436 surface derivatization, 435–437 Single particle tracking (SPT) method. See mRNA tracking, living cell Singular-value decomposition (SVD) based image filtering artifacts, photobleaching, 105 noise reduction, 102–104 pseudocode, 104–105 Spatiotemporal image correlation spectroscopy (STICS), 324 Spin casting, 2, 7–9 Spin coating flow diagram, 8–9 homogeneity, 7–8 Staudinger–Bertozzi ligation, bioorthogonal labeling ˚ Alexa647-phosphine20 A synthesis, 25–26 ˚ 24 A synthesis, 26 Alexa647-phosphine Alexa488-phosphine synthesis, 23 applications, 20 azide-specific labeling, 26–27 Cy3B-phosphine synthesis, 25 fluorescent probes, 21 materials, 21–23 phosphine derivatives, 21 quantitation labeling efficiency, 27 labeling specificity, 27–28 reaction, 20 reversed-phase HPLC, 22 strategies, 20–21 Surface grafting, 2–4 Surface immobilization, nanovesicle

476

Subject Index

Surface immobilization, nanovesicle (cont.) bovine serum albumin (BSA) coating, 48 lipid bilayer coating, 47–48 polyethylene glycol (PEG) coating, 48–49 a-Synuclein (aS) conformation, 110 disordered state GdnHCl-induced changes, 194–195 noncooperative protein transition, 195 Parkinson’s disease, 193 smFRET measurement, Nanodiscs, 111–112 sodium dodecyl sulfate (SDS) binding modes, 196–197 F conformation, 198, 199 labeling, 195–196 structural dynamics, 197–198 T Tertiary contacts, RNA folding cooperativity measurement, 217 folding equilibrium, 209–210 identification, 210, 212 ionic condition, 217–218 knocking out, 212 single molecule constructs design considerations, 212–213 equilibrium constant, FRET, 216 FRET pair incorporation, 213 heterogeneity, smFRET, 217 P4-P6 cooperativity, 215–216 surface tether, 214–215 synthetic RNA, 214 variability, 215 thermodynamic cooperativity complex chemical equilibrium, 207 free energy diagram, 208–209 generic thermodynamic cycle, 207–208 Testicular orphan receptor 4 (TR 4), 355 Tetramethylrhodamine (TAMRA), 107–109 Tetramethylrhodamine (TMR), 128, 370 Time-resolved fluorescence anisotropy, EGFP, 81–82 Toeprinting assays EF-G(GTP), 238 IF1, 235–236 IF3, 236 IF2g activity, 235 initiation reaction, 234

Lys-tRNALys ternary complex, 238 Phe-tRNAPhe ternary complex, 237–238 primer labeling reaction, 233 Total internal reflection fluorescence microscopy (TIRFM), 183, 395 description, 295 membrane proteins, 95–96 microfluidic flowcells and injection system, 295–297 simple prism-type, 295 Transcriptome quantitation, digital gene expression cDNA blocking, 427–428 poly-A tailing, cDNA determination, 427 reaction, 426–427 RNA Seq, 423–424 single-stranded cDNA preparation cDNA synthesis, 424 quantification, 426 RNA digestion, 425–426 sample cleanup, 426 smsDGE reproducibility, 428 Transition density plot (TDP), 163 Two brightness state flickering, 350, 351 long-lived states, 351–353 short-lived states and flickering, 350 Two-dimensional Gaussian fitting, 288 U Uridine, 33, 37, 38 Y Yeast prion protein Sup35, 191–193 Z Zero-mode waveguides (ZMW), DNA sequencing fabrication, 435, 436 principle, 433–434 surface derivatization, enzyme immobilization aluminum-clad, 435–436 biotin PEG silane, 437 neutravidin binding, 436–437