METHODS IN ENZYMOLOGY Editors-in-Chief
JOHN N. ABELSON AND MELVIN I. SIMON Division of Biology California Institute of ...
22 downloads
835 Views
13MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
METHODS IN ENZYMOLOGY Editors-in-Chief
JOHN N. ABELSON AND MELVIN I. SIMON Division of Biology California Institute of Technology Pasadena, California, USA 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 2009 Copyright # 2009, 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: 978-0-12-380922-3 ISSN: 0076-6879 Printed and bound in United States of America 09 10 11 12 10 9 8 7 6 5 4 3 2 1
CONTRIBUTORS
Benjamin M. Akiyama Department of Molecular, Cell, and Developmental Biology, and Center for Molecular Biology of RNA, University of California, Santa Cruz, California, USA Nathan A. Baker Computational and Molecular Biophysics Program, and Department of Biochemistry and Molecular Biophysics; Center for Computational Biology, Washington University in St. Louis, St. Louis, Missouri, USA Robert T. Batey Department of Chemistry and Biochemistry, University of Colorado at Boulder, Boulder, Colorado, USA Dana A. Baum Department of Chemistry, Saint Louis University, St. Louis, Missouri, USA Thu Betteridge Department of Biochemistry and Molecular Biology, Thomas Jefferson University, Philadelphia, Pennsylvania, USA Eric B. Brauns Department of Chemistry, University of Idaho, Moscow, Idaho, USA Pavol Cekan Department of Chemistry, Science Institute, University of Iceland, Reykjavik, Iceland Alan A. Chen Computational and Molecular Biophysics Program, and Center for Computational Biology, Washington University in St. Louis, St. Louis, Missouri, USA Shi-Jie Chen Department of Physics and Astronomy, and Department of Biochemistry, University of Missouri, Columbia, Missouri, USA Sebastian Doniach Department of Applied Physics, and Department of Physics, Stanford University, Stanford, California, USA David E. Draper Department of Biophysics, and Department of Chemistry, Johns Hopkins University, Baltimore, Maryland, USA xiii
xiv
Contributors
R. Brian Dyer Department of Chemistry, Emory University, Atlanta, Georgia, USA Laura E. Easton MRC Laboratory of Molecular Biology, Cambridge, United Kingdom Kenneth D. Finkelstein Cornell High Energy Synchrotron Source, Cornell University, Ithaca, New York, USA Andrew D. Garst Department of Chemistry and Biochemistry, University of Colorado at Boulder, Boulder, Colorado, USA Max Greenfeld Department of Chemical Engineering and Biochemistry, Stanford University, Stanford, California, USA Kathleen B. Hall Department of Biochemistry and Molecular Biophysics, Washington University School of Medicine, St. Louis, Missouri, USA Daniel Herschlag Departments of Biochemistry and Chemistry, Stanford University, Stanford, California, USA Ya-Ming Hou Department of Biochemistry and Molecular Biology, Thomas Jefferson University, Philadelphia, Pennsylvania, USA Amanda Y. Keel Department of Biochemistry and Molecular Genetics, and University of Colorado Denver, Aurora, Colorado, USA Jeffrey S. Kieft Howard Hughes Medical Institute, and Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, Colorado, USA Eda Koculi Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, Evanston, Illinois, USA Dominic Lambert Department of Chemistry, Johns Hopkins University, Baltimore, Maryland, USA Desirae Leipply Department of Biophysics, Johns Hopkins University, Baltimore, Maryland, USA
Contributors
xv
David M. J. Lilley Cancer Research UK Nucleic Acid Structure Research Group, MSI/WTB Complex, The University of Dundee, Dundee, United Kingdom Jan Lipfert Kavli Institute of Nanoscience, Delft University of Technology, Delft, The Netherlands Cuiping Liu Department of Biochemistry and Molecular Biology, Thomas Jefferson University, Philadelphia, Pennsylvania, USA Peter J. Lukavsky MRC Laboratory of Molecular Biology, Cambridge, United Kingdom Marcelo Marucho Department of Biochemistry and Molecular Biophysics, and Center for Computational Biology, Washington University in St. Louis, St. Louis, Missouri, USA Somdeb Mitra Department of Biochemistry, Albert Einstein College of Medicine, Bronx, New York, USA Suzette A. Pabit School of Applied and Engineering Physics, Cornell University, Ithaca, New York, USA Rohit V. Pappu Computational and Molecular Biophysics Program, and Center for Computational Biology; Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, Missouri, USA Lois Pollack School of Applied and Engineering Physics, Cornell University, Ithaca, New York, USA Peter Z. Qin Department of Chemistry, University of Southern California, Los Angeles, California, USA Francis E. Reyes Department of Chemistry and Biochemistry, University of Colorado at Boulder, Boulder, Colorado, USA Olav Schiemann Centre for Biomolecular Sciences, Centre of Magnetic Resonance, University of St Andrews, St Andrews, United Kingdom Xuesong Shi Department of Biochemistry, Stanford University, Stanford, California, USA
xvi
Contributors
Snorri Th. Sigurdsson Department of Chemistry, Science Institute, University of Iceland, Reykjavik, Iceland Scott K. Silverman Department of Chemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA Sergey Solomatin Department of Biochemistry, Stanford University, Stanford, California, USA Michael D. Stone Center for Molecular Biology of RNA, and Department of Chemistry and Biochemistry, University of California, Santa Cruz, California, USA Zhi-Jie Tan Department of Physics, Wuhan University, Wuhan, Hubei, China Sarah A. Woodson T.C. Jenkins Department of Biophysics, Johns Hopkins University, Baltimore, Maryland, USA Xiaojun Zhang Department of Chemistry, University of Southern California, Los Angeles, California, USA
PREFACE
After the discovery of catalytic RNA nearly 30 years ago, and after the initial excitement wore off, RNA was viewed predominantly as an ancient biological macromolecule with vestigial, albeit critical, functions in modernday biology. Thus, while important and informative, studies of RNA behavior and function took a back seat to the interrogation of transcription factors, protein kinases, and other molecules directly involved in the regulation of gene expression. But more recent discoveries have clearly illuminated the central importance of RNA to modern-day biology—the discovery of fewer genes but vastly more alternative spliced gene products than anticipated in the human genome, the discovery of RNA regulatory elements in the form of riboswitches, the finding of many noncoded RNAs, the finding of functional groupings of RNAs by RNA binding proteins, and, of course, the discovery of RNA interference (RNAi) and its likely role in regulation of about half of all human genes. Thus, it is now clear that we need to understand these molecules in order to understand how biology works, and applications to understanding and curing diseases, while still largely remote, are ultimately likely to become common. Fortunately, in the years since the discovery of catalytic RNA, many incisive and powerful chemical, biochemical, and biophysical tools have been developed to study the folding and conformational behavior of RNA. Here, we have assembled many of these together, in this two volume series. Descriptions of two of the most common and powerful techniques are not covered, NMR and X-ray crystallography (although there is a chapter on preparation of RNA for crystallography), as these approaches can warrant volumes on their own and have been treated separately (see, e.g., Volume 394 on Biological NMR). I believe that this compilation is particularly important, not just because of the importance of the subject matter, but because the tools and contributors span previously distinct fields from molecular biology, biochemistry, chemistry, and physics. Now students and researchers in each of these areas can get a sense of, as well as detailed protocols for, the entire gambit of approaches. I hope that the efforts of the many contributors will inspire new students and investigators to join the search for understanding of these most fascinating macromolecules. And finally I want to thank a phenomenal group of contributors for agreeing to contribute and then coming through, even sometimes on time, with chapters of uniform high quality. DAN HERSCHLAG xvii
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
xix
xx
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
xxi
xxii
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
xxiii
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
xxiv
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
xxv
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
xxvi
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
xxvii
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
xxviii
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
xxix
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
xxx
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
xxxi
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
xxxii
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
xxxiii
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
xxxiv
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
xxxv
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
xxxvi
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
xxxvii
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
xxxviii
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
xxxix
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
xl
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
xli
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
xlii
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
xliii
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
xliv
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
xlv
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,
xlvi
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
C H A P T E R
O N E
Large-Scale Native Preparation of In Vitro Transcribed RNA Amanda Y. Keel,* Laura E. Easton,† Peter J. Lukavsky,† and Jeffrey S. Kieft*,‡ Contents 4 5 6 6 9 12 14 15 16 16 19 20 24 24
1. Introduction 2. Native Purification of RNA: Affinity Chromatography Method 2.1. Materials 2.2. Preparation of HMM protein 2.3. Preparation of DNA templates by PCR 2.4. Transcription and purification of RNA 3. Native Purification of RNA: Anion-Exchange Chromatography 3.1. Materials 3.2. Cloning of the plasmid DNA template 3.3. Cell culture and plasmid purification 3.4. In vitro RNA transcription 3.5. Weak anion-exchange FPLC Acknowledgments References
Abstract Biophysical studies of RNA require concentrated samples that are chemically and structurally homogeneous. Historically, the most widely used methods for preparing these samples involve in vitro transcription, denaturation of the RNA, purification based on size, and subsequent refolding. These methods are useful but are inherently slow and do not guarantee that the RNA is properly folded. Possible mis-folding is of particular concern with large, complexly folded RNAs. To address these problems, we have developed methods for purifying in vitro transcribed RNAs in their native, folded states. These methods also have the advantage of being rapid and readily scaled to virtually any size RNA or transcription amount. Two methods are presented: the first is an affinity * Department of Biochemistry and Molecular Genetics, University of Colorado Denver, Aurora, Colorado, USA MRC Laboratory of Molecular Biology, Cambridge, United Kingdom { Howard Hughes Medical Institute, University of Colorado Denver, Aurora, Colorado, USA {
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69001-7
#
2009 Elsevier Inc. All rights reserved.
3
4
Amanda Y. Keel et al.
chromatography approach and the second is a weak ion-exchange chromatography approach. Both use equipment and materials readily available to almost any lab and hence should provide flexibility for those seeking alternate approaches to large-scale purification of RNA in the folded state.
1. Introduction Biophysical and structural studies of RNA often require concentrated, highly pure samples of a specific RNA molecule. Techniques such as X-ray crystallography or nuclear magnetic resonance (NMR) spectroscopy often demand a large amount of RNA, and thus obtaining suitable samples is often challenging and time-consuming. In addition, the RNA in these samples must not only be chemically pure, but structurally homogenous. Traditional methods to prepare large amounts of RNA have relied largely on in vitro transcriptions followed by denaturation of the RNA molecule and then purification based on size (Doudna, 1997; Milligan et al., 1987). Although useful, these methods require the RNA being refolded into its native structure before experiments are conducted. For many smaller, less structurally complex RNAs, this is accomplished readily. However, for large and more structurally complex RNAs, refolding into a structurally homogenous population can be difficult (Uhlenbeck, 1995). In addition, denaturing purification methods are inherently slow and time-consuming, and thus preparation of a pure sample is the rate-limiting step for many biophysical and structural studies (Doudna, 2000). In recent years, there has been considerable interest and experimentation toward developing higher throughput, nondenaturing methods for purifying large amounts of RNA for structural and biophysical studies (Batey and Kieft, 2007; Easton and Lukavsky, 2009; Kieft and Batey, 2004; Kim et al., 2007; Lukavsky and Puglisi, 2004; McKenna et al., 2007). Several useful methods have emerged and are now being employed and optimized. In some cases, these methods have resulted in RNA that subsequently has been used to solve high-resolution structures of RNA or RNA–protein complexes, demonstrating their utility (Batey et al., 2004; Lukavsky et al., 2003; Seif and Hallberg, 2009). Here, we present two methods that are used in our labs to purify RNA in a ‘‘native’’ or nondenatured state. Both methods use DNA templates for large-scale in vitro transcriptions, but differ in how the RNA is purified. The first method uses an affinity chromatography approach, the second an ionexchange chromatography approach. Both methods are designed to yield a highly pure, structurally homogeneous, concentrated RNA sample that can be taken directly to biophysical or structural studies. Also, both methods use standard laboratory reagents and equipment and thus should be accessible
5
Native Purification of RNAs
to most. Each method has advantages and disadvantages, and thus we leave it to the reader to assess which is most suitable for their application. We also emphasize that like all methods, it is important to optimize the protocol for each individual lab’s equipment and use.
2. Native Purification of RNA: Affinity Chromatography Method This method relies on the addition of a specific RNA sequence at the 30 -end of the RNA transcript that serves as an affinity tag by binding to a specific protein, which then binds to a column matrix. This tag is eliminated during purification (Fig. 1.1). The advantages of this method are that it is rapid, and many RNAs can be purified in parallel using a set of reusable, commercially available gravity-flow columns. The procedure can also be scaled down to make use of disposable ‘‘spin columns.’’ Furthermore, we have developed a ‘‘cloning-free’’ PCR-based method of generating DNA templates for in vitro transcription, which further increases the throughput of RNA purification. Disadvantages of this method include the requirement that the HMM protein be expressed and purified first. Additional information regarding this method can be found in other references (Batey and Kieft, 2007; Edwards et al., 2009). 1. DNA template T7 promoter RNA X 5¢
glmS ribozyme
MS2
3¢
MS2
3¢
2. RNA from in vitro transcription 5¢ RNA X
glmS ribozyme
3. Affinity immobilization 5¢ RNA X
glmS ribozyme MS2
MS2/ 3' MBP
6XHis NiNTA
4. GlcN6P cleavage and elution 5¢
RNA X
6XHis Ni+ 5¢ glmS ribozyme MS2 MS2/ 3' NTA MBP
5. Column regeneration 5¢ glmS ribozyme MS2 MS2/ 3' MBP
6XHis
NiNTA
Figure 1.1 Affinity purification scheme. ‘‘RNA X’’ denotes the desired target RNA to be purified; ‘‘6 His’’ indicates the hexahistidine tag that is added to the MS2/MBP fusion protein to yield protein HMM. Figure is adapted from Batey and Kieft (2007).
6
Amanda Y. Keel et al.
2.1. Materials LB-kanamycin agar plates LB-kanamycin media IPTG MilliQ water (ddH2O) and RNase free water Sodium phosphate (mono- and dibasic) NaCl MES Tween 20 detergent Imidazole Glycerol Supplies for SDS–PAGE (acrylamide, SDS, etc.) T7 RNA polymerase (6 His-tagged) Magnesium chloride HEPES KCl Ammonium sulfate Magnesium sulfate Triton X-100 Taq and Pfu polymerase 1.25 mM dNTP mix 1 M Tris–HCl, pH 8.1 at 37 C Spermidine Dithiothreitol (DTT) 100 mM stocks of each NTP mix pH 7.5 (Sigma) Inorganic pyrophosphatase (Sigma) Ni-NTA resin (Qiagen) Amicon spin concentrators Glucosamine-6 phosphate (Sigma) (GluN6P) Supplies for denaturing urea polyacrylamide acrylamide, etc.)
gels
(urea,
TBE,
2.2. Preparation of HMM protein The His-tagged MBP-MS2 coat fusion protein (HMM) used in this method was created to allow the protein to be immobilized on both Ni2þ or amylose affinity resins. HMM is a 59-kDa protein containing an N-terminal hexahistidine (6 His) tag, a central maltose-binding protein (MBP) domain, and a C-terminal MS2 coat protein containing the V29/dIFG mutations, which prevent protein multimerization and increase its affinity for RNA (Lim and Peabody, 1994). The protein is expressed in E. coli from plasmid pHMM, which confers kanamycin resistance, and purified using affinity chromatography.
Native Purification of RNAs
7
2.2.1. Buffers and media Bacteria growth media: LB þ 10 mg/ml kanamycin Lysis buffer: 50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 0.5% Tween 20, 10 mM imidazole, 10% glycerol Elution buffer: 50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 0.5% Tween 20, 250 mM imidazole, 10% glycerol Sepharose column dialysis buffer: 25 mM Na–MES, pH 6.0, 25 mM NaCl Sepharose column elution buffers: 25 mM Na–MES, pH 6.0, with either 25 mM or 1 M NaCl Storage buffer: 25 mM Na–HEPES, pH 7.5, 200 mM NaCl, 10% glycerol 2.2.2. Procedures 1. Transform chemically competent Rosetta (BL21)/pLysS cells with the pHMM plasmid and plate on LB þ kanamycin agar plates, then incubate at 37 C overnight (or until colonies appear). Using a picked colony, begin a 50 ml ‘‘starter culture’’ with LB þ 10 mg/ml kanamycin and grow at 37 C with vigorous shaking (220 rpm) overnight. 2. Inoculate 1 l of LB þ 10 mg/ml kanamycin with 5 ml of starter culture and incubate at 37 C with vigorous shaking until the OD600 nm is at 0.6–0.7 AU. It is helpful to check the OD600 nm at least once an hour, and more often as the desired OD600 nm approaches. Once the desired OD600 nm is reached, induce expression by adding IPTG to 0.5 mM. Allow the cells to grow for 3 h at 37 C. 3. Harvest cells by centrifuging at 5000g for 15 min at 4 C. Remove the supernatant and resuspend the cell pellet in 50 ml of lysis buffer þ 500 ml of bacterial protease inhibitor cocktail (Sigma). At this point, the cells can be frozen and stored at 80 C. 4. Lyse the cells using a sonicator at 80% power using 15 s bursts with 45 s intervals between bursts. Repeat this five times, keeping the cells on ice. Transfer the lysate to centrifuge tubes and pellet the cell debris by centrifugation in a JA-20 rotor (Beckman) for 30 min at 35,000g. Immediately transfer the supernatant to a fresh beaker or tube. 5. Apply the supernatant onto a Ni-NTA affinity column (10 ml volume) that has been thoroughly equilibrated with 10 column volumes (CV) of lysis buffer. If desired, the flow-though can be saved and analyzed later by SDS–PAGE to verify protein binding to the column. 6. Wash the column with at least 10 CV of lysis buffer, then elute the protein in 10 ml fractions (1 CV each) of elution buffer. Collect at least 4 fractions, and then analyze each fraction using SDS–PAGE (12% acrylamide) to check for the presence of the protein. In general, the majority of the protein elutes in the first 3 fractions. 7. Pool the fractions containing the HMM protein and dialyze against Sepharose column dialysis buffer overnight at 4 C. Apply the protein
8
Amanda Y. Keel et al.
to a Hi-Prep 16/10 SP-Sepharose column (GE Healthcare) and wash the column with dialysis buffer until the A280 returns to baseline, then elute with a gradient of 0.025–1 M NaCl over a 200-ml volume (the protein elutes around 0.2–0.3 M NaCl). Collect fractions and use 12% SDS–PAGE to identify those containing HMM protein. 8. Pool fractions containing HMM and dialyze exhaustively against storage buffer at 4 C, then divide into 1 ml aliquots and store at 20 C. Determine concentration by measuring A280 and use a molar extinction coefficient of 83,310 M1 cm1. Typical yields (averaged over five individual preparations of the protein on the 2–4 l cell culture size) are 100–120 mg/l culture (Fig. 1.2). Notes and hints:
To avoid repeated transformations of competent cells and the need for a starter culture, a frozen cell stock can be created after step 1. To do this, combine 250 ml of the starter culture with 250 ml of storage media (LB þ kanamycine, 30% glycerol) and store at 80 C. This can be used to inoculate future 1 l cultures. 1
2
3
MW
120 100
*
80 70 60 50 40 30
20 15 10
Figure 1.2 Stained 4–12% SDS–PAGE of expression and purification of HMM protein. Lane 1 is the total soluble fraction from induced and harvested cells, lane 2 is the elution from the Ni-NTA column, and lane 3 is the sample after SP-sepharose purification. An asterisk indicates the HMM protein. Figure is reprinted with permission from Batey and Kieft (2007).
9
Native Purification of RNAs
Because HMM contains an MBP domain, affinity chromatography using amylose resin also could be used as an alternate or additional purification step, although we have not done this.
2.3. Preparation of DNA templates by PCR In vitro transcriptions using T7 RNA polymerase are the standard way to generate large amounts of a specific RNA sequence (Milligan et al., 1987). DNA templates for this reaction can be generated in several ways. We have optimized a method to produce DNA templates by PCR that is used directly in transcription reactions to produce RNA that can be purified by affinity chromatography (Fig. 1.3) (Edwards et al., 2009). The advantage of this approach is that many DNA templates can be generated rapidly and in parallel, without the need for cloning. Our method uses two rounds of PCR comprising three reactions, which in the end yields a DNA template with a T7 promoter that encodes the desired RNA product, the activatable glmS ribozyme, and the MS2 hairpin affinity tag. In a later section, we present an alternate method to produce and use linearized DNA plasmids as templates in transcription reactions. 2.3.1. Reagents and buffers Universal DNA primers: 50 -GEN- 50 -GCGCGCGAATTCTAATACGACTCACTATAG-30 50 -GLMS- 50 -AGCGCCCGAACTACCGGT-30 30 -MS250 -CAGACCCTGATGGTGTCTGAA-30 0 3 -TAG- 50 -ACCGGTACCGGTAGTTCGGGCGCT-30 pRAV23 plasmid Primers A and B, specific to desired RNA product (Fig. 1.3) MW 1a 1b
2
First PCR step: Reaction 1a
Reaction 1b
T7 promoter RNA 5¢-GEN Primer A
glmS 5¢-GLMS
Primer B Second PCR step: (Reaction 2): 5¢-GEN 5¢ T7 promoter
glmS ribozyme
2 ´ MS2 3¢-MS2
3¢-TAG Joint
RNA
3¢-MS2
glmS ribozyme Cleavage site
2 ´ MS2
Figure 1.3 At left is a schematic of the PCR-based method of generating DNA templates for in vitro transcription. The three-reaction method yields the DNA template shown at the bottom. Transcription from this DNA results in the desired RNA product linked to the affinity purification tag. At right is an ethidium bromide stained agarose gel showing representative results from reactions 1a, 1b, and 2. Reprinted with permission from Edwards et al. (2009).
10
Amanda Y. Keel et al.
10 HEPES PCR buffer: 200 mM Na–HEPES, pH 8.8, 100 mM KCl, 100 mM ammonium sulfate, 20 mM magnesium sulfate, 0.1% Triton X-100 2.3.2. Procedures 1. PCR reaction 1a (50 ml): This reaction generates a DNA product that contains the sequence of the desired RNA under control of the T7 promoter, as well as a ‘‘joint’’ on the 30 -end that allows it to be combined with the product of reaction 1b (in reaction 2) to generate the full-length product. To generate this first product, two universal primers are combined with two primers specific for the desired RNA. These two specific primers are designed as partially overlapping DNA sequences that produce the final sequence (Fig. 1.3).
5 ml 10 HEPES PCR buffer 10 ml dNTP mix (1.25 mM each dNTP) 1 ml 50 -GEN primer (100 mM stock) 1 ml 30 -TAG primer (100 mM stock) 1 ml specific primer A (2 mM stock) 1 ml specific primer B (2 mM stock) 1 ml Pfu polymerase 30 ml ddH2O
Thermocycler protocol: Initial denaturing at 94 C for 5 min; 25 cycles of 94 C for 30 s, 55 C for 30 s, 68 C for 1 min; final extension 72 C for 7 min. 2. PCR reaction 1b (50 ml): This reaction generates a DNA product that contains the glmS ribozyme and the MS2 hairpin affinity tag, as well as a ‘‘joint’’ on the 50 -end that allows it to combine with the product of reaction 1a (in reaction 2) to generate full-length product. The reaction uses two universal primers and the pRAV23 plasmid as the template of the reaction. The product of this reaction can be used with any desired RNA sequence, and hence can be stored as a ‘‘stock’’ solution. 5 ml 10 HEPES PCR buffer 10 ml dNTP mix (1.25 mM each dNTP) 1 ml 50 -GLMS primer (100 mM stock) 1 ml 30 -MS2 primer (100 mM stock) 2 ml pRAV23 plasmid DNA, from a standard 5 ml ‘‘miniprep’’ diluted 1:10 1 ml Pfu polymerase 30 ml ddH2O
Thermocycler protocol: Initial denaturing at 94 C for 5 min; 25 cycles of 94 C for 30 s, 55 C for 30 s, 68 C for 1 min; final extension 72 C for 7 min.
Native Purification of RNAs
11
3. PCR reaction 2 (1 ml): In this final reaction, the products from reactions 1a and 1b join via the ‘‘joint’’ sequences, and this joined DNA is amplified by two universal primers, resulting in the complete DNA template for in vitro transcription. In general, this reaction is done in very large scale, with at least 1 ml of reaction volume, in order to generate enough DNA template for a large-scale transcription.
100 ml 10 HEPES PCR buffer 200 ml dNTP mix (1.25 mM each dNTP) 20 ml 50 -GEN primer (100 mM stock) 20 ml 30 -MS2 primer (100 mM stock) 20 ml product from reaction 1a 1 ml product from reaction 1b 20 ml Taq or Pfu polymerase 619 ml ddH2O
Thermocycler protocol: Initial denaturing at 94 C for 5 min; 25 cycles of 94 C for 30 s, 55 C for 30 s, 68 C for 1 min; final extension 72 C for 7 min. After the thermocycling protocol is complete, 5–10 ml of the reaction is analyzed on an agarose gel with an appropriate size marker ladder to verify that a product of the desired length has been produced and to verify that the product is pure (Fig. 1.3). The desired DNA product will be 200 bp larger than the size of the final desired RNA product (i.e., for an RNA product of 100 nt, the size of the DNA template made by this method will be 300 bp). Notes and hints:
It is important that Tris-containing buffers not be used in these reactions, particularly reaction 2. The reason is that Tris will induce slow selfcleavage of the glmS ribozyme (Batey and Kieft, 2007; McCarthy et al., 2005; Roth et al., 2006). Small amounts of Tris introduced into the transcription reaction from the PCR reaction should be avoided and can be alleviated by using HEPES-based buffers. Note that most commercial PCR buffers contain Tris. Although agarose gel purification or commercial ‘‘PCR clean-up’’ kits can be used to process the DNA produced in these PCR reactions, we have found this to be unnecessary in most cases. However, if yields of DNA are very low, or substantial amounts of products other than the desired products are obtained from the reactions, purification of the DNA products may be necessary. Both Taq and Pfu polymerases are suitable for this protocol. In general, Taq gives more consistent and greater yields but can also introduce undetected mutations. Pfu is less error prone, but we find it to be more sensitive to specific DNA sequences and conditions.
12
Amanda Y. Keel et al.
2.4. Transcription and purification of RNA 2.4.1. Reagents and buffers 10 HEPES transcription buffer: 300 mM Na–HEPES, pH 8.0, 100 mM DTT, 20 mM spermidine, 0.1% Triton X-100 100 mM each NTP, pH adjusted to 7.4 with NaOH T7 RNA polymerase (10–15 mg/ml) DNA from PCR reaction 2 of Section 2.4 1 mM MgCl2 RNA column wash buffer: 50 mM K–HEPES, pH 7.5, 150 mM NaCl, 10 mM MgCl2, 10 mM imidazole RNA column regeneration buffer: 50 mM K–HEPES, pH 7.5, 150 mM NaCl, 10 mM MgCl2, 250 mM imidazole 2.4.2. Procedures 1. RNA is made by using the PCR-generated DNA template directly in an in vitro transcription. We generally conduct a transcription reaction of 5 ml final volume using the 1 ml PCR reaction 2 DNA from above, using the following conditions:
1 HEPES transcription buffer 32 mM MgCl2 4 mM each NTP T7 RNA polymerase to a final concentration of 50 mg/ml 1 unit/ml inorganic pyrophosphatase (optional) 1 ml DNA template from PCR reaction 2, above RNase-free water to a final volume of 5 ml
2. The reaction is assembled in 15 ml conical tubes, and then is incubated for 2–3 h at 37 C. If no inorganic pyrophosphatase is added, the solution will turn cloudy over time as pyrophosphate is released. If pyrophosphatase is not used, then at the completion of the reaction, this precipitate must be pelleted in a tabletop clinical centrifuge for 10 min and the supernatant immediately removed. 3. To prepare the transcription reaction for affinity purification, 1.6 mg of HMM protein is added directly to the reaction and allowed to incubate for 10 min on ice to allow the protein to bind to the MS2 hairpin affinity tag. This reaction is applied to a gravity flow column containing 1 ml Ni-NTA resin (QIAGEN) per 2 ml of transcription reaction at room temperature, and the solution passed through with a slow drip. To facilitate complete binding of the RNA/protein complex to the column, the flow-through can be passed through a second time. 4. Wash the column four times, each wash containing 4 CV of RNA column wash buffer, to remove excess protein, nucleotides and RNA abortive transcription products.
13
Native Purification of RNAs
5. To elute the product RNA from the column, add 1 CV of RNA column wash buffer þ 1 mM glucosamine-6-phosphate (GlcN6P) to the column and allow this to pass through, then close the column valve and allow the column to sit for 10 min at room temperature. Apply a second CV of column wash buffer þ 1 mM GlcN6P, open the column valve, and collect this elution fraction. Collect two subsequent 1 CV elutions. Usually, most of the RNA comes out in the first two elution fractions (Fig. 1.4). 6. The HMM 30 -tag complex is removed from the column using regeneration buffer to regenerate the Ni-NTA resin. Add 5 CV and allow column to sit for 10 min. Drain column and repeat with another 5 CV of this buffer. Rinse the column with 4 CV RNA column wash buffer to prepare for another use. A
Tx FT W1 W2 W3 E1 E2
S a b c
B
Tx FT W1 W2 W3 E1 E2 E3 S
a b
c
Figure 1.4 Results from the affinity purification method. (A) Purification of a 94-nt RNA from a 100-ml transcription reaction, using a QIAGEN Ni-NTA spin column. Lane ‘‘Tx’’: the raw transcription; lane ‘‘FT’’: column flow-through; lanes ‘‘W1’’– ‘‘W3’’: column washes; lanes ‘‘E1’’ and ‘‘E2’’: column elutions; lane ‘‘s’’: imidazole regeneration. Bands ‘‘a’’, ‘‘b’’, and ‘‘c’’ indicate the full-length transcript, 30 -tag, and product RNA, respectively. (B) Purification of the same RNA as panel (A), but from a larger-scale (3.25 ml) reaction using 3 ml of Ni-NTA resin in a gravity flow column. Labeled identically to (A), but with a third elution step. Figure is reprinted with permission from Batey and Kieft (2007).
14
Amanda Y. Keel et al.
7. To determine which elution fractions contain RNA, a small amount of each fraction (20 ml) can be analyzed by denaturing PAGE and staining with ethidium bromide. Combine RNA-containing fractions and then concentrate the RNA and exchange into the desired buffer using Amicon spin concentrators of the appropriate molecular weight. In general, we concentrate to >10 mg/ml in RNase-free water and store the RNA at 20 C. Notes and hints:
It is highly recommended (especially the first few times this method is used) that fractions be saved from each step of the purification procedure and small aliquots of these fractions analyzed by denaturing PAGE. Problems with the quality of the HMM protein, the transcription yields, binding to the column, etc. can be detected and appropriate troubleshooting conducted. If binding of the RNA to the column appears to be weak, the above procedure can be adjusted. Transcription and addition of HMM protein are done as described above, but then the solution is passed over the NiNTA resin at 4 C and allowed to sit for 5–10 min; to ensure complete binding of the RNA/protein complex to the column, the flow-through can be passed through a second time. In the cold room, the column is washed four times with 4 CV of cold RNA column buffer. To cleave the product RNA from the tag and elute from the column, move the columns to room temperature then add 2 CV of room temperature RNA column buffer þ 1 mM GlcN6P. Seal the column well and place it on slushy ice for 10 min. Then remove the column from the ice and open it to collect the eluate (repeat). Two subsequent elutions with cold buffer were allowed to flow through the column to remove further RNA. Always make sure to thoroughly clean the columns after each use, and if columns are not to be used in the next few days, store them in 20–30% ethanol at 4 C. It is best to keep several Ni-NTA columns that are dedicated to RNA purification. Using these columns for protein purification can introduce many foreign proteins to the columns, potentially including RNases.
3. Native Purification of RNA: Anion-Exchange Chromatography Two other powerful methods for native RNA purification, one based on size-exclusion chromatography and the other using affinity purification, have been published in the past. The former employs separation of the different species in a transcription reaction (NTPs, small abortive transcripts, the transcribed RNA, and the plasmid DNA) based on size rather than charge (Kim et al., 2007; Lukavsky and Puglisi, 2004; McKenna et al., 2007).
Native Purification of RNAs
15
Similar to our protocol based on anion-exchange chromatography, this method allows efficient recycling of unincorporated rNTPs and separation of mono- and multimeric RNA species, but still requires removal of the T7 polymerase from the reaction mixture using tedious phenol/chloroform extraction and desalting of the sample prior to chromatography, which takes about 1–2 h. In our protocol, the T7 polymerase is simply separated from the RNA during chromatography, which is more convenient and saves time. In the latter protocol, the desired RNA transcript is fused to an activatable ribozyme and an affinity tag for purification (Batey and Kieft, 2007; Kieft and Batey, 2004). This elegant method, which is described in detail in the first protocol, allows production of milligram amounts of native RNA with homogenous 30 -ends, which can be crucial for crystallographic applications. In contrast to size-exclusion or weak anion-exchange chromatography, no separation of mono- and oligomeric species generated during transcription is achieved using this protocol, but this can be easily achieved using either method as a subsequent purification step. For NMR spectroscopic application though, this method is less desirable, since a significant amount (>100 nt) of the transcript comprises the ribozyme and affinity tag resulting in a significantly lower final RNA yield, which in turn increases the cost of isotopically labeled RNA oligonucleotides. The next sections describe a detailed protocol for RNA preparation by in vitro transcription from linearized DNA plasmids and RNA purification using weak anion-exchange chromatography.
3.1. Materials PCR supermix (Invitrogen) or similar PCR reagents pUC18 plasmid Restriction enzymes—HindIII, BbsI, and EcoRI T4 DNA ligase DH5a cells TYE-ampicillin agar plates 2 TY-ampicillin or -carbenicillin media QIAfilter plasmid MEGA or GIGA kit 70% ethanol RNase free water 4.9 M magnesium chloride (Sigma) 25 mM NTP mix (Sigma) T7 RNA polymerase (6 His-tagged) 1 M Tris–HCl, pH 8.1 at 37 C 100 mM spermidine 500 mM dithiothreitol 10% Triton X-100 Inorganic pyrophosphatase (Sigma)
16
Amanda Y. Keel et al.
8% acrylamide 8 M urea gels 0.1% toluidine blue solution 500 mM EDTA, pH 8.1 at 37 C
3.2. Cloning of the plasmid DNA template In order to achieve a high yield of the plasmid DNA template coding for the RNA oligonucleotide of interest, a high copy number vector such as pUC18 should be used. The DNA construct should be engineered with a 50 HindIII restriction site preceded by a 4-nt overhang to allow efficient cleavage by the enzyme, the T7 promoter sequence (TAATACGACTCACTATA), the coding sequence of the desired RNA followed by two spacer nucleotides (usually TT), and finally BbsI and EcoRI restriction sites again with a 4-nt overhang for efficient enzyme cleavage as described (Lukavsky and Puglisi, 2004). The 2-nt linker between the RNA coding sequence and the BbsI site is necessary for cleavage 2 and 6 nucleotides upstream of the restriction site, thereby cutting directly at the last nucleotide position of the DNA template coding for the desired RNA sequence, which is necessary to perform ‘‘runoff’’ transcriptions. If there is an internal BbsI restriction site in the DNA template sequence, the BsaI restriction site can be used as an alternative, cutting 1 and 5 nucleotides upstream and therefore requiring only a 1-nt linker between DNA template sequence and restriction site. The DNA construct can be prepared using standard polymerase chain reaction (PCR) methods and for longer DNA constructs overlapping primers can be used as described elsewhere (Lukavsky and Puglisi, 2004). The resulting DNA fragment is then digested with HindIII and EcoRI, ligated into the pUC18 vector digested with the same enzymes, and the plasmid is transformed into DH5a cells, grown at 37 C and spread on a TYE-ampicillin agar plate. The plasmids from any resulting colonies should be sequenced using a primer further upstream of the T7 promoter (PLUHIII ¼ CTTCGCTATTACGCCAG) to be able to confirm the correct T7 promoter sequence, before beginning the preparation of milligram quantities of the plasmid required for in vitro RNA transcription.
3.3. Cell culture and plasmid purification 1. Streak a TYE-ampicillin agar plate with cells containing the correct plasmid DNA construct and grow at 37 C overnight. Inoculate 2 5 ml 2TYampicillin or -carbenicillin (50 mg/l) media with single colonies and grow in an orbital incubator at 37 C to 0.5 OD600. Inoculate 2 1 l 2TYampicillin or carbenicillin with 1 ml of the 0.5 OD600 culture and grow for 16–18 h in an orbital incubator at 37 C, harvest the cells by centrifugation and either store at 20 C immediately or begin the plasmid extraction and purification.
Native Purification of RNAs
17
When growing the cells for the plasmid preparation, it is advised not to over-grow the culture as this leads to lower plasmid yield due to cell lysis which can be observed in the cell pellet as dark veins. The plasmid can be extracted and purified from the harvested cells using the following protocol based on QIAfilter plasmid MEGA and GIGA protocols. The volumes of buffers used should always correspond to the GIGA protocol. However, the lysate can also be loaded onto a QIAGEN-tip 2500 intended for the MEGA protocol, which is sufficient to purify up to 6 mg of pure plasmid DNA at a cheaper price than the GIGA QIAGEN-tip 10000. The washing procedures should be performed according to the GIGA protocol (300–400 ml buffer QC), but elution and precipitation volumes again correspond to the MEGA protocol. All procedures should be performed at room temperature, which should not exceed 25 C. 2. Add 125 ml chilled buffer P1 to the cells and gently resuspend using a 25 ml pipette to remove the cells from the wall of the tube and to remove any lumps of cells. Lyse the cells by adding 125 ml buffer P2, gently mix by inverting four to six times and incubate at room temperature for 5 min. Add 125 ml chilled buffer P3 and mix well by inverting four to six times to neutralize buffer P2 and stop lysis. Pour the lysate into a MEGA–GIGA filter cartridge and leave to settle for 5–10 min. 3. During this period equilibrate a QIAGEN-tip 2500 with 35 ml buffer QBT by gravity flow. 4. Filter the solution to clear the cell lysate from the cell debris. When the cell debris reaches the filter, switch off the vacuum, add 50 ml wash buffer FWB, and gently stir into the cell debris using a plastic pipette, then continue to filter until the cell debris strains against the filter, usually 350–400 ml. 5. Apply the lysate to the QIAGEN-tip 2500 by gravity flow, then wash with 6–8 QIAGEN-tip 2500 volumes of buffer QC. Place the QIAGEN-tip 2500 on a clean centrifuge tube and elute the DNA by adding 35 ml buffer QF. Once all the buffer has passed add 24.5 ml room temperature isopropanol to precipitate the DNA. 6. Centrifuge at 15,000g, 4 C for 15 min. Set a low brake so that the DNA pellet remains attached to the tube wall after the spin. Carefully decant the supernatant and wash the pellet with 7 ml 70% ethanol before centrifuging as before for 10 min, decant the supernatant and repeat the 70% ethanol wash. This step is especially important, as washing the DNA pellet twice with 70% ethanol reduces the salt content of the final plasmid DNA, which could interfere with the restriction enzyme digestion for plasmid linearization. 7. Decant the supernatant, pipette off any excess ethanol and air dry at room temperature for 1 h or at 4 C overnight.
18
Amanda Y. Keel et al.
The DNA pellet dries colorless, which might be difficult to see when trying to resuspend, therefore, it is useful to mark on the outside of the tube the position of the DNA pellet while it is still visible. 8. Add 2 ml RNase-free water to the pellet and allow softening of the DNA pellet at room temperature for 10 min before resuspending, then transfer into a 15-ml tube and rinse the centrifuge tube with another 1 ml of RNase-free water and add to the 2 ml resuspension. 9. Measure the concentration and dilute to 700 mg/ml by adjusting the volume with appropriate volumes of the recommended 10 BbsI buffer and RNase-free water. Linearize the plasmid by incubating with 50 U/ml BbsI at 37 C overnight. 10. Analyze the digestion by loading 10 ml of 7 mg/ml undigested and digested samples on a 1% agarose gel. The linearized plasmid now can be used directly to perform in vitro RNA transcription. Notes and hints:
The buffers (except P1 and P3) and QIAGEN-tips 2500 required for the large-scale plasmid purification should be stored at room temperature (15–25 C). Temperatures exceeding 25 C lead to cell debris passing through the filter which leads to cloudy lysates potentially congesting the QIAGEN-tip 2500. Lower storage temperature causes SDS precipitation in buffer P2, but this can be rectified by warming the buffer or by preparing fresh buffer according to the manufacturer’s instructions. Storing buffers for a long period of time can also lead to salt precipitation in buffers QC and QF and fresh buffers should be prepared according to manufacturer’s instructions. We also noticed that plasmids prepared with the QIAGEN MEGA kit at temperatures exceeding 25 C are poorly linearized and often partially degrade during the enzyme digest, probably due to higher salt concentration and nuclease contamination. It is also important to add the isopropanol to the completely eluted sample rather than allowing the sample to elute into the isopropanol, since the initially very high isopropanol concentration once again can lead to increased salt precipitation interfering with the subsequent plasmid linearization. Problems with plasmid linearization can be caused by several additional factors. The plasmid yield might be higher than measured or the activity of BbsI might be lower than anticipated and therefore the plasmid remains partially undigested after the overnight incubation period at 37 C. This can be resolved easily by addition of more restriction enzyme and continued incubation for a few hours and also by storage of BbsI at 80 C until needed. Another problem might be that the salt concentration of the plasmid DNA solution is too high, because the isopropanol used for precipitation was not at room temperature or both 70% ethanol
Native Purification of RNAs
19
washes were not performed. If problems with incomplete and slow digestion occur repeatedly, small-scale trial digestions should be performed to determine the optimal salt concentration for the large-scale cleavage of the remaining plasmid.
3.4. In vitro RNA transcription It is desirable to maximize the RNA yield for each sample and this can be achieved by determining the optimal magnesium chloride concentration required for in vitro transcription. A series of 13 small-scale (25 ml) reactions with a range of magnesium chloride concentrations from 4 to 52 mM, increasing by 4 mM increments each, are typically used for magnesium chloride optimization. 1. Prepare the series of reactions containing 4 mM each NTP, 70 mg/ml linearized plasmid, 1200 U/ml T7 RNA polymerase, 40 mM Tris–HCl (pH 8.1 at 37 C), 1 mM spermidine, 5 mM dithiothreitol, 0.1% Triton X-100, 1 U/ml inorganic pyrophosphatase and 4–52 mM magnesium chloride and incubate in a 37 C water bath for 1 h. 2. Analyze the yield by loading 2 ml of each reaction on a denaturing PAGE (8% acrylamide, 8 M urea) and visualizing by UV shadowing or staining with 0.1% toluidine blue solution. A large-scale reaction to produce milligram quantities of RNA can now be performed in a 20 ml reaction using the optimal magnesium chloride concentration. 3. Prepare the reaction using the same conditions as the small-scale reactions and incubate in a 37 C water bath for 2–4 h before comparing the yield against the optimal small-scale reaction by loading 2 ml of each on a denaturing PAGE. 4. Stop the reaction by the adding 0.5 M EDTA (pH 8.1 at 37 C) to a final concentration of 50 mM. The transcription reaction can be stored frozen at 20 C or directly purified. Notes and hints:
Typically we use NTPs from Sigma, and most RNA transcriptions give the best yield in the range of 20–25 mM final magnesium chloride concentration. NTPs from other sources might require different concentrations and isotopically labeled NTPs prepared in-house following published methods (Batey et al., 1992) are usually obtained as magnesium salts and therefore much lower magnesium chloride concentrations are required (8–12 mM ) for optimal transcription yield. The inorganic pyrophosphate (PPi) which builds up during the transcription traps magnesium at a molar ratio of 2:1 of Mg:PPi and thereby
20
Amanda Y. Keel et al.
1
2
3
4
5
6
DNA RNA
NTPs
Figure 1.5 Addition of pyrophosphatase improves transcription yield. Denaturing PAGE analysis of large scale transcriptions in the absence (lanes 1–3) or presence (lanes 4–6) of 1 U/ml pyrophosphatase. The yield is analyzed by loading 2 ml of the transcription reaction after 45 (lanes 1 and 4) and 90 min (lanes 2, 3, 5, 6) without or with addition of magnesium after 45 min (compare lanes 2 and 3 with 5 and 6). RNA bands are visualized by staining with 0.1% toluidine blue. The best yield is obtained in the presence of pyrophosphatase without addition of magnesium (lane 5). Reprinted with permission from Easton and Lukavsky (2009).
inhibits the rate of transcription (Kern and Davis, 1997). In the past, we therefore added additional magnesium to the reaction after 45 min, which slightly improved the final yield (Fig. 1.5, compare lanes 3 and 5). Lanes 2 and 3 should be compared for +/- additional MgCl effect without inorganic pyrophosphatase. The better choice though is to perform the transcription in the presence of inorganic pyrophosphatase (1U/ml) without further addition of magnesium, which improves the yield by about 20–30% (Fig. 1.5, compare lanes 4 and 6). Prolonged incubation times (up to 4 h) of the in vitro RNA transcription reactions can help to improve the yield. However, the likelihood of hydrolysis will also increase.
3.5. Weak anion-exchange FPLC For fast and simple purification of RNA oligonucleotides from crude transcription reactions, we use an AKTA prime FPLC system equipped with a 50-ml superloop and three 5 ml HiTrap diethylaminoethyl (DEAE) sepharose FastFlow columns (GE Healthcare) connected in series. The DEAE columns are equilibrated with 3 CV of buffer A (50 mM sodium phosphate, pH 6.5, 150 mM sodium chloride, and 0.2 mM EDTA) at room temperature. Buffer B contains the same components with 2 M sodium chloride. Both buffers can be prepared in large quantities, sterile filtered and stored at 4 C (buffer A) or room temperature (buffer B) to avoid precipitation of sodium chloride.
21
Native Purification of RNAs
1. Load the stopped transcription reaction into the 50 ml superloop and perform weak anion-exchange chromatography using the following gradient collecting 10 ml fractions: 0–70 ml 70–100 ml 100–380 ml 380–410 ml 410–455 ml 455–485 ml
0% B to 10% B to 30% B to 100% B 100% B to 0% B
1 ml/min 2 ml/min 2 ml/min 4 ml/min 4 ml/min 4 ml/min
2. The fractions are analyzed by denaturing PAGE (8% acrylamide, 8 M urea) loading 5 ml of each fraction. Unincorporated NTPs and the T7 polymerase usually elute in fractions 3–7 and small abortive oligonucleotides in fractions 8–12 (Fig. 1.3). RNA oligonucleotides elute depending on the overall phosphate charge per molecule starting with fractions 15–16 (30 nt, 400 mM NaCl) up to fraction 28 (500 nt, 570 mM NaCl), while the plasmid DNA template elutes later over several fractions (>630–700 mM NaCl). Therefore, a very shallow gradient is required especially for larger RNAs. 3. Pool the fractions corresponding to the clean RNA oligonucleotide and concentrate using 15 ml Centriprep centrifugal devices with 10 kDa MWCO-cutoff for RNAs larger than 30 nt or 3 kDa cutoff for smaller RNA oligonucleotides. Concentrate the RNA samples to 1 ml and then equilibrate them into the appropriate buffer using three to five consecutive 15 ml buffer exchanges. Concentrate the RNA to 1 ml at each step to maximize the efficiency of the buffer exchange (two spins) and then thoroughly mix with 14 ml fresh buffer for another round of centrifugation. The final RNA sample is best stored frozen at 20 C. Purification of RNA using weak anion-exchange chromatography relies on the difference in the overall phosphate charge per molecule. Therefore, unincorporated NTPs and small abortive transcripts bind only weakly to the DEAE matrix while RNA binds with medium affinity depending on the size and the plasmid DNA binds strongly and elutes as a broad peak towards the end of the shallow gradient (Fig. 1.6). Using this method, we routinely purify RNA samples ranging from 30 to 500 nt in length. For small RNAs (30–40 nt), we sometimes observe slight contamination with abortive transcripts, but they are removed easily during the buffer exchanges with Centriprep centrifugal devices. When purifying RNA samples larger than 500 nt, care has to be taken not to pool fractions contaminated with the linearized plasmid DNA and therefore only the utmost peak fractions (3–5 fractions) are pooled avoiding the long tail fractions, which are more likely to contain the linearized plasmid. In all cases, the RNA is
22
Amanda Y. Keel et al.
A A260 (O.D./ml) and fraction buffer B (%)
100 80 60
C
Small aborts
20
DNA
RNA
3
4
5
5
10
98 RNA
DNA
15 20 25 30 35 Fraction number
40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
B
2
188
40
0
MW 1
NTPs
45
62 49 38 28 17 14 6 3
Figure 1.6 RNA purification using weak anion-exchange FPLC. (A) Elution profile of a 48-nt RNA obtained from a 20-ml in vitro transcription reaction using DEAE sepharose chromatography. Unincorporated NTPs, small abortive transcripts, and the plasmid DNA are well separated from the desired RNA product. (B) Denaturing PAGE analysis loading 5 ml of the eluted fractions. RNA and DNA bands are visualized by staining with 0.1% toluidine blue. The crude transcription reaction is shown in lane 1, NTPs and small abortive transcripts are loaded in lanes 2–8, the eluted RNA fractions are in lanes 9–12, and the eluted plasmid DNA is in lanes 13–15 visible as a very faint band in lanes 14 and 15. (C) Denaturing SDS–PAGE analysis of the eluted fractions shows that the purified RNA is free of T7 RNA polymerase. Resuspended pellets from TCA-precipitation of 1 ml of the crude transcription reaction (lane 2), the pooled flow-through (lane 3), abortive transcripts (lane 4), and the pooled RNA fractions (lane 5) are loaded. T7 RNA polymerase (lane 1) and a molecular weight marker are loaded as references and the molecular weight is indicated on the left. T7 RNA polymerase bands are visualized by coomassie-staining. Reprinted with permission from Easton and Lukavsky (2009).
free of the T7 RNA polymerase, which does not bind to DEAE sepharose matrix at salt concentrations higher than 100 mM (Fig. 1.6). This method also helps to reduce the preparation cost of isotopically labeled RNA samples for NMR spectroscopic studies, since the expensive unincorporated NTPs can be pooled, lyophilized, and desalted using boronate-affinity chromatography as described (Batey et al., 1992). An additional benefit of this purification method is that different RNA species can be separated, if the charge per molecule is significantly different, as for instance between mono- and multimeric RNA species. During purification of a 74-nt RNA, four-way junction from the
A260 (O.D./ml) and fraction buffer B (%)
A
B
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
DNA
80 60 40
RNA
NTPs Small aborts
20
RNA DNA
0
5
10
15
20 25 30 35 Fraction number
40
45
Figure 1.7 Separation of monomeric from oligomeric RNA species using weak anion-exchange FPLC. (A) Elution profile of an 80-nt RNA obtained from a 20-ml in vitro transcription reaction using DEAE sepharose chromatography. The RNA product elutes in two distinct peaks at lower and higher NaCl concentration. (B) Denaturing PAGE analysis loading 5 ml of the eluted fractions. RNA bands are visualized by staining with 0.1% toluidine blue. The crude transcription reaction is shown in lane 1, NTPs and small abortive transcripts in lanes 2–4, and the eluted RNA fractions corresponding to fractions 20–30 from Fig. 1.7A are loaded in lanes 5–15. While all the fractions contain the clean RNA product, only fractions 20–22 contain the monomeric RNA species (lanes 5–7). Other fractions (lanes 8–15) represent mixtures of mono-, di-, and tetrameric RNA species as judged by gel filtration (data not shown). Reprinted with permission from Easton and Lukavsky (2009).
24
Amanda Y. Keel et al.
classical swine fever virus (Pestova et al., 1998), we observed that the RNA eluted in two distinct peaks along the shallow salt gradient (Fig. 1.7). Analysis of the individual fractions by gel filtration revealed that the main peak contains the monomeric RNA species while the fractions eluted at higher salt concentration represent di- and tetrameric species, which could be separated efficiently from the desired monomer based on their overall charge difference (Fig. 1.7).
ACKNOWLEDGMENTS Development of native purification methods in the Kieft Lab has been supported by a grant from the Butcher Foundation, a grant from the University of Colorado Technology Transfer Office, and NIH grant R03 AI072187. This work in the Lukavsky Lab has been supported by MRC and a HFSP grant RGP0024/2008-C. JSK is a Howard Hughes Medicinal Institute Early Career Scientist.
REFERENCES Batey, R. T., and Kieft, J. S. (2007). Improved native affinity purification of RNA. RNA 13, 1384–1389. Batey, R. T., Inada, M., Kujawinski, E., Puglisi, J. D., and Williamson, J. R. (1992). Preparation of isotopically labeled ribonucleotides for multidimensional NMR spectroscopy of RNA. Nucleic Acids Res. 20, 4515–4523. Batey, R. T., Gilbert, S. D., and Montange, R. K. (2004). Structure of a natural guanineresponsive riboswitch complexed with the metabolite hypoxanthine. Nature 432, 411–415. Doudna, J. A. (1997). Preparation of homogeneous ribozyme RNA for crystallization. Methods Mol. Biol. 74, 365–370. Doudna, J. A. (2000). Structural genomics of RNA. Nat. Struct. Biol. 7(Suppl.), 954–956. Easton, L. E., and Lukavsky, P. J. (2009). Rapid, nondenaturing purification of RNA using weak anion-exchange FPLC. RNA (in press). Edwards, A. L., Garst, A. D., and Batey, R. T. (2009). Determining structures of RNA aptamers and riboswitches by X-ray crystallography. Methods Mol. Biol. 535, 135–163. Kern, J. A., and Davis, R. H. (1997). Application of solution equilibrium analysis to in vitro RNA transcription. Biotechnol. Prog. 13, 747–756. Kieft, J. S., and Batey, R. T. (2004). A general method for rapid and nondenaturing purification of RNAs. RNA 10, 988–995. Kim, I., McKenna, S. A., Viani Puglisi, E., and Puglisi, J. D. (2007). Rapid purification of RNAs using fast performance liquid chromatography (FPLC). RNA 13, 289–294. Lim, F., and Peabody, D. S. (1994). Mutations that increase the affinity of a translational repressor for RNA. Nucleic Acids Res. 22, 3748–3752. Lukavsky, P. J., and Puglisi, J. D. (2004). Large-scale preparation and purification of polyacrylamide-free RNA oligonucleotides. RNA 10, 889–893. Lukavsky, P. J., Kim, I., Otto, G. A., and Puglisi, J. D. (2003). Structure of HCV IRES domain II determined by NMR. Nat. Struct. Biol. 10, 1033–1038. McCarthy, T. J., Plog, M. A., Floy, S. A., Jansen, J. A., Soukup, J. K., and Soukup, G. A. (2005). Ligand requirements for glmS ribozyme self-cleavage. Chem. Biol. 12, 1221–1226.
Native Purification of RNAs
25
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. Pestova, T. V., Shatsky, I. N., Fletcher, S. P., Jackson, R. J., and Hellen, C. U. (1998). A prokaryotic-like mode of cytoplasmic eukaryotic ribosome binding to the initiation codon during internal translation initiation of hepatitis C and classical swine fever virus RNAs. Genes Dev. 12, 67–83. Roth, A., Nahvi, A., Lee, M., Jona, I., and Breaker, R. R. (2006). Characteristics of the glmS ribozyme suggest only structural roles for divalent metal ions. RNA 12, 607–619. Seif, E., and Hallberg, B. M. (2009). RNA-protein mutually induced fit: structure of Escherichia coli isopentenyl-tRNA transferase in complex with tRNA(Phe). J. Biol. Chem. 284, 6600–6604. Uhlenbeck, O. C. (1995). Keeping RNA happy. RNA 1, 4–6.
C H A P T E R
T W O
Assembly of Complex RNAs by Splinted Ligation Benjamin M. Akiyama*,† and Michael D. Stone†,‡ Contents 1. Introduction 2. General Considerations for Splinted RNA Ligation 3. Preparation of Unmodified RNA Ligation Precursor Molecules 3.1. In vitro transcription of telomerase RNA 3.2. Targeted RNase H cleavage of telomerase RNA 4. Preparation of Modified (Dye Labeled) RNA Ligation Precursor Molecules 4.1. Protocol 3: Dye-labeling and HPLC purification of synthetic RNA oligonucleotides 5. RNA Ligation Methods 5.1. Protocol 4: Splinted RNA ligation method for producing FRET-labeled telomerase RNA 6. Application: Single-Molecule FRET Measurements Acknowledgments References
28 29 31 31 33 37 39 40 42 44 45 45
Abstract Mechanistic studies of RNA enzymes (ribozymes) and ribonucleoprotein (RNP) complexes such as the ribosome and telomerase, often seek to characterize RNA structural features, either dynamic or static, and relate these properties to specific catalytic functions. Many experimental techniques that probe RNA structure–function relationships rely upon site-specific incorporation of chemically modified ribonucleotides into the RNA of interest, often in the form of chemical cross-linkers to probe for sites of protein–RNA interaction or small organic fluorophores to measure dynamic structural properties of RNAs. The ability to arbitrarily modify any RNA molecule has been greatly enabled by modern RNA synthesis techniques; however, there remains a practical size * { {
Department of Molecular, Cell, and Developmental Biology, University of California, Santa Cruz, California, USA Center for Molecular Biology of RNA, University of California, Santa Cruz, California, USA Department of Chemistry and Biochemistry, University of California, Santa Cruz, California, USA
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69002-9
#
2009 Elsevier Inc. All rights reserved.
27
28
Benjamin M. Akiyama and Michael D. Stone
limitation ( 70 bases). Consequently, experimental approaches involving specific chemical modifications of larger RNAs require the use of RNA ligation methods. The aim of this chapter is to describe a general approach for covalently joining multiple site-specifically modified RNA fragments, drawing from our fluorescence-based structural studies of telomerase RNA as an example.
1. Introduction The essential role of RNA as a cellular messenger of genetic information has been well appreciated since the early 1960s. However, recent discoveries in the fields of RNA structure, RNA-mediated catalysis, and ribonucleoproteins (RNP) have demonstrated a far more diverse set of physiological functions for RNA. For example, catalytic RNAs termed ribozymes, such as the group I self splicing intron from Tetrahymena, can speed up the rate of chemical reactions by many orders of magnitude (Cech, 1990). Functional contributions of RNA within the context of large multisubunit ribonucleoprotein complexes have also been demonstrated. One striking example comes from the recent tour de force determination of the high-resolution structure of the ribosome, which verified the central role of RNA during protein synthesis (Cate et al., 1999). Since the first demonstration of the catalytic properties of RNA, intense effort has been focused on the development of methods which probespecific structural characteristics of biological RNAs. Researchers have long utilized direct UV cross-linking methods to study RNA–protein interactions (Pinol-Roma et al., 1989); however, this approach can often be extremely inefficient. Advances in RNA synthesis techniques have opened new avenues for RNA research by facilitating the construction of RNA substrates possessing site-specific chemical modifications that may be used to probe the structure and conformational dynamics of RNA molecules (Scaringe et al., 1998). Some of the more common chemical modification strategies enlisted in structural studies of RNA include: (1) directed cleavage of RNA by site-specifically tethered EDTA–Fe (Han and Dervan, 1994); (2) the use of specific thiol moieties to form disulfide cross-links under nonreducing conditions (Cohen and Cech, 2001); (3) the incorporation of 4-thiouridine as an intrinsically UV-activated cross-linker (Yu, 2000); and (4) labeling with small organic fluorophores which provide a means to study dynamic structural properties of RNA (Hengesbach et al., 2008). Despite the obvious utility of using synthetic RNA in structural studies, there exists a practical size limit (70 bases) for RNA generated by routine chemical synthesis. Many biological RNA molecules of interest are much larger than this, leading to a general demand for robust methodologies that produce large RNA molecules harboring site-specific modifications.
Assembly of Complex RNAs by Splinted Ligation
29
Here, we describe an optimized method for generating site-specifically modified RNAs of arbitrary length. This method builds upon previously described RNA ligation techniques (Moore and Query, 2000), and includes several modifications to existing protocols that have substantially improved ligation efficiencies in our hands. The protocol described herein should provide experimentalists with large quantities of modified RNA substrates for a wide variety of applications. We demonstrate the utility of the method with a description of the design and fabrication of a 177-nucleotide multiply dye-labeled RNA construct derived from the Tetrahymena thermophila telomerase RNA. Our protocol for constructing telomerase RNA for fluorescence analysis requires a three part splinted RNA ligation reaction to covalently couple 50 - and 30 -terminal dye-labeled fragments onto an in vitro transcribed RNA insert fragment (Stone et al., 2007). In Section 6, we briefly present representative data from single molecule Fo¨rster resonance energy transfer (smFRET) experiments conducted on dye-labeled RNA molecules prepared by splinted RNA ligation.
2. General Considerations for Splinted RNA Ligation RNA ligation approaches typically employ one of several naturally occurring RNA ligase enzymes to catalyze the ATP-dependent joining of a ‘‘donor’’ fragment possessing a 50 -monophosphate and an ‘‘acceptor’’ fragment terminating in a 30 -hydroxyl group. All ligase enzymes produce products with a naturally occurring 30 –50 -phosphodiester bond at the ligation site. T4 RNA ligase was identified early on as a candidate enzyme for producing oligoribonucleotides of defined sequence, due to its activity on a wide variety of RNA substrates (England and Uhlenbeck, 1978; Walker et al., 1975). However, the strong preference of T4 RNA ligase for single-stranded substrates can produce a variety of unwanted side-products such as RNA circles and homodimers (see Section 5) (Romaniuk and Uhlenbeck, 1983). In contrast, splinted RNA ligation techniques have a fundamental advantage over T4 RNA ligase-catalyzed joining reactions, due to the dependence upon a properly annealed preligation complex, and therefore provide a powerful quality control step which biases the reaction products toward the desired RNA sequence and topology (Moore and Query, 2000; Moore and Sharp, 1992, 1993). For this reason, splinted RNA ligation has become the method of choice for generating large complex RNA molecules. The assembly of complex RNA molecules, harboring site-specific modifications, is a multistep procedure that can be subdivided into three stages (Fig. 2.1): (i) preparation of unmodified RNA ligation precursor fragments, (ii) preparation of modified RNA ligation precursor fragments, and
30
Benjamin M. Akiyama and Michael D. Stone
Preparation of unmodified RNA ligation precursor fragments Clone telomerase RNA gene behind T7 RNA polymerase promoter
PCR amplify telomerase RNA gene with T7 RNA polymerase promoter
In vitro transcribe telomerase RNA using PCR amplicons as template
Preparation of modified RNA ligation precursor fragments Chemically synthesize RNA with 5-N-U modification at desired labeling site
Dye-label synthetic RNA fragments
Purify telomerase RNA transcripts with desalting column
Deprotect dye-labeled RNA fragments
Targeted RNase H digestion
PAGE purify dye-labeled RNA fragments
PAGE purify RNase H digestion products
HPLC purify dye-labeled RNA
RNA ligation Splinted RNA ligation of unmodified and dye-labeled RNA fragments
PAGE purify RNA ligation products
Verify by diagnostic PAGE and quantify by UV–vis spectrometry
Figure 2.1 General procedure for preparation of dye-labeled Telomerase RNA.
(iii) splinted RNA ligation and product purification. Efficient splinted RNA ligation requires that each precursor fragment possess the desired sequence and terminal functional groups. Therefore, purification procedures and
Assembly of Complex RNAs by Splinted Ligation
31
diagnostic analysis should be performed at each stage of the procedure. Of course, one pays a price for purity by taking losses at each purification step. Therefore, to ensure sufficient quantities of final product, each reaction step should be scaled up enough to account for the expected losses.
3. Preparation of Unmodified RNA Ligation Precursor Molecules There are two preferred methods for producing unmodified RNA ligation precursor fragments: chemical synthesis and in vitro transcription. For shorter RNA molecules, wherein ligation precursor fragments do not exceed 60–70 ribonucleotides, it is possible to construct complex RNA molecules entirely from chemically synthesized precursor fragments. However, while commercially available RNA synthesis has become more affordable, it still remains prohibitively expensive for some research groups. In these situations, as well as when RNA ligation precursor fragments must be greater than 70 nucleotides in length, in vitro transcription becomes the method of choice for producing large quantities of unmodified RNA ligation precursor molecules.
3.1. In vitro transcription of telomerase RNA Run-off transcription by one of several phage polymerases can readily produce milligram quantities of RNA fragments in excess of 1 kb in length (Fig. 2.2). The most common phage polymerase used for in vitro transcription is from the bacteriophage T7, which can be purchased commercially or overexpressed and purified from E. coli (Zawadzki and Gross, 1991). Detailed description of in vitro transcription techniques exist elsewhere (Wyatt et al., 1991); therefore, in this section we only highlight a few important points. Duplex DNA templates for T7 RNA polymerase reactions must possess the T7-specific promoter sequence, and can be generated either by linearization of plasmid DNA or via the polymerase chain reaction (PCR). We prefer to produce transcription templates by PCR with a highfidelity polymerase because it facilitates the production of RNA transcripts terminating at an arbitrary sequence. The efficiency of any given transcription reaction will vary significantly and should be optimized by titrating the concentration of Mg2þ, rNTPs, and DNA template. 3.1.1. Protocol 1: In vitro transcription of full-length telomerase RNA 3.1.1.1. Reagents
Phusion Hot Start High Fidelity Polymerase, New England Biolabs (Finnzymes) (F-540L) 5 Phusion HF reaction buffer (F-518)
32
Benjamin M. Akiyama and Michael D. Stone
Telomerase RNA gene T7 RNA polymerase promoter
DNA plasmid
PCR amplify
In vitro transcribe
Purify on desalting column
Early fractions: Telomerase RNA
Late fractions: Free NTPs
Figure 2.2 Schematic diagram of in vitro transcription reaction. The telomerase RNA gene is cloned behind a T7 RNA promoter. PCR is used to generate linear DNA and the PCR amplicons are in vitro transcribed with T7 RNA polymerase. The RNA is purified away from free nucleotide on a desalting column.
Plasmid DNA PCR template in 10 mM Tris–HCl (pH 8) Custom DNA primers (100 mM ), Integrated DNA Technologies dNTP mix (40 mM stock), New England Biolabs (N0447L) Nuclease Free Water for PCR, Ambion (AM9932) Qiagen MiniElute PCR Purification Kit (28004) T7 RNA polymerase, New England Biolabs (M0251L) or laboratory purified T7 RNA polymerase reaction buffer (B9012S) (6 mM MgCl2, 10 mM dithiothreitol, 2 mM spermidine, 40 mM Tris–HCl, pH 7.9) rNTPs (80 mM), New England Biolabs (N0466S)
Assembly of Complex RNAs by Splinted Ligation
33
100% ethanol 3.3 M Na acetate (pH 5.2) PD10 desalting columns (for removing excess rNTPs after transcription reaction), GE Biosciences
3.1.1.2. Preparation of DNA template for transcription reaction by PCR Prepare PCR DNA template using standard plasmid purification techniques. In a total reaction volume of 500 ml prepare the following reaction mix: 50 ng template DNA, 1 mM forward PCR primer, 1 mM reverse PCR primer, 0.2 mM dNTP mix, 5 ml Taq Phusion enzyme in 1 Phusion HF buffer, adding the enzyme last. Aliquot into five PCR reaction tubes (100 ml/tube) and perform the PCR reaction in a thermocycler. 3.1.1.3. In vitro transcription Transcription reactions require 25 ng/ml linear DNA as a template for T7 RNA polymerase. In a total reaction volume of 500 ml, add 12.5 mg PCR template DNA, 0.5 mM NTP mix, and 50 ml T7 RNA polymerase in 1 T7 RNA polymerase buffer. During preparation the reaction mixture and enzyme should be kept on ice. Incubate the reaction at 37 C for 2 h and stop reaction with 10 ml 0.5 M EDTA. 3.1.1.4. Purification of telomerase RNA transcripts Due to the presence of short abortive transcripts within in vitro transcription reactions, it is typically advisable to gel purify desired RNA products by polyacrylamide gel electrophoresis (PAGE). However, the losses incurred during PAGE purification of large RNAs are significant (>50%); therefore, we try to minimize the number of gel purification steps during the construction of complex RNA molecules. Since full-length telomerase RNA transcripts will be further processed by targeted RNase H digestion (see below), we simply purify in vitro transcription reaction products by ethanol precipitation with 0.1 M NaOAc (pH 5.2), followed by a desalting column (PD-10, GE Biosciences) to remove residual unincorporated rNTPs.
3.2. Targeted RNase H cleavage of telomerase RNA Although extremely powerful, the use of in vitro transcription techniques for producing RNA ligation precursor fragments is limited by several intrinsic properties of phage RNA polymerases. First, the efficiency of the transcription reaction is very sensitive to the 50 -sequence of the nascent RNA transcript, typically requiring a guanosine at the þ1 position. Second, the propensity of T7 RNA polymerase to append nontemplated ribonucleotides onto the 30 -ends of nascent transcripts results in a heterogeneous distribution of transcript lengths (so called N þ 1 products). Several approaches have been employed to circumvent these shortcomings, with
34
Benjamin M. Akiyama and Michael D. Stone
the aim of producing RNA fragments of arbitrary sequence with homogeneous 50 - and 30 -termini (Fig. 2.3). The use of cis- and trans-acting ribozymes to remove 50 - and 30 -heterogeneities from large scale RNA preparations has been described (FerreD’Amare and Doudna, 1996; Price et al., 1995) (Fig. 2.3A). This method has the benefit of low cost and facile increase in scale, but suffers from being sequence sensitive and the laborious design/fabrication process for each construct to be produced. As many applications for site-specifically modified A
Telomerase RNA 5¢
Ribozyme cleavage Cis-ribozyme
DNAzyme cleavage
RNase H cleavage
DNAzyme
Chimeric 2¢-O-me RNA/DNA targeting oligo
Trans-ribozyme 3¢
5¢
3¢
5¢
5¢
3¢ 5¢
3¢
3¢
RNase H
Cleavage product RNA 5¢
3¢
B
C 3¢
2¢-O-me RNA DNA 2¢-O-me RNA 5¢
5¢
3¢
1
2
3
4
+
Site of Roche RNase H Site of NEB RNase H cleavage cleavage
Figure 2.3 Preparation of unmodified RNA ligation precursor fragments. (A) Schematic diagram of cleavage strategies for generating RNA ligation precursor fragments of specific length. Open triangles indicate desired site of RNA cleavage. (B) Cleavage site specificity of RNase H enzymes from different vendors. Enzyme obtained from Roche cleaves between the third and fourth base pair of the RNA–DNA duplex (read 50 –30 from the RNA). RNase H from New England Biolabs cleaves after the fourth base pair. (C) Denaturing PAGE gel of RNase H digestion products stained with ethidium bromide. Lane 1: No chimeric targeting oligonucleotide added. Lanes 2 and 3: Single chimeric targeting oligonucleotide added. Lane 4: Double cleavage directed by two chimeric targeting oligonucleotides.
Assembly of Complex RNAs by Splinted Ligation
35
RNA molecules require the production of several different RNA variants, we prefer more versatile approaches for postprocessing products of in vitro transcription reactions. Two alternatives to ribozyme-based approaches for site-specifically cleaving RNA transcripts are targeted cleavage by laboratory engineered DNAzymes (Breaker and Joyce, 1994; Silverman, 2008) or by ribonuclease H (RNase H) (Lapham and Crothers, 1996) (Fig. 2.3A). DNAzymes are oligodeoxyribonucleic acids that have been engineered to possess RNA hydrolyzing activity, with a cleavage specificity targeted by base pair complementarity between the DNAzyme and the RNA target. Although DNAzymes are both cost-effective and elegant, quantitative cleavage of target RNA substrates requires multiple annealing/cleavage cycles which can often result in detectable RNA degradation. For this reason, we rely upon targeted RNase H cleavage to generate large quantities of unmodified RNA ligation precursor fragments with homogeneous termini. RNase H is capable of cleaving RNA site-specifically using a 20 -Omethyl RNA/DNA chimeric oligonucleotide to direct the site of cleavage via Watson–Crick base pairing with the target RNA (Inoue et al., 1987). Thus, one can readily alter the cleavage specificity by simply redesigning a short synthetic 20 -O-methyl RNA/DNA chimeric oligonucleotide (Fig. 2.3A and C). Our general design for RNase H targeting oligonucleotides includes four deoxynucleotide bases flanked by ten 20 -O-methyl ribonucleotides on each side. RNase H can be purchased from a number of commercial sources. However, a cryptic variation in cleavage position specificity has been observed by our group and several others, depending on the enzyme source (Lapham et al., 1997). To circumvent this confusion in the past, we have processed all RNA ligation precursor fragments (unmodified and modified) by RNase H cleavage using a single source of enzyme. However, more recently we have mapped cleavage site preferences for two available sources of RNase H enzymes (Fig. 2.3B), eliminating the requirement for RNase H cleavage of chemically synthesized RNA ligation precursor fragments. The extent of site-specific RNA cleavage will vary with the efficiency of annealing between the chimeric RNase H targeting oligonucleotide and the RNA substrate. Sites with a high degree of secondary structure are typically more difficult to cleave to completion; however, in our experience slight modifications in the position of the cleavage site, length of the chimeric RNase H targeting oligonucleotide or salt concentration during the annealing step usually results in greater than 80% cleavage. Furthermore, we have found that once an efficient RNase H cleavage site has been determined, the efficiency of subsequent splinted RNA ligation at this site is also high. To construct dye-labeled telomerase RNA molecules for our FRET measurements, we first prepare an 89 nucleotide long RNA insert fragment by
36
Benjamin M. Akiyama and Michael D. Stone
targeting RNase H cleavage to two different sites within full-length telomerase RNA produced by in vitro transcription as described above. 3.2.1. Protocol 2: Targeted RNase H cleavage of unmodified RNA ligation precursor fragment 3.2.1.1. Reagents
Ribonuclease H, New England Biolabs (M0297L) RNase H reaction buffer, New England Biolabs (B0297S) (75 mM KCl, 3 mM MgCl2,10 mM dithiothreitol, 50 mM Tris–HCl, pH 8.3) Custom synthetic chimeric 20 -O-methyl RNA/DNA oligonucleotides, Integrated DNA Technologies, Inc. Full-length telomerase RNA prepared by in vitro transcription (see above) RNasin Plus RNase inhibitor, Promega (N2615) Urea loading buffer (8 M urea, 25 mM EDTA, 5 mM Tris, pH 8.0) Fluor-Coated TLC Plates, Ambion (AM10110) KONTES RNase-Free Pellet Pestle, VWR (749521-1590)
3.2.1.2. Annealing chimeric RNase H targeting oligos and telomerase RNA Set up the annealing reaction with 100 mg of RNA from the previous transcription reaction. Add to this a fivefold molar excess of your custom synthetic chimeric 20 -O-methyl RNA/DNA oligonucleotides. Bring up to a total volume of 250 ml with nuclease-free H2O. Denature the reaction for 4 min at 90 C. Incubate for 10 min at 37 C to allow the chimeric oligonucleotides to anneal to the RNA. 3.2.1.3. RNase H cutting reaction In a separate tube on ice add 50 ml 10 RNase H buffer, 12.5 ml RNasin Plus RNase inhibitor, and 50 ml RNase H enzyme to 137.5 ml nuclease-free H2O. Add this to the annealing reaction. Incubate at 37 C for 2 h. Stop the reaction with 1/20th volume 0.5 M EDTA, pH 8.0. 3.2.1.4. Purification of RNase H cleavage products Ethanol precipitate the RNA and resuspend the pellet in 1 urea loading buffer. Run on an 8% denaturing PAGE gel. Transfer the gel to a sheet of saran wrap and place over a fluorescent TLC plate. Excise the band containing the desired product by UV shadowing and place in a 15-ml conical tube. Add 5 ml gel elution buffer (10 mM Tris–HCl, pH 8.0 þ 100 ml phenol:chloroform: isoamyl Alcohol) and crush the excised gel slice with an RNase-free pellet pestle or a pipet tip. Agitate in a shaker or tape to a vortexer left on overnight. Spin down tube and transfer as much solution as possible to a new tube, avoiding the crushed pieces of acrylamide. Butanol extract the volume down to about 200 ml. Ethanol precipitate the RNA and save the pellet.
Assembly of Complex RNAs by Splinted Ligation
37
4. Preparation of Modified (Dye Labeled) RNA Ligation Precursor Molecules In our laboratory, we have primarily utilized splinted RNA ligation methods to generate fluorescently labeled telomerase RNA molecules (Fig. 2.4A); however, the method outlined in this chapter is applicable to the construction of large RNA molecules for use in a variety of RNA structure probing experiments. In principle, any base modification that can be introduced through chemical synthesis can be site-specifically incorporated into a larger RNA construct via RNA ligation approaches. If it is necessary to perform further chemical treatment of modified synthetic RNA fragments (i.e., dye labeling) it is desirable to purify away the fraction of RNA that was not properly functionalized. This can often be done using a HPLC system with the appropriate column. Once purified, the modified synthetic RNA fragments may be ligated to in vitro transcribed RNA to generate full-length RNAs with one or more site-specific labels. For the purposes of this chapter, we describe the labeling of telomerase RNA with two dye labels, Cy3 and Cy5, for use in single-molecule FRET experiments. To generate RNA ligation precursor fragments harboring site-specific dye modifications, synthetic RNAs are purchased possessing a 5-amino-allyluridine base at the desired position. Under basic conditions, monoreactive Cy3 and Cy5 dyes (GE Life Sciences) containing the dye conjugated to an N-hydroxysuccinimide-activated carboxylic acid, will react with amines and amides to covalently attach the RNA to the dye. As the dye packs are reactive with primary amines, buffers that contain amines such as Tris will quench the coupling reaction and should therefore be avoided. It is important to note that any synthetic oligoribonucleotide that will serve as a ‘‘donor’’ during the ligation step must have a 50 -monophosphate introduced, which may be directly incorporated during synthesis or added enzymatically using T4 polynucleotide kinase. We have found it unnecessary to PAGE purify synthetic RNA fragments prior to dye-labeling, as there will be a subsequent PAGE purification step. Furthermore, to ensure stability of the RNA during the labeling reaction, we perform the dye coupling reaction with protected RNA, and deprotect after the RNA is labeled. In our experience, shorter synthetic RNA oligonucleotides with a minimum of predicted secondary structure have a higher labeling efficiency. The length of synthetic RNA ligation precursors will typically also be dictated by other factors such as the size of the target RNA and the desired position of the modification. Lastly, ligation efficiencies will vary significantly from one position to another; therefore, it may be necessary to extend the length of a synthetic fragment in order to improve yields. The removal of unmodified RNA fragments following the dye-labeling step is accomplished by reverse-phase high-pressure liquid chromatography (HPLC) (Fig. 2.4B) (Walter, 2003). Reversed-phase HPLC uses a column
38
Benjamin M. Akiyama and Michael D. Stone
B Absorbance (A.U.)
A Chemically synthesize RNA with 5-N-U modification at desired labeling site
Dye-label 5-N-U labeled RNA with amino-reactive Cy3 or Cy5
EtOH precipitate to remove excess dye
1500
Cy5-RNA
1000
ABS650
500 0
ABS260
8
10
12
14
16
18
mL
C Cy5-RNA
Cy3-RNA
Deprotect dye-labeled RNA Cy5 EtOH precipitate Cy3
PAGE purify and gel extract
Butanol extract, EtOH precipitate, and resuspend in HPLC buffer A (100 mM TEAA, pH 7.0)
HPLC purify dye-labeled RNA by reverse phase chromatography
Figure 2.4 Preparation of modified (dye labeled) RNA ligation precursor fragments. (A) General protocol for dye-labeling synthetic RNA fragments. (B) HPLC trace of purification for a Cy5-labeled RNA oligonucleotide. Unlabeled synthetic telomerase RNA fragment elutes 10 ml (ABS260 peak) and Cy5-labeled RNA elutes at 15 ml (overlapping ABS260 and ABS650 peaks) (C) Diagnostic denaturing PAGE gel of HPLCpurified Cy5-RNA and Cy3-RNA fragments imaged on Typhoon scanner (GE Life Sciences).
with a hydrophobic stationary phase and an aqueous mobile phase. As the mobile phase is gradually combined with a nonpolar solvent, molecules are separated by their polarity in the column with hydrophobic molecules eluting later in the gradient. Since Cy5 and Cy3 add a large hydrophobic surface to the RNA, reversed-phase HPLC is ideal for separating labeled from unlabeled RNA.
Assembly of Complex RNAs by Splinted Ligation
39
4.1. Protocol 3: Dye-labeling and HPLC purification of synthetic RNA oligonucleotides 4.1.1. Reagents
3.3 M NaOAc, pH 5.2 Sodium bicarbonate Custom synthesized RNA oligonucleotides, Thermo Scientific/ Dharmacon Cy5 and Cy3 monoreactive dye packs, GE Healthcare (PA25001 and PA23001) 100% ethanol RNA deprotection buffer, Thermo Scientific/Dharmacon (B-001000DP-018)—100 mM acetic acid, adjusted to pH 3.8 with TEMED KONTES RNase-Free Pellet Pestle, VWR (749521-1590) Urea loading buffer (8 M urea, 25 mM EDTA, 5 mM Tris, pH 8.0) 100 mM triethylammonium acetate (TEAA), pH 7.0 100% acetonitrile Eclipse XDB-C8 HPLC column, Agilent (990967-906) AKTA purifier
4.1.2. Dye-labeling reaction Resuspend 200–400 mg synthetic RNA in 200 ml 10 mM Tris, pH 8.0, and ethanol precipitate synthetic RNA with 0.1 M NaOAc, pH 5.2. Resuspend the pellet in 200 ml 0.1 M sodium bicarbonate solution. Add solution to Cy5 or Cy3 monoreactive dye pack tube. Incubate for 1 h at 37 C and ethanol precipitate. Resuspend dye-labeled RNA pellets in 200 ml deprotection buffer. Incubate at 60 C for 30 min and ethanol precipitate. Resuspend pellet in 100 ml 1 urea loading buffer containing no loading dye. Run on an 8% denaturing PAGE gel. Excise the cyan- or pink-colored band, avoiding any smears below, and transfer to a 15 ml conical tube. Crush the gel slice in 5 ml TE using a pipet tip or RNase-free pellet pestle. Agitate in a shaker or tape to a vortexer left on overnight. Spin down tube and transfer as much solution as possible to a new tube, avoiding the crushed pieces of acrylamide. Butanol extract volume down to about 200 ml. Ethanol precipitate RNA and save pellet. 4.1.3. HPLC purification of Dye-Labeled RNA Equilibrate column in 100% acetonitrile until pressure stabilizes ( 20 ml acetonitrile). Run a gradient to 100% 100 mM TEAA, pH 7.0 over 5 min and stay at 100% 100 mM TEAA. Wait for pressure to stabilize. Resuspend labeled RNA pellets from previous PAGE purification in 60 ml 100 mM TEAA, pH 7.0. Load sample into a 100 ml loop, and inject. Set a gradient to 100% acetonitrile over 35 column volumes. Begin collecting fractions when
40
Benjamin M. Akiyama and Michael D. Stone
the UV 260 absorbance begins to climb. The unlabeled RNA should elute first (Fig. 2.4B), whereas the second large peak should contain the labeled RNA with additional absorbance at 650 nm for Cy5 or 550 nm for Cy3. If the HPLC model does not have a built-in UV–vis detector, a spectrophotometer can be used to identify dye-labeled fractions. Dye-labeled RNA-containing fractions should be pooled and butanol extracted to a volume of 200 ml. They should then be ethanol precipitated and the pellet should be stored at 20 C for ligation to the in vitro transcribed precursor RNA fragment. RNA labeling and purification may be confirmed by UV spectrophotometry and diagnostic PAGE analysis (Fig. 2.4C).
5. RNA Ligation Methods In our laboratory, we have used three different enzymes to catalyze the ATP-dependent ligation of a ‘‘donor’’ RNA fragment possessing a 50 monophosphate to an ‘‘acceptor’’ RNA fragment having a 30 -hydroxyl group: T4 RNA ligase, T4 DNA ligase, and T4 RNA ligase 2 (Fig. 2.5A). The robust ability of T4 RNA ligase to covalently join individual oligoribonucleotides makes this enzyme an obvious choice for the assembly of complex RNA molecules (England and Uhlenbeck, 1978; Walker et al., 1975). However, the utility of T4 RNA ligase for this purpose is limited by its strong preference for single-stranded RNA fragments, which can often lead to unwanted side products including RNA circles and dimers (Fig. 2.5A, left) (Romaniuk and Uhlenbeck, 1983). To circumvent this shortcoming, researchers more typically employ T4 DNA ligase, which can covalently couple two RNA fragments when annealed to a complementary DNA splint (Fig. 2.5A, right) (Moore and Query, 2000; Moore and Sharp, 1992). Splinted RNA ligations have several important advantages over T4 RNA ligase-catalyzed joining reactions. First, the requirement for a complementary DNA splint virtually eliminates formation of unwanted side products, making possible more complex multisite ligation reactions. Second, the precise geometry required to promote T4 DNA ligase-mediated ligation imposes strict constraints upon putative ligation complexes and provides an important level of quality control which bypasses any residual heterogeneity in the lengths of RNA ligation precursor fragments. Despite these clear advantages, the efficiency of splinted RNA ligation reactions catalyzed by T4 DNA ligase is typically limited by the amount of enzyme one can introduce into the reaction. This problem is derived from the poor turnover of T4 DNA ligase on DNA/RNA heteroduplexes. More recently, a variant of T4 RNA ligase, called T4 RNA ligase 2 (Nandakumar et al., 2004), has been identified. This enzyme has a pronounced preference
41
Assembly of Complex RNAs by Splinted Ligation
A
RNA ligation precursor fragments Cy5-labeled RNA fragment
Unmodified RNA fragment
Cy5
Cy3-labeled RNA fragment Cy3
T4 RNA ligase
DNA-splinted RNA ligation
Target product
DNA splint
DNA splint T4 DNA ligase or T4 RNA ligase 2
+ Side products RNA circles Dimers
B
No ligase
T4 RNA T4 DNA ligase 2 ligase
* 1
2
3
4
5
6
Figure 2.5 Strategies for RNA ligation. (A) (left) T4 RNA ligase joins single-stranded RNA and can generate undesired side products such as RNA circles and dimers. (Right) T4 DNA ligase and T4 RNA ligase 2 prefer double-stranded nucleic acid. When directed by a DNA splint, they show greatly reduced side product formation. (B) T4 RNA ligase 2 is highly active. Lanes 1 and 2 are the same gel lane containing all RNA ligation precursor fragments but no ligase enzyme. Lane 1: Typhoon scanner image of ethidium bromide staining of in vitro transcribed telomerase RNA precursor fragments (asterisk indicates cross-excitation of Cy3-RNA fragment in Typhoon scanner). Lanes 3–6 are a titration of commercially available high-concentration preparations of T4 RNA ligase 2 and T4 DNA ligase imaged on Typhoon scanner. Cy3-RNA- and Cy5RNA-labeled fragments are present in twofold excess over in vitro transcribed unmodified RNA fragment. Lane 3: 1% (v/v) T4 RNA ligase 2. Lane 4: 10% (v/v) T4 RNA ligase 2. Lane 5: 1% (v/v) T4 DNA ligase. Lane 6: 10% (v/v) T4 DNA ligase.
42
Benjamin M. Akiyama and Michael D. Stone
for joining RNA ends within a DNA/RNA duplex, and turns over products much faster than T4 DNA ligase. Thus, when using T4 RNA ligase 2 one can use far less enzyme to achieve comparable levels of RNA ligation to that of T4 DNA ligase-catalyzed reactions, making this approach significantly more cost-effective (Fig. 2.5B). For these reasons, we typically use T4 RNA ligase 2 for our splinted RNA ligation reactions. However, it is important to note that T4 RNA ligase 2 retains a detectable level of activity on single-stranded RNA fragments, and thus produces more side products than T4 DNA ligase. These side products can usually be eliminated by gel purification of the desired RNA ligation products. However, for certain joining reactions we have found it necessary to use the less efficient T4 DNA ligase enzyme to suppress unwanted reaction products. Before performing a large scale RNA ligation reaction, we recommend setting up a diagnostic reaction in a 20-ml reaction volume. Once the ligation is determined to be efficient, the reaction can then be scaled up linearly. When designing DNA splints, several factors should be considered. We find that DNA splints that are 25–30 bases long are typically sufficient to form a stable preligation complex. If possible, the site of ligation should be chosen so as to avoid stable secondary structure within the DNA splint strand. If annealing of the RNA ligation precursor fragments to the DNA splint is found to be inefficient, the length of the DNA splint may be increased. In addition, two ligation sites that are sufficiently close to each other may be annealed to a single DNA splint. The relative concentrations of the DNA splint and RNA precursor fragments will have a significant impact on the efficiency of preligation complex formation. While it is possible to ‘‘drive’’ the annealing reaction by increasing the amount of one of the RNA ligation precursor fragments, it is important that the molar concentration of DNA splint not exceed the concentration of the most abundant RNA ligation precursor fragment. The presence of a molar excess of DNA splint will inhibit the reaction by titrating away RNA ligation precursor fragments.
5.1. Protocol 4: Splinted RNA ligation method for producing FRET-labeled telomerase RNA 5.1.1. Reagents
T4 DNA ligase, New England Biolabs (M0202M) T4 RNA ligase 2, New England Biolabs (M0239L) T4 DNA ligase buffer, New England Biolabs (B0202S) (10 mM MgCl2,10 mM dithiothreitol, 1 mM ATP, 50 mM Tris–HCl, pH 7.5) T4 RNA ligase 2 buffer New England Biolabs (B0239S) (2 mM MgCl2, 1 mM DTT, 400 mM ATP, 50 mM Tris–HCl, pH 7.5)
Assembly of Complex RNAs by Splinted Ligation
43
Unmodified RNA ligation precursor fragment (see Section 3) Dye-labeled RNA ligation precursor fragment (see Section 4) Custom DNA ‘‘splint’’ oligos, Integrated DNA Technologies, Inc. RNasin Plus RNase Inhibitor, Promega (N2615) Urea Loading buffer (8 M urea, 25 mM EDTA, 5 mM Tris, pH 8.0) KONTES RNase-Free Pellet Pestle, VWR (749521-1590) 25:24:1 phenol:chloroform:isoamyl alcohol Chloroform Glycogen for Molecular Biology, Roche (10901393001)
5.1.2. Annealing of DNA Splints to precursor RNA fragments The conditions listed in this protocol are for a diagnostic scale reaction. Once successful, these same conditions may be scaled up linearly to produce larger quantities of RNA ligation products. In a total reaction volume of 10 ml, add 20 pmol of the unmodified RNA ligation precursor fragment (from Section 3), 40 pmol Cy5-labeled RNA ligation precursor fragment (from Section 4), 40 pmol Cy3-laveled RNA ligation precursor fragment (from Section 4), and 40 pmol each of the corresponding DNA splint oligos. Denature at 90 C for 3 min, and anneal at 30 C for 10 min. 5.1.3. Ligation reaction In a separate tube, create a 10-ml mixture containing 2 ml T4 DNA ligase or T4 RNA ligase 2 and 1 ml RNasin Plus RNase inhibitor in 1 concentration T4 DNA ligase or T4 RNA ligase 2 buffer. Add this mixture to the annealing reaction above. Incubate at 30 C for 2 h. Bring to 200 ml with 10 mM Tris, pH 8.0, phenol–chloroform extract, and ethanol precipitate, adding 1 ml glycogen to help visualize the pellet. 5.1.4. PAGE purification Resuspend the pellet in 20 ml urea loading buffer and run on an 8% denaturing PAGE gel. Cover the gel while it runs to avoid photobleaching of the dyes. Excise the purple band on the gel corresponding to the Cy5 and Cy3 double-labeled RNA. Confirm the success of your excision using the fluorescence setting on a Typhoon gel scanner. Crush the excised band in 1 ml 10 mM TE pH 8.0 using a pipet tip or an RNase-free pellet pestle. Agitate in a shaker or tape to a vortexer left on overnight. Spin down to pellet the acrylamide and remove as much supernatant as possible, avoiding the acrylamide. Butanol extract supernatant down to about 200 ml. Ethanol precipitate and resuspend the pellet in 5 ml 10 mM Tris, pH 8.0. Quantify RNA concentration using a Nanodrop spectrophotometer.
44
Benjamin M. Akiyama and Michael D. Stone
6. Application: Single-Molecule FRET Measurements We have developed the splinted RNA ligation procedure outlined in this chapter to generate site-specifically dye-labeled telomerase RNA constructs. These modified telomerase RNA constructs may be used to characterize dynamic RNA structural properties using Forster resonance energy transfer (FRET) (Stone et al., 2007). Our laboratory specializes in single molecule FRET measurements, which facilitates the direct observation of transient RNA structural states. The details of single molecule FRET A
Prism Quartz slide Sample FRET-labeled Telomerase RNA
Objective Fluorescence Acceptor (Cy5)
FRET =
IA IA + ID CCD
Donor (Cy3)
Donor Acceptor intensity (ID) intensity (IA)
C Intensity (A.U.)
B
250 200
1500 1000 500 0
150 100 FRET
Number of molecules
300
50 0 0
0.2
0.4 0.6 FRET
0.8
1.0
0
20
40
60 Time (s)
80
100
120
0
20
40
60 Time (s)
80
100
120
1.0 0.8 0.6 0.4 0.2 0
Figure 2.6 Application of splinted RNA ligation procedure: single molecule FRET. (A) Diagram of prism-type total internal reflection fluorescence microscope (TIRFM) for single molecule FRET measurements. (B) Distribution of single-molecule FRET values for dye-labeled telomerase RNA molecules generated by splinted RNA ligation. (C) Dye intensity and FRET traces of a single telomerase RNA molecule: Cy3 emission (green), Cy5 emission (red), FRET ratio (blue).
Assembly of Complex RNAs by Splinted Ligation
45
techniques have been described elsewhere (Roy et al., 2008), so we will only briefly discuss the technique to demonstrate one of many possible applications of splinted RNA ligation techniques. The protocols described in the previous sections enable us to produce extremely pure samples of Cy3–Cy5-labeled telomerase RNA. In addition, we often engineer the 50 -end of the RNA to possess a 15-base sequence extension that is annealed to a complementary DNA strand containing a 30 -biotin modification (Fig. 2.6A). The biotin moiety serves to immobilize the FRET-labeled telomerase RNA on a streptavidin-coated microscope slide. In this way, individual telomerase RNA molecules can be imaged onto a CCD camera using total internal reflection microscopy. A typical single molecule FRET measurement is conducted by illuminating the sample with a 532-nm laser, which directly excites the Cy3 (donor) dye. If the Cy3 dye is sufficiently close (2–9 nm) to the Cy5 (acceptor) dye, energy transfer can occur resulting in a mixture of emitted light coming from the donor and acceptor dyes. Because the emission spectra of Cy3 and Cy5 are sufficiently separated, fluorescence coming from each dye may be separated using a dichroic mirror, and the FRET ratio, defined as the intensity of the acceptor dye (IA) divided by the sum of the acceptor plus donor dyes (IA þ ID), may be determined. In this experimental geometry, FRET measurements can be conducted for many hundreds of telomerase RNAs simultaneously, allowing one to rapidly construct a histogram of observed FRET values (Fig. 2.6B). However, the true power of single molecule approaches is the ability to analyze FRET trajectories for individual RNA molecules (Fig. 2.6C), which has been used characterize structural dynamics of ribozymes (Zhuang et al., 2000) and ribonucleoprotein complexes (Cornish et al., 2008; Stone et al., 2007).
ACKNOWLEDGMENTS We thank our colleagues Xiaowei Zhuang, Mariana Mihalusova, John Y. Wu, and Kathleen Collins for their contributions to the work described in this chapter. We apologize to those scientists whose work was not cited due to space considerations. The research described in this chapter was supported by the National Institutes of Health, Howard Hughes Medical Institute, and University of California at Santa Cruz startup funds.
REFERENCES Breaker, R. R., and Joyce, G. F. (1994). A DNA enzyme that cleaves RNA. Chem. Biol. 1, 223–229. Cate, J. H., et al. (1999). X-ray crystal structures of 70 S ribosome functional complexes. Science 285, 2095–2104. Cech, T. R. (1990). Self-splicing of group I introns. Annu. Rev. Biochem. 59, 543–568. Cohen, S. B., and Cech, T. R. (2001). Engineering disulfide cross-links in RNA using thioldisulfide interchange chemistry. Curr. Protoc. Nucleic Acid Chem. Chapter 5, Unit 5 1.
46
Benjamin M. Akiyama and Michael D. Stone
Cornish, P. V., et al. (2008). Spontaneous intersubunit rotation in single ribosomes. Mol. Cell. 30, 578–588. England, T. E., and Uhlenbeck, O. C. (1978). Enzymatic oligoribonucleotide synthesis with T4 RNA ligase. Biochemistry 17, 2069–2076. Ferre-D’Amare, A. R., and Doudna, J. A. (1996). Use of cis- and trans-ribozymes to remove 50 and 30 heterogeneities from milligrams of in vitro transcribed RNA. Nucleic Acids Res. 24, 977–978. Han, H., and Dervan, P. B. (1994). Visualization of RNA tertiary structure by RNA-EDTA. Fe(II) autocleavage: Analysis of tRNA(Phe) with uridine-EDTA.Fe(II) at position 47. Proc. Natl. Acad. Sci. USA 91, 4955–4959. Hengesbach, M., et al. (2008). RNA intramolecular dynamics by single-molecule FRET. Curr. Protoc. Nucleic Acid Chem. Chapter 11, Unit 11 12. Inoue, H., et al. (1987). Sequence-dependent hydrolysis of RNA using modified oligonucleotide splints and RNase H. FEBS Lett. 215, 327–330. Lapham, J., and Crothers, D. M. (1996). RNase H cleavage for processing of in vitro transcribed RNA for NMR studies and RNA ligation. RNA 2, 289–296. Lapham, J., et al. (1997). The position of site-directed cleavage of RNA using RNase H and 20 -O-methyl oligonucleotides is dependent on the enzyme source. RNA 3, 950–951. Moore, M. J., and Query, C. C. (2000). Joining of RNAs by splinted ligation. Methods Enzymol. 317, 109–123. Moore, M. J., and Sharp, P. A. (1992). Site-specific modification of pre-mRNA: The 20 -hydroxyl groups at the splice sites. Science 256, 992–997. Moore, M. J., and Sharp, P. A. (1993). Evidence for two active sites in the spliceosome provided by stereochemistry of pre-mRNA splicing. Nature 365, 364–368. Nandakumar, J., et al. (2004). RNA substrate specificity and structure-guided mutational analysis of bacteriophage T4 RNA ligase 2. J. Biol. Chem. 279, 31337–31347. Pinol-Roma, S., et al. (1989). Ultraviolet-induced cross-linking of RNA to proteins in vivo. Methods Enzymol. 180, 410–418. Price, S. R., et al. (1995). Crystallization of RNA-protein complexes. I. Methods for the large-scale preparation of RNA suitable for crystallographic studies. J. Mol. Biol. 249, 398–408. Romaniuk, P. J., and Uhlenbeck, O. C. (1983). Joining of RNA molecules with RNA ligase. Methods Enzymol. 100, 52–59. Roy, R., et al. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Scaringe, S. A., et al. (1998). Novel RNA synthesis method using 50 -O-silyl-20 -O-orthoester protecting groups. J. Am. Chem. Soc. 120, 11820–11821. Silverman, S. K. (2008). Catalytic DNA (deoxyribozymes) for synthetic applications-current abilities and future prospects. Chem. Commun. (Camb.) 14(30), 3467–3485. Stone, M. D., et al. (2007). Stepwise protein-mediated RNA folding directs assembly of telomerase ribonucleoprotein. Nature 446, 458–461. Walker, G. C., et al. (1975). T4-induced RNA ligase joins single-stranded oligoribonucleotides. Proc. Natl. Acad. Sci. USA 72, 122–126. Walter, N. G. (2003). Probing RNA structural dynamics and function by fluorescence resonance energy transfer (FRET). Curr. Protoc. Nucleic Acid Chem. Chapter 11, Unit 11 10. Wyatt, J. R., et al. (1991). Synthesis and purification of large amounts of RNA oligonucleotides. Biotechniques 11, 764–769. Yu, Y. T. (2000). Site-specific 4-thiouridine incorporation into RNA molecules. Methods Enzymol. 318, 71–88. Zawadzki, V., and Gross, H. J. (1991). Rapid and simple purification of T7 RNA polymerase. Nucleic Acids Res. 19, 1948. Zhuang, X., et al. (2000). A single-molecule study of RNA catalysis and folding. Science 288, 2048–2051.
C H A P T E R
T H R E E
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes Sergey Solomatin* and Daniel Herschlag† Contents 1. Introduction 2. Design of Labeled RNA Constructs 2.1. Selection of labeling sites and construct assembly methods 2.2. Design of the construct assembly 2.3. Selection of dyes for single molecule fluorescence studies 3. Dye Labeling of RNA Fragments 4. Notes on In Vitro Transcription with T7 RNA Polymerase 5. Assembly of Labeled RNA Constructs 6. Examples of Protocols 6.1. Protocol 1: Labeling of RNA oligos by with fluorescent dyes (NHS ester form) 6.2. Protocol 2: Ligation of large RNAs with T4 DNA ligase Acknowledgments References
48 49 49 49 51 53 55 56 58 58 61 66 66
Abstract Single molecule fluorescence techniques offer unique insights into mechanisms of conformational changes of RNA. Knowing how to make fluorescently labeled RNA molecules and understanding potential limitations of different labeling strategies is essential for successful implementation of single molecule fluorescence techniques. This chapter offers a step by step overview of the process of obtaining RNA constructs ready for single molecule measurements. Several alternative methods are described for each step, and ways of troubleshooting the most common problems, in particular, splinted RNA ligation, are suggested.
* Department of Biochemistry, Stanford University, Stanford, California, USA Departments of Biochemistry and Chemistry, Stanford University, Stanford, California, USA
{
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69003-0
#
2009 Elsevier Inc. All rights reserved.
47
48
Sergey Solomatin and Daniel Herschlag
1. Introduction Over the past two decades, an ever increasing appreciation of multiple roles RNAs play in biology has led to an increasing interest in understanding the fundamental behavior of RNA. A broad variety of experimental approaches has been applied to studying structure and dynamics of RNA, including native gel electrophoresis, NMR, small angle X-ray scattering (SAXS), chemical structure probing, atomic force microscopy, UV, and fluorescence spectroscopy. Single molecule methods have become the latest frontier in studies of RNA dynamics (Weiss, 2000; Zhuang, 2005), revealing unique information about the behavior of RNA molecules hidden from bulk experiments by ensemble averaging (Downey et al., 2006; Ha et al., 1999; Hodak et al., 2005; Lee et al., 2007; Qu et al., 2008; Russell et al., 2002; Xie et al., 2004; Zhuang et al., 2000, 2002). Fluorescence-based techniques have two characteristics that make them particularly useful in single molecule implementation. First, they allow increased throughput of single molecule measurements through simultaneous observation of hundreds of individual molecules at the same time. Analysis of data from thousands of individual molecules is essential for bridging the gap between observations made on individual molecules and traditional bulk measurements made with ensembles of 1020 molecules. Second, an ability to label different parts of RNA molecules allows one to study the dynamics at a submolecular level and obtain increasingly detailed information about the RNA structure and motions. To take full advantage of these traits, one needs to incorporate fluorescent dyes site-specifically into an RNA of interest and to be able to do this at different selected position on an RNA molecule. This chapter outlines general strategies of preparing dye-labeled RNA constructs, in particular concentrating on double-labeled constructs for single molecule fluorescence resonance energy transfer (smFRET) measurements (Ha, 2001; Roy et al., 2008). These constructs are designed such that the dynamics of interest is revealed through changes of the distance between two dye labels, donor and acceptor, which in turn result in changes of the energy transfer efficiency from the donor to the acceptor. Anticorrelated changes of the donor and acceptor fluorescence, resulting from changes in energy transfer efficiency, are easy to distinguish from uncorrelated fluctuations of the fluorescence intensity arising from multiple possible sources. Because of this trait, FRET is the technique that has been most widely used for studying RNA dynamics (Weiss, 2000; Zhuang, 2005). A step by step overview of the process of obtaining RNA constructs ready for single molecule FRET is presented here. For each step, several alternative methods described in the literature are suggested. Means of
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
49
troubleshooting of the most common problems, in particular, splinted RNA ligation, are suggested based on the literature and on the authors’ personal experience.
2. Design of Labeled RNA Constructs 2.1. Selection of labeling sites and construct assembly methods Design of a labeled RNA construct is aimed at achieving two goals: 1. Obtaining the best possible FRET signal. 2. Minimally perturbing the behavior of the RNA. Observation of a FRET signal requires two dyes to be positioned on the RNA within the range of efficient energy transfer determined by the Fo¨rster distance for a particular dye pair (typically, 3–6 nm). A crystal structure is a great starting point for identifying appropriate labeling positions. A simple heuristic rule is to choose labeling sites that are remote in the secondary structure, but close in the tertiary structure. However, one should be careful not to perturb residues that might be involved in forming long-range tertiary interactions (e.g., Brion and Westhof, 1997, and references therein). Such residues are likely to conform to the rule above, but modifying them can severely destabilize the native structure of the RNA. On the other hand, labeling base paired regions adjacent to residues that make long-range tertiary contacts is safer, and it ensures that FRET will be observed. If a crystal structure is not available, phylogenetic ( Jaeger et al., 1994; Michel and Westhof, 1990), cross-linking (Chen et al., 1998), and biochemical (Lehnert et al., 1996) data can help guide the search for best labeling sites.
2.2. Design of the construct assembly Labeling with two different dyes typically requires the final RNA construct to be assembled from at least two fragments (Scheme 3.1). Depending on the choice of labeling sites, one of the following assembly strategies can be pursued: 1. RNA is made (synthetically or by in vitro transcription) in one piece, and labeled by base pairing to complementary-labeled oligos: (a) Two oligos are bound at the 50 and 30 ends (b) One of RNA ends is labeled directly, and an oligo is bound at the other end (c) One oligo is bound at an end and the other at an internal site
50
Sergey Solomatin and Daniel Herschlag
1
2
a)
b)
c)
d)
a)
b) Ligation
e)
Circular permutation
Scheme 3.1 Design strategies for making labeled RNA constructs. Numeration corresponds to the outline numbering in the text. Brown lines designate RNA, blue lines designate DNA oligos, stars designate fluorescent dyes.
(d) Two oligos are bound at internal sites. (e) New 50 and 30 ends are designed by circular permutations, and an oligo is bound at an internal site. 2. RNA is split into several fragments: (a) RNA is reassembled by base pairing of the fragments. (b) RNA is reassembled by covalent joining of the fragments. Assembly method 2.b (Lee et al., 2007; Sattin et al., 2008) reproduces exactly the same RNA as of the original unlabeled construct and can be considered the least perturbing method of labeling, as long as the positions of the dyes were appropriately chosen. However, it also remains the most technically challenging method, and it imposes the strictest requirements on the purity of RNA fragments, as discussed below. Other methods of assembly (1.a–e and 2.a) are easier to implement, but they restrict the choice of labeling sites to the ends of the molecule (1.a–c and 2.a), and/or require assumptions that modifications of the sequence— that is, binding of oligos at the ends and at internal sites (1.b–d), circular permutations (1.e) or nicks in the continuous backbone (2.a)—do not affect the dynamics of interest. Testing such assumptions can be nontrivial. The modular architecture of RNA structure lends some support to the assumption that adding duplexes at the ends of the molecule will not generally perturb its behavior (Brion and Westhof, 1997; Tinoco and Bustamante, 1999). Single molecule studies demonstrated that the overall folding rate, substrate docking rate and catalysis by the Tetrahymena group I
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
51
ribozyme were the same for the construct labeled by and oligo annealed to a 30 end extension as for unmodified construct (Russell et al., 2002; Zhuang et al., 2000, 2002). This method of labeling has been most widely used, but it has an obvious limitation of placing dyes in the vicinity of the ends of the molecule. In principle, the ends of an RNA molecule can be moved by making circular permutations (Pan, 2000). As it is known that circular permutations can strongly affect folding mechanisms of RNA (Lease et al., 2007; Pan et al., 1999), this approach may be mostly useful for studying structure and local dynamics of RNA. Also, after a circular permutation the ends of the molecule are expected to be right next to each other, so that the second dye has to be placed at some internal position. To place a dye at internal positions without breaking the RNA backbone, the Pan lab developed a method that involves replacing nonessential hairpin loops within RNA sequences with larger loops with specific sequences that are hybridized to labeled DNA oligos (Smith et al., 2005). These modifications had little effect on structure, as assayed by chemical footprinting, or catalytic efficiency of the catalytic domain of RNase P. The same method was successfully employed for studying ribosome dynamics (Dorywalska et al., 2005).
2.3. Selection of dyes for single molecule fluorescence studies Single molecule FRET experiments push the limits of sensitivity and time resolution of the detection systems, because the goal of these experiments is to get as much information as possible from the weakest possible light source. Good photophysical properties of dyes are essential for getting the most out of these experiments. The following properties are highly desirable for smFRET dyes: (1) high extinction coefficient and quantum yield (i.e., most of the excitation light is converted into useful signal); (2) high stability against photobleaching (i.e., each molecule can be observed for a long time); (3) stable fluorescent signal (i.e., no chemical or conformational transformations of the dye leading to large fluctuations of fluorescence such as blinking); (4) good spectral overlap of donor emission and acceptor excitation (allowing high maximum FRET efficiency); (5) good spectral separation of donor and acceptor emissions (i.e., easy to optically separate two signals and calculate the actual value of FRET); (6) donor and acceptor emission in the range of high quantum efficiency of the detection systems (e.g., for some CCDs quantum efficiency falls off sharply outside of the 450–850 nm window). Currently, one can choose from a broad variety of organic fluorophores covering the entire optical spectrum from UV to near IR that are commercially available (see Table 3.1). A lot of early smFRET work was performed
52
Sergey Solomatin and Daniel Herschlag
Table 3.1 Spectral properties of several selected fluorophores for RNA labeling Dye
Cyanine fluorphores Cy2 Cy3 Cy5 Cy5.5 Alexa fluorophores Alexa 350 Alexa 430 Alexa 488 Alexa 532 Alexa 555 Alexa 647 Alexa 700 ATTO fluorophores ATTO 425 ATTO 532 ATTO 647 ATTO 700 Other fluorophores Fluorescein (0.1 M NaOH) Tetramethyl rhodamine Texas Red
lex (nm)
lem (nm)
e (M 1 cm 1)
F
489 548 649 675
506 562 670 694
150,000 150,000 250,000 190,000
0.12 0.16–0.39 0.28 0.23
346 430 494 530 555 651 702
445 545 517 555 572 672 723
19,000 15,000 73,000 81,000 155,000 270,000 205,000
– – 0.92 0.61 0.1 0.33 0.25
436 532 645 700
484 553 669 719
45,000 115,000 120,000 120,000
0.90 0.90 0.20 0.25
495
519
79,000
0.79–0.95
557
576
103,000
0.2
589
615
139,000
0.9
with the cyanine dyes Cy3 and Cy5, and, while several lines of dyes were marketed recently as superior to cyanine dyes, the Cy dyes are still widely used because they have a highly desirable combination of properties. The fluorescence signal of Cy dyes is strong, long-lived and stable in oxygen-depleted environments in the presence of stabilizing agents,1 and they have well-separated emission spectra in the range ideal for most detection systems. Manufacturers of Alexa (Invitrogen), Dylight (Thermo Fisher Scientific), and ATTO (ATTO-Tec, also available from Sigma-Aldrich) dyes offer a broad choice of fluorophores covering the entire visible spectrum in small increments. Red dyes from these lines are reported to be significantly 1
Trolox (Rasnik et al., 2006) is an exceptionally good one, but other compounds, such as b-mercaptoethanol, n-propyl gallate, ascorbic acid, or chloramphenicol, have been used (Widengren et al., 2007).
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
53
more photostable than Cy5.2 Furthermore, ATTO dyes display exceptionally high brightness and have low intersystem crossing rates, which should favorably reflect on the stability of their fluorescence signal. Furthermore, some of the older, nonbranded, dyes, for example, Texas Red (Ha et al., 1996), fluorescein (Xie et al., 2004), or tetramethyl rhodamine (Lang et al., 2004), can work perfectly well in single molecule applications, and these dyes typically cost less than those noted above. All of the above dyes can be purchased in N-succinimide activated form and used for labeling of amino groups (see below). The choice of fluorophores are more limited for labeling of thiol groups with maleimide derivatives of dyes, and especially for full synthesis of oligos using dye-labeled phosphoramidites.
3. Dye Labeling of RNA Fragments Any strategy of assembly of a labeled RNA construct for smFRET requires obtaining two dye-labeled RNA fragments, typically two oligos. Dye-labeled RNA fragments can be obtained in one of the following ways: 1. Commercially labeled oligos. 2. In-house labeling of commercially synthesized oligos with amino- or thiol groups. 3. Full in-house oligonucleotide synthesis with 50 end incorporation of dye phosphoramidites. 4. Direct RNA labeling. One should keep in mind that not every labeling method is compatible with assembly strategies that require ligation of the oligos, for example enzymatic ligation with T4 DNA ligase would require 50 monophosphate group and a free hydroxyl group at the 30 of each junction to be ligated, whereas RNAs labeled by dye phosphoramidite incorporation or direct labeling at the 50 and 30 ends will have different end groups most likely incompatible with the ligation. 1. Commercially labeled oligos. Several commercial oligos manufacturers (e.g., International DNA Technology, or IDT, www.idtdna.com, Dharmacon, www.dharmacon.com, Gene Link, www.genelink.com) provide an option of purchasing a custom RNA oligos with a variety of dyes incorporated at the 50 end, 30 end or internally (currently only at uracils). While undeniably convenient, this strategy often is not the most 2
These comparisons (by the manufacturers) were probably made in oxygen-rich environments, as the differences appear to be small in oxygen-depleted solutions in the presence of Trolox (unpublished observations).
54
Sergey Solomatin and Daniel Herschlag
cost-effective because of higher prices and low yields of supplied oligos. Furthermore, the choice of dyes is limited to those few that available from the RNA manufacturer, and currently the selection is narrow for dyes to be incorporated at 30 and internal positions. 2. Labeling of amino- or thiol-modified oligos. In-house labeling of commercially synthesized oligos carrying reactive groups at specified positions combines the convenience of not having to synthesize the full oligo sequence with much broader flexibility of choosing one’s favorite dyes and placing them at any position along the oligo sequence. These factors, along with the lower cost of synthesis, makes in house labeling strategy our current favorite. Most fluorescent dyes are available as N-hydroxysuccinimide (NHS) or maleimide derivatives, ready for conjugation to primary amino- or thiol groups, respectively. As natural RNAs lack strongly nucleophilic aliphatic primary amino groups, the use of NHS-derivatives is typically easier and more straightforward than thiol modification by maleimides, which may require working in reducing or oxygen-free environments. Amino groups can be incorporated into custom RNA oligos during synthesis at either of the ends (through an aliphatic linker), internally at the uridine base as 5-aminoallyl uridine or at the backbone (Uni-LinkTM by IDT).3 The labeling reaction (Scheme 3.2) is easy to perform (see below), as long as certain precautions are taken. It is necessary to make sure that amines are not present in the reaction mix (e.g., TRIS-based buffers are incompatible with NHS labeling) and that the pH is optimal for labeling. At low pH most of the amino groups are protonated and their reactivity is low, but at high pH NHS esters can be hydrolyzed faster than they can react with the oligo, resulting in low yields. Carbonate or phosphate-based buffers at pH 8–9 typically work well. R2 NH2
O R1 O
N
R1
N R2 H Dye-labeled RNA
O
OH
O + O
O R1
OH Inactivated dye
O
OH
O H2O
N
+ O
N
O
Scheme 3.2 The labeling reaction of amino-modified RNA (R2) with NHS-activated dye (R1) and a competing side reaction of NHS hydrolysis.
3
20 -Amino ribose can also be used, but amino group in this position is much less reactive than a typical aliphatic amine.
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
55
Purification of the labeled oligo from excess dye can be accomplished by a combination of ethanol precipitation and PAGE or HPLC purification. Changes in the oligo mobility resulting from dye labeling are usually sufficient to purify the labeled oligo from the unlabeled material. Accomplishing this purification is beneficial for single molecule experiments, as it decreases the number of molecules that are not labeled, or labeled with a single dye only. 3. Full in-house oligonucleotide synthesis with 50 end incorporation of dye phosphoramidites. Due to wide availability of commercially synthesized oligonucleotides, full synthesis of RNA oligos in individual labs is not frequently carried out these days and choices of commercially available dye phosphoramidites are limited. Thermo Fisher Scientific (www. thermo.com) provides a selection of DyLight phosphoramidites that have spectral properties analogous to Cy3, Cy5, and Cy5.5 dyes. 4. Direct RNA labeling. Several methods of labeling the directly labeling transcribed RNA molecules been reported in the literature, but have not yet found wide applications. Dyes at the 50 were incorporated cotranscriptionally by initiating with a dye–guanosine conjugate (Fang et al., 1999), at the 30 end through oxidation of the terminal ribose to aldehyde form by sodium periodate with subsequent reaction with hydrazine derivatives of dyes (Proudnikov and Mirzabekov, 1996), and at the 20 hydroxyl in the middle of an RNA by deoxyribozyme-catalyzed ligation (Baum and Scott, 2007).
4. Notes on In Vitro Transcription with T7 RNA Polymerase RNA fragments longer than 40–50 nucleotides are most often synthesized by in vitro transcription with T7 RNA polymerase. Heterogeneity of transcripts obtained by this method must be recognized by researchers aiming to use such transcripts for ligations, as it can lead to serious artifacts that are especially notable in single molecule experiments. T7 RNA polymerase strongly prefers guanosine at the first and second transcribed positions, otherwise the transcription yields drastically decrease (Milligan and Uhlenbeck, 1989). However, if T7 polymerase encounters four or more guanosines in a row at the start site, it generates heterogeneity at the 50 end (Pleiss et al., 1998). A proper choice of the starting sequence appears to be sufficient to avoid this problem. However, heterogeneity at the 30 end is the rule rather than the exception for any RNA sequence transcribed by T7 polymerase. Run-off in vitro transcripts almost invariably contain a significant fraction of nontemplated nucleotides at the 30 end, with n þ 1, n þ 2, and n þ 3
56
Sergey Solomatin and Daniel Herschlag
transcripts being the main contaminants (Milligan and Uhlenbeck, 1989). For transcripts that are larger than 100 nt this contamination is not easy to recognize and essentially impossible to purify away. To obtain clean 30 ends, it is best to extend the transcribed sequence beyond the intended end and then cleave the RNA at the desired site. RNA can be extended with (a) a sequence that encodes one of small ribozymes (Ferre-D’Amare and Doudna, 1996; Price et al., 1995); (b) a recognition sequence for a DNAzyme (Santoro and Joyce, 1997); or (c) with a sequence complementary to a DNA oligo with subsequent cleavage of the hybrid by RNase H (Stone et al., 2007, see Akiyama and Stone, Chapter 2, this volume). Extending the 30 end sequence with a cis-cleaving hammerhead ribozyme is an easy and efficient way to obtain ‘‘clean’’ 30 ends, as this gives cotranscriptional cleavage. However, it does place certain constraints on the sequence at cleavage site (Birikh et al., 1997). If the desired 30 end sequence is not compatible with the hammerhead cleavage, another small ribozyme or other methods mentioned above can be used. Small ribozymes and DNAzymes leave 20 –30 cyclic phosphate at the site of cleavage, and this group must be removed before the ligation. This task can be accomplished by treating cleaved products with polynucleotide kinase (PNK) in the absence of ATP (Schurer et al., 2002).
5. Assembly of Labeled RNA Constructs 1. Noncovalent assembly by Watson–Creek base pairing. Assembly of larger RNA constructs through base pairing of complementary oligos, or oligos and in vitro transcripts, is the most widely used method in single molecule fluorescence field (Ha et al., 1999; Hodak et al., 2005; Xie et al., 2004; Zhuang et al., 2002). Its biggest advantage is the ease of the procedure, which usually involves simple mixing of the solutions of oligonucleotides and annealing via some combination of heating and cooling steps. Upon annealing, the efficiency of construct assembly can be tested using nondenaturing acrylamide gels. Purification of the fully assembled constructs can also be done using nondenaturing PAGE. 2. Covalent incorporation of labeled oligos. Full-length-labeled RNA molecules of essentially any size can be obtained by joining together labeled RNA oligos (typically obtained by synthesis) and either synthetic or in vitro transcribed RNA fragments comprising the rest of the sequence. Joining several RNA fragments into a single chain can be done using (i) protein ligases, (ii) deoxyribozyme ligase, and (iii) chemical ligations. (a) Enzymatic ligation using T4 DNA ligase (Moore and Sharp, 1992), remains the most often used process, offering significant advantages
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
57
over RNA ligase I as discussed by Moore and Query (2000). Recently discovered RNA ligase II (Ho and Shuman, 2002) does not suffer from many of the limitations of RNA ligase I and may become an efficient alternative to T4 DNA ligase. Heterogeneity of fragments that need to be ligated is one of the biggest obstacles for enzymatic ligations. Heterogeneity of in vitro transcripts was discussed above, and it must be avoided. Synthetic oligos are always contaminated by shorter products (n 1, n 2, etc.) because of <100% coupling efficiency in synthesis, and these contaminants must also be purified away. This purification is typically accomplished together with purification of labeled from unlabeled material, but it is best to check the purified material on a denaturing gel. Templated nature of the reactions with DNA ligase and RNA ligase II ensures that these products would not be ligated efficiently (Moore and Query, 2000). However, T4 DNA ligase does not perfectly exclude nontemplated nucleotides at the end from the ligation (K. Travers, W. Zhao, D. Herschlag, unpublished data), which results in incorporation of extra bases in the middle of the sequence—a very undesirable outcome. The exact specificity of T4 RNA ligase II toward templated nucleotides is not known to the authors. Even if they are not ligated, the presence of shorter products will limit the annealing efficiency, increasing the amount of enzyme consumed in the reaction and decreasing the yields. Even with purified fragments, ligation efficiency is often low, especially when multiple fragments need to be ligated. In many cases, the yield of the ligation is limited by the extent of formation of correctly annealed complexes between all RNA fragments and DNA splints. The most likely explanation for the poor annealing efficiency is intramolecular structure formation by the RNA fragments. Long splints (Kurschat et al., 2005), ‘‘disrupter’’ oligos (Strobel and Cech, 1993), and optimization of the annealing protocols can be tried for improving annealing efficiency, but, in some cases, choosing different sites for the ligation junctions is the only method that results in yield improvement. Purification of the full-length-labeled RNA can be performed by PAGE. Even small amount of the ligated product can be detected by fluorescent scanners, and detection of fluorescence from both dyes can be used to identify the correct band.4 (b) Deoxyribozymes (see S. Silverman, Chapter 5, this volume) that are capable of efficient ligation of RNA fragments have recently been evolved in the Silverman laboratory (Purtha et al., 2005). These 4
As a rule rather than an exception, multiple products of incomplete ligations will be observed on the gel. It is therefore not recommended to have either of the terminal fragments very short and unlabeled, as incomplete products missing these fragments would be very difficult to distinguish and separate from the full-length RNA, especially for longer molecules.
58
Sergey Solomatin and Daniel Herschlag
deoxyribozymes catalyze the formation of the natural 30 –50 linkages and have reasonably broad sequence requirements at the site of cleavage, so they have a large potential as alternatives to protein ligases. (c) Chemical ligations employ reagents such as carbodiimide, or bromocyane (BrCN, also known as cyanogen bromide) (Dolinnaya et al., 1988; Fedorova et al., 1996). These ligations typically work better with DNA than with RNA substrates and are prone to formation of nonnatural linkages.
6. Examples of Protocols 6.1. Protocol 1: Labeling of RNA oligos by with fluorescent dyes (NHS ester form) 6.1.1. Protocol outline 1. Prepare reagents (a) Dry dimethyl sulfoxide (DMSO) (b) Buffer exchange (and concentrate) RNA oligos (c) Pour the purification gel 2. Run the labeling reaction 3. Purify products (a) (b) (c) (d)
Precipitation PAGE purification5 Second precipitation and/or desalting step Analysis
Reagents RNA oligo (5-aminoallyl uridine and 50 -monophosphate modified)a Dye, NHS-derivative (powder) Dimethyl sulfoxide (DMSO) Molecular sieves, class 3A Phosphate buffer (500 mM, pH 8) Sodium acetate (3 M, pH 5.2) Ethanol (95% or 200 proof ) 5
HPLC purification can be used instead of PAGE.
Equipment Microcon columns (YM-3 or YM-10, Millipore) Table top centrifuge Floor centrifuge Centrifuge tubes (e.g., Oakridge tubes fro Nalgene) Gel electrophoresis box Sterile scalpels (alternatively, one can use flame sterilized razor blades) Glass rods
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
59
Loading solution (90% formamide, trace amounts of bromphenol blue and xylene cyanol) Acrylamide (29:1, 20%, 7 M urea) TBE buffer (100 mM Tris–base, 83 mM boric acid, 1 mM EDTA), autoclaved a
50 -Monophosphate is only needed if the 50 end of the oligo will be ligated.
6.1.2. Protocol 1. Preparation of the reagents 1.1. Drying DMSO. Dyes are expensive, typically, $300 per 1 mg, and rarely available packaged in small amounts that are convenient for test reactions or small scale labeling. Labeling of 100 nmol or less of an RNA oligo will require only a fraction of 1 mg of dye, and the excess dye solution can be stored for future use. The presence of residual water in DMSO can cause hydrolysis of the NHS ester, so it is important to keep DMSO anhydrous. This task is accomplished by drying DMSO over molecular sieves 3A.6 1.2. Buffer exchanging and concentrating the oligos. If the RNA oligos arrived from the manufacturer in lyophilized form and were deprotected before use, they may contain residual TEMED that will interfere with the labeling reaction. If the oligos are supplied in ‘‘lab-ready’’ form, the buffer might be TRIS based, which will also interfere with labeling. It is advised to desalt and buffer-exchange the oligos using Microcon columns (Millipore) or the equivalent. If the oligo concentration is less than 1 mM, it is also strongly advised to concentrate the solution at this step to 1 mM or higher. Higher concentrations of RNA result in better yields at the same dye-tooligo ratio, as the labeling reaction is second order and can more effectively compete with hydrolysis of the active dye (see Scheme 3.2). 1.3. Pour a denaturing polyacrylamide gel. Choose the percentage of the gel and its length according to the length of the oligo. For some dyes (e.g., Cy3 and Cy5), the labeled material has electrophoretic mobility approximately equal to an oligo that is longer by a single nucleotide. It is strongly suggested to run a test gel with a small aliquot of labeling reaction and an unlabeled oligo marker. The labeled and unlabeled bands can be visualized by staining using, for 6
Molecular sieves should be activated according to manufacturer’s instructions (typically by heating at 250 C for 2 h or longer).
60
Sergey Solomatin and Daniel Herschlag
example, StainsAll (Sigma-Aldrich) and the percentage of acrylamide and the length of the gel can then be adjusted to obtain proper separation. For <30-mer oligos, 20% gels with 20 cm plates provide adequate separation; 1.5 mm gels with 10–20 mm wells provide a good balance between band visibility and resolution on 10–30 nmol (oligo) scale; and 3 mm gels and/or wider wells should be used for labeling reactions on larger scales. 2. A sample labeling reaction for 100 nmol of an RNA oligo Stock reagent Volume 1 mM amino-modified RNA 100 ml 500 mM phosphate buffer, pH ¼ 8 20 ml DMSO (anhydrous) 100 ml Dye (1 mg) Total volume 220 ml
Final concentration 500 mM 50 mM 45% 5 mM
2.1. Combine RNA and phosphate buffer in a 2-ml Eppendorf tube. 2.2. Add DMSO to the tube of dye and pipette up and down vigorously to dissolve it. 2.3. Add the needed amount of dye solution to the RNA solution and mix well. 2.4. If there is dye remaining, it should be quickly frozen on dry ice and stored at 80 C to prevent hydrolysis of the NHS ester. 2.5. Incubate at 22 C for 1 h. 3. Purification of the products 3.1. Precipitate RNA by adding 0.11 volumes of 3 M sodium acetate (pH 5.2), and 4 volumes of cold ethanol and incubate on dry ice for 30 min or longer. 3.2. Centrifuge at 14,000 rpm for 40 min at 4 C, decant and save the supernatant, allow the pellet to dry under air. 3.3. Resuspend the pellet in a minimal volume of water, add 1 volume of the loading solution, load the gel and run as long as needed to achieve good separation of the labeled product from the unlabeled oligo (see 1.3). 3.4. Identify the correct band and excise it with a sterile scalpel.7 3.5. Place excised gel pieces into ice-chilled collection tubes and crush them thoroughly with glass rods, then quickly freeze on dry ice. 7
Note that the excess free dye often aggregates and runs as multiple bands on the gel. To avoid misidentification of the bands, it is essential to perform the precipitation step (3.1–3.4), removing the excess dye, and to run a test gel (1.3) to establish the mobility of the labeled oligo.
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
61
3.6. Add water (3 volume of the gel) and repeat freeze-thaw cycle three times. 3.7. Centrifuge for 10 min at 4000g to remove the gel fragments and take out the supernatant. Perform the second round of elution overnight on a shaker at 4 C. 3.8. Combine the supernatants from the first and the second elution rounds, filter through 0.2 mm filters and precipitate RNA as in (3.1)–(3.4). 3.9. Resuspend in water or in a storage buffer of choice, and estimate the concentration of the dye by measuring the UV absorbance in the range of dye absorption, and of the RNA oligo by measuring the UV absorbance at 260 nm. The labeling efficiency can be calculated from the ratio of the dye concentration to the RNA concentration. Alternatively, the concentration of the dye can be estimated by measuring its fluorescence intensity and comparing it to a standard curve measured for a series of known dye concentrations. 3.10. It is recommended to test the purity of the product at the end of the purification procedure by running a denaturing PAGE gel and (1) scanning it on a fluorescent scanner (e.g., Typhoon system, Molecular Dynamics) to determine the extent of degradation (and contamination by shorter, for example, n 1, n 2, etc., labeled oligos), and (2) staining it with StainsAll to determine the extent of contamination with the unlabeled oligo.
6.2. Protocol 2: Ligation of large RNAs with T4 DNA ligase Preparation of the labeled constructs by enzymatic ligation involves several steps of preparing ligation components, as shown in Scheme 3.3. This protocol describes only the last step in the overall procedure, as the preceding steps are standard in RNA biochemistry and corresponding protocols can be found elsewhere (see, e.g., Akiyama and Stone, Chapter 2, this volume). 6.2.1. Outline of the protocol 1. Anneal RNA fragments and DNA splints 2. Run the ligation reaction (a) Test the results of the ligation 3. Purify products by PAGE
62
Sergey Solomatin and Daniel Herschlag
Construct design
Transcription of RNA fragments
Labeling of aminomodified RNA oligos
3¢ end processing Purification of labeled oligos from unlabeled and truncated oligos
3¢ end cyclic phosphate removal
Enzymatic ligation
Product purification
Scheme 3.3 Flow-chart for making labeled RNA constructs by enzymatic (T4 DNA ligase) ligation. Denaturing PAGE purifications after each step of RNA transcript preparation are performed, but now shown for clarity.
Reagents TE buffer (10 mM Tris–HCl, pH 8, 1 mM EDTA) NaCl (5 M ) T4 DNA Ligase 10 ligase buffer (660 mM Tris–HCl, pH 7.6, 80 mM MgCl2, 100 mM DTT, 1 mM ATP, 0.04% Triton X-100) RNA fragmentsa (10 mM is desirable) DNA splint(s) a
Equipment PCR machine (Thermocycler) (or heating block) for annealing Table top centrifuge Gel electrophoresis box Sterile scalpels (or flame sterilized razor blades) Glass rods
50 -Ends that will be joined must be phosphorylated, and 30 ends must have free hydroxyl groups.
6.2.2. Protocol 1. Annealing Annealing efficiency is higher when all the components are in approximately stoichiometric amounts, and the concentration of each component is
63
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
higher than 1 mM (10 mM is preferred). Excess amounts of middle fragments or splints will markedly decrease ligation efficiency because instead of a single splint joining the two ends at the junction, each end can bind a splint, so that no joining occurs. Therefore, concentrations of each component must be carefully measured. To achieve 10 mM final concentrations in the annealing mixture, solutions of RNA fragments might need to be concentrated, especially for multijunction ligations. This can be done using Microcon columns with an appropriate molecular weight cut-off (Table 3.2). Mix the components together in a tube, place into a PCR machine and run the following program: 95 C for 5 min; ramp down to 22 C at 0.1 C/min; keep at 4 C. Alternatively, heat to 95 C in heat block for 5 min, then let cool to room temp by leaving in turned off heat block. It will take 1.5–2 h to reach room temperature; can be varied if needed. 2. Run ligation reaction Mix the components and incubate at room temperature overnight. Run a test gel with a small aliquote of the reaction mix (1 ml should be sufficient, as the Typhoon scanner can detect <10 fmol of dye) before stopping the ligation. Some junctions might require significantly longer incubation times, increase of the yield over as many as 72 h have been observed (unpublished data). RNA degradation may become a limiting factor over such long incubation times, so maintaining RNase-free conditions is essential. Nonenzymatic RNA degradation can be limited by lowering the pH (Table 3.3). 3. Purify products as described in the Protocol 1, Part 3.
Table 3.2 Annealing at 1 nmol scale for ligating N fragments with M splints
a
Stock reagent
Volume
Final concentration
RNA fragments (100 mM each) DNA splints (100 mM each) TE buffer NaCl (5 M ), optional H2O Total volume
N 10 ml
10 mM each
M 10 ml
10 mM each
10 ml 0–20 ml 100 (N þ M) 10 VNaCl
0.1 0–1 Ma 100 ml
The amount of NaCl added during the annealing step can be varied to achieve optimal annealing. However, T4 DNA ligase is inhibited at >200 mM NaCl, and consequently NaCl concentration in the ligation step should be <200 mM.
64
Sergey Solomatin and Daniel Herschlag
Table 3.3 Ligation at 1 nmol scale for N fragments Stock reagent
Volume
Final concentration
Annealed complex (10 mM ) Ligase buffer Superas*In (20 U/ml), optional T4 DNA ligase ( 200 mM ) H2O Total volume
100 ml 100 ml 5 ml (N 1) 5 ml 800 (N 1) 5
1 mM each 10 mM each 0.1 U/ml (N 1) mM 1 ml
If the scale of the ligation is more than 10 nmol, the band of the full-length product on the gel will most likely be visible to a naked eye. If the ligation is performed on a smaller scale, the position of the ligated full-length RNA can be identified by scanning the gel on a Typhoon scanner in the fluorescent mode. 6.2.3. Troubleshooting of ligations Typical yields of full-length ligated products after gel purification are 40% for two-fragment ligations, 20% for three-fragment ligations, and 10% for four- and five-fragments ligations. If none of the full-length product is formed or the yield is too low, conditions of the annealing and ligation reaction can be optimized (see below). As multiple factors may contribute to low ligation yields, it is recommended that individual parameters affecting ligation efficiency are systematically tested. Figure 3.1 provides an example of systematic optimization of the ligation reaction that produced no product before optimization primarily due to a single junction not being ligated, most likely because of structure formation in the 30 -end fragment. Several reaction parameters ([ATP], [Mg2þ], reaction time) had to be optimized to achieve ligation of the full-length product. 1. Vary ligation conditions: (a) Concentration of the ligase (1 mol of ligase per junction is typically required) (b) Concentration of ATP. Although ATP is required, excess ATP may inhibit the reaction (Cherepanov and de Vries, 2003) (c) Incubation time (d) Incubation temperature 2. Test ligation of individual junctions (a) Analysis of a test gel for the full length ligation often gives a clear indication as to which junction is not ligated efficiently. Identify which product corresponds to each band by its mobility and dye color.
65
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
DNA splints
A
Cy3 B
1 2
RNA1
RNA2
Cy5 3
4 C
% ligated
25 20 15 10 5 0
ATP
Mg2+
Time
Figure 3.1 Optimization of the four-fragment ligations of double-labeled Tetrahymena group I ribozyme. (A) Design of the labeled construct. (B) Fluorescence scans of a denaturing PAGE of the ligation reactions before optimization (left) and after optimization (right). It is obvious that at the first junction Cy3-labeled oligo is not ligated to RNA1 at all. Lanes 1 and 3 correspond to 12 h ligation time, lanes 2 and 4 correspond to 36 h ligation times. (C) Optimization of the ligation efficiency of the first junction. Parameters indicated on the x-axis were varied as follows: [ATP] was 10 mM (clear), 100 mM (hatched), 1 mM (black), for all reactions [Mg2þ] was 12 mM, reaction time 12 h; [Mg2þ] was 10 mM (clear), 20 mM (hatched), 30 mM (black), for all reactions [ATP] was 100 mM, reaction time 12 h; reaction time was 12 h (clear), 24 h (hatched), 36 h (black), for all reactions [ATP] was 100 mM, [Mg2þ] was 20 mM.
(b) If a particular ‘‘difficult’’ junction is identified, test if the 50 side has the phosphate and the 30 side has the free OH group (cyclic phosphate was efficiently removed). The presence of the 50 -phosphate in RNA oligos is easiest to confirm by the mass spectrometry. The presence of the free 30 -OH group can be detected by, for example, a primer extension assay with the Klenow fragment. 3. Test annealing by running the annealed complex on a nondenaturing gel and visualizing by fluorescent scanning. If low amount of fully annealed complex is observed, suggesting poor annealing efficiency, the following parameters should be optimized in parallel small scale reactions: (a) Concentration of RNA and DNA splints (middle fragments should NEVER be added in excess, since this reduces the concentration of the complexes.) (b) NaCl concentration (c) Additional salts to aid annealing (e.g., MgCl2) (d) Rate of cooling, or several cycles of heating and cooling
66
Sergey Solomatin and Daniel Herschlag
ACKNOWLEDGMENTS This work was supported by an NIH Program Project Grant (PO1-GM-066275). We thank W. Zhao, K. Travers, B. Sattin, and M. Forconi for sharing their experimental expertise.
REFERENCES Baum, DanaA., and Scott, K. S. (2007). Deoxyribozyme-catalyzed labeling of RNA. Angew. Chem. 46, 3502–3504. Birikh, K. R., Heaton, P. A., and Eckstein, F. (1997). The structure, function and application of the hammerhead ribozyme. Eur. J. Biochem. 245, 1–16. Brion, P., and Westhof, E. (1997). Hierarchy and dynamics of RNA folding. Annu. Rev. Biophys. Biomol. Struct. 26, 113–137. Chen, J.-L., Nolan, J. M., Harris, M. E., and Pace, N. R. (1998). Comparative photocrosslinking analysis of the tertiary structures of Escherichia coli and Bacillus subtilis RNAse P RNAs. EMBO J. 17, 1515–1525. Cherepanov, A. V., and de Vries, S. (2003). Kinetics and thermodynamics of nick sealing by T4 DNA ligase. Eur. J. Biochem. 270, 4315. Dolinnaya, N. G., Sokolova, N. I., Gryaznova, O. I., and Shabarova, Z. A. (1988). Sitedirected modification of DNA duplexes by chemical ligation. Nucleic Acid Res. 16, 3721–3738. 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 Acid Res. 33, 182–189. Downey, C. D., Fiore, J. L., Stoddard, C. D., Hodak, J. H., Nesbitt, D. J., and Pardi, A. (2006). Metal ion dependence, thermodynamics, and kinetics for intramolecular docking of a GAAA tetraloop and receptor connected by a flexible linker. Biochemistry 45, 3664–3673. Fang, X., Pan, T., and Sosnick, T. R. (1999). A thermodynamic framework and cooperativity in the tertiary folding of a Mg2þ-dependent ribozyme. Biochemistry 38, 16840–16846. Fedorova, O. A., Gottikh, M. B., Oretskaya, T. S., and Shabarova, Z. A. (1996). Cyanogen bromide-induced chemical ligation: Mechanism and optimization of the reaction conditions. Nucleosides Nucleotides 15, 1137–1147. Ferre-D’Amare, A. R., and Doudna, J. A. (1996). Use of cis- and trans-ribozymes to remove 50 and 30 heterogeneities from milligrams of in vitro transcribed RNA. Nucleic Acid Res. 24, 977–978. Ha, T. (2001). Single-molecule fluorescence resonance energy transfer. Methods 25, 78–86. 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. Electrochem. Soc. 93, 6264–6268. Ha, T., Zhuang, X., Kim, H. D., Orr, J. W., Williamson, J. R., and Chu, S. (1999). Ligandinduced conformational changes observed in single RNA molecules. Proc. Natl. Acad. Sci. USA 96, 9077–9082. 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.
Methods of Site-Specific Labeling of RNA with Fluorescent Dyes
67
Hodak, J. H., Fiore, J. L., Nesbitt, D. J., Downey, C. D., and Pardi, A. (2005). Docking kinetics and equilibrium of a GAAA tetraloop-receptor motif probed by single-molecule FRET. Proc. Natl. Acad. Sci. USA 102, 10505–10510. Jaeger, L., Michel, F., and Westhof, E. (1994). Involvement of a GNRA tetraloop in long-range RNA tertiary interactions. J. Mol. Biol. 236, 1271–1276. Kurschat, W. C., Mueller, J., Wombacher, R., and Helm, M. (2005). Optimizing splinted ligation of highly structured small RNAs. RNA 11, 1909–1914. Lang, M. J., Fordyce, P. M., Engh, A. M., Neuman, K. C., and Block, S. M. (2004). Simultaneous, coincident optical trapping and single-molecule fluorescence. Nat. Methods 1, 133–139. Lease, R. A., Adilakshmi, T., Heilman-Miller, S., and Woodson, S. A. (2007). Communication between RNA folding domains revealed by folding of circularly permuted ribozymes. J. Mol. Biol. 373, 197–210. Lee, T.-H., Lapidus, L. J., Zhao, W., Travers, K. J., Herschlag, D., and Chu, S. (2007). Measuring the folding transition time of single RNA molecules. Biophys. J. 92, 3275–3283. Lehnert, V., Jaeger, L., Michele, F., and Westhof, E. (1996). New loop-loop tertiary interactions in self-splicing introns of subgroup IC and ID: A complete 3D model of the Tetrahymena thermophila ribozyme. Chem. Biol. 3, 993–1009. Michel, F., and Westhof, E. (1990). Modelling of the three-dimensional architecture of group I catalytic introns based on comparative sequence analysis. J. Mol. Biol. 216, 585–610. Milligan, J. F., and Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods Enzymol. 180, 51–62. Moore, M., and Query, C. C. (2000). Joining of RNAs by splinted ligation. Methods Enzymol. 317, 109–123. Moore, M. J., and Sharp, P. A. (1992). Site-specific modification of pre-mRNA: The 20 -hydroxyl groups at the splice sites. Science 256, 992–997. Pan, T. (2000). Probing RNA structure and function by circular permutations. Methods Enzymol. 317, 313–330. Pan, T., Fang, X., and Sosnick, T. (1999). Pathway modulation, circular permutation and rapid RNA folding under kinetic control. J. Mol. Biol. 286, 721–731. Pleiss, J. A., Derrick, M. L., and Uhlenbeck, O. C. (1998). T7 RNA polymerase produces 50 end heterogeneity during in vitro transcription from certain templates. RNA 4, 1313–1317. Price, S. R., Ito, N., Oubridge, C., Avis, J. M., and Nagai, K. (1995). Crystallization of RNA-protein complexes I. Methods for the large-scale preparation of RNA suitable for crystallographic studies. J. Mol. Biol. 249, 398–408. Proudnikov, D., and Mirzabekov, A. (1996). Chemical methods of DNA and RNA fluorescent labeling. Nucleic Acid Res. 24, 4535–4542. Purtha, W. E., Coppins, R. L., Smalley, M. K., and Silverman, S. K. (2005). General deoxyribozyme-catalyzed synthesis of native 30 –50 RNA linkages. J. Am. Chem. Soc. 127, 13124–13125. Qu, X., Smith, G. J., Lee, K. T., Sosnick, T. R., Pan, T., and Scherer, N. F. (2008). Singlemolecule nonequilibrium periodic Mg2þ-concentration jump experiments reveal details of the early folding pathways of a large RNA. Proc. Natl. Acad. Sci. USA 105, 6602–6607. Rasnik, I., McKinney, S. A., and Ha, T. (2006). Nonblinking and long-lasting singlemolecule fluorescence imaging. Nat. Methods 3, 891–893. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Russell, R., Zhuang, X., Babcock, H. P., Millett, I. S., Doniach, S., Chu, S., and Herschlag, D. (2002). Exploring the folding landscape of a structured RNA. Proc. Natl. Acad. Sci. USA 99, 155–160.
68
Sergey Solomatin and Daniel Herschlag
Santoro, S. W., and Joyce, G. F. (1997). A general purpose RNA-cleaving DNA enzyme. Proc. Natl. Acad. Sci. USA 94, 4262–4266. Sattin, B. D., Zhao, W., Travers, K., Chu, S., and Herschlag, D. (2008). Direct measurement of tertiary contact cooperativity in RNA folding. J. Am. Chem. Soc. 130, 6085–6087. Schurer, H., Lang, K., Schuster, J., and Morl, M. (2002). A universal method to produce in vitro transcripts with homogeneous 30 ends. Nucleic Acid Res. 30, e56. Smith, G. J., Sosnick, T. R., Scherer, N. F., and Pan, T. A. O. (2005). Efficient fluorescence labeling of a large RNA through oligonucleotide hybridization. RNA 11, 234–239. Stone, M. D., Mihalusova, M., O’Connor, C. M., Prathapam, R., Collins, K., and Zhuang, X. (2007). Stepwise protein-mediated RNA folding directs assembly of telomerase ribonucleoprotein. Nature 446, 458–461. Strobel, S. A., and Cech, T. R. (1993). Tertiary interactions with the internal guide sequence mediate docking of the P1 helix into the catalytic core of the Tetrahymena ribozyme. Biochemistry 32, 13593–13604. Tinoco, I., and Bustamante, C. (1999). How RNA folds. J. Mol. Biol. 293, 271–281. Weiss, S. (2000). Measuring conformational dynamics of biomolecules by single molecule fluorescence spectroscopy. Nat. Struct. Mol. Biol. 7, 724–729. Widengren, J., Chmyrov, A., Eggeling, C., Lofdahl, P.-A., and Seidel, C. A. M. (2007). Strategies to improve photostabilities in ultrasensitive fluorescence spectroscopy. J. Phys. Chem. 111, 429–440. Xie, Z., Srividya, N., Sosnick, T. R., Pan, T., and Scherer, N. F. (2004). Single-molecule studies highlight conformational heterogeneity in the early folding steps of a large ribozyme. Proc. Natl. Acad. Sci. USA 101, 534–539. Zhuang, X. (2005). Single-molecule RNA science. Annu. Rev. Biophys. Biomol. Struct. 34, 399–414. Zhuang, X., Bartley, L., 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. 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
F O U R
Fluorophore Labeling to Monitor tRNA Dynamics Cuiping Liu, Thu Betteridge, and Ya-Ming Hou Contents 70 75
1. Introduction 2. Methodology 3. Fluorescent Labeling of D Residues in Native and Transcripts of tRNAs 3.1. Preparation of a target tRNA transcript 3.2. The U to D conversion in the target tRNA transcript 3.3. Fluorescent labeling of D residues 3.4. Experimental considerations 4. Fluorescent Labeling of the CCA Sequence 4.1. Incorporation of PyC to position 75 4.2. Experimental considerations 4.3. Incorporation of 2AP to position 76 4.4. Experimental considerations 5. Conclusions Acknowledgments References
75 77 78 80 81 82 83 87 88 89 89 90 90
Abstract Transfer RNA (tRNA) molecules mediate translation of the nucleic acid genetic code into the amino acid building blocks of proteins, thus ensuring the survivability of cells. The dynamic properties of tRNA molecules are crucial to their functions in both activity and specificity. This chapter summarizes two methods that have been recently developed or improved upon previous protocols to introduce fluorophores to site-specific positions in tRNA. One method enables incorporation of fluorophores carrying a primary amine (such as proflavin or rhodamine) to dihydrouridine (D) residues in the tRNA tertiary core, and a second method enables incorporation of pyrroloC and 2-aminopurine to positions 75 and 76, respectively, of the CCA sequence at the 30 end. These site-specific fluorophore labeling methods utilize tRNA transcripts as the Department of Biochemistry and Molecular Biology, Thomas Jefferson University, Philadelphia, Pennsylvania, USA Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69004-2
#
2009 Elsevier Inc. All rights reserved.
69
70
Cuiping Liu et al.
substrates to provide the versatility with both wild-type and mutant sequences for examining their motions in space and time during the process of decoding genetic information.
1. Introduction Transfer RNAs (tRNAs) are a family of ribonucleic acids to which amino acids are joined. This joining is the underlying basis of decoding, where amino acids are specifically related to the anticodon triplet sequences in tRNAs and in turn to codons in mRNAs. The tRNA molecules are dynamic for interactions with multiple partners in the decoding process. They are transcribed as precursors, with 50 leader and 30 trailer sequences and some even with introns, and thus they must undergo a maturation process, involving removal of precursor sequences, modification at specific base and backbone groups, and tagging with the ubiquitous CCA sequence at the 30 end. Each mature tRNA is recognized by its cognate aminoacyltRNA synthetase (aaRS) for joining with the correct amino acid. The synthesized aminoacyl-tRNA (aa-tRNA) is carried to the A (aminoacyltRNA) site of the ribosome by elongation factor EF-Tu in bacteria (or EF-1a in eukaryotes and archaea) to a codon position complementary to the anticodon of the tRNA. Once accepted at the ribosome A site, the aminoacyl group of the aa-tRNA is used for peptide bond formation, while the tRNA is in constant motion, transiting from the A site to the P (peptidyl-tRNA) site and finally to the E (exit) site. In each of these transitions, the tRNA makes contact with an ensemble of proteins and RNAs in the ribosome and changes partners as it performs the steps through the cycle of protein synthesis (Korostelev et al., 2006; Selmer et al., 2006). The conformational space that each tRNA can explore is necessarily large in order to adapt to different binding partners in a particular step, while delivering the activity and specificity required for decoding (<1 error in 3000 amino acids in vivo) (Bouadloun et al., 1983; Loftfield and Vanderjagt, 1972). The kinetics of movements between different conformational states is also critical so as to promote rapid decoding (20 amino acids per sec in vivo) that is necessary to support cell growth (Dennis et al., 2004). These demands emphasize the importance of tRNA dynamics, which is crucial to tRNA functions in both specificity and activity. However, despite this importance, our understanding of tRNA dynamics and regulation of the dynamics for a specific decoding step is still rather limited. Dynamics of tRNA is a process that changes intrinsic tRNA motions in response to binding to one or more partners. The intrinsic tRNA motions arise from the covalent and noncovalent restraining forces that hold the tRNA structure together. There are typically 76 nucleotides in each
71
Fluorescent tRNA
tRNA molecule, with regions of internal self-complementarity that fold into a compact cloverleaf secondary structure (Fig. 4.1A). A set of conserved and semiconserved nucleotides in the dihydrouridine (D) loop, variable (V)
A
A C C A G C C G C U C
Aminoacylation
site C G Acceptor stem G C G A T-stem AT-loop D-loop G U A D-stem U UC UCC C G G U A A Dus1p GGAGG CGCG U UC U C site GCGCA U G A 18 G U A UG G V-loop C G G U A Primer extension G C Anticodon stem G C stop site A U U G U G Anticodon loop G Anticodon E. coli tRNAPro/UGG
B
C
CCA end
T-loop
– + + – – Dus1p tRNA (mg) – 0.6 1.2 0.6 1.2 Primer + + + + +
D-loop
G18 Anticodon loop
Primer (21 mer)
Figure 4.1 E. coli tRNAPro/UGG for modification by Dus1p. (A) Sequence and cloverleaf structure, showing regions of the acceptor stem (red), D stem and loop (blue), anticodon stem and loop (purple), V loop (cyan), T stem and loop (magenta), the UGG anticodon triplet (indicated by a black bracket), and the CCA tag as the amino acid attachment site. Dus1p modifies U17a, which is mapped by primer extension, using a primer complementary to the sequence indicated by an arrow. (B) The L shaped tRNA structure, showing regions of the T loop, D loop, anticodon loop, and the CCA end in the same colors as those in (A). (C) Primer extension analysis of transcript of E. coli tRNAPro/UGG by a denaturing 12% PAGE/7 M urea gel, showing the Dus1p-dependent major stop at G18 (Betteridge et al., 2007).
72
Cuiping Liu et al.
loop, and TCC (T) loop facilitates tertiary folding of the cloverleaf structure into an L shape (Fig. 4.1B), where the acceptor stem and T stem are coaxially stacked into one helical stem, while the D stem and anticodon stem are coaxially stacked into another helical stem (Kim et al., 1974). The L shape highlights three regions that are hot spots for interactions with binding partners: the CCA tag at one end, the anticodon triplet at the other end, and a tertiary core at the elbow region mediated by long-range hydrogen bonding and stacking interactions between the D, V, and T loops. Each of these three hot spots fluctuates among many states on evolutionarily selected energy landscapes. Binding with a partner at one spot can shift the energy landscapes among different states. This signal of binding can be localized or relayed to a distal region, depending on the sequence of tRNA and the nature of the binding interaction. Both local and distal changes are mediated by tRNA dynamics, which has far-reaching impacts on reaction specificity and activity. An example of tRNA dynamics is in the recognition of tRNACys for aminoacylation by the Escherichia coli cysteine-specific tRNA synthetase (CysRS). The enzyme recognizes all three bases of the tRNA anticodon, using a reorganized anticodon-binding domain to perform direct readout of base-specific functional groups (Hauenstein et al., 2004). Anticodon recognition in turn alters the backbone shape and conformation of the D-anticodon stem as compared to that in the enzyme-free tRNA, suggesting a signaling event from enzyme binding to the anticodon bases to affect distant regions in the D-anticodon stem up to the tertiary core. Mutations in any of the anticodon bases typically reduce the catalytic efficiency (kcat/Km) of aminoacylation by 102–103-fold (Komatsoulis and Abelson, 1993; Zhang et al., 2008b), suggesting a relay of the signaling from the anticodon to the CCA end, where aminoacylation takes place. Importantly, E. coli CysRS also recognizes specific backbone groups in an unusual shape of the tertiary core mediated by a unique 15–48 bp (Hauenstein et al., 2004). Mutations of these backbone groups reduce the catalytic efficiency of aminoacylation by 102-fold (Hamann and Hou, 1997, 2000; Hou, 1994; Hou et al., 1993), suggesting a separate signaling event from the tertiary core to the CCA end to influence the activity of aminoacylation. In these cases, the signals of multiple interactions (at the anticodon and at the tertiary core) must be integrated to control the dynamics of the CCA sequence for aminoacylation. In a second example, recognition of tRNA by E. coli CCA enzyme, which catalyzes the step-wise nucleotide addition of the CCA sequence to a damaged 30 end, is focused on backbone groups of the acceptor-T stem domain (Tomita et al., 2006; Xiong and Steitz, 2004). However, backbone damage in the anticodon loop reduces the enzyme activity (Dupasquier et al., 2008), indicating a long-distance signaling between the two ends of the tRNA L shape. In this case, enzyme binding to the acceptor end appears to have triggered a signaling network that senses the dynamic changes of the
Fluorescent tRNA
73
anticodon end, which in turn predisposes the acceptor end to form the correct conformation for the CCA enzyme. The third example is recognition of an aa-tRNA for entry into the A site of the ribosome, which is inspected by the small ribosomal subunit for anticodon–codon base pairing interaction (Wimberly et al., 2000). The signal of correct base pairing is transmitted to the large ribosomal subunit to accommodate the aa-tRNA and to catalyze peptide bond formation. Importantly, mutations localized outside of the base pairing interaction, such as in the 32–38 positions of the anticodon loop (Ledoux et al., 2009; Murakami et al., 2009) or at the 24 position of the D stem (Cochella and Green, 2005; Cochella et al., 2007; Hirsh, 1971), are known to affect the accuracy of this signaling, providing yet another example of integrating multiple signals as tRNA samples its conformational space within the ribosome en route to developing specificity. These examples illustrate the fundamental importance of signaling in the activity and specificity of tRNA functions and raise the awareness of tRNA dynamics in mediating the complex interacting pathways during signaling. Understanding how tRNA dynamics is controlled in space and time in response to signals of binding partners requires appropriate tools. Although there are ample tRNA crystal structures, including those in complexes with aaRSs, modification enzymes, and with the ribosome, each of these structures captures only a static image of tRNA without giving insight into the ensemble of accessible states or the time scales required for interconversion between states in the course of dynamic motions. A further limitation of the structural approach is that it cannot readily or precisely predict how a specific mutation might affect tRNA dynamics and signaling. The value of studying a specific tRNA mutation cannot be overemphasized, because it is only through comparison between a wild-type and mutant tRNA that specific insight into the origin and control of signaling and dynamics can be elucidated. For example, in the example of aminoacylation of tRNA by E. coli CysRS, it is necessary to analyze tRNA dynamics in terms of both the kinetics and amplitude, and to compare these terms between the wild-type and an anticodon mutant of tRNACys, to establish the necessary basis to address if the mutational effect is to slow down the kinetics of tRNA dynamics, or to shift the CCA end to an incorrect conformational state, or both. Resolution between these possibilities will prove highly significant for understanding the molecular underpinnings of tRNA dynamics for the reaction of interest. An effective approach to study tRNA dynamics that is amenable to measurements of changes in both kinetics and amplitude is by fluorescence labeling at a site-specific position as a probe for local environment changes, either within the tRNA molecule or in the contact with a binding partner. The clear advantage of the fluorescence approach is the ability to monitor environmental changes continuously and in real time. Such fluorescently labeled tRNAs were first used to study tRNA reactions on the ribosome
74
Cuiping Liu et al.
(Paulsen and Wintermeyer, 1986; Robertson and Wintermeyer, 1981; Robertson et al., 1986; Rodnina et al., 1994, 1997) and have remained highly popular for tRNA-ribosome studies. Recent work has further demonstrated the utility of fluorescent tRNAs in single-molecule detection of tRNA-ribosome dynamics (Blanchard et al., 2004a,b; Fei et al., 2008; Lee et al., 2007; Munro et al., 2007). In contrast to ensemble experiments, which monitor molecular motions in averaged time scales, single-molecule experiments directly monitor the time-dependent motions and report conformational fluctuations between different states in high spatial and time resolution, thus offering more information into details of the ribosome translation mechanism. There are two strategies in the design and selection of the fluorescent probes for site-specific introduction to a tRNA molecule. One strategy is tailored to the specific sequence composition of the target tRNA. For example, the wybutine base at position 37 of yeast tRNAPhe can be selectively removed under acid condition and replaced by a nucleophilic fluorophore, such as proflavin (Schleich et al., 1978). The acp3 (3-(3amino-3-carboxypropyl) uridine) modification of the U47 base of yeast tRNAPhe can be specifically tagged with the Cy3 or Cy5 fluorophore in a reaction mediated by NHS esterification (Blanchard et al., 2004a,b; Munro et al., 2007). The e-amino group of Lys-tRNALys can react with BODIPYderivatized fluorophores (Woolhead et al., 2004), while the a-amino group of Met-tRNAfMet can react with a series of fluorophores, such as pyrene, eosin, and coumarin (McIntosh et al., 2000). In each of these cases, the fluorophore is introduced to a unique feature of the target tRNA or aa-tRNA, which cannot be generalized. The second strategy is tailored to the general features of tRNAs and will have broader application. For example, the modified base s4U at position 8 (s4U8) is common to all tRNAs as a defense mechanism against UV irradiation (Kramer et al., 1988). Fluorophore labeling at s4U8 has been successfully achieved for a number of different tRNAs (Blanchard et al., 2004a,b; Fei et al., 2008; Munro et al., 2007), although such labeling can interfere with some reactions on the ribosome, possibly because U8 is involved in the tertiary folding of the tRNA L shape. A less invasive fluorescent labeling is to the 5,6-dihydrouridine (D) residues (Wintermeyer and Zachau, 1974, 1979), which are derivatized from U residues by the action of dihydrouridine synthases (Xing et al., 2002) and are present in the D loop as a landmark of the tRNA tertiary core. The vast majority of D residues occur at positions 16, 17, 20, 20a, and 20b of tRNAs, although some occur at the base of the V loop at position 47. The D residues are nonaromatic and thus are rarely involved in stacking or hydrogen-bonding interactions in the tRNA L shape but instead are flexibly flipped out into solution, facilitating fluorophore labeling while allowing tRNA to retain functions. Indeed, fluorophore labeling at D residues provided the groundwork for earlier
Fluorescent tRNA
75
success in elucidating many tRNA reaction steps on the ribosome before the availability of high-resolution crystal structures of tRNA–ribosome complexes (Paulsen and Wintermeyer, 1986; Robertson and Wintermeyer, 1981; Robertson et al., 1986; Rodnina et al., 1994, 1997). In addition, recent work has developed methods of fluorescent labeling to the universal CCA sequence at the tRNA 30 end (Ling et al., 2009; Zhang et al., 2008a), opening up new possibilities to probe the dynamics of tRNA at or near the site of aminoacylation. This chapter describes methods of fluorophore labeling to D residues with improvements over the previous ones, and methods to introduce fluorophores to the CCA sequence. Both sequence motifs are general to tRNAs and are located in the hot spots for binding interactions, suggesting that these site-specific fluorescent labeling methods will have a broad range of utility in generating important information on tRNA dynamics.
2. Methodology 1. Fluorescent labeling of D residues in native and transcripts of tRNAs. 2. Fluorescent labeling of the CCA sequence in transcripts of tRNAs.
3. Fluorescent Labeling of D Residues in Native and Transcripts of tRNAs The heterocyclic ring of D residues is subject to reductive cleavage by sodium borohydride, yielding 3-ureidopropanol bound to the ribose C-10 position (Cerutti and Miller, 1967), which is a facile leaving group that is readily replaced by fluorophores bearing a primary amino group (Wintermeyer and Zachau, 1974), such as proflavin or rhodamine (Fig. 4.2, steps 1–3). To introduce a fluorophore to D residues, the target tRNA must harbor the appropriate U ! D conversion, which can be achieved in one of three ways. The easiest is if the target tRNA is commercially available in its native form, which would be modified with U ! D conversions as specified by the gene sequence during cellular biosynthesis and maturation. At present, Sigma (St. Louis, MO) and Chemical Block (Moscow, Russia) are the two primary sources that offer native tRNAs, which collectively provide tRNAPhe of yeast and E. coli, the initiator tRNAfMet of E. coli, and elongator tRNALys, tRNAVal, tRNATyr, tRNAGlu, and tRNAArg of E. coli. However, the commercial supply for native tRNAs is rather limited, and an additional drawback is that these tRNA species all have wild-type sequences and cannot be used to study mutational effects. The second way to obtain a D-containing tRNA is to purify the
76
Cuiping Liu et al.
OH
O
O
H
H HN
5¢-tRNA
HN
5¢-tRNA +
O
O
+ NAD /NADP
N O
H
H
H
O
H OH
O
O
NADH/NADPH
N
FAD Dus1p
H
Step 1
H
H
O
H OH 3¢-tRNA
3¢-tRNA
O
H
O
O
Step 2
H
H
H
O
H OH
−
Cl −O
R=
O
NH2 +
CH3
3S
5¢-tRNA
Rhodamine 110
O
HN
CH3
SO− 3
Cy3 hydrazide
+ N
H
H
H
O
H OH 3¢-tRNA
CH3
O
R O
CH3 CH3 N
H
Proflavine, rhodamine, Cy3 hydrazide, Cy5 hydrazide
Step 3
OH O
N
3¢-tRNA
NH2
N
H H
O
NaBH4
Proflavine HN
NH2
5¢-tRNA
H H
NH-HN −O
3
CH3
S
CH3 N O
CH3
SO− 3
CH3
Cy5 hydrazide
+ N CH3
NH-HN
Figure 4.2 The scheme of fluorophore labeling of a tRNA transcript. In step 1, a tRNA transcript is modified by Dus1p. In step 2, the Dus1p-modified tRNA transcript is reduced by NaBH4 to open the dihydrouracil ring to form a ureidopropanal group. In step 3, replacement of the ureidopropanal group with a fluorophore (R) bearing a primary amine, such as proflavin, rhodamine 110, or Cy3- and Cy5-hydrazides, which are shown in their chemical structures.
target tRNA from the crude mixture of total cellular tRNAs. Methods to purify a specific tRNA from a mixture have been described (Kothe et al., 2006), involving aminoacylation of the tRNA by its cognate aaRS, chemical modification of the aminoacyl group, HPLC separation of the modified aa-tRNA from all nonaminoacylated tRNA species, and removal of the aminoacyl group from the purified aa-tRNA by hydrolysis in mild base. Preferably, the target tRNA is overexpressed in E. coli or yeast under the control of a strong promoter. However, while the advantage of the method is that the target tRNA can be designed as either the wild-type or a mutant sequence, allowing investigation of a specific mutational effect, the limitation of the method is that it cannot adequately separate the mutant sequence from the endogenous wild-type sequence by aminoacylation or by
Fluorescent tRNA
77
chromatography. The only way to effectively address the mutant issue is by the third approach (Betteridge et al., 2007) (Fig. 4.2), which uses in vitro transcription to generate the target tRNA as a transcript possessing the wildtype or a mutant sequence. This target tRNA is then modified by a recombinant dihydrouridine synthase to introduce site-specific U ! D conversions (step 1 in Fig. 4.2), which provide the sites for NaBH4 reduction (step 2) and for subsequent fluorophore labeling by replacement of the reduced D residues (step 3). The details of the third approach are provided below.
3.1. Preparation of a target tRNA transcript The target tRNA is prepared as a transcript based on run-off transcription of the desired DNA template of a wild-type or mutant sequence. The DNA template can be constructed in a plasmid, which is linearlized by digestion with a restriction enzyme (Sampson et al., 1989), or it can be constructed by annealing of overlapping oligonucleotides. The template sequence is usually placed behind the promoter of the highly processive T7 RNA polymerase, because transcription by this enzyme gives the best yield for the range of tRNA lengths (70–90 nucleotides) (Milligan and Uhlenbeck, 1989). However, because T7 RNA polymerase initiates transcription only with G (guanosine, GMP, or GTP), tRNA sequences that do not start with G need to have a modification in their templates, which involves insertion of a hammerhead ribozyme sequence between the promoter and the first nucleotide of the tRNA gene (Fechter et al., 1998). Transcription of this construct gives rise to a fusion of the ribozyme attached to the target tRNA, where the self-cleaving activity of the ribozyme liberates the target tRNA with a 50 -OH end. Another consideration is the propensity of T7 RNA polymerase to generate heterogeneous 30 -ends, which is a problem for the aminoacylation reaction or reactions on the ribosome. This problem can be reduced, but not completely eliminated, by two modifications. One is to use the ‘‘foot’’ mutant of T7 RNA polymerase, which lacks the C-terminal F882-A883 residues of the wild-type enzyme and as a result exhibits reduced processivity (Mookhtiar et al., 1991). The other is to introduce two consecutive 20 -O-methyl backbone modifications to the 50 -terminus of the noncoding strand of the DNA template, which also function to reduce the processivity of T7 RNA polymerase (Kao et al., 1999). Notably, introduction of the 20 -O-methyl backbone modifications are not easily achieved with plasmid DNA templates, but are readily accommodated by chemical synthesis of oligonucleotides for construction of DNA templates. The synthesized tRNA transcripts are usually separated from DNA templates by denaturing 12% polyacylamide/7 M urea gels (12% PAGE/urea), identified by UV shadowing, and extracted from gel materials into the TE buffer (10 mM Tris–HCl, pH 8.0, 1 mM EDTA).
78
Cuiping Liu et al.
3.2. The U to D conversion in the target tRNA transcript Transcription by T7 RNA polymerase in vitro uses normal NTPs as the nucleotide substrates, and as such the tRNA transcript contains no base or backbone modifications. We have used the yeast dihydrouridine synthase enzyme Dus1p (Xing et al., 2004) to introduce U to D conversion specifically to positions 16 and 17 in several tRNA transcripts (Betteridge et al., 2007), including transcripts of E. coli tRNAPro/UGG (Fig. 4.1), tRNAfMet, tRNAPhe, and a mutant of tRNACys. Yeast has three other Dus enzymes (Xing et al., 2004), each with a positional specificity distinct from that of Dus1p: Dus2p for position 20, Dus3p for position 47, and Dus4p for positions 20a and 20b. Recombinant clones of these Dus enzymes encoding a C-terminal His tag have been produced and are well expressed in, and easily purified from, E. coli (Xing et al., 2004). The U to D conversion reaction catalyzed by the Dus enzymes utilizes FAD as a cofactor and NADH and NADPH as electron donors. 1. The target tRNA transcript (1 nmol) should be heat-cooled before use and dissolved in 20 mM for a 50 ml reaction with Dus1p. The reaction buffer consists of 100 mM Tris–HCl, pH 8.0, 100 mM NH4Ac, 5 mM MgCl2, 2 mM dithiothreitol (DTT), 0.1 mM EDTA, 1 mM b-NADH (Sigma N1161), 1 mM NADPH (Sigma N6505), 250 mM FAD (Sigma F6625), and purified Dus1p (5 mM). After incubation at 30 C for at least 40 min, the tRNA transcript is purified by phenol extraction and ethanol precipitation. 2. To quantify the number of U to D conversions, the Dus1p-modified tRNA transcript is dissolved in 100 ml of water at a series of concentrations, mixed with 10 ml of 1 M KOH, and incubated at 37 C for 30 min. This reaction is to cleave the heterocyclic D ring to generate an acyclic ureido group, which is quantified by the development of an intense reddish pink color upon reaction with FeCl3 in concentrated H2SO4. The colorimetric assay was developed by Jacobson and Hedgcoth (1970), which specifically monitors the ureido group derived from the D residues and thus is more accurate than simply monitoring the loss of U residues by the decrease of absorbance at 235 nm, although both methods have yielded identical results (Xing et al., 2002). 3. To prepare for the colorimetric assay, a saturated semidine solution is made by mixing 200 mg of semidine (N-phenyl-p-phenylenediamineHCl, Fluka #07920) with 10 ml of 95% ethanol, followed by dilution to 100 ml with double-distilled water (ddH2O). After filtering to remove the undissolved particles, the saturated semidine solution is stable at 4 C for up to 6 weeks. A 3% diacetyl monoxime solution is made by dissolving 3 g of 2,3-butanedione 2-oxime (Fluka #31550) in 100 ml
Fluorescent tRNA
79
of ddH2O or ethanol, and is stored at 4 C in a dark plastic bottle. Prior to the assay, mix one volume of the saturated semidine solution with one volume of the 3% diacetyl monoxime solution. The 1 mM FeCl3 solution is made by mixing 1 ml of 0.1 M FeCl3 with 99 ml of concentrated H2SO4. 4. To perform the colorimetric assay, the tRNA concentration series with the exposed ureido group from step 2 are each neutralized by adding 50 ml of concentrated H2SO4, followed by 100 ml of a 1:1 mixture (v/v) of saturated semidine and 3% diacetyl monoxime. Take caution to add H2SO4 directly into the center of the solution. Cap the tubes and incubate the series in a heat block at 95 C for 5 min in a fume hood and then at 50 C for 5 min. A solution of 100 ml of 1 mM FeCl3 in concentrated H2SO4 is added to react with the ureido group. After the reaction is cooled to room temperature, measure the absorption at 550 nm against a control sample without tRNA, which is prepared in parallel and used as a blank for the absorption reading. The linear relationship between the absorption and concentration is established by using dihydrouracil (Sigma, D7628) as a standard for the ureido group. Based on this assay, E. coli bulk tRNA is estimated to have 1.5 0.5 residues of D per tRNA, similar to the previously determined value of 1.4 0.1 ( Jacobson and Hedgcoth, 1970), while the Dus1p-modified transcript of E. coli tRNAPro/UGG is estimated to have 1.0 0.2 residues of D (Betteridge et al., 2007), indicating that only one of the three U residues (at positions 17, 17a, and 20) is converted. 5. To determine the site of the U to D conversion, the Dus1p-modified tRNA transcript (320 pmol) is denatured in 2 ml of 0.1 M KOH at 37 C for 20 min to convert the D residue to a ureido group, which is then mapped by primer extension. The denatured tRNA is neutralized with an equal volume of 5 annealing buffer (250 mM Tris–HCl, pH 8.3, 150 mM NaCl, 50 mM DTT) and is hybridized with a 32P-end labeled primer (25 pmol), which is designed with the 30 terminus placed with 4–5 nucleotides downstream from the anticipated site of the U to D conversion (e.g., Fig. 4.1A). Primer extension is catalyzed by AMV reverse transcriptase (2 units, Roche) for 30 min at 42 C, terminated by adding formamide to 65% and EDTA to 6.5 mM, and the products are analyzed by electrophoresis through a 12% PAGE/urea gel. This analysis was used to identify position G18 as the major stop site in primer extension of the Dus1p-modified transcript of E. coli tRNAPro/UGG (Betteridge et al., 2007) (Fig. 4.1C), indicating that U17a is the site of the U to D conversion. Notably, if the colorimetric assay indicates the presence of more than one U to D conversion, then the primer extension stop site represents the first conversion encountered from the 30 side.
80
Cuiping Liu et al.
3.3. Fluorescent labeling of D residues Both proflavin and rhodamine 110 have been incorporated into native or Dus1p-modified transcripts of tRNAs by replacement of D residues (Fig. 4.2, step3) as the fluorescent reporters for successfully monitoring reactions on the ribosome (Betteridge et al., 2007). In the case of the native yeast tRNAPhe, which contains D residues at positions 16 and 17, the labeling stoichiometry is 2.0 for proflavin and 1.0 for rhodamine. In the case of the Dus1p-modified transcript of E. coli tRNAPro/UGG, which harbors a single D at position 17a, the labeling stoichiometry is 1.0 for proflavin and 0.5 for rhodamine. The higher stoichiometry of proflavin incorporation is likely due to the higher nucleophilicity of the dye as compared with rhodamine, given the pKa value of 9.6 for proflavin and pKa of 4.3 for rhodamine. Also, the primary amine of proflavin is more accessible than that of rhodamine (Fig. 4.2). However, proflavin is easily photobleached, rendering proflavin-labeled tRNA unsuitable for single-molecule experiments in which fluorescent probes are subject to high light fluxes. In contrast, rhodamine is resistant to photobleaching and has higher emission intensity than proflavin, and as such is commonly used in single-molecule studies. The procedure below is applicable for both proflavin and rhodamine. To introduce proflavin or rhodamine, the target tRNA (1800 pmol in 40 mM Tris–HCl, pH 7.5) is subject to reduction of the D residue by NaBH4 (100 mg/ml in 10 mM KOH) in 20 ml at a final NaBH4 concentration of 10 mg/ml for 1 h at 25 C on a shaker in the dark. The reaction is terminated by gradually lowering the pH to 4–5 with 6 ml of 6 M acetic acid and the tRNA is ethanol precipitated three times to completely remove NaBH4. 1. The tRNA is resuspended in 5 ml ddH2O and mixed with 85 ml of 0.1 M sodium formate, pH 3.0, followed by 10 ml of a 20–22 mM solution of proflavin (Sigma Aldrich #131105) or rhodamine 111 (Fluka #83695) in a total volume of 100 ml. The fluorophore labeling reaction is incubated at 37 C for 45–60 min for proflavin and 90 min for rhodamine. Terminate the labeling reaction with addition of 2 M Tris–HCl, pH 8.5 (8 ml) to give a final pH of 7.5. The tRNA is extracted with watersaturated phenol (pH 4.3), ethanol precipitated, and resuspended in ddH2O. Repeat the phenol extraction three more times to completely remove the fluorescent dye from the labeling reaction. Adjust the pH of the tRNA to 7.5 if necessary. 2. Resuspend the tRNA in 10–20 ddH2O. Concentration is determined by absorption at 260 nm for tRNA, at 462 nm for proflavin, and at 512 nm for rhodamine. Labeling stoichiometry is determined from the molar ratio of proflavin or rhodamine to tRNA. 3. If the labeling stoichiometry is less than quantitative (such as in the case of rhodamine), the unlabeled fraction can be removed by an RNase H-mediated cleavage reaction (Hou et al., 2006). For example, after the labeling reaction with rhodamine, the transcript of E. coli tRNAPro/UGG
Fluorescent tRNA
81
(880 pmol) is hybridized to an oligonucleotide probe (1000 pmol) complementary to the site of the label in 40 ml of the RNase H buffer. This oligonucleotide probe contains 20 -O-methyl backbones except for the four central nucleotides with 20 -deoxy backbones, which flank the RHase H cleavage site. Because the unlabeled transcript is accessible to hybridization with the probe whereas the labeled transcript is not, subsequent RNase H digestion (at 50 mM, 37 C, 1 h) selectively cleaves the unlabeled transcript. After separation from the cleaved transcript by a denaturing PAGE, the fluorophore-labeled transcript is extracted from the gel, ethanol precipitated, and resuspended in the desired buffer for characterization of the emission spectrum and intensity. Note that the efficiency of the RNase H cleavage reaction depends on the tRNA hybridization with the oligonucleotide, which is sequence-specific. If the hybridization efficiency in the RNase H buffer is poor, we recommend that removal of Mgþ 2 ion from the RNase H buffer can improve the efficiency.
3.4. Experimental considerations In addition to proflavin and rhodamine, the photobleaching-resistant Cy3 and Cy5 fluorophores are also frequently used in single-molecule experiments and have been incorporated in the form of hydrazide derivatives into tRNAs via D residues (Pan et al., 2009) (Fig. 4.2). However, quantitative uptake of these hydrazide dyes requires modification of three reaction parameters: higher concentrations of the hydrazide dyes (40 mM) than that required for proflavin or rhodamine (22 mM), pH 3.7 rather than pH 3.0, and 2 h reaction time instead of 45–90 min. The requirement of higher concentration is to promote formation of hydrazide adduct, while the slightly elevated pH prevents hydrolysis of the adduct, which is acid labile. Thus, while the labeling method can be adapted to incorporate new fluorophores besides proflavin and rhodamine, it is prudent to systematically evaluate for the fluorophores under consideration for coupling efficiency as a function of dye concentration, pH, and reaction time. For tRNAs that are modified with more than one D residue, the labeling procedure may not be quantitative, leading to heterogeneous fluorophore labeling in more than one position. If this heterogeneity is problematic, it may require the development of high-resolution HPLC chromatography to resolve the different structural isomers. HPLC chromatography offers the further prospect of separating charged aa-tRNA from its uncharged counterpart. This separation is without a doubt the most challenging problem, because the two forms of tRNA differ by only the aminoacyl group. If the aminoacyl group is hydrophobic, separation can be accomplished using a reverse phase (RP)-HPLC column. For example, separation of phe-tRNAPhe from tRNAPhe has been
82
Cuiping Liu et al.
successfully achieved by using a phenyl-based RP-HPLC column (4.6 250 mm, Alltech) with a gradient of 0–24% methanol in 20 mM NH4Ac, pH 5.5, and 50 mM NaCl (Pan et al., 2009). A key consideration is that formation of the charged aa-tRNA and its separation from the uncharged counterpart should be performed after the NaBH4 reduction step, but before the fluorophore-labeling step. This consideration is based on concerns that the base-labile aminoacyl group could be hydrolyzed during the pH 7.5 NaBH4 reduction step and that charging by aaRS enzyme could be inhibited by the presence of a bulky fluorophore in the tertiary core. The latter concern is particularly relevant for aaRS enzymes that make contact with the tertiary core of their cognate tRNA during aminoacylation, such as PheRS (Goldgur et al., 1997), ArgRS (Delagoutte et al., 2000), SerRS (Biou et al., 1994), ValRS (Fukai et al., 2000), E. coli CysRS (Hou et al., 1993), and ProRS (Liu et al., 1995), but may be less of an issue for E. coli MetRS (Pan et al., 2009). The step-wise labeling procedure of an aa-tRNA, starting with NaBH4 reduction of the tRNA, followed by aminoacylation and purification of the aa-tRNA, and ending with fluorophore labeling of the purified aa-tRNA, has been established for the native yeast phetRNAPhe. The procedure incorporated 1.2–1.3 molecules of Cy3 (Pan et al., 2009) to replace the D residues at position 16 or 17. The final product of Cy3-labeled yeast phe-tRNAPhe can be further purified by anion exchange on an FPLC monoQ column to generate a high-quality probe that has proved useful for several reactions on the ribosome (Pan et al., 2009). It should be noted that many tRNA dynamics experiments do not require the aa-tRNA form, such as those investigating tRNA interactions with processing, modification, and aaRS enzymes. In these cases, the fluorophore-labeled tRNA can be used directly. Even for reactions on the ribosome that require the aa-tRNA form, separation of the aa-tRNA from the uncharged counterpart is not strictly necessary. This is because the entry of an aa-tRNA to the ribosome is controlled by EF-Tu selection, which discriminates against noncharged tRNA by at least 100-fold in binding affinity ( Janiak et al., 1990). Indeed, the exceptional selectivity of EF-Tu has been previously exploited as a method to purify a charged aa-tRNA (Derwenskus et al., 1984), and this selectivity may be further exploited to purify a fluorescent labeled aa-tRNA from its uncharged counterpart for biochemical studies.
4. Fluorescent Labeling of the CCA Sequence The CCA sequence is conserved at the tRNA 30 end as a single stranded motif with a wide range of dynamic conformations that are essential for tRNA functions. The CCA sequence is synthesized and
Fluorescent tRNA
83
maintained by step-wise nucleotide addition reactions of the CCA enzyme, (Deutscher, 1982). The development of a molecular probe for the CCA sequence is obviously highly desirable, particularly for reactions involving tRNA aminoacylation and tRNA–ribosome interactions. In an earlier attempt (Ott et al., 1989), transcripts of tRNATyr and tRNAPhe were synthesized up to C74 and were extended with 2-thiocytidine (s2C) at position 75 and the normal A76 by the yeast CCA enzyme. The s2C analog was then alkylated with the fluorophore N-iodoacetyl-N0 -(5-sulfo-1naphthyl) ethylenediamine (1,5-I-AEDANS), which however turned out to be too bulky for aminoacylation by aaRS enzymes, preventing application of this probe to monitor reactions that require charged aa-tRNA. Recent work has identified two alternative and less bulky fluorophores for the CCA sequence. One is the pyrrolo C (PyC) analog of CTP for incorporation to position 75 (Fig. 4.3A) (Zhang et al., 2008a), while the other is the 2-aminopurine (2AP) analog of ATP for incorporation to position 76 (Fig. 4.4A) (Ling et al., 2009). Both probes are incorporated into tRNA transcripts, rather than native tRNAs, taking advantage of the ease with which to generate an appropriate terminus in tRNA transcripts and the flexibility with which to study sequence variants. Two versions of C-terminal His-tagged recombinant CCA enzyme have been successfully used for incorporation of fluorophores, one encoded by Bacillus stearothermophilus (Bst) (reaction temperature at 60 C) (Cho et al., 2003) and the other encoded by E. coli (37 C) (Dupasquier et al., 2008). The reactions described below can be accomplished with either enzyme with similar efficiencies.
4.1. Incorporation of PyC to position 75 PyCTP (Glen Research) maintains the imine N3 and exocyclic N4 of the normal CTP as the H-bond donor and acceptor, respectively. PyC has been shown to be an excellent probe for protein–nucleic acid interactions (Berry et al., 2004; Tinsley and Walter, 2006), because it can be selectively excited at 350 nm, well separated from the abosorbance at 260 nm of natural nucleotides, and has an emission maximum of 460 nm, far removed from that of protein tryptophans (emission ¼ 340 nm). In addition, the fluorescence emission of PyC is sensitive to local environment, which has been exploited to study dynamics and stabilities of DNA helices, RNA duplexes, and DNA/RNA hybrids (Liu and Martin, 2002; Tinsley and Walter, 2006). We have incorporated PyC into at least two different tRNA sequence frameworks as a molecular probe for the activity and dynamics of the CCA sequence (Zhang et al., 2008a,b). For example, E. coli tRNACys harboring PyC at position 75 is a functional substrate for aminoacylation and has been used to monitor the binding interaction with the cognate CysRS, reporting a Kd value (2.2 0.2 mM) (Zhang et al., 2008a) similar to
84
Cuiping Liu et al.
A HN
N PyC O
−O
P
P
P
−O
O
−O
N
O
O
O
O
−O
O
O H
H
OH
H OH
H PyCTP
A76 PyC75 C74
B C74 5¢
5¢ CCA enzyme PyCTP, ATP
tRNA-C74PyC75A76
tRNA-C74 C
32P-tRNA-C
ATP (1 mM) CTP (1 mM) PyCTP (0.5 mM)
74
− − −
+ − −
− + −
+ + −
− − +
+ − +
1
2
3
4
5
6
tRNA-A76 tRNA-C74
Figure 4.3 Incorporation of PyC to position 75 in a tRNA transcript by CCA enzyme. (A) Chemical structure of PyCTP, where inward and outward arrows denote hydrogen bond acceptors and donors, respectively. (B) A scheme showing the cloverleaf of a tRNA-C74 transcript as the substrate for CCA addition, using PyCTP and ATP as the nucleotide donors. (C) Denaturing gel analysis (12% PAGE/7 M urea) of extension of the tRNA-C74 transcript by CCA enzyme, showing the tRNA transcript in lane 1, lack of extension of the transcript without CTP or PyCTP in lane 2, extension with polyC in lane 3, extension with C75–A76 in lane 4, extension with polyPyC in lane 5, and extension with PyC75–A76 in lane 6 (Zhang et al., 2008a).
that obtained by monitoring the intrinsic fluorescence of the enzyme (1.9 0.1 mM) (Hauenstein et al., 2004). The charged cys-tRNACys harboring PyC at position 75 has been used to monitor the tRNA translocation step within the ribosome (Zhang et al., 2008a), reporting the kinetics for the
A
B
N
N
5¢
2AP
−O
O
O
O H2N
P
P
P
−O
O
−O
O
−O
2AP76 C75
C75 5¢
N
N
CCA enzyme
O
O
2APTP
H
H
OH
H OH
H
tRNA-C75
2APTP C
tRNA-C752AP76
6.5
D ATP
tRNA-A76
30
0
10
20
30
45
60
90
120 180
Fluorescence
Time (min)
2APTP 6.0 5.5 Kd = 12 ± 3 nM
5.0
tRNA-C75 4.5
0
200 100 EF-Tu (nM)
300
Figure 4.4 Incorporation of 2AP to position 76 in a tRNA transcript by CCA enzyme. (A) Chemical structure of 2APTP, where inward and outward arrows denote hydrogen bond acceptors and donors, respectively. (B) A scheme showing the cloverleaf of a tRNA-C75 transcript as the substrate for CCA addition, using 2APTP as the nucleotide donor. (C) A denaturing gel analysis by 12% PAGE/7 M urea of 2AP incorporation to a 32P-labeled tRNA transcript in a time course, showing complete incorporation in 45–60 min under the reaction conditions. (D) An equilibrium binding titration of 2AP-labeled E. coli cys-tRNACys with preactivated E. coli EF-Tu in binding buffer (50 mM HEPES pH 7.5, 50 mM KCl, 100 mM NH4Cl, 10 mM MgCl2, 5 mM DTT, 1.5 mM GTP), showing a Kd of 12 3 nM.
86
Cuiping Liu et al.
formation and decay of a translocation intermediate (19.0 1.9 s 1, 3.4 2.0 s 1) similar to the kinetics reported by monitoring the proflavin probe incorporated to replace the D residues in the native yeast tRNAPhe (9 1 s 1, 3.4 0.3 s 1) (Pan et al., 2007). These results demonstrate the utility of PyC as a versatile probe for the dynamics of the CCA sequence in reactions ranging from tRNA aminoacylation to those on the ribosome. 1. The target tRNA should be prepared as a transcript up to C74 so as to allow incorporation of only one PyC into the CCA sequence (Fig. 4.3B). An analytical scale of PyC incorporation should be examined to establish the conditions prior to conducting reactions of larger scales. The analytical reaction uses tRNA at 10 mM with a trace amount of 32P-labeled tRNA as the reaction marker, which is prepared by bodylabel of the tRNA in the transcription reaction. The incorporation of PyC is initiated by addition of 0.1–0.5 mM Bst CCA enzyme to a reaction buffer containing 100 mM glycine–NaOH, pH 9.0, 10 mM MgCl2, 1 mM DTT, 1 mM ATP, and 1 mM PyCTP. After incubation at 60 C for 30 min, the target tRNA is analyzed by a denaturing sequencing gel (12% PAGE/7 M urea) in TBE buffer (90 mM Tris–HCl, pH 8.3, 90 mM boric acid, and 2 mM EDTA) at 1800 V for 4 h at 60 C to achieve single-nucleotide resolution. The gel is dried and visualized by a phosphorimager. 2. The incorporation of PyC is detected as extension of the target tRNA from C74 to A76 in a PyC-dependent manner (Fig. 4.3C). Notably, because of the aberrant polyC activity of the CCA enzyme (Hou, 2000; Seth et al., 2002), which adds Cs to both positions 75 and 76 in the absence of ATP (lane 3 in Fig. 4.3C), incorporation of multiple PyC nucleotides is possible (lane 5) and should be avoided by inclusion of a saturating concentration of ATP (1 mM) in the extension reaction (lane 6). 3. Upon extension of the target tRNA to A76, indicating successful incorporation of PyC to position 75, the reaction is scaled up with proportional increases in tRNA and enzyme. The reaction product tRNA-C74PyC75A76 is separated from the substrate tRNA-C74 by RP-HPLC on a C8 column (Alltech Alltima, 0.5 25 cm, pore diam˚ , particle diameter 3 mM) with a 0–60% linear gradient of eter 300 A methanol in 20 mM NH4Ac, pH 5.5, 50 mM NaCl. In the extension reaction of the transcript of E. coli tRNACys, the un-incorporated nucleotides PyCTP and ATP are eluted at 5–10% methanol (v/v), the substrate tRNA-C74 is eluted at 20%, and the product tRNAC74PyC75A76 is eluted at 30% (Zhang et al., 2008a). Separation of the tRNA substrate from product is facilitated by the increased hydrophobilicty of the product due to extension of two nucleotides, although
Fluorescent tRNA
87
different tRNA species may display a different elution profile. The tRNA-C74PyC75A76 product peak is ethanol precipitated and should be tested for aminoacylation as an indicator for the presence of a complete CCA sequence. We have digested the product peak of E. coli tRNACys to mononucleosides by nuclease P1 and shrimp alkaline phosphatase and shown that it indeed contains PyC by mass spectrometry (Zhang et al., 2008a). 4. The target tRNA after PyC incorporation is phenol extracted, ethanol precipitated, and used directly for aminoacylation, followed by the same RP-HPLC chromatography conditions to purify the aa-tRNA. In the case of the transcript of E. coli tRNACys, the aa-tRNA peak elutes at 15% methanol (v/v), ahead of the tRNA-C74 substrate peak and the tRNAC74PyC75A76 product peak (Zhang et al., 2008a). The identity of the aa-tRNA peak can be confirmed if the peak coelutes with the radioactivity of the amino acid that is used in the aminoacylation reaction. Notably, each PyC-containing tRNA or aa-tRNA may exhibit a distinct RP-HPLC chromatography, which should be determined by empirical analysis to sample various RP-HPLC columns of increasing hydrophobicity.
4.2. Experimental considerations We focus on position 75 to limit PyC incorporation to one well-defined location. In certain cases, it may be beneficial to incorporate PyC to both positions C74 and C75 in order to enhance the fluorescence emission signal of the probe. This is achieved by using a tRNA primer terminated at position 73 and by extending the primer with PyCTP and ATP as the nucleotide substrates. In the absence of normal CTP, complete extension of the primer from position 74 to 76, as visualized by denaturing gel analysis (e.g., Fig. 4.3C), is an indication that PyC has been incorporated at both positions. However, due to the rapid reaction of the CCA enzyme to synthesize consecutive C74 and C75 (Dupasquier et al., 2008), it would be difficult to incorporate PyC to just position 74. As a comparison, fluorescent labeling of tRNA with PyC is achieved in one step by the CCA enzyme, and thus is conceptually and technically simpler than labeling of tRNA with proflavin, rhodamine, or Cy3 and Cy5hydrazides via D residues. However, the fluorescence emission intensity of PyC is not as high as those of the other fluorophores and thus may not be suitable for single-molecule experiments. Nonetheless, enzymatic labeling of tRNA with PyC is easy to implement and should be applicable to all tRNA sequences (both wild type and mutants), which can be generated by in vitro transcription without the requirement for a specific modification or for native tRNA species.
88
Cuiping Liu et al.
4.3. Incorporation of 2AP to position 76 The 2AP base has been widely used as an effective fluorophore (excitation at 315 nm and emission maximum at 360 nm) to probe changes in basestacking interactions in nucleic acid structure and dynamics. A high level of fluorescence intensity is observed for the free 2AP nucleotide, but its intensity is quenched when 2AP resides in single- or double-stranded DNA or RNA. Enzyme interactions with 2AP-labeled nucleic acids that disrupt 2AP base-stacking interactions, however, can produce large increases in fluorescence intensity (Mandal et al., 2002). Recent work has incorporated 2AP to position 76 of tRNA (Ling et al., 2009), replacing the normal A76, as a probe to monitor the dynamics of the CCA sequence in the binding interaction of phe-tRNAPhe with EF-Tu. Because position 76 is the site to which the specific amino acid is attached in each tRNA, the 2AP labeling of this position is particularly useful to study tRNA interactions with proteins that focus on the CCA end, such as the aaRS enzymes and the ribosome peptidyl transferase center that catalyzes peptide bond formation of the aminoacyl group. The published work of incorporation of 2AP to tRNA involved synthesis of a tetranucleotide RNA fragment A73CC2AP76, representing the last four nucleotides in tRNAPhe, and joining of this fragment by T4 RNA ligase to a transcript of the remaining tRNA sequence (Ling et al., 2009). Because chemical synthesis of RNA fragments, containing 2AP, can be costly, we have developed a more costeffective approach for incorporating 2AP to position 76, using the CCA enzyme. Key details for this alternative approach are provided below. 1. The target tRNA is prepared as a transcript up to C75 as a primer for incorporation of 2AP to position 76 (Fig. 4.4B). An analytical scale of 2AP incorporation should be established prior to a scaled-up reaction. The analytical reaction uses the target tRNA at 10 mM and a trace amount of 32P-labeled target tRNA as the reaction marker. The incorporation reaction consists of 100 mM glycine–NaOH, pH 9.0, 10 mM MgCl2, 3 mM DTT, 2 mM 2AP triphosphate (TriLink Biotechnologies, CA), and is initiated by addition of 1.0 mM E. coli CCA enzyme. After incubation at 37 C for 60–90 min, the target tRNA is analyzed by a denaturing sequencing gel (12% PAGE/7 M urea) run in TBE buffer at 1800 V for 4 h at 60 C to achieve single-nucleotide resolution. The gel is dried and visualized by a phosphorimager. 2. The incorporation of 2AP is detected as a one nucleotide extension of the target tRNA from C75 to 2AP76 and is quantitative under the experimental conditions (Fig. 4.4C). Upon confirmation of the 2AP incorporation, the reaction can be scaled up with proportional increases in tRNA and enzyme. The labeling stoichiometry is determined by the ratio of absorbance at 315 nm for 2AP over the absorbance at 260 nm for tRNA.
Fluorescent tRNA
89
4.4. Experimental considerations The stoichiometric labeling of 2AP suggests that the product of the labeling reaction can be directly used for experiments. For example, we have shown that the 2AP-labeled transcript of E. coli tRNACys is aminoacylated to 12% as compared to 20% for the unlabeled transcript. The reduced plateau of aminoacylation of the labeled transcript is expected, because position 76 in the tRNA is the site of aminoacylation and the terminal base must make direct contact with the cognate CysRS for proper positioning of the reaction active site. However, despite the incomplete aminoacylation level, we have used the mixture (containing the 2AP-labeled form of both cys-tRNACys and tRNACys) to determine the equilibrium affinity for the binding interaction between cys-tRNACys and EF-Tu. The experiment determined the Kd of the binding interaction as 12.2 0.2 nM (Fig. 4.4D), which is well within the range of the reported Kd values of 10–20 nM of EF-Tu binding to aa-tRNAs (Sanderson and Uhlenbeck, 2007). The accurate determination of Kd was feasible, because of the effective affinity discrimination by EF-Tu against the uncharged tRNA ( Janiak et al., 1990). In principle, this rationale can be applied to study the dynamics of aa-tRNA in binding interactions with EF-Tu and with the ribosome during protein synthesis.
5. Conclusions Dynamics of tRNA motions is of fundamental importance for tRNA functions. This chapter describes fluorophore labeling to monitor two of the most prominent regions of tRNA dynamics: the D residues in the tertiary core region and the CCA sequence at the 30 end. The common theme of these methods is the emphasis on the utility of tRNA transcripts as the substrates for labeling in order to have the flexibility to work with both wild-type and mutant sequences. Comparison of tRNA dynamics for a pair of wild-type and mutant sequences will provide the necessary basis to elucidate the sequence determinants for the dynamics of a specific reaction. The methods described here will eventually allow elucidation of how tRNA dynamics responds to binding signals and how different signals are regulated and integrated together to control the specificity and kinetics of tRNA reactions. Such information is limited at present but will likely expand substantially with the easy accessibility of these methods and with continued development of new methods to incorporate additional fluorophores to tRNA.
90
Cuiping Liu et al.
ACKNOWLEDGMENTS We thank Prof. Barry S. Cooperman of University of Pennsylvania and members of the Cooperman lab and the Hou lab for their contribution in developing the protocols described here. We also thank Prof. Howard Gamper of Thomas Jefferson University for carefully reading the manuscript. This work was supported by grants from NIH GM68561 and GM81601 to YMH.
REFERENCES Berry, D., Jung, K., Wise, D., Sercel, A., Pearson, W., Mackie, H., Randolph, J., and Somers, R. (2004). Pyrrolo-dC and pyrrolo-C: Fluorescent analogs of cytidine and 2’-deoxycytidine for the study of oligonucleotides. Tetrahedron Lett. 45, 2457–2461. Betteridge, T., Liu, H., Gamper, H., Kirillov, S., Cooperman, B. S., and Hou, Y. M. (2007). Fluorescent labeling of tRNAs for dynamics experiments. RNA 13, 1594–1601. Biou, V., Yaremchuk, A., Tukalo, M., and Cusack, S. (1994). The 2.9 A crystal structure of T. thermophilus seryl-tRNA synthetase complexed with tRNA(Ser). Science 263, 1404–1410. 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. Bouadloun, F., Donner, D., and Kurland, C. G. (1983). Codon-specific missense errors in vivo. EMBO J. 2, 1351–1356. Cerutti, P., and Miller, N. (1967). Selective reduction of yeast transfer ribonucleic acid with sodium borohydride. J. Mol. Biol. 26, 55–66. Cho, H. D., Oyelere, A. K., Strobel, S. A., and Weiner, A. M. (2003). Use of nucleotide analogs by class I and class II CCA-adding enzymes (tRNA nucleotidyltransferase): Deciphering the basis for nucleotide selection. RNA 9, 970–981. Cochella, L., and Green, R. (2005). An active role for tRNA in decoding beyond codon: anticodon pairing. Science 308, 1178–1180. Cochella, L., Brunelle, J. L., and Green, R. (2007). Mutational analysis reveals two independent molecular requirements during transfer RNA selection on the ribosome. Nat. Struct. Mol. Biol. 14, 30–36. Delagoutte, B., Moras, D., and Cavarelli, J. (2000). tRNA aminoacylation by arginyl-tRNA synthetase: Induced conformations during substrates binding. EMBO J. 19, 5599–5610. Dennis, P. P., Ehrenberg, M., and Bremer, H. (2004). Control of rRNA synthesis in Escherichia coli: A systems biology approach. Microbiol. Mol. Biol. Rev. 68, 639–668. Derwenskus, K. H., Fischer, W., and Sprinzl, M. (1984). Isolation of tRNA isoacceptors by affinity chromatography on immobilized bacterial elongation factor Tu. Anal. Biochem. 136, 161–167. Deutscher, M. P. (1982). tRNA nucleotidyltransferase. Enzymes 15, 183–215. Dupasquier, M., Kim, S., Halkidis, K., Gamper, H., and Hou, Y. M. (2008). tRNA integrity is a prerequisite for rapid CCA addition: Implication for quality control. J. Mol. Biol. 379, 579–588. Fechter, P., Rudinger, J., Giege, R., and Theobald-Dietrich, A. (1998). Ribozyme processed tRNA transcripts with unfriendly internal promoter for T7 RNA polymerase: Production and activity. FEBS Lett. 436, 99–103.
Fluorescent tRNA
91
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. Fukai, S., Nureki, O., Sekine, S., Shimada, A., Tao, J., Vassylyev, D. G., and Yokoyama, S. (2000). Structural basis for double-sieve discrimination of L-valine from L-isoleucine and L-threonine by the complex of tRNA(Val) and valyl-tRNA synthetase. Cell 103, 793–803. Goldgur, Y., Mosyak, L., Reshetnikova, L., Ankilova, V., Lavrik, O., Khodyreva, S., and Safro, M. (1997). The crystal structure of phenylalanyl-tRNA synthetase from thermus thermophilus complexed with cognate tRNAPhe. Structure 5, 59–68. Hamann, C. S., and Hou, Y. M. (1997). An RNA structural determinant for tRNA recognition. Biochemistry 36, 7967–7972. Hamann, C. S., and Hou, Y. M. (2000). Probing a tRNA core that contributes to aminoacylation. J. Mol. Biol. 295, 777–789. Hauenstein, S., Zhang, C. M., Hou, Y. M., and Perona, J. J. (2004). Shape-selective RNA recognition by cysteinyl-tRNA synthetase. Nat. Struct. Mol. Biol. 11, 1134–1141. Hirsh, D. (1971). Tryptophan transfer RNA as the UGA suppressor. J. Mol. Biol. 58, 439–458. Hou, Y. M. (1994). Structural elements that contribute to an unusual tertiary interaction in a transfer RNA. Biochemistry 33, 4677–4681. Hou, Y. M. (2000). Unusual synthesis by the Escherichia coli CCA-adding enzyme. RNA 6, 1031–1043. Hou, Y. M., Westhof, E., and Giege, R. (1993). An unusual RNA tertiary interaction has a role for the specific aminoacylation of a transfer RNA. Proc. Natl. Acad. Sci. USA 90, 6776–6780. Hou, Y. M., Li, Z., and Gamper, H. (2006). Isolation of a site-specifically modified RNA from an unmodified transcript. Nucleic Acids Res. 34, e21. Jacobson, M., and Hedgcoth, C. (1970). Determination of 5,6-dihydrouridine in ribonucleic acid. Anal. Biochem. 34, 459–469. Janiak, F., Dell, V. A., Abrahamson, J. K., Watson, B. S., Miller, D. L., and Johnson, A. E. (1990). Fluorescence characterization of the interaction of various transfer RNA species with elongation factor Tu.GTP: Evidence for a new functional role for elongation factor Tu in protein biosynthesis. Biochemistry 29, 4268–4277. Kao, C., Zheng, M., and Rudisser, S. (1999). A simple and efficient method to reduce nontemplated nucleotide addition at the 3 terminus of RNAs transcribed by T7 RNA polymerase. RNA 5, 1268–1272. Kim, S. H., Sussman, J. L., Suddath, F. L., Quigley, G. J., McPherson, A., Wang, A. H., Seeman, N. C., and Rich, A. (1974). The general structure of transfer RNA molecules. Proc. Natl. Acad. Sci. USA 71, 4970–4974. Komatsoulis, G. A., and Abelson, J. (1993). Recognition of tRNA(Cys) by Escherichia coli cysteinyl-tRNA synthetase [published erratum appears in Biochemistry (1993), 32(48), 13374]. Biochemistry 32, 7435–7444. Korostelev, A., Trakhanov, S., Laurberg, M., and Noller, H. F. (2006). Crystal structure of a 70 S ribosome-tRNA complex reveals functional interactions and rearrangements. Cell 126, 1065–1077. Kothe, U., Paleskava, A., Konevega, A. L., and Rodnina, M. V. (2006). Single-step purification of specific tRNAs by hydrophobic tagging. Anal. Biochem. 356, 148–150. Kramer, G. F., Baker, J. C., and Ames, B. N. (1988). Near-UV stress in Salmonella typhimurium: 4-Thiouridine in tRNA, ppGpp, and ApppGpp as components of an adaptive response. J. Bacteriol. 170, 2344–2351. Ledoux, S., Olejniczak, M., and Uhlenbeck, O. C. (2009). A sequence element that tunes Escherichia coli tRNA(Ala)(GGC) to ensure accurate decoding. Nat. Struct. Mol. Biol. 16, 359–364.
92
Cuiping Liu et al.
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. Ling, J., So, B. R., Yadavalli, S. S., Roy, H., Shoji, S., Fredrick, K., Musier-Forsyth, K., and Ibba, M. (2009). Resampling and editing of mischarged tRNA prior to translation elongation. Mol. Cell 33, 654–660. Liu, C., and Martin, C. T. (2002). Promoter clearance by T7 RNA polymerase. Initial bubble collapse and transcript dissociation monitored by base analog fluorescence. J. Biol. Chem. 277, 2725–2731. Liu, H., Peterson, R., Kessler, J., and Musier-Forsyth, K. (1995). Molecular recognition of tRNA(Pro) by Escherichia coli proline tRNA synthetase in vitro. Nucleic Acids Res. 23, 165–169. Loftfield, R. B., and Vanderjagt, D. (1972). The frequency of errors in protein biosynthesis. Biochem. J. 128, 1353–1356. Mandal, S. S., Fidalgo da Silva, E., and Reha-Krantz, L. J. (2002). Using 2-aminopurine fluorescence to detect base unstacking in the template strand during nucleotide incorporation by the bacteriophage T4 DNA polymerase. Biochemistry 41, 4399–4406. McIntosh, B., Ramachandiran, V., Kramer, G., and Hardesty, B. (2000). Initiation of protein synthesis with fluorophore-Met-tRNA(f) and the involvement of IF-2. Biochimie 82, 167–174. Milligan, J. F., and Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods Enzymol. 180, 51–62. Mookhtiar, K. A., Peluso, P. S., Muller, D. K., Dunn, J. J., and Coleman, J. E. (1991). Processivity of T7 RNA polymerase requires the C-terminal Phe882-Ala883-COO- or ‘‘foot’’ Biochemistry 30, 6305–6313. 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. Murakami, H., Ohta, A., and Suga, H. (2009). Bases in the anticodon loop of tRNA(Ala) (GGC) prevent misreading. Nat. Struct. Mol. Biol. 16, 353–358. Ott, G., Faulhammer, H. G., and Sprinzl, M. (1989). Interaction of elongation factor Tu from Escherichia coli with aminoacyl-tRNA carrying a fluorescent reporter group on the 30 terminus. Eur. J. Biochem. 184, 345–352. Pan, D., Kirillov, S. V., and Cooperman, B. S. (2007). Kinetically competent intermediates in the translocation step of protein synthesis. Mol. Cell 25, 519–529. Pan, D., Qin, H., and Cooperman, B. S. (2009). Synthesis and functional activity of tRNAs labeled with fluorescent hydrazides in the D-loop. RNA 15, 346–354. Paulsen, H., and Wintermeyer, W. (1986). tRNA topography during translocation: Steady-state and kinetic fluorescence energy-transfer studies. Biochemistry 25, 2749–2756. 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. Robertson, J. M., Paulsen, H., and Wintermeyer, W. (1986). Pre-steady-state kinetics of ribosomal translocation. J. Mol. Biol. 192, 351–360. Rodnina, M. V., Fricke, R., and Wintermeyer, W. (1994). Transient conformational states of aminoacyl-tRNA during ribosome binding catalyzed by elongation factor Tu. Biochemistry 33, 12267–12275. Rodnina, M. V., Savelsbergh, A., Katunin, V. I., and Wintermeyer, W. (1997). Hydrolysis of GTP by elongation factor G drives tRNA movement on the ribosome. Nature 385, 37–41. Sampson, J. R., DiRenzo, A. B., Behlen, L. S., and Uhlenbeck, O. C. (1989). Nucleotides in yeast tRNAPhe required for the specific recognition by its cognate synthetase. Science 243, 1363–1366.
Fluorescent tRNA
93
Sanderson, L. E., and Uhlenbeck, O. C. (2007). Exploring the specificity of bacterial elongation factor Tu for different tRNAs. Biochemistry 46, 6194–6200. Schleich, H. G., Wintermeyer, W., and Zachau, H. G. (1978). Replacement of wybutine by hydrazines and its effect on the active conformation of yeast tRNAPhe. Nucleic Acids Res. 5, 1701–1713. Selmer, M., Dunham, C. M., Murphy, F. V. IV, Weixlbaumer, A., Petry, S., Kelley, A. C., Weir, J. R., and Ramakrishnan, V. (2006). Structure of the 70 S ribosome complexed with mRNA and tRNA. Science 313, 1935–1942. Seth, M., Thurlow, D. L., and Hou, Y. M. (2002). Poly(C) synthesis by class I and class II CCA-adding enzymes. Biochemistry 41, 4521–4532. Tinsley, R. A., and Walter, N. G. (2006). Pyrrolo-C as a fluorescent probe for monitoring RNA secondary structure formation. RNA 12, 522–529. Tomita, K., Ishitani, R., Fukai, S., and Nureki, O. (2006). Complete crystallographic analysis of the dynamics of CCA sequence addition. Nature 443, 956–960. Wimberly, B. T., Brodersen, D. E., Clemons, W. M. Jr., Morgan-Warren, R. J., Carter, A. P., Vonrhein, C., Hartsch, T., and Ramakrishnan, V. (2000). Structure of the 30 S ribosomal subunit. Nature 407, 327–339. Wintermeyer, W., and Zachau, H. G. (1974). Replacement of odd bases in tRNA by fluorescent dyes. Methods Enzymol. 29, 667–673. Wintermeyer, W., and Zachau, H. G. (1979). Fluorescent derivatives of yeast tRNAPhe. Eur. J. Biochem. 98, 465–475. Woolhead, C. A., McCormick, P. J., and Johnson, A. E. (2004). Nascent membrane and secretory proteins differ in FRET-detected folding far inside the ribosome and in their exposure to ribosomal proteins. Cell 116, 725–736. Xing, F., Martzen, M. R., and Phizicky, E. M. (2002). A conserved family of Saccharomyces cerevisiae synthases effects dihydrouridine modification of tRNA. RNA 8, 370–381. Xing, F., Hiley, S. L., Hughes, T. R., and Phizicky, E. M. (2004). The specificities of four yeast dihydrouridine synthases for cytoplasmic tRNAs. J. Biol. Chem. 279, 17850–17860. Xiong, Y., and Steitz, T. A. (2004). Mechanism of transfer RNA maturation by CCA-adding enzyme without using an oligonucleotide template. Nature 430, 640–645. Zhang, C. M., Liu, C., Christian, T., Gamper, H., Rozenski, J., Pan, D., Randolph, J. B., Wickstrom, E., Cooperman, B. S., and Hou, Y. M. (2008a). Pyrrolo-C as a molecular probe for monitoring conformations of the tRNA 30 end. RNA 14, 2245–2253. Zhang, C. M., Liu, C., Slater, S., and Hou, Y. M. (2008b). Aminoacylation of tRNA with phosphoserine for synthesis of cysteinyl-tRNA(Cys). Nat. Struct. Mol. Biol. 15, 507–514.
C H A P T E R
F I V E
Use of Deoxyribozymes in RNA Research Scott K. Silverman* and Dana A. Baum† Contents 1. Introduction 2. Deoxyribozymes for RNA Cleavage 2.1. Deoxyribozymes available for different RNA cleavage-site sequences 2.2. Experimental procedures 2.3. Further efforts needed to develop deoxyribozymes for RNA cleavage 3. Deoxyribozymes for RNA Ligation: Synthesis of Linear RNA Products 3.1. Deoxyribozymes available for 30 –50 RNA ligation 3.2. Deoxyribozymes available for 20 –50 RNA ligation 3.3. Experimental procedures 3.4. Further efforts needed to develop deoxyribozymes for linear RNA ligation 4. Deoxyribozymes for RNA Ligation: Synthesis of Branched RNA Products 4.1. Deoxyribozymes available for 20 ,50 -branched and lariat RNA synthesis 4.2. Experimental procedures 4.3. Further efforts needed to develop deoxyribozymes for branched RNA synthesis 5. Deoxyribozyme-Catalyzed Labeling (DECAL) of RNA 5.1. Overview of DECAL approach 5.2. Experimental procedures for preparing the labeled tagging RNA 5.3. Experimental procedures for DECAL using the labeled tagging RNA
* {
96 97 97 99 101 101 102 104 105 106 107 108 110 110 110 111 111 113
Department of Chemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA Department of Chemistry, Saint Louis University, St. Louis, Missouri, USA
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69005-4
#
2009 Elsevier Inc. All rights reserved.
95
96
Scott K. Silverman and Dana A. Baum
5.4. Further efforts needed to develop deoxyribozyme-catalyzed labeling (DECAL) of RNA Acknowledgments References
114 114 114
Abstract Since their first identification by in vitro selection in 1994, deoxyribozymes have been developed to catalyze a variety of chemical reactions. The first DNAcatalyzed reaction was cleavage of a ribonucleotide linkage within an oligonucleotide substrate. In subsequent years, growing collections of deoxyribozymes have been developed for several reactions that have practical utility for RNA research. These deoxyribozymes are useful for site-specific RNA cleavage as well as ligation to form linear, branched, and lariat RNA products. An application related to RNA ligation is deoxyribozyme-catalyzed labeling of RNA (DECAL), which is used to attach a biophysical tag to a desired RNA sequence at a specific position. With current achievements and likely future developments, deoxyribozymes are a useful contributor to the toolbox of RNA research methods.
1. Introduction Deoxyribozymes (also called DNA enzymes or DNAzymes) are specific sequences of DNA that have catalytic activity. All currently known deoxyribozymes have been identified by in vitro selection from large random-sequence DNA pools ( Joyce, 2004; Silverman, 2009). The catalytic range of DNA encompasses both oligonucleotide and nonoligonucleotide substrates (Baum and Silverman, 2008; Silverman, 2008). This report focuses on deoxyribozymes that are useful for reactions of RNA substrates, especially to assist studies of RNA structure, folding, and catalysis. For anyone who wishes to use a deoxyribozyme as a practical RNA cleavage or ligation catalyst by following the procedures described in this chapter, we recommend that a positive control experiment should be performed in parallel, using an RNA substrate that is known to be tolerated well by an analogous deoxyribozyme. For any nucleic acid enzyme that binds to its oligonucleotide substrate via extensive Watson–Crick interactions, it is impossible in practice to validate catalytic activity with all possible substrate sequences. Therefore, rates and yields with any particular substrate can vary from the values reported for other substrates. If the ‘‘experimental’’ sample fails to show the desired reactivity, then the results with the positive control will distinguish specific failure of the particular deoxyribozyme– substrate combination from a more general problem with the overall
Use of Deoxyribozymes in RNA Research
97
application of deoxyribozymes (e.g., a problem with buffers, metal ions, oligonucleotide purification, and so on).
2. Deoxyribozymes for RNA Cleavage The first deoxyribozyme was described in 1994 and catalyzes the Pb2þ-dependent cleavage of a DNA oligonucleotide strand at a particular embedded ribonucleotide (Breaker and Joyce, 1994). Since that first report, a large number of RNA-cleaving deoxyribozymes have been identified, including many that function well with oligonucleotide substrates that are made entirely out of RNA (Silverman, 2005). The cleavage reaction is shown in Fig. 5.1A and forms 20 ,30 -cyclic phosphate and 50 -OH RNA termini, which are the same products as formed by RNase A and other ribonuclease protein enzymes. At present, nearly all possible sites within an arbitrarily chosen RNA target sequence can be cleaved by a deoxyribozyme, simply by choosing a suitable combination of catalytic region and Watson–Crick ‘‘binding arms’’ to interact with the RNA target (Fig. 5.1B). Essentially all of the useful RNA-cleaving deoxyribozymes require Mg2þ and/ or Mn2þ for their catalytic activity. In addition to their preparative utility, RNA-cleaving deoxyribozymes can also be applied as analytical tools; for example, for mapping RNA branch sites (Pyle et al., 2000) or for revealing the presence of chemical modifications on long RNAs (Silverman, 2004).
2.1. Deoxyribozymes available for different RNA cleavage-site sequences Santoro and Joyce (1997) identified the 10–23 and 8–17 deoxyribozymes, which respectively cleave at R#Y and A#G dinucleotide junctions (# denotes the cleavage site; R, purine; Y, pyrimidine). Figure 5.1B depicts the sequences of the 10–23 and 8–17 deoxyribozymes, along with several other subsequently discovered variants of 8–17 that allow cleavage of other RNA linkages. Both 10–23 and 8–17 are active with 40–100 mM Mg2þ alone as the divalent metal ion cofactor, although in practice the cleavage rates are substantially higher if 10 mM Mn2þ is also included. Only if Mn2þinduced RNA degradation is a concern (e.g., for a particularly long RNA substrate) should Mn2þ be omitted. Li and coworkers have described numerous 8–17 variants that collectively enable cleavage of 14 out of the 16 possible RNA dinucleotide junctions. Their in vitro selection experiments were performed with substrates that contained only a single ribonucleotide linkage embedded within an otherwise all-DNA strand (i.e., DNA–rX–DNA). Therefore, separate
98
Scott K. Silverman and Dana A. Baum
Figure 5.1 Deoxyribozyme-catalyzed RNA cleavage. (A) The cleavage reaction, which forms 20 ,30 -cyclic phosphate and 50 -OH RNA termini. (B) Individual deoxyribozymes and their target sequences for efficient cleavage of all-RNA substrates. N, any nucleotide; R, purine; Y, pyrimidine. Outside of the explicitly indicated nucleotides, any RNA sequence is tolerated as long as Watson–Crick RNA:DNA covariation is maintained.
testing was required to assess the cleavage activities of the 8–17 variants with all-RNA substrates (Mi Zhang, D.A.B., and S.K.S., unpublished results). The two deoxyribozymes E1111 and E5112 depicted in Fig. 5.1B (Cruz et al., 2004) cleave either N#G or N#A linkages, respectively, within allRNA substrates. E1111 can also cleave R#U RNA substrates. The Li laboratory reported additional 8–17 variants named 8–17NG and 8–17NA (Schlosser et al., 2008a) that also cleave N#G or N#A linkages within all-RNA substrates. We are unaware of any reported deoxyribozymes that efficiently cleave all-RNA substrates at either Y#U or Y#C linkages.
Use of Deoxyribozymes in RNA Research
99
Indeed, Y#U linkages in any type of oligonucleotide have been refractory to efficient cleavage by any deoxyribozyme (Schlosser et al., 2008b). Unfortunately, in our hands neither E2112 (Cruz et al., 2004) nor 8–17NC (Schlosser et al., 2008a) works well with an all-RNA N#C substrate, although these deoxyribozymes were reported to cleave N#C at a single embedded ribonucleotide. In Section 2.2 are provided experimental procedures for using the RNA-cleaving deoxyribozymes shown in Fig. 5.1B. The analytical-scale procedure is appropriate for a picomole-scale amount of a 50 -32P-radiolabeled RNA substrate. The preparative-scale procedure is appropriate for a nanomole-scale amount of an unradiolabeled RNA substrate. In all cases, we typically design the deoxyribozyme binding arms each to have at least 12 kcal/mol of DNA:RNA binding energy, as readily computed using published parameters (Sugimoto et al., 1995). In practice, this is usually equivalent to binding arms with 10–14 bp, although the exact length depends on the nucleotide composition. It should be noted that multiple turnover is generally not required, and the deoxyribozyme is used in excess relative to the RNA substrate; typically the deoxyribozyme is less costly (in both time and money) than the RNA substrate. We recommend an analytical-scale check of the overall cleavage activity of a given deoxyribozyme–substrate combination before committing large amounts of RNA to the preparative-scale procedure. For those RNA substrates that are very long or have intrinsically strong secondary structure, use of 20 -OMe or LNA binding arms has been successful for disrupting secondary structure within RNA targets (Schubert et al., 2004; Silverman, 2005). However, because oligonucleotides with these chemical modifications are more expensive than unmodified DNA, it is reasonable to perform an initial check of whether the appropriate unmodified deoxyribozyme is capable of useful RNA cleavage. If not, then preparing and testing the chemically modified variant would be the next step. Alternatively, discrete ‘‘disruptor oligonucleotides’’ complementary to key regions of the target RNA secondary structure can be included so that the desired cleavage site is accessible to the deoxyribozyme. The design of such disruptor oligonucleotides is beyond the scope of this chapter; for an example that should assist design for any particular large RNA, see Wang and Silverman (2005b).
2.2. Experimental procedures 2.2.1. General annealing procedure All procedures involving deoxyribozymes begin with an annealing step to assist proper binding interactions between the deoxyribozyme and the RNA substrate(s). The deoxyribozyme and RNA substrate(s) are mixed with 5 mM Tris, pH 7.5, 15 mM NaCl, and 0.1 mM EDTA (HEPES may
100
Scott K. Silverman and Dana A. Baum
be used instead of Tris). The sample is annealed by heating at 95 C for 3 min and cooling on ice for 5 min. The annealed sample is then used for subsequent reaction steps. 2.2.2. Analytical-scale RNA cleavage by a deoxyribozyme In our experience, incubation conditions related to those reported by Cruz et al. (2004) provide optimal RNA cleavage activity for all 8–17 deoxyribozyme variants. Our standard incubation conditions (1) for 8–17 cleavage include 50 mM Tris, pH 7.5, 10 mM MgCl2, 10 mM MnCl2, and 150 mM NaCl at 37 C. A typical incubation time ranges from 1 h to overnight, depending on which specific 8–17 variant and RNA substrate sequence are used. Suitable buffer compounds such as HEPES can be used in place of Tris. For RNA cleavage by the 10–23 deoxyribozyme, our standard incubation conditions are the same as for 8–17, except that including Mg2þ is unnecessary (although not deleterious). Reagents
50 -32P-Radiolabeled RNA substrate to be cleaved (prepared by reaction of the RNA with g-32P-ATP and T4 polynucleotide kinase; alternatively, 30 -32P-radiolabeling with 32P-pCp and T4 RNA ligase can be used) 8–17 Deoxyribozyme, designed with binding arms complementary to the RNA substrate as shown in Fig. 5.1B 10 annealing buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA) 10 cleavage buffer (500 mM Tris, pH 7.5, 1.5 M NaCl) 10 Mg2þ/Mn2þ mix (100 mM MgCl2, 100 mM MnCl2)—note that solutions containing Mn2þ should be stored at 20 C to suppress oxidation of the metal ion
Procedure The deoxyribozyme is used in excess to the RNA substrate. Commonly at least threefold excess of deoxyribozyme is used to ensure that all of the RNA substrate has the opportunity to be cleaved. Typical mole amounts are 0.1–1 pmol RNA substrate and 3 pmol deoxyribozyme in 10 ml final volume, providing final concentrations of 10–100 nM RNA substrate and 300 nM deoxyribozyme. The deoxyribozyme and RNA substrate are annealed as described in Section 2.2.1. The cleavage reaction is initiated by addition of 10 cleavage buffer followed by 10 Mg2þ/Mn2þ mix to give final incubation conditions of 50 mM Tris, pH 7.5, 10 mM MgCl2, 10 mM MnCl2, and 150 mM NaCl. The sample is incubated at 37 C for several hours. Aliquots are removed at appropriate times, quenched onto stop solution (80% aqueous formamide, 1 TB [89 mM each Tris and boric acid, pH 8.3], 50 mM EDTA, 0.25% each bromophenol blue and xylene cyanol), and analyzed by denaturing PAGE (e.g., 20%).
Use of Deoxyribozymes in RNA Research
101
2.2.3. Preparative-scale RNA cleavage by a deoxyribozyme The preparative-scale RNA cleavage procedure is essentially the same as that used on the analytical scale, except the RNA substrate and deoxyribozyme concentrations are higher to avoid unreasonably large reaction volumes. Typically, 1 nmol of RNA substrate and 2 nmol of deoxyribozyme are used in a final volume of 100 ml (10 mM RNA substrate and 20 mM deoxyribozyme). The amount of deoxyribozyme may be decreased to as little as 1.05 equiv. of the RNA substrate, although usually the deoxyribozyme is less precious than the RNA substrate and therefore a larger excess can be used. After incubation, the nucleic acids are precipitated with ethanol and separated by PAGE. Care should be taken to ensure that the lengths of the desired RNA product and deoxyribozyme are sufficiently different to allow resolution by PAGE. If necessary, additional noncomplementary nucleotides may be included on either end of the deoxyribozyme to shift its PAGE mobility away from that of the desired RNA cleavage product. In addition, consideration should be given to the relative sizes of the two RNA cleavage products, which may have similar PAGE migration rates depending on location of the cleavage site. The deoxyribozyme may be isolated and reused in subsequent RNA cleavage experiments.
2.3. Further efforts needed to develop deoxyribozymes for RNA cleavage With RNA-cleaving deoxyribozymes available as depicted in Fig. 5.1B, the remaining work needed to obtain a complete set of RNA-cleaving deoxyribozymes is to identify catalysts for efficient cleavage of Y#U and Y#C junctions in all-RNA substrates. In the absence of new experimental efforts in this direction, it is impossible to know whether such deoxyribozymes are chemically possible, but this seems likely, especially for Y#C. Future studies could focus on evaluating metal ion cofactors other than Mg2þ and Mn2þ, which may allow higher cleavage rate constants even for difficult RNA cleavage sequences.
3. Deoxyribozymes for RNA Ligation: Synthesis of Linear RNA Products The ability to ligate two RNA substrates enables studies of RNA structure, folding, and catalysis. For example, site-specific chemical modifications are often incorporated into large RNA molecules by initial chemical synthesis of a short RNA oligonucleotide bearing one or more modifications. This modified oligonucleotide is then ligated to one or more RNAs segments that are prepared by either chemical synthesis or
102
Scott K. Silverman and Dana A. Baum
in vitro transcription. Many RNA ligation reactions have been achieved by enzymatic ‘‘splint ligation,’’ in which a DNA oligonucleotide splint is used to hold together the two RNA substrates for joining by T4 DNA ligase (Moore and Query, 2000; Moore and Sharp, 1992). In some cases, T4 RNA ligase has instead been used, either in splinted fashion with unpaired RNA nucleotides at the ligation junction (Bain and Switzer, 1992; Stark et al., 2006) or at an RNA secondary structure element where no external splint is necessary (Bruce and Uhlenbeck, 1978; Pan et al., 1991). However, for some particular RNA substrates and ligation sites, an approach based on T4 DNA ligase or T4 RNA ligase does not work well empirically, and in general it is difficult to predict success or failure in advance of experiment. Therefore, alternative RNA ligation approaches would be valuable. Deoxyribozymes offer one such alternative strategy in which the ligation splint is also the ligation catalyst. In 2003, we reported the first deoxyribozyme for RNA ligation (FlynnCharlebois et al., 2003), and our subsequent efforts have been reviewed (Silverman, 2008). This section provides information on deoxyribozymes that may be useful for RNA ligation with formation of linear products. For the ligation reaction, two combinations of RNA functional groups are possible (Fig. 5.2). First, joining a 20 ,30 -cyclic phosphate with a 50 -OH group results in either a native 30 –50 linkage or a nonnative 20 –50 linkage (Fig. 5.2A); when the 30 –50 linkage is formed, the reaction is the reverse of that shown in Fig. 5.1A. Second, ligating a 20 ,30 -diol with a 50 -triphosphate also provides either a 30 –50 or 20 –50 linkage (Fig. 5.2B). The 20 ,30 -cyclic phosphate RNA substrate can be prepared either using a deoxyribozyme as described in Section 2 or by other means, such as cleavage of a precursor RNA with a ribozyme (Ferre´-D’Amare´ and Doudna, 1996; Grosshans and Cech, 1991) or with RNase H and a helper oligonucleotide (Lapham and Crothers, 1996; Lapham et al., 1997). The 50 -triphosphate substrate is generally prepared by in vitro transcription (Milligan and Uhlenbeck, 1989; Milligan et al., 1987).
3.1. Deoxyribozymes available for 30 –50 RNA ligation A substantial thrust of our efforts with DNA-catalyzed RNA ligation has been to achieve synthesis of native 30 –50 linkages. Some of our efforts have been described elsewhere (Silverman, 2008). Here, we focus upon the most useful final deoxyribozymes. The key considerations for a practical RNA ligation catalyst are the rate constant, yield, and substrate sequence requirements. Identifying deoxyribozymes with favorable properties in all three aspects has been challenging, but in several cases we have succeeded. In particular, by exerting a suitable selection pressure to enforce formation of 30 –50 linkages (Wang and Silverman, 2005a), we found two specific deoxyribozymes that are useful for practical RNA ligation (Fig. 5.3) (Purtha et al.,
Use of Deoxyribozymes in RNA Research
103
Figure 5.2 Reactions catalyzed by deoxyribozymes that ligate RNA to form linear products. (A) Reaction of a 20 ,30 -cyclic phosphate with a 50 -OH group, leading to either a native 30 –50 linkage or a nonnative 20 –50 linkage. (B) Reaction of a 20 ,30 -diol with a 50 -triphosphate, again leading to either a 30 –50 or 20 –50 linkage.
2005). The Mg2þ-dependent 9DB1 deoxyribozyme joins two RNA sequences that match the D#RA motif (D ¼ A, G, or U), whereas the Zn2þ-dependent 7DE5 deoxyribozyme requires A#R. Both deoxyribozymes achieve their highest ligation yield within 2–4 h; the yield is 50–80% for 9DB1 and 40–50% for 7DE5. While variations in both rate and yield are observed for different RNA substrate sequences, our systematic surveys suggested that the indicated sequence requirements near the ligation site are the only ones. In several other studies, we also found many deoxyribozymes that create native 30 –50 RNA linkages, using either the 20 ,30 -cyclic phosphate substrate combination of Fig. 5.2A (Kost et al., 2008; Purtha et al., 2005) or the
104
Scott K. Silverman and Dana A. Baum
Figure 5.3 Individual deoxyribozymes for linear 30 –50 RNA ligation. R, purine; D, one of A, G, or U. Outside of the explicitly indicated nucleotides, any sequence for either the left-hand (L) or right-hand (R) RNA substrate is tolerated as long as Watson–Crick RNA:DNA covariation is maintained.
50 -triphosphate substrate combination of Fig. 5.2B (Wang and Silverman, 2005a). In all such cases, however, we either know or suspect that these DNA enzymes require enough particular RNA nucleotides so that their sequence generality is poor. Such deoxyribozymes might be useful in practical RNA ligation, as long as the precise RNA sequence at the ligation site is not constrained.
3.2. Deoxyribozymes available for 20 –50 RNA ligation RNA oligonucleotides with an internal nonnative 20 –50 linkage can be prepared by solid-phase synthesis using an appropriate ribonucleotide 20 -phosphoramidite, which is commercially available for the four standard RNA nucleotides. Alternatively, either of the DNA-catalyzed RNA ligation reactions of Fig. 5.2 can be used, even when one or both of the RNA substrates are much larger than can be synthesized by solid-phase methods. We have reported numerous deoxyribozymes that create 20 –50 linkages via the reaction of Fig. 5.2A (Flynn-Charlebois et al., 2003; Hoadley et al., 2005; Kost et al., 2008; Semlow and Silverman, 2005). In all cases, there are likely to be substantial restrictions on the RNA sequences near the ligation site. Nevertheless, for some applications the mere presence of the 20 –50 linkage in the RNA ligation product may be the most important consideration regardless of the precise sequence, and in such cases our deoxyribozymes provide a viable experimental approach. For the alternative ligation chemistry of Fig. 5.2B, again we have reported several deoxyribozymes that
Use of Deoxyribozymes in RNA Research
105
form 20 –50 linkages, albeit again with probable RNA sequence requirements (Wang and Silverman, 2005a).
3.3. Experimental procedures The standard incubation conditions (1) for 9DB1 ligation include 50 mM CHES, pH 9.0, 40 mM MgCl2, 150 mM NaCl, and 2 mM KCl at 37 C. The standard incubation conditions for 7DE5 ligation include 50 mM Tris, pH 7.5, 1 mM ZnCl2, 150 mM NaCl, and 2 mM KCl at 37 C. For both deoxyribozymes, the typical incubation time is 2–4 h. For 9DB1, the pH 9.0 buffer may be replaced with Tris at pH 7.5. The ligation rate will be much lower, but the reduced pH will result in less nonspecific RNA degradation, which may be important for especially long or sensitive RNA substrates. 3.3.1. Analytical-scale RNA ligation by a deoxyribozyme One of the two RNA substrates, either L or R (see Fig. 5.3), must be 32P-radiolabeled. The L substrate may be 50 -32P-radiolabeled by reaction with g-32P-ATP and T4 polynucleotide kinase, or the R substrate may be 30 -32P-radiolabeled by reaction with 32P-pCp and T4 RNA ligase. A typical incubation time for maximal ligation activity is 2–4 h, depending on the precise RNA substrate sequences that are used. Reagents 50 -32P-Radiolabeled L RNA substrate (or unradiolabeled) R RNA substrate (or 30 -32P-radiolabeled) 9DB1 or 7DE5 deoxyribozyme, designed with binding arms complementary to the RNA substrates as shown in Fig. 5.3 10 annealing buffer (9DB1: 50 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA; 7DE5: 50 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA) 10 ligation buffer (9DB1: 500 mM CHES, pH 9.0, 1.5 M NaCl, 20 mM KCl; 7DE5: 500 mM Tris, pH 7.5, 1.5 M NaCl, 20 mM KCl) 10 metal (9DB1: 400 mM MgCl2; 7DE5: 10 mM ZnCl2)—the 10 Zn2þ solution is prepared by diluting 1 volume of 100 mM ZnCl2 in 200 mM HNO3 with 2 volumes of 1 M Tris, pH 7.5, and 7 volumes of water
Procedure The 50 -32P-radiolabeled L RNA substrate (*L), deoxyribozyme (E), and R RNA substrate are used in approximate ratio <1:3:6. The key point is that *L < E < R, so that all *L is saturated with E and all E is saturated with R (if 30 -32P-radiolabeled R is prepared, then *R < E < L is used). Typical mole amounts are 0.1–1 pmol L, 3 pmol E, and 6 pmol R in 10 ml final volume, providing final concentrations of 10–100 nM L, 300 nM E, and 600 nM R. The deoxyribozyme and two RNA substrates are annealed
106
Scott K. Silverman and Dana A. Baum
as described in Section 2.2.1. The ligation reaction is initiated by addition of 10 ligation buffer followed by 10 metal to give final incubation conditions of 50 mM CHES, pH 9.0, 40 mM MgCl2, 150 mM NaCl, and 2 mM KCl (9DB1) or 50 mM Tris, pH 7.5, 1 mM ZnCl2, 150 mM NaCl, and 2 mM KCl (7DE5). The sample is incubated at 37 C for several hours. Aliquots are removed at appropriate times, quenched onto stop solution (80% aqueous formamide, 1 TB [89 mM each Tris and boric acid, pH 8.3], 50 mM EDTA, 0.25% each bromophenol blue and xylene cyanol), and analyzed by denaturing PAGE (e.g., 20%). 3.3.2. Preparative-scale RNA ligation by a deoxyribozyme The preparative-scale RNA ligation procedure is essentially the same as that used on the analytical scale, except the concentrations of the two RNA substrates and deoxyribozyme are higher to avoid unreasonably large reaction volumes. Typically, 1.0 nmol of L, 1.1 nmol of E, and 1.2 nmol of R are used in a final volume of 100 ml (10 mM L and slightly higher for each of E and R). After incubation, the nucleic acids are precipitated with ethanol and separated by PAGE. Care should be taken to ensure that the lengths of the desired RNA product and deoxyribozyme are sufficiently different to allow resolution by PAGE. If necessary, additional noncomplementary nucleotides may be included on either end of the deoxyribozyme to shift its PAGE mobility away from that of the desired RNA ligation product. The deoxyribozyme may be isolated and reused in subsequent RNA ligation experiments.
3.4. Further efforts needed to develop deoxyribozymes for linear RNA ligation For 30 –50 RNA ligation using the 50 -triphosphate substrate combination of Fig. 5.2B, our unpublished work (D. A. B. and S. K. S.) has shown that identifying deoxyribozymes with minimal RNA sequence requirements necessitates imposing a selection pressure directly for this generality. This goal can be approached by systematically changing the RNA substrate sequences in successive rounds of selection. Ongoing efforts are focused on implementing this selection pressure to identify a full set of general RNAligating deoxyribozymes. In addition, the requirement for a 50 -triphosphate on the R substrate means in practice that R must be an in vitro transcript. Although a method for appending a 50 -triphosphate to a synthetic RNA oligonucleotide via solid-phase synthesis has been reported (Paul et al., 2006), in our hands this approach is technically demanding and difficult to reproduce. Alternatively, an activated phosphorus at the RNA 50 -terminus can be provided by a 50 -adenylate, which can be installed onto readily obtained 50 -phosphorylated RNA using T4 RNA ligase (Silverman, 2004) or T4 DNA ligase (Wang and Silverman, 2006b). RNA-ligating deoxyribozymes that readily accept a 50 -adenylated R substrate still await development.
Use of Deoxyribozymes in RNA Research
107
For 30 –50 RNA ligation using the 20 ,30 -cyclic phosphate substrate combination of Fig. 5.2A, a suitable combination of selection design aspects (including use of Zn2þ as the metal ion cofactor) leads to the desired native linkages (Kost et al., 2008). However, as is the case for the 50 -triphosphate substrate combination, a selection pressure for sequence generality must again be imposed to identify useful deoxyribozymes. For 20 –50 RNA ligation using any substrate combination, no efforts have yet been made to identify deoxyribozymes that are sequence-general. Presumably the same approaches intended in this regard for 30 –50 ligation would be applicable for 20 –50 ligation as well.
4. Deoxyribozymes for RNA Ligation: Synthesis of Branched RNA Products 20 ,50 -Branched RNA molecules are formed naturally during spliceosomal pre-mRNA processing (Wahl et al., 2009) and during self-splicing by group II introns (Michel and Ferat, 1995). Although chemical (solid-phase) approaches to synthesis of branched RNA have been developed (Damha et al., 1992), such methods are tedious and impractical for many biochemists. Alternatively, we have identified a variety of deoxyribozymes that create branched RNA by catalyzing attack of a 20 -OH group into a 50 -triphosphate (Fig. 5.4). Because these deoxyribozymes are found by in vitro selection, they are not constrained by the same substrate sequence requirements as the natural splicing enzymes, thereby offering the potential for synthesis of unnatural branched RNA molecules. This section describes the deoxyribozymes that have been used to synthesize 20 ,50 -branched RNAs. The biological pre-mRNA splicing intermediates are a specific subclass of branched RNAs termed ‘‘lariats.’’ In a lariat RNA, the two RNA strands that emerge from the 50 - and 20 -oxygen atoms of the branch-site nucleotide are covalently connected (Fig. 5.4, dashed loop) (Domdey et al., 1984; Padgett et al., 1984; Rodriguez et al., 1984; Ruskin et al., 1984; Zeitlin and Efstratiadis, 1984). Lariats are even more challenging to synthesize than branches. Some of the deoxyribozymes developed for branched RNA synthesis can also be used for direct one-step lariat synthesis (Wang and Silverman, 2005b). Alternatively, a lariat may be formed in two steps, by closure of the loop of an initially synthesized branch (Wang and Silverman, 2006a). Although this section does not directly address lariat synthesis by either approach, the methodology to synthesize branches is directly useful for making lariats as well. It should be noted that deoxyribozymes are capable of synthesizing branched nucleic acids in which one or both of the oligonucleotide strands are DNA rather than RNA (Mui and Silverman, 2008; Zelin and Silverman, 2007). (When the strand providing the 20 -OH nucleophile
108
Scott K. Silverman and Dana A. Baum
Figure 5.4 Synthesis of 20 ,50 -branched and lariat RNA by reaction of a 20 -OH group with a 50 -triphosphate. The product is a lariat when the dashed loop is present connecting the two substrates.
is DNA, the particular nucleotide with the 20 -OH is still of course a ribonucleotide.) Such deoxyribozymes are found by in vitro selection approaches similar to those used in identification of DNA enzymes for branched RNA synthesis, and these deoxyribozymes are used to create branched nucleic acids via analogous procedures.
4.1. Deoxyribozymes available for 20 ,50 -branched and lariat RNA synthesis The selection strategies and outcomes for identifying branch-forming deoxyribozymes have been described in part elsewhere (Silverman, 2008). Here, the most useful DNA enzymes are briefly described. First, the Mg2þdependent 7S11 deoxyribozyme creates 20 ,50 -branched RNA with a
Use of Deoxyribozymes in RNA Research
109
preference for a branch-site adenosine nucleotide (Coppins and Silverman, 2005; Coppins and Silverman, 2004) (Fig. 5.5A). A branch-site guanosine is tolerated, albeit with about 50-fold lower ligation rate constant than with branch-site adenosine. The 50 -triphosphorylated RNA nucleotide that reacts with the 20 -OH can be either A or G. 50 -Adenylated C (Wang and Silverman, 2006b) is also accepted with lower rate and yield; 50 -adenylated U shows little reactivity. Other than these two nucleotides directly at the ligation junction, the two RNA substrates may have nearly any sequence while still allowing substantial ligation rate and yield. Because 7S11 works very quickly (kobs of 0.6 min–1) with the RNA substrate sequences used during the selection process, even significant decreases in rate due to sequence changes still allow branch formation in preparatively useful fashion. We subsequently identified a series of deoxyribozymes related to 7S11, of which 10DM24 (Fig. 5.5B) was examined in the most detail (Zelin et al., 2006). 10DM24 is similar to 7S11 in its substrate acceptance, with slightly improved tolerance of different RNA sequences. We recommend that 10DM24 should be tested first for any new substrate combination. During
Figure 5.5 Individual deoxyribozymes for branched RNA synthesis. (A) 7S11 deoxyribozyme. Note the four Watson–Crick paired regions denoted P1–P4. (B) 10DM24 deoxyribozyme. (C) 6CE8 deoxyribozyme. (D) 6BX22 deoxyribozyme.
110
Scott K. Silverman and Dana A. Baum
the same selection experiment, many other deoxyribozymes were identified, some of which have substantial ligation yields with branch-site C or U. Therefore, for synthesizing branched RNA with a pyrimidine branch-site nucleotide, one of these 10DM24-related deoxyribozymes may be useful. Several other deoxyribozymes are also useful for branched RNA synthesis. The Mn2þ-dependent 6CE8 deoxyribozyme accepts any branch-site nucleotide, although it requires the sequence 50 -GAR for the RNA substrate that reacts with the branch-site 20 -OH group (Fig. 5.5C) (Pratico et al., 2005). The Mn2þ-dependent 6BX22 deoxyribozyme also works with any branch-site nucleotide, although A is best (Fig. 5.5D) (Wang and Silverman, 2005b). A feature of 6BX22 is that it also works especially well to form lariat RNAs.
4.2. Experimental procedures The analytical- and preparative-scale branch-forming procedures are essentially equivalent to those used for linear RNA ligation (Section 3.3). Appropriate pH and metal ion are used (7S11 and 10DM24: CHES, pH 9.0, with 40 mM Mg2þ, or HEPES, pH 7.5, with lower rate; 6CE8 and 6BX22: Tris, pH 7.5, with 20 mM Mn2þ).
4.3. Further efforts needed to develop deoxyribozymes for branched RNA synthesis As described in Section 4.1 and depicted in Fig. 5.5, the currently available branch-forming deoxyribozymes each have modest but nonzero restrictions on the RNA substrate sequences that can be used. It should be possible to identify new branch-forming deoxyribozymes that have even more permissive sequence requirements. Selection experiments to identify such deoxyribozymes will likely use approaches analogous to those mentioned in Section 3.4 for linear RNA ligation.
5. Deoxyribozyme-Catalyzed Labeling (DECAL) of RNA One specific application of branch-forming deoxyribozymes is named DECAL, which stands for deoxyribozyme-catalyzed labeling of RNA (Baum and Silverman, 2007). This section describes the DECAL approach and its implementation.
Use of Deoxyribozymes in RNA Research
111
5.1. Overview of DECAL approach In DECAL, the 10DM24 deoxyribozyme that forms branched RNA is exploited to place a biophysical tag (label) such as a fluorophore or biotin onto a target RNA of interest (Fig. 5.6). To accomplish this, a ‘‘tagging RNA’’ is created in two steps by first using T7 RNA polymerase to transcribe a short RNA oligonucleotide that has a 5-aminoallyl-C nucleotide at its second position. The primary amino group on the aminoallylRNA is then chemically modified with the desired biophysical tag; for example, by using an appropriate N-hydroxysuccinimidyl ester (Fig. 5.6A). The target RNA is then covalently modified with the tagging RNA using the 10DM24 deoxyribozyme (Fig. 5.6B). This section describes the methods for preparing the labeled tagging RNA (Section 5.2) as well as for attaching the tagging RNA to the target RNA (Section 5.3).
5.2. Experimental procedures for preparing the labeled tagging RNA The two steps of the preparative procedure are each straightforward.
Figure 5.6 Deoxyribozyme-catalyzed labeling (DECAL) of RNA. (A) Synthesis of the tagging RNA. The unlabeled tagging RNA is prepared by in vitro transcription using commercially available 5-aminoallyl-CTP and T7 RNA polymerase. The sequence is designed such that the modified C is the only such nucleotide in the entire 19 nt sequence. (B) Attachment of the tagging RNA to the target RNA by the 10DM24 deoxyribozyme.
112
Scott K. Silverman and Dana A. Baum
5.2.1. Transcription of the unlabeled tagging RNA The transcription reaction is performed using T7 RNA polymerase according to the general protocol of Milligan et al. (1987), using a DNA template strand that incorporates 20 -OMe modifications at each of the first two nucleotides at the 50 -end to suppress formation of n þ 1 transcription products (Kao et al., 1999). The sequence of the 19-mer unlabeled tagging RNA is 50 -GCaaAAGAGAUGGUGAUGGGA-30 , where Caa denotes 5-aminoallyl-C. Reagents
Double-stranded DNA template comprising T7 RNA polymerase promoter sequence (50 -ACGCACGCTGTAATACGACTCACTATA-30 ; promoter italicized) and reverse complement of the tagging RNA (50 -UCCCATCACCATCTCTTGCTATAGTGAGTCGTATTAC AGCGTGCGT-30 ; boldface portion complementary to tagging RNA sequence; two underlined nucleotides 20 -OMe) Standard buffers and reagents (1 M Tris, pH 8.0, 1 M MgCl2, 250 mM DTT, and 100 mM spermidine) 100 mM each ATP, GTP, and UTP 5-(Aminoallyl)cytidine 50 -triphosphate (5-aminoallyl-CTP; TriLink Biotechnologies or ChemCyte) T7 RNA polymerase (prepared by expression of a His6-tagged construct, or purchased from a commercial supplier)
Procedure The double-stranded DNA template is prepared by combining the two component strands in 5 mM Tris, pH 8.0, 15 mM NaCl, and 0.1 mM EDTA; the sample is annealed by heating at 95 C for 3 min and cooling on ice for 5 min. The annealed DNA template is added to a transcription reaction, with final concentrations of 40 mM Tris, pH 8.0, 30 mM MgCl2, 10 mM DTT, 2 mM spermidine, 4 mM each ATP, GTP, and UTP, 2 mM 5-aminoallyl-CTP, and T7 RNA polymerase (e.g., 1/20 by volume of a suitably diluted prepared stock). The transcription sample is incubated at 37 C for 5 h and quenched by addition of sufficient 0.5 M EDTA, pH 8.0 to chelate all of the Mg2þ. The crude transcript is precipitated with ethanol and purified by 20% denaturing PAGE. For a 100 ml transcription reaction volume containing 100 pmol of each DNA template strand, a typical yield of purified unlabeled tagging RNA is 1.4–3.0 nmol. Concentrations of 5-aminoallyl-CTP lower than 2 mM lead to reduced yield of the unlabeled tagging RNA transcript. In contrast, increasing the concentration above 2 mM does not improve the yield. If 20 -OMe groups are not included at the first two 50 -nucleotides of the reverse complement strand of the DNA template, then the yield of the correct length of this particular RNA transcript is decreased considerably.
Use of Deoxyribozymes in RNA Research
113
5.2.2. Labeling of the tagging RNA The optimal reaction conditions for derivatizing the tagging RNA with a biophysical label depend on the NHS ester used (Fig. 5.6A). Derivatizations with NHS-biotin or NHS-TAMRA can be performed under the same conditions; derivatization with NHS-fluorescein was optimal with different conditions. Reagents
1 M sodium phosphate, pH 8.0 (for biotin or TAMRA), or 0.5 M NaHCO3, pH 9.0 (for fluorescein) 5 mM EDTA, pH 8.0 NHS ester reagent in DMSO (50 mM for biotin or TAMRA; 70 mM for fluorescein) Procedure Related procedures are used for all labeling reagents. For biotin or TAMRA, 1 nmol of the unlabeled tagging RNA is incubated in 200 ml with final concentrations of 5 mM RNA, 100 mM sodium phosphate, pH 8.0, 0.2 mM EDTA, 5 mM NHS ester, and 50% (v/v) DMSO. For fluorescein, 1 nmol of the unlabeled tagging RNA is incubated in 100 ml with final concentrations of 10 mM RNA, 70 mM NaHCO3, pH 9.0, 0.2 mM EDTA, 21 mM NHS ester, and 30% (v/v) DMSO. The sample is incubated at 37 C for 24 h (biotin) or 3 h (TAMRA). For labeling with fluorescein, in our experience the reaction was preparatively useful only using the conditions described here; 3 h incubation leads to lower labeling yield but minimal RNA degradation, whereas 12 h incubation leads to higher labeling yield at the expense of modest RNA degradation. The labeled tagging RNA is precipitated with ethanol and purified by 20% denaturing PAGE. A typical yield of purified labeled tagging RNA after gel extraction (Wang and Silverman, 2003) and precipitation is 270 pmol (biotin), 150 pmol (TAMRA), or 200 pmol (fluorescein). Some inefficiency is observed in the extraction process for the labeled RNAs, especially with the TAMRA label, leading to reduced preparative yields as compared with analytical-scale yield estimates.
5.3. Experimental procedures for DECAL using the labeled tagging RNA The analytical- and preparative-scale DECAL procedures are essentially equivalent to those used for linear RNA ligation by the 9DB1 deoxyribozyme (Section 3.3) as well as for branched RNA formation. If degradation of a large target RNA is a concern, then the CHES, pH 9.0, 40 mM Mg2þ conditions may be replaced with HEPES, pH 7.5, 40 mM Mg2þ; a longer incubation time will be required.
114
Scott K. Silverman and Dana A. Baum
If desired, 11 out of 19 nucleotides may be cleaved from the 30 -end of the tagging RNA after its attachment to the target RNA. This may be accomplished using a suitable 10–23 deoxyribozyme according to the procedures in Section 2.2 (Baum and Silverman, 2007).
5.4. Further efforts needed to develop deoxyribozymecatalyzed labeling (DECAL) of RNA Because DECAL depends upon branched RNA formation as catalyzed by 10DM24, this approach faces similar challenges related to undesired RNA target sequence requirements as indicated in Section 4.3 for conventional branched RNA synthesis. The presence of the biophysical label on the tagging RNA also presents a challenge; depending on the nature of this label, the catalytic ability of the deoxyribozyme may be affected. Therefore, selection experiments in which the label is specifically present on the selection substrate may be needed to obtain fully functional deoxyribozymes. The current approach uses a tagging RNA that incorporates 5-aminoallyl-C. We anticipate that analogous selection efforts should enable the use of 5-aminoallyl-U, which is more widely available as its NTP.
ACKNOWLEDGMENTS Research on deoxyribozymes in the Silverman laboratory is supported by the National Institutes of Health, the Defense Threat Reduction Agency, and the David and Lucile Packard Foundation, with previous support from the Burroughs Wellcome Fund, the March of Dimes, the ACS Petroleum Research Fund, and the University of Illinois. Research on deoxyribozymes in the Baum laboratory is supported by Saint Louis University.
REFERENCES Bain, J. D., and Switzer, C. (1992). Regioselective ligation of oligoribonucleotides using DNA splints. Nucleic Acids Res. 20, 4372. Baum, D. A., and Silverman, S. K. (2007). Deoxyribozyme-catalyzed labeling of RNA. Angew. Chem. Int. Ed. 46, 3502–3504. Baum, D. A., and Silverman, S. K. (2008). Deoxyribozymes: Useful DNA catalysts in vitro and in vivo. Cell. Mol. Life Sci. 65, 2156–2174. Breaker, R. R., and Joyce, G. F. (1994). A DNA enzyme that cleaves RNA. Chem. Biol. 1, 223–229. Bruce, A. G., and Uhlenbeck, O. C. (1978). Reactions at the termini of tRNA with T4 RNA ligase. Nucleic Acids Res. 5, 3665–3677. Coppins, R. L., and Silverman, S. K. (2004). A DNA enzyme that mimics the first step of RNA splicing. Nat. Struct. Mol. Biol. 11, 270–274. Coppins, R. L., and Silverman, S. K. (2005). A deoxyribozyme that forms a three-helixjunction complex with its RNA substrates and has general RNA branch-forming activity. J. Am. Chem. Soc. 127, 2900–2907.
Use of Deoxyribozymes in RNA Research
115
Cruz, R. P. G., Withers, J. B., and Li, Y. (2004). Dinucleotide junction cleavage versatility of 8–17 deoxyribozyme. Chem. Biol. 11, 57–67. Damha, M. J., Ganeshan, K., Hudson, R. H., and Zabarylo, S. V. (1992). Solid-phase synthesis of branched oligoribonucleotides related to messenger RNA splicing intermediates. Nucleic Acids Res. 20, 6565–6573. Domdey, H., Apostol, B., Lin, R. J., Newman, A., Brody, E., and Abelson, J. (1984). Lariat structures are in vivo intermediates in yeast pre-mRNA splicing. Cell 39, 611–621. Ferre´-D’Amare´, A. R., and Doudna, J. A. (1996). Use of cis- and trans-ribozymes to remove 50 and 30 heterogeneities from milligrams of in vitro transcribed RNA. Nucleic Acids Res. 24, 977–978. Flynn-Charlebois, A., Wang, Y., Prior, T. K., Rashid, I., Hoadley, K. A., Coppins, R. L., Wolf, A. C., and Silverman, S. K. (2003). Deoxyribozymes with 20 -50 RNA ligase activity. J. Am. Chem. Soc. 125, 2444–2454. Grosshans, C. A., and Cech, T. R. (1991). A hammerhead ribozyme allows synthesis of a new form of the Tetrahymena ribozyme homogeneous in length with a 30 end blocked for transesterification. Nucleic Acids Res. 19, 3875–3880. Hoadley, K. A., Purtha, W. E., Wolf, A. C., Flynn-Charlebois, A., and Silverman, S. K. (2005). Zn2þ-dependent deoxyribozymes that form natural and unnatural RNA linkages. Biochemistry 44, 9217–9231. Joyce, G. F. (2004). Directed evolution of nucleic acid enzymes. Annu. Rev. Biochem. 73, 791–836. Kao, C., Zheng, M., and Rudisser, S. (1999). A simple and efficient method to reduce nontemplated nucleotide addition at the 30 terminus of RNAs transcribed by T7 RNA polymerase. RNA 5, 1268–1272. Kost, D. M., Gerdt, J. P., Pradeepkumar, P. I., and Silverman, S. K. (2008). Controlling regioselectivity and site-selectivity in RNA ligation by Zn2þ-dependent deoxyribozymes that use 20 ,30 -cyclic phosphate RNA substrates. Org. Biomol. Chem. 6, 4391–4398. Lapham, J., and Crothers, D. M. (1996). RNase H cleavage for processing of in vitro transcribed RNA for NMR studies and RNA ligation. RNA 2, 289–296. Lapham, J., Yu, Y. T., Shu, M. D., Steitz, J. A., and Crothers, D. M. (1997). The position of site-directed cleavage of RNA using RNase H and 20 -O-methyl oligonucleotides is dependent on the enzyme source. RNA 3, 950–951. Michel, F., and Ferat, J. L. (1995). Structure and activities of group II introns. Annu. Rev. Biochem. 64, 435–461. Milligan, J. F., and Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods Enzymol. 180, 51–62. 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. Moore, M. J., and Query, C. C. (2000). Joining of RNAs by splinted ligation. Methods Enzymol. 317, 109–123. Moore, M. J., and Sharp, P. A. (1992). Site-specific modification of pre-mRNA: The 20 -hydroxyl groups at the splice site. Science 256, 992–997. Mui, T. P., and Silverman, S. K. (2008). Convergent and general one-step DNA-catalyzed synthesis of multiply branched DNA. Org. Lett. 10, 4417–4420. Padgett, R. A., Konarska, M. M., Grabowski, P. J., Hardy, S. F., and Sharp, P. A. (1984). Lariat RNA’s as intermediates and products in the splicing of messenger RNA precursors. Science 225, 898–903. Pan, T., Gutell, R. R., and Uhlenbeck, O. C. (1991). Folding of circularly permuted transfer RNAs. Science 254, 1361–1364. Paul, N., Springsteen, G., and Joyce, G. F. (2006). Conversion of a ribozyme to a deoxyribozyme through in vitro evolution. Chem. Biol. 13, 329–338.
116
Scott K. Silverman and Dana A. Baum
Pratico, E. D., Wang, Y., and Silverman, S. K. (2005). A deoxyribozyme that synthesizes 20 ,50 -branched RNA with any branch-site nucleotide. Nucleic Acids Res. 33, 3503–3512. Purtha, W. E., Coppins, R. L., Smalley, M. K., and Silverman, S. K. (2005). General deoxyribozyme-catalyzed synthesis of native 30 –50 RNA linkages. J. Am. Chem. Soc. 127, 13124–13125. Pyle, A. M., Chu, V. T., Jankowsky, E., and Boudvillain, M. (2000). Using DNAzymes to cut, process, and map RNA molecules for structural studies or modification. Methods Enzymol. 317, 140–146. Rodriguez, J. R., Pikielny, C. W., and Rosbash, M. (1984). In vivo characterization of yeast mRNA processing intermediates. Cell 39, 603–610. Ruskin, B., Krainer, A. R., Maniatis, T., and Green, M. R. (1984). Excision of an intact intron as a novel lariat structure during pre-mRNA splicing in vitro. Cell 38, 317–331. Santoro, S. W., and Joyce, G. F. (1997). A general purpose RNA-cleaving DNA enzyme. Proc. Natl. Acad. Sci. USA 94, 4262–4266. Schlosser, K., Gu, J., Sule, L., and Li, Y. (2008a). Sequence-function relationships provide new insight into the cleavage site selectivity of the 8–17 RNA-cleaving deoxyribozyme. Nucleic Acids Res. 36, 1472–1481. Schlosser, K., Gu, J., Lam, J. C., and Li, Y. (2008b). In vitro selection of small RNA-cleaving deoxyribozymes that cleave pyrimidine-pyrimidine junctions. Nucleic Acids Res. 36, 4768–4777. Schubert, S., Furste, J. P., Werk, D., Grunert, H. P., Zeichhardt, H., Erdmann, V. A., and Kurreck, J. (2004). Gaining target access for deoxyribozymes. J. Mol. Biol. 339, 355–363. Semlow, D. R., and Silverman, S. K. (2005). Parallel selections in vitro reveal a preference for 20 -50 RNA ligation by deoxyribozyme-mediated opening of a 20 ,30 -cyclic phosphate. J. Mol. Evol. 61, 207–215. Silverman, S. K. (2004). Practical and general synthesis of 50 -adenylated RNA (50 -AppRNA). RNA 10, 731–746. Silverman, S. K. (2005). In vitro selection, characterization, and application of deoxyribozymes that cleave RNA. Nucleic Acids Res. 33, 6151–6163. Silverman, S. K. (2008). Catalytic DNA (deoxyribozymes) for synthetic applications— Current abilities and future prospects. Chem. Commun. 3467–3485. Silverman, S. K. (2009). Artificial functional nucleic acids: Aptamers, ribozymes, and deoxyribozymes identified by in vitro selection. In ‘‘Functional Nucleic Acids for Analytical Applications,’’ (Y. Li and Y. Lu, eds.), pp. 47–108. Springer ScienceþBusiness Media, LLC, New York. 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. Sugimoto, N., Nakano, S., Katoh, M., Matsumura, A., Nakamuta, H., Ohmichi, T., Yoneyama, M., and Sasaki, M. (1995). Thermodynamic parameters to predict stability of RNA/DNA hybrid duplexes. Biochemistry 34, 11211–11216. Wahl, M. C., Will, C. L., and Luhrmann, R. (2009). The spliceosome: Design principles of a dynamic RNP machine. Cell 136, 701–718. Wang, Y., and Silverman, S. K. (2003). Characterization of deoxyribozymes that synthesize branched RNA. Biochemistry 42, 15252–15263. Wang, Y., and Silverman, S. K. (2005a). Directing the outcome of deoxyribozyme selections to favor native 30 -50 RNA ligation. Biochemistry 44, 3017–3023. Wang, Y., and Silverman, S. K. (2005b). Efficient one-step synthesis of biologically related lariat RNAs by a deoxyribozyme. Angew. Chem. Int. Ed. 44, 5863–5866. Wang, Y., and Silverman, S. K. (2006a). A general two-step strategy to synthesize lariat RNAs. RNA 12, 313–321.
Use of Deoxyribozymes in RNA Research
117
Wang, Y., and Silverman, S. K. (2006b). Efficient RNA 50 -adenylation by T4 DNA ligase to facilitate practical applications. RNA 12, 1142–1146. Zeitlin, S., and Efstratiadis, A. (1984). In vivo splicing products of the rabbit b-globin pre-mRNA. Cell 39, 589–602. Zelin, E., and Silverman, S. K. (2007). Allosteric control of ribozyme catalysis using DNA constraints. ChemBioChem. 8, 1907–1911. Zelin, E., Wang, Y., and Silverman, S. K. (2006). Adenosine is inherently favored as the branch-site RNA nucleotide in a structural context that resembles natural RNA splicing. Biochemistry 45, 2767–2771.
C H A P T E R
S I X
Strategies in RNA Crystallography Francis E. Reyes, Andrew D. Garst, and Robert T. Batey Contents 1. Introduction 2. RNA Selection and Initial Characterization 2.1. Characterize the folded state of the RNA 2.2. Consider native purification of the RNA 3. Construction of Library of RNAs for Crystallization Trials 3.1. Choosing phylogenetic variants of conserved function but variable periphery domains from a sequence alignment 3.2. Consideration of intermolecular and intramolecular RNA packing 3.3. Variation of peripheral helical lengths 4. Improving Crystal Quality Through Postcrystal Analysis 4.1. Cation additives to improve crystal quality 4.2. Determine and vary the residues involved in crystal packing 4.3. Directed mutagenesis using a crystal structure 4.4. Analyze the crystal for RNA cleavage 5. Phasing Methods 5.1. Include heavy metal cations and heavy nucleotide derivatives 5.2. Molecular replacement with RNA fragments 6. A Case Study in the Crystallization of Lysine Riboswitch Regulatory Element 7. Concluding Remarks Acknowledgments References
120 122 122 123 124 124 125 126 127 127 128 129 130 130 131 132 132 134 135 135
Abstract A number of RNAs ranging from small helices to large megadalton ribonucleoprotein complexes have been solved to atomic resolution using X-ray crystallography. As with proteins, RNA crystallography involves a number of screening trials in which the concentration of macromolecule, precipitant, salt, and temperature are varied, an approach known as searching ‘‘condition space.’’ In contrast to proteins, the nature of base pairing in nucleic acids creates Department of Chemistry and Biochemistry, University of Colorado at Boulder, Boulder, Colorado, USA Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69006-6
#
2009 Elsevier Inc. All rights reserved.
119
120
Francis E. Reyes et al.
predictable secondary structure that facilitates the rational design of RNA variants, allowing ‘‘sequence space’’ to be screened in parallel. This chapter reviews RNA-specific techniques and considerations for RNA crystallography and presents a complete workflow used by our laboratory for solving RNA structures starting with initial library construction, methods to investigate and improve RNA crystal quality, and finally phase determination and structure solution.
1. Introduction The RCSB Protein Databank currently holds crystallographic data for over 300 RNA-only structures and over 600 RNA–protein complexes (Berman et al., 2000). These structures are powerful models that enable detailed genetic and biochemical experiments that probe the role of RNA in diverse biological functions including gene regulation, viral infection, and antibiotic resistance (Magnet and Blanchard, 2005; Sharp, 2009). Despite increasing success of RNA crystallographic efforts and the importance of the resulting structural information, RNA still suffers from the belief that it is an inherently difficult molecule to crystallize deters some structural studies. However, in our experience, many of the perceived difficulties of RNA can be overcome by a careful biochemical characterization of the RNA prior to initiating a crystallization effort. RNA crystallography has the same fundamental requirements as protein crystallography: the ability to grow an ordered crystal capable of diffracting X-rays. The basic approach is the same as protein crystallography in that a macromolecule is subjected to a variety of solution conditions and temperatures, otherwise known as ‘‘condition space,’’ in an effort to grow a diffraction-quality crystal. RNA as a macromolecule has features that can be exploited in the crystallographic effort including a reduced alphabet of building blocks (4 bases as opposed to 20 amino acids) and the dominance of base pairing and the A-form helix. In addition, almost any RNA sequence less than 400 nucleotides in length is easy to prepare in large quantities, refold into the native state and is soluble at the high concentrations (10–25 mg/ml) necessary for crystal growth. These properties are routinely exploited in RNA crystallography in a robust strategy that involves exploration of ‘‘sequence space’’ in concert with condition space. This chapter describes a set of established techniques (Edwards et al., 2009; Golden and Kundrot, 2003; Holbrook et al., 2001; Ke and Doudna, 2004) commonly used to solve RNA crystal structures to high resolution that we consider when embarking upon the determination of an RNA structure by X-ray crystallography. Specifically, we think about the design of RNA sequences (which we refer to as ‘‘constructs’’) to yield initial
121
Strategies in RNA Crystallography
crystals (Berman et al., 2000), application of RNA-specific methods to improve diffraction when a crystal form is obtained (Magnet and Blanchard, 2005), and RNA-specific phasing techniques in a workflow that seek to minimize the time needed to solve the RNA of interest (Sharp, 2009) (Fig. 6.1). Our strategies rely upon the fact that RNA can be manipulated in predictable ways due to its regular secondary structure, and are supported with lessons learned from the last decade of RNA
Initial characterization
RNA of interest Assess solution behavior
Multimeric
Alternate purification or refolding scheme
Monomeric
Construct design
Use phylogeny and engineering to guide library design
No
Construct refinement to improve crystal contacts or reduce intramolecular disorder
Sparse matrix screening
Cation/additive screening
Initial crystals
Refining crystal quality
RNA analysis from crystal
Yes Shape crystal contact mapping Optimize condition space
Diffraction
Post-crystallization analysis Low quality or low diffracting crystals
Yes
Phasing Low quality phases Engineer experimental phasing modules
De novo molecular replacement
Structure solution
Figure 6.1 Flowchart illustrating the workflow for RNA crystallography. In its linear form, it is composed of four parts: initial characterization, construct design, crystal quality refinement, and experimental phasing.
122
Francis E. Reyes et al.
structural research and crystallographic studies. We tailor our overall approach to each new RNA with a central philosophy in mind: a construct that is readily crystallizable will yield some form of crystal in a variety of conditions. Thus, each individual sequence is not initially tested under a broad set of conditions. Instead, each is tested against 200 conditions at one temperature (typically 25 C) for 1 week. If the construct yields little or no crystals of any quality, no further exploration of that RNA is performed. We rely heavily on a continuous pipeline of new constructs entering into our crystallization trials such that the failure of any one RNA (or more likely, multiple RNAs) is not an issue. In this fashion, we can rapidly interrogate a library of designed RNA constructs (typically 20–50 in our experience) with hopes of identifying several candidates that can be further explored in condition space in order to optimize the size and quality of crystals for diffraction studies.
2. RNA Selection and Initial Characterization 2.1. Characterize the folded state of the RNA Characterizing the conformation of the RNA of interest is a crucial first step in any RNA crystallography effort. RNAs are susceptible to misfolding into local energy minima or into multiple states, otherwise known as the ‘‘alternative conformer hell’’ (Fedor and Westhof, 2002; Uhlenbeck, 1995). RNA conformational heterogeneity from RNA misfolding can inhibit the crystallization process or yields a biologically irrelevant structure, and should be assessed using a variety of established techniques such as native polyacrylamide gel electrophoresis, size exclusion chromatography, and dynamic light scattering (Ferre´-D’Amare´ and Burley, 1994; Ferre´D’Amare´ et al., 1998). Biological activity assays are also important if such an assay available, particularly in cases where the RNA has been altered in an effort to promote crystallizability (see Section 3). These include kinetic assays in the case of catalytic RNAs (DeRose, 2002), isothermal titration calorimetry in the case of metabolite binding RNAs (Batey et al., 2004; Gilbert et al., 2006, 2008), or native polyacrylamide gel mobility shifts for RNA–protein complexes (or RNAs that undergo a large conformational change that is associated with their function). As the fold of the RNA is critical, the folding step after purification of the RNA is important. The most commonly used folding protocols involve heat-cooling the RNA in a metal-free buffer such as Tris–EDTA and subsequently adding metals (Ke and Doudna, 2004) such as monovalent or divalent cations, most commonly Na/Kþ and Mg2þ. Slow renaturation of the folded RNA from denaturing conditions may also be effective, particularly in RNA–protein systems. Both the human d virus ribozyme
Strategies in RNA Crystallography
123
(Ke et al., 2004) and the signal recognition particle (SRP) M domain–4.5 S RNA complex (Batey et al., 2001) required complete denaturation in 8 M urea and refolding into the native form by slow dialysis to native conditions. Some RNAs in our experience have required extensive screening of ionic conditions, temperature protocols, and use of denaturants to find conditions that yield a near-homogeneous population.
2.2. Consider native purification of the RNA In some instances, an exhaustive exploration of refolding conditions fails to produce a highly pure native conformation; under these circumstances a native purification protocol should be considered. Native purification also has the advantage of obtaining higher yields from the transcription reaction—it has been observed that larger RNAs can form irreversible aggregates or incompletely denature resulting in reduced yields using denaturing polyacrylamide gel electrophoresis (Lukavsky and Puglisi, 2004). For smaller RNAs (<250 nucleotides), a native RNA purification system has been developed that is capable of producing sufficient RNA for crystallography screening (Batey and Kieft, 2007). In this method, the RNA of interest is attached to the 50 -side of a purification tag comprising the glmS ribozyme and a small stem-loop RNA motif capable of binding the MS2 coat protein. The protein is hexahistidine tagged such that it binds to nickel affinity chromatographic resin, and thereby immobilizes full length transcript obtained from an in vitro transcription reaction. Addition of glucosamine-6-phosphate activates the ribozyme, resulting in cleavage and liberation of the target RNA from the column. For RNAs larger than 250 nucleotides, an alternative protocol is to treat the transcription reaction with DNase I (to remove PCR templates and RNA/DNA hybrids) and then with proteinase K (to remove T7 RNA polymerase, DNase I, and inorganic pyrophosphatase). The protein-free RNA transcription is then buffer exchanged and concentrated centrifugal concentrators with a nominal molecular weight limit greater than 30 kDa (for proteinase K). This approach was used to purify and crystallize the 411-nucleotide self-splicing group II intron (Toor et al., 2008). Size exclusion chromatography provides another route in which RNA can be purified natively (Doudna et al., 1993; Lukavsky and Puglisi, 2004). After the transcription is complete, the reaction is extracted with a phenol: chloroform solution to remove proteins, and applied to a size exclusion column of an exclusion limit of approximately 350 base pairs (bp) to separate the template DNA plasmid from the target RNA. In comparison to the previous method, the phenol:chloroform extraction to remove proteins prevents prolonged incubation of the target RNA in magnesiumcontaining buffers, which may prevent degradation of the desired product.
124
Francis E. Reyes et al.
Ultimately, a number of different protocols should be attempted, and assessed to maximize RNA yield and quality.
3. Construction of Library of RNAs for Crystallization Trials 3.1. Choosing phylogenetic variants of conserved function but variable periphery domains from a sequence alignment One means of fully exploring the sequence space approach to crystallography is to test a series of different orthologs of the RNA of interest—a wellvalidated crystallographic approach that exploits the structural diversity found in nature while retaining the desired functional characteristics. Choosing RNAs that will be screened for crystallizability should be guided by phylogenetic sequence alignments, which are generally good predictors of the core functional elements of the RNA in the absence of experimental data. The RNA Families (Rfam) database provides an excellent resource as it contains high-quality member and seed alignments for all known Rfam (Griffiths-Jones et al., 2005). From the sequence alignment, one can infer: (1) a secondary structure of the RNA, (2) possible tertiary contacts (in cases where the sequence composition of the RNA motif is defined such as a tetraloop receptor), and (3) residues whose identity or presence vary to serve as starting points for RNA engineering. The secondary structure then serves as a template to determine which residues can be altered as they are not well conserved1 and provide a set of phylogenetic variants in which to attempt crystallization trials2 (see Section 3.2). Several recent RNA structural studies illustrate the importance of considering phylogeny in RNA crystallization trials. Interestingly, the glmS ribozyme (Klein and Ferre´-D’Amare´, 2006), many riboswitches (Garst et al., 2008; Montange and Batey, 2006; Spitale et al., 2009), RNAse P RNA (Kazantsev et al., 2005), and several ribosome structures (Ban et al., 1999; Selmer et al., 2006) were all derived from thermophilic organisms. This occurrence mimics a trend often seen in protein crystallography where thermophiles are used because they are predicted to have fewer disordered 1
2
The tertiary and secondary structures for a number of RNAs were deduced from their sequence alignments (Brown, 1998; Chen et al., 2000; Damberger and Gutell, 1994; Gutell, 1993; Larsen et al., 1998; Michel et al., 1989; Romero and Blackburn, 1991; Schnare et al., 1996; Szymanski et al., 1998) and experimental data showed such predictions are accurate. A number of bioinformatic programs can calculate the distance between RNA secondary structures and perform clustering analysis to identify ensembles of related structures (Liu et al., 2008; Torarinsson et al., 2007). Partitioning the alignment in a set of ensembles thus reduces the total number of phylogenetic variants to be considered and single representative from each ensemble can be subjected to crystallization trials.
Strategies in RNA Crystallography
125
residues (Savchenko et al., 2003). The size of the RNA has shown to be an important factor as well. RNA sequences from organisms with relatively small genomes may have been selected by evolution to optimize function. For example, the flavin mononucleotide (FMN) riboswitch was solved using a representative from Fusobacterium nucleatum, an organism with a minimized genome (Serganov et al., 2009). The structure of the S-adenosylmethionine riboswitch (SAM-II) was solved by screening at least 13 phylogenetic variants and represents the smallest sequence of this family (Gilbert et al., 2008). Thus, it is important to consider a number of variants from various organisms whose RNAs may display different properties.
3.2. Consideration of intermolecular and intramolecular RNA packing The most distinguishable feature of RNA is the dominance of the A-form helix as the basis for secondary structure. Simple nucleic acid helices, as initially characterized in crystallographic structures of DNA, tend to coaxially stack to form a pseudo-continuous helix in the crystal lattice (Drew et al., 1981). The modes of stacking are generally limited to end-on-end blunt stacking or the packing of the blunt end into the minor groove of a neighboring helix (Batey et al., 1999). Helices containing overhanging (nonpaired) nucleotides at their ends can coaxially stack via Watson–Crick or base triples between adjacent helices (Mooers, 2009). These interaction modes are also observed to be important for the close packing of RNA helices (Strobel and Doudna, 1997). Therefore, one of the most often used strategies to create sequence variants is to explore different terminal ends of helices (such as the introduction of blunt or staggered ends). As coaxial stacking of helices in adjacent RNA molecules in the lattice is commonplace, utmost consideration must be made to the composition and homogeneity of the 50 and 30 ends of the RNA, whether a blunt or staggered end configuration is desired. Although T7 RNA polymerase produces robust transcription yields in vitro, it has the tendency to add nontemplated nucleotides at the 50 (Pleiss et al., 1998) and 30 (Milligan et al., 1987) ends. A number of methods have been developed to combat these forms of heterogeneity such as the incorporation of alternative T7 promoters for improved 50 -homogeneity (Coleman et al., 2004), cis-cleaving RNA ribozymes for both 50 - and 30 -homogeneity (Price et al., 1995), or the use of 20 -O-methyl terminated templates to be used in transcription for 30 homogeneity (Kao et al., 2001; Martick and Scott, 2006). In our laboratory, we favor the use of 20 -O-methylated templates to minimize 30 -end heterogeneity and limiting transcription times to no more than 2 h to prevent excessive 50 -end heterogeneity. The terminal loops of RNA present an additional source of generating construct variety. In cases where the sequence of a particular terminal loop is
126
Francis E. Reyes et al.
poorly conserved, it can be substituted with known RNA motifs known to promote intermolecular contacts. The two approaches most often employed are the inclusion of the GNRA tetraloop or the 21 nucleotide U1A hairpin to the human U1A RNA binding protein. The GNRA (where N is any nucleotide and R is any purine) tetraloop is a common motif in natural RNAs that promotes internal RNA helical packing. For instance, the GAAA tetraloop can dock into an A-form RNA helix into the minor groove, forming a set of interactions called A-minor triples (Costa and Michel, 1995; Doherty et al., 2001; Ferre´-D’Amare´ et al., 1998; Pley et al., 1994a,b), or dock with a defined internal loop motif called a tetraloop receptor (Cate et al., 1996; Murphy and Cech, 1994). The versatility of this motif in mediating RNA packing has lead to its use as a general module for RNA hairpin loops in crystallography. A study in which the P4–P6 domain of the self-splicing Tetrahymena group I intron was extensively mutagenized and screened for crystallizability suggested the location of the GNRA tetraloop motif on specific ends of helices were critical to obtaining high quality crystals (Golden et al., 1997). Furthermore, the identity of the nucleotides in the GNRA tetraloop contributed to dramatically different crystal forms as well (Golden and Kundrot, 2003). These studies suggest that the composition and the location of the GNRA tetraloop is an important factor in crystallization trials. The human U1A RNA binding domain (RBD) is a small 11 kDa protein capable of binding with high affinity to a conserved 21 nucleotide hairpin (Oubridge et al., 1994). As the contacts with the protein are limited to the terminal loop of the hairpin, it can be used as a motif to cap the ends of helices in larger RNAs (Ferre´-D’Amare´ et al., 1998). In this method, purified U1A protein is bound to the folded RNA and the resulting complex used in crystallization trials. Association of the U1A RNA binding protein with the target RNA is assessed via size-exclusion chromatography or native gel electrophoresis. This protein has the added benefit that it often promotes intermolecular contacts, as has been observed for the glmS ribozyme (Cochrane et al., 2007), the self-splicing group I intron (Adams et al., 2004), and the human d virus ribozyme (Ke et al., 2004). The U1A protein can also be derivatized with selenomethionine, thereby serving as a source of experimental phases via multiple-wavelength anomalous dispersion (MAD) (Ferre´-D’Amare´ et al., 1998). For these reasons, the U1A protein has proven itself to be a robust tool in RNA crystallography.
3.3. Variation of peripheral helical lengths With a sequence alignment and secondary structure, one can reasonably assess the length requirement for each helix, as this is an important parameter in creating an initial library of variants. For example, crystals of a l
Strategies in RNA Crystallography
127
repressor–operator complex were obtained after systematic variation of the DNA helix length ( Jordan et al., 1985). The authors reasoned that changing the length of the helix not only extends the helix but also changes the azimuthal relationship of protein molecules bound to the helix. Another documented case of helix length as a critical variable is the SRP ribonucleoprotein complex. In this example, the length of the RNA hairpin was critical for crystallizability as variations of the helix length either failed to produce crystals or lowered the resolution limit (Batey et al., 2001). Examination of the lattice revealed that the proper helical length resulted in GAAA tetraloops pointing in the same direction in a pseudo-dimer. This allowed the RNA to orient in a fashion that promoted GAAA-minor groove packing interactions between adjacent molecules in the lattice.
4. Improving Crystal Quality Through Postcrystal Analysis At this stage, a particular sequence may reproducibly yield welldefined crystals that either diffract to low resolution, are highly mosaic or twinned, or yield poor or partial experimental phases. Before abandoning such a construct, several methods to improve intermolecular packing or reduce intramolecular disorder should be considered, as well as analyzing the contents of the crystal to determine changes in the RNA. Small changes within the construct, such as point mutations to residues that form the lattice contacts, can have dramatic consequences on the ordering of the crystal lattice. The following section discusses a number of postcrystallization techniques and approaches toward modifying the crystallization process (both in construct and condition space) in hopes of improving crystal quality.
4.1. Cation additives to improve crystal quality Crystallization additives are a common method to improve crystal quality in proteins (McPherson and Cudney, 2006). With RNA, it is common to screen a series of different cations, as their role in RNA structure and catalysis is well documented for many RNA and RNA–protein systems (Pyle, 2002). Since various cations will interact with the RNA differently, our lab uses a cation screen comprised simple metal cations and polyamines (Table 6.1). Each solution in this table is a 10 stock that is added to an optimized condition (or one found in a sparse matrix) and the set of 24 conditions assessed for potential improvement in crystal quality.
128
Francis E. Reyes et al.
Table 6.1 Additive screen for RNA No.
Additive (10 stock)
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.
0.1 M nickel chloride 0.1 M magnesium chloride 0.1 M cobalt(II) chloride 0.1 M cadmium chloride 0.1 M zinc(II) chloride 0.1 M calcium chloride 0.1 M strontium chloride 0.1 M barium chloride 0.1 M manganese(II) chloride 0.05 M hexamine cobalt(III) chloride 0.05 M samarium(II) acetate 0.05 M terbium(III) chloride 0.05 M dysprosium(III)chloride 0.05 M gadolinium(III) chloride 1.0 M lithium chloride 1.0 M potassium chloride 1.0 M cesium chloride 0.1 M imidazole 0.2 M spermine tetrahydrochloride 0.2 M spermidine trihydrochloride 0.2 M cystamine dihydrochloride 0.2 M putrescine dihydrochloride 0.05 M hexamine iridium(III) chloride Water (control)
4.2. Determine and vary the residues involved in crystal packing In designing a strategy for altering an RNA sequence that favors a more ordered lattice, choosing the location and type of mutation to be made is often quite ambiguous. As discussed above, the 50 - and 30 -ends of the RNA or terminal hairpin loops are usually involved in crystal contacts and are obvious sites of further variation. For example, we typically construct an initial library in which all RNAs have a single adenosine overhang at the 30 -end, the identity of the nucleotide (or importance of its presence) is tested by making overhangs with the other three nucleotides and deleting the overhang. It is also likely that residues internal to the RNA are involved in crystal packing that may not be easily inferred. Hence, one is faced with either systematic blind screening or identification of a crystal contact and subsequent focused engineering to improve that crystal contact. Fortunately, RNA is amenable to structural probing revealing the relative chemical environment for each nucleotide in the RNA.
Strategies in RNA Crystallography
129
One means of determining sites of potential lattice contacts without diffraction data is chemical probing, which allows one to assess the relative reactivity of each nucleotide in the RNA when the RNA is in solution (ideally, the mother liquor) versus the crystal lattice (Vicens et al., 2007). The preferred reagent is N-methylisatoic anhydride (NMIA) because it does not have a preference for any base, and mostly reacts with structurally dynamic 20 -hydroxyl functional groups (Gherghe et al., 2008). Nucleotides involved in crystal contacts are expected to be in a restricted conformation relative to solution conditions and hence have lower reactivity to the probing agent in the crystal lattice compared with that of the RNA free in solution. In a study of the crystals of the P4–P6 domain of the self-splicing Tetrahymena group I intron, Cech and coworkers were able to verify 14 out of 22 (64%) nucleotides involved in crystal packing by this procedure (Vicens et al., 2007). Each crystal contact then defines a region as a target for mutational analysis.
4.3. Directed mutagenesis using a crystal structure In some cases, an initial structure can be obtained at lower resolution (3.0–4.0 A˚ resolution), but requires higher resolution data before details of interest emerge such a ligands or the role of metal ions. The initial model provides specific information about the sites of the RNA that forms intermolecular contacts and thus a highly directed set of mutations can be tested that may improve the packing. For example, the structure SAM-I riboswitch was refined to 2.9 A˚ resolution, which did not allow for certain details of the RNA–ligand complex to be unambiguously determined (Montange and Batey, 2006). To improve the resolution of this structure and reveal additional features, a series of variants were made that targeted sites of lattice contacts. As a result of over 20 directed changes, we observed two mutations that improved the diffraction limit of this RNA: changing the 30 -overhanging residue from adenosine to guanosine improved the coaxial stacking of the P1 helix with the adjacent molecule, and a mutation of a bulged uridine to a cytidine in the kink-turn (manuscript in preparation). Reassuringly, over half of the mutants tested still yielded crystals, most diffracting to around the original resolution limit, indicating that a lattice may be quite forgiving to small conservative changes in the RNA sequence. In making subtle mutations in the RNA of interest, it is important to note that the absence of crystal growth in the same conditions as before the mutation is not indicative of failure to improve a crystal contact. The mutation may have introduced an intermolecular contact that does not crystallize in the same conditions. Therefore, with every mutation, it is recommended that the construct be rescreened against commercially available matrix screens in addition to the same condition that produced a crystal before the mutation was made.
130
Francis E. Reyes et al.
4.4. Analyze the crystal for RNA cleavage RNA degradation via hydrolysis of the phosphodiester backbone is generally viewed as an impediment to crystallization, particularly in the RNA preparation step where the uncontrolled hydrolysis of an RNA leads to chemical heterogeneity. However, controlled hydrolysis of the RNA backbone during crystallization, otherwise known as ‘‘in-drop digestion,’’ can be advantageous. This is analogous to limited proteolytic degradation in several crystallization efforts that lead to crystals of a degradation product (Campbell et al., 2002; Sawaya et al., 1994). Hydrolysis of the RNA is controlled because of the stereochemical restraints of the hydrolysis reaction: the 20 -hydroxyl group of the ribose sugar attacks the phosphate group in an in-line configuration to yield a cleavage of the backbone immediately 30 of the attacking hydroxyl group. The attack does not occur with bases in an A-form helix or a restricted conformation (as the hydroxyl must be able to sample the in-line configuration) and hence occurs in regions that are conformationally flexible. Sites of backbone cleavage can be readily resolved by P labeling the crystallized RNA and separation of the products on a denaturing polyacrylamide gel. Isolation of each cleavage product and enzymatic sequencing reveals the location of the cleavage event. Managing sites of conformational flexibility as revealed by limited hydrolysis may lead to better diffracting crystals as a result of reduced disorder of the flexible area or creating a conformationally homogeneous site of a lattice contact. A popular approach is to design a two-piece RNA system defined by the cleavage position. The separate pieces are annealed postpurification and subjected to crystallography trials. Golden and Cech realized that the P4–P6 domain contained cleavages due to crystallization and used a two-piece system as a means to introduce heavy atom nucleotide derivatives by chemically synthesizing a small oligonucleotide with 5-iodouridine (Golden et al., 1996). The glmS ribozyme (Klein and Ferre´D’Amare´, 2009) and the FMN riboswitch (Serganov et al., 2009) specifically used two-piece systems to improve crystallizability and diffraction resolution. In the case of FMN, the loss of the loop capping helix 4 allowed for improved lattice contacts. In both cases, the use of a two-piece system may have relieved backbone restraint and allowed for an alternate conformation that was necessary for establishing a tighter crystal packing.
5. Phasing Methods Computation of the electron density map from X-ray diffraction data requires knowledge of the intensities and the phase angle of each measured reflection. The lack of phase angle information in recorded diffraction
Strategies in RNA Crystallography
131
images serves as the basis of the ‘‘phase problem’’ in X-ray crystallography. A number of classical methods have been developed to solve the phase problem including the use of heavy atom derivatives, anomalous scatterers, or model phases from a molecular replacement solution. Unlike in vivo incorporation of selenium-labeled methionine amino acids that has greatly aided protein crystallography, the soaking of heavy metal cations into fortuitous sites still dominates in RNA crystallography. Fortunately, a number of experimental methods specific for RNA have been developed to address this issue.
5.1. Include heavy metal cations and heavy nucleotide derivatives Preparation of a heavy atom derivative of an RNA crystal is typically achieved by cocrystallizing or soaking it in the presence of a heavy metal cation. As RNAs have a high affinity and usually require magnesium, heavy metal divalent cations such as zinc (Ennifar et al., 2001) and barium (Tereshko et al., 2003) can be often be substituted. More exotic metals such as ytterbium (Toor et al., 2008), europium (Guo et al., 2004), and dysprosium (Guo et al., 2004) have been used as well. By far, the most successful method for introducing heavy metal atoms into RNA crystals for phasing is the use of hexammine salts of iridium(III) and osmium(III). Although these salts were initially used as part of multiple isomorphous replacement strategies (Cate et al., 1996), they have anomalous scattering properties as well. To further extend the utility of these compounds, a ‘‘phasing module’’ was developed to promote heavy metal binding within the major groove of a standard A-form helix (Keel et al., 2007). This motif has the ability to bind metal cations due to the electron rich major groove face of a central GU wobble pair and therefore serves as a site-specific phasing module for RNA crystallography. Cobalt(III) hexammine can also be useful as a derivitizing agent because it is commercially available and anomalously scatters on Cu-Ka sources for ‘‘in-house’’ single-wavelength anomalous dispersion (SAD) phasing (Keel et al., 2007). For increased anomalous scattering properties, iridium(III) hexammine and its synthesis has been published (Edwards et al., 2009). A number of recent crystal structures owe their experimental phases to the use of hexammine salts such as the glmS ribozyme (Cochrane et al., 2007), the purine (Batey et al., 2004), SAM-II (Batey et al., 2004), SAM-I (Montange and Batey, 2006), and lysine (Garst et al., 2008) riboswitches, the group II intron (Toor et al., 2008), and the 30 S ribosomal subunit (Wimberly et al., 2000). Heavy atom incorporation for phasing can also be accomplished through the use of modified nucleosides. Specifically, 5-bromouridine (Adams et al., 2004; Baugh et al., 2000; Kieft et al., 2002; Martick and Scott, 2006), 5-iodouridine (Klein et al., 2009), 50 -a-P-seleno-triphosphates (Brandt et al., 2006),
132
Francis E. Reyes et al.
20 -methylseleno-phosphoramadites (Carrasco et al., 2004; Ho¨bartner and Micura, 2004; Ho¨bartner et al., 2005) have been used to covalently derivatize RNA for phasing. These analogs can be incorporated into RNAs via solid phase synthesis, in the case of phosphoramadite analogs, or enzymatic synthesis, in cases where the analogs are triphosphates. Synthetic oligonucleotides can be annealed to the parent RNA in a two-piece system as was used in the P4–P6 domain of the group I intron (Golden et al., 1996). These methods have the advantage over the direct soak strategy described above in that it assures incorporation of the heavy atom in a site-specific manner. Site-specific incorporation can also be helpful in structure solution as its location can aid in establishing the proper register in model building for lower resolution data.
5.2. Molecular replacement with RNA fragments A common practice in DNA crystallography is to use B-form helices as models for phasing by molecular replacement (Ramakrishnan and Sundaralingam, 1993). Although RNA contains regular helical secondary structure, it is complicated by tertiary interactions, making molecular replacement with A-form helices difficult because the organization of the helices is unknown. However, advances in molecular replacement routines have reduced this challenge. In a recent determination of the structure of a ligase ribozyme, an iterative molecular replacement procedure was used to obtain suitable phases by using A-form helices capped with GNRA tetraloops as the initial search model (Robertson and Scott, 2008). A more recent example is the structure solution of the 34-nucleotide preQ1 riboswitch (Klein et al., 2009). Molecular replacement solutions were chosen based on fragment packing, the model refined, and manually edited to include or omit nucleotides based on electron density observed in 2Fo–Fc maps. The phases of the resulting molecular replacement solution were then treated as an experimental phase set and density modified producing a pseudoexperimental map in which the entire model was rebuilt. This approach illustrates how iterative molecular replacement can be used to phase RNA crystals in cases where obtaining experimental phases is difficult.
6. A Case Study in the Crystallization of Lysine Riboswitch Regulatory Element Riboswitches are RNA regulatory elements capable of binding small molecule metabolites to regulate the mRNA in which they are embedded (Montange and Batey, 2008; Winkler and Breaker, 2005). Our lab recently solved the crystal structure of the lysine riboswitch ligand-binding domain
133
Strategies in RNA Crystallography
from the Thermotoga maritima asd mRNA (Garst et al., 2008). The lysine binding riboswitch is an interesting structural target because resistance to antimicrobial lysine analogs such as S-(2-aminoethyl)-L-cysteine in Escherichia coli and Bacillus subtilis is the result of mutations within the regulatory element (Lu et al., 1992; Patte et al., 1998). The ligand-binding domain of this riboswitch has a conserved core centered about a five-way helical junction (Fig. 6.2A). The flanking helices contain a number of known RNA motifs such as a kink-turn (Klein et al., 2001), a sarcin/ricin loop (Szewczak et al., 1993) motif, and a kissing–loop interaction between helices P2 and P3. Along with several phylogenetically conserved adenosine residues in the terminal loop of P4, the observed pattern of conservation limits variation of peripheral elements primarily to P1 and P5. Initial construct screening implemented the sequence space approach that we described above. Four lysine riboswitches from the Rfam database were considered from the following organisms: T. maritima, Haemophilus influenzae, E. coli, and B. subtilis. Due to the rigidity in helix length in P2, P3, and P4 helices, helix variants in the P1 and P5 helices were made (Fig. 6.2B). A
B
G G G A C
U A
U C
A U
90 - U G A
P3
G
U
0
G G
A
C C
−4
Kissing loops
G G A G
G
P2a
-
50 -
Kink turn
C C C U G U
A
C
30
G
C C C
C
G
G
C
G
U-
A
G
U
A
G A
A
G U G
Loop E A
A
A
C
G
G
120 - G
C C
A
G
C
C
G-
C
G
UU A
C
A C
U G
C
G
A G U
G A
C U
G
C
G
C
C
G
G
C
G
U G
G
C
G
A
G
C
C
G
80 - G
C
A
U
0 −11 G
P1
C G
A
G
G
C
G
C
C
G
A
U
G
C
1-G
G
U
A
G
P5 130
P4
A
CA
A
A GGCGG A G UCA CGG A GGU G C C
A A G A A
GGUG C C U C A
U
G
C-
140
C
G
P5
G
C
A
160
C C G U C G G C C A
C G
50 -
G G C A G C C G G
G
U C A C GGAG
Lys
A G
C C G U C C C A
A A G UCGCC A
U
G
A
10 - c G
G G C A G G G
GGC G
G
70
C C G U C C A
G
P2
G G C A G G
A
60
A G G
C U −100
C C G C C
UCGC
20 - A G
P1
G G C G G
G A A
Figure 6.2 (A) The tertiary architecture of the Thermotoga maritima lysine riboswitch. (B) The different helix variants used in our initial crystallization library are shown for P1 and P5 (highlighted in red). (C) A crystal from a P1 helix length of six and P5 helix of four.
134
Francis E. Reyes et al.
Each construct was cloned into a vector containing a cis-cleaving HdV ribozyme that yields an overhanging adenosine on the 30 -end, and in vitro transcribed using established methods (Edwards et al., 2009). The transcription reaction was purified via denaturing polyacrylamide gel electrophoresis and the RNA refolded in the presence of 1 mM lysine. Each construct was subjected to commercially available sparse matrix screens (Natrix, CrystalScreen-I, and PEG/Ion from Hampton Research) using a Fluidigm TOPAZ Protein Crystallization system (Hansen et al., 2002) at 30 C. The initial screening process yielded a number of crystals for the species T. maritima, B. subtilis and H. influenzae, all with similar lengths in the P1 and P5 helices. Conventional hanging drop trays (24 well) were made for each condition that produced a crystal in the initial screening, except the concentration of salt, precipitant, and RNA around the initial hit were varied using a grid screen approach. Crystals from these focused screens were tested for diffraction. Ultimately, the construct from T. maritima containing 6 bp in P1 and 4 bp in P5 (based on Rfam phylogenetic consensus models) was grown in 2 M Li2SO4, 5 mM MgCl2, 10 mM Na-HEPES, pH 7.0, and 60 mM iridium hexamine diffracted to high ˚ ) on a home X-ray source. resolution (<3 A The crystals were then regrown in the presence of cobalt(III) and iridium(III) hexammine and the former screened for anomalous signal on the home source. Anomalous difference Patterson maps revealed heavy atom positions suggesting that cobalt(III) hexammine had bound to the RNA and maybe suitable for structure solution via SAD techniques. A full SAD dataset to ˚ was collected at the Brookhaven National Laboratory X29 beamline 2.8 A using iridium derivatized crystals. Heavy atom positions for iridium and SAD phasing were calculated using SHELX (Sheldrick, 2008). A total of four heavy atom positions were found, two of which had an occupancy greater than 0.5. The SAD electron density map had clear features for the phosphate backbone and was able to distinguish between purines and pyrimidines. The entire RNA was then built in COOT (Emsley and Cowtan, 2004) and refined using PHENIX (Bru¨nger et al., 1998). Simulated annealing omit maps of the ligand pocket showed clear density for the entire lysine ligand, indicating that the atomic-level details of lysine recognition by a natural RNA could be inferred from the crystallographic model.
7. Concluding Remarks In this chapter, a general workflow has been presented that provides a guide from initial construct design to structure solution. Each step employs techniques specific to RNA, from sequence engineering, crystal analysis,
Strategies in RNA Crystallography
135
and subsequent mutations to improve crystal quality and obtaining phases for structure solution. It is a goal of this protocol to streamline the RNA crystallographic process, relying on blind screening in the initial phases of the project but also using approaches that will hopefully facilitate being able to rapidly identification of an initial crystal hit. This methods protocol has relied heavily on decades of RNA structural research and has been successfully used by our laboratory to solve a number of novel structures. However, in our experience, each RNA structural effort presents novel challenges in solving a structure and therefore requires a flexible experimental strategy that can incorporate a variety of these approaches and tailor them toward the specific RNA of interest and the unique problems faced during the crystallization effort.
ACKNOWLEDGMENTS The authors thank Quentin Vicens and Jennifer Pfingsten for thoughtful reviews and suggestions to improve this chapter. This work was supported by a grant to R. T. B. from the National Institutes of Health (GM 083953).
REFERENCES Adams, P. L., Stahley, M. R., Kosek, A. B., Wang, J., and Strobel, S. A. (2004). Crystal structure of a self-splicing group I intron with both exons. Nature 430, 45–50. Ban, N., Nissen, P., Hansen, J., Capel, M., Moore, P. B., and Steitz, T. A. (1999). Placement of protein and RNA structures into a 5 A-resolution map of the 50S ribosomal subunit. Nature 400, 841–847. Batey, R. T., and Kieft, J. S. (2007). Improved native affinity purification of RNA. RNA 13, 1384–1389. Batey, R., Rambo, R., and Doudna, J. (1999). Tertiary motifs in RNA structure and folding. Angew. Chem. Int. Ed. Engl. 38, 2326–2343. Batey, R. T., Sagar, M. B., and Doudna, J. A. (2001). Structural and energetic analysis of RNA recognition by a universally conserved protein from the signal recognition particle. J. Mol. Biol. 307, 229–246. Batey, R. T., Gilbert, S. D., and Montange, R. K. (2004). Structure of a natural guanineresponsive riboswitch complexed with the metabolite hypoxanthine. Nature 432, 411–415. Baugh, C., Grate, D., and Wilson, C. (2000). 2.8 A crystal structure of the malachite green aptamer. J. Mol. Biol. 301, 117–128. Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N., and Bourne, P. E. (2000). The protein data bank. Nucleic Acids Res. 28, 235–242. Brandt, G., Carrasco, N., and Huang, Z. (2006). Efficient substrate cleavage catalyzed by hammerhead ribozymes derivatized with selenium for X-ray crystallography. Biochemistry 45, 8972–8977. Brown, J. W. (1998). The ribonuclease P database. Nucleic Acids Res. 26, 351–352. Bru¨nger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., GrosseKunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J.,
136
Francis E. Reyes et al.
Rice, L. M., et al. (1998). Crystallography and NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921. Campbell, E. A., Muzzin, O., Chlenov, M., Sun, J. L., Olson, C. A., Weinman, O., TresterZedlitz, M. L., and Darst, S. A. (2002). Structure of the bacterial RNA polymerase promoter specificity sigma subunit. Mol. Cell 9, 527–539. Carrasco, N., Buzin, Y., Tyson, E., Halpert, E., and Huang, Z. (2004). Selenium derivatization and crystallization of DNA and RNA oligonucleotides for X-ray crystallography using multiple anomalous dispersion. Nucleic Acids Res. 32, 1638–1646. Cate, J. H., Gooding, A. R., Podell, E., Zhou, K., Golden, B. L., Kundrot, C. E., Cech, T. R., and Doudna, J. A. (1996). Crystal structure of a group I ribozyme domain: Principles of RNA packing. Science 273, 1678–1685. Chen, J. L., Blasco, M. A., and Greider, C. W. (2000). Secondary structure of vertebrate telomerase RNA. Cell 100, 503–514. Cochrane, J. C., Lipchock, S. V., and Strobel, S. A. (2007). Structural investigation of the GlmS ribozyme bound to its catalytic cofactor. Chem. Biol. 14, 97–105. Coleman, T. M., Wang, G., and Huang, F. (2004). Superior 50 homogeneity of RNA from ATP-initiated transcription under the T7 phi 2.5 promoter. Nucleic Acids Res. 32, e14. Costa, M., and Michel, F. (1995). Frequent use of the same tertiary motif by self-folding RNAs. EMBO J. 14, 1276–1285. Damberger, S. H., and Gutell, R. R. (1994). A comparative database of group I intron structures. Nucleic Acids Res. 22, 3508–3510. DeRose, V. J. (2002). Two decades of RNA catalysis. Chem. Biol. 9, 961–969. Doherty, E. A., Batey, R. T., Masquida, B., and Doudna, J. A. (2001). A universal mode of helix packing in RNA. Nat. Struct. Biol. 8, 339–343. Doudna, J. A., Grosshans, C., Gooding, A., and Kundrot, C. E. (1993). Crystallization of ribozymes and small RNA motifs by a sparse matrix approach. Proc. Natl. Acad. Sci. USA 90, 7829–7833. Drew, H. R., Wing, R. M., Takano, T., Broka, C., Tanaka, S., Itakura, K., and Dickerson, R. E. (1981). Structure of a B-DNA dodecamer: Conformation and dynamics. Proc. Natl. Acad. Sci. USA 78, 2179–2183. Edwards, A. L., Garst, A. D., and Batey, R. T. (2009). Determining structures of RNA aptamers and riboswitches by X-ray crystallography. Methods Mol. Biol. 535, 135–163. Emsley, P., and Cowtan, K. (2004). Coot: Model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132. Ennifar, E., Walter, P., and Dumas, P. (2001). An efficient method for solving RNA structures: MAD phasing by replacing magnesium with zinc. Acta Crystallogr. D Biol. Crystallogr. 57, 330–332. Fedor, M. J., and Westhof, E. (2002). Ribozymes: The first 20 years. Mol. Cell 10, 703–704. Ferre´-D’Amare´, A. R., and Burley, S. K. (1994). Use of dynamic light scattering to assess crystallizability of macromolecules and macromolecular assemblies. Structure 2, 357–359. Ferre´-D’Amare´, A. R., Zhou, K., and Doudna, J. A. (1998). A general module for RNA crystallization. J. Mol. Biol. 279, 621–631. Garst, A. D., He´roux, A., Rambo, R. P., and Batey, R. T. (2008). Crystal structure of the lysine riboswitch regulatory mRNA element. J. Biol. Chem. 283, 22347–22351. Gherghe, C. M., Shajani, Z., Wilkinson, K. A., Varani, G., and Weeks, K. M. (2008). Strong correlation between SHAPE chemistry and the generalized NMR order parameter (S2) in RNA. J. Am. Chem. Soc. 130, 12244–12245. Gilbert, S. D., Stoddard, C. D., Wise, S. J., and Batey, R. T. (2006). Thermodynamic and kinetic characterization of ligand binding to the purine riboswitch aptamer domain. J. Mol. Biol. 359, 754–768. Gilbert, S. D., Rambo, R. P., Tyne, D. V., and Batey, R. T. (2008). Structure of the SAM-II riboswitch bound to S-adenosylmethionine. Nat. Struct. Mol. Biol. 15, 177–182. Golden, B. L., and Kundrot, C. E. (2003). RNA crystallization. J. Struct. Biol. 142, 98–107.
Strategies in RNA Crystallography
137
Golden, B. L., Gooding, A. R., Podell, E. R., and Cech, T. R. (1996). X-ray crystallography of large RNAs: Heavy-atom derivatives by RNA engineering. RNA 2, 1295–1305. Golden, B. L., Podell, E. R., Gooding, A. R., and Cech, T. R. (1997). Crystals by design: A strategy for crystallization of a ribozyme derived from the Tetrahymena group I intron. J. Mol. Biol. 270, 711–723. Griffiths-Jones, S., Moxon, S., Marshall, M., Khanna, A., Eddy, S. R., and Bateman, A. (2005). Rfam: Annotating non-coding RNAs in complete genomes. Nucleic Acids Res. 33, D121–D124. Guo, F., Gooding, A. R., and Cech, T. R. (2004). Structure of the Tetrahymena ribozyme: Base triple sandwich and metal ion at the active site. Mol. Cell 16, 351–362. Gutell, R. R. (1993). Collection of small subunit (16S- and 16S-like) ribosomal RNA structures. Nucleic Acids Res. 21, 3051–3054. Hansen, C. L., Skordalakes, E., Berger, J. M., and Quake, S. R. (2002). A robust and scalable microfluidic metering method that allows protein crystal growth by free interface diffusion. Proc. Natl. Acad. Sci. USA 99, 16531–16536. Ho¨bartner, C., and Micura, R. (2004). Chemical synthesis of selenium-modified oligoribonucleotides and their enzymatic ligation leading to an U6 SnRNA stem-loop segment. J. Am. Chem. Soc. 126, 1141–1149. Ho¨bartner, C., Rieder, R., Kreutz, C., Puffer, B., Lang, K., Polonskaia, A., Serganov, A., and Micura, R. (2005). Syntheses of RNAs with up to 100 nucleotides containing sitespecific 20 -methylseleno labels for use in X-ray crystallography. J. Am. Chem. Soc. 127, 12035–12045. Holbrook, S. R., Holbrook, E. L., and Walukiewicz, H. E. (2001). Crystallization of RNA. Cell. Mol. Life Sci. 58, 234–243. Jordan, S. R., Whitcombe, T. V., Berg, J. M., and Pabo, C. O. (1985). Systematic variation in DNA length yields highly ordered repressor-operator cocrystals. Science 230, 1383–1385. Kao, C., Ru¨disser, S., and Zheng, M. (2001). A simple and efficient method to transcribe RNAs with reduced 30 heterogeneity. Methods 23, 201–205. Kazantsev, A. V., Krivenko, A. A., Harrington, D. J., Holbrook, S. R., Adams, P. D., and Pace, N. R. (2005). Crystal structure of a bacterial ribonuclease P RNA. Proc. Natl Acad. Sci. USA 102, 13392–13397. Ke, A., and Doudna, J. A. (2004). Crystallization of RNA and RNA-protein complexes. Methods 34, 408–414. Ke, A., Zhou, K., Ding, F., Cate, J. H. D., and Doudna, J. A. (2004). A conformational switch controls hepatitis delta virus ribozyme catalysis. Nature 429, 201–205. Keel, A., Rambo, R., Batey, R., and Kieft, J. (2007). A general strategy to solve the phase problem in RNA crystallography. Structure 15, 761–772. Kieft, J. S., Zhou, K., Grech, A., Jubin, R., and Doudna, J. A. (2002). Crystal structure of an RNA tertiary domain essential to HCV IRES-mediated translation initiation. Nat. Struct. Biol. 9, 370–374. Klein, D. J., and Ferre´-D’Amare´, A. R. (2006). Structural basis of glmS ribozyme activation by glucosamine-6-phosphate. Science 313, 1752–1756. Klein, D. J., and Ferre´-D’Amare´, A. R. (2009). Crystallization of the glmS ribozymeriboswitch. Methods Mol. Biol. 540, 129–139. Klein, D. J., Schmeing, T. M., Moore, P. B., and Steitz, T. A. (2001). The kink-turn: A new RNA secondary structure motif. EMBO J. 20, 4214–4221. Klein, D. J., Edwards, T. E., and Ferre´-D’Amare´, A. R. (2009). Cocrystal structure of a class I preQ1 riboswitch reveals a pseudoknot recognizing an essential hypermodified nucleobase. Nat. Struct. Mol. Biol. 16, 343–344. Larsen, N., Samuelsson, T., and Zwieb, C. (1998). The signal recognition particle database (SRPDB). Nucleic Acids Res. 26, 177–178.
138
Francis E. Reyes et al.
Liu, Q., Olman, V., Liu, H., Ye, X., Qiu, S., and Xu, Y. (2008). RNACluster: An integrated tool for RNA secondary structure comparison and clustering. J. Comput. Chem. 29, 1517–1526. Lu, Y., Shevtchenko, T. N., and Paulus, H. (1992). Fine-structure mapping of cis-acting control sites in the lysC operon of Bacillus subtilis. FEMS Microbiol. Lett. 71, 23–27. Lukavsky, P. J., and Puglisi, J. D. (2004). Large-scale preparation and purification of polyacrylamide-free RNA oligonucleotides. RNA 10, 889–893. Magnet, S., and Blanchard, J. S. (2005). Molecular insights into aminoglycoside action and resistance. Chem. Rev. 105, 477–498. Martick, M., and Scott, W. G. (2006). Tertiary contacts distant from the active site prime a ribozyme for catalysis. Cell 126, 309–320. McPherson, A., and Cudney, B. (2006). Searching for silver bullets: An alternative strategy for crystallizing macromolecules. J. Struct. Biol. 156, 387–406. Michel, F., Umesono, K., and Ozeki, H. (1989). Comparative and functional anatomy of group II catalytic introns—A review. Gene 82, 5–30. Milligan, J., Groebe, D., Witherell, G., and Uhlenbeck, O. (1987). Oligoribonucleotide synthesis using T7 RNA polymerase and synthetic DNA templates. Nucleic Acids Res. 15, 8783. Montange, R. K., and Batey, R. T. (2006). Structure of the S-adenosylmethionine riboswitch regulatory mRNA element. Nature 441, 1172–1175. Montange, R. K., and Batey, R. T. (2008). Riboswitches: Emerging themes in RNA structure and function. Annu. Rev. Biophys. 37, 117–133. Mooers, B. H. M. (2009). Crystallographic studies of DNA and RNA. Methods 47, 168–176. Murphy, F. L., and Cech, T. R. (1994). GAAA tetraloop and conserved bulge stabilize tertiary structure of a group I intron domain. J. Mol. Biol. 236, 49–63. Oubridge, C., Ito, N., Evans, P. R., Teo, C. H., and Nagai, K. (1994). Crystal structure at 1.92 A resolution of the RNA-binding domain of the U1A spliceosomal protein complexed with an RNA hairpin. Nature 372, 432–438. Patte, J. C., Akrim, M., and Me´jean, V. (1998). The leader sequence of the Escherichia coli lysC gene is involved in the regulation of LysC synthesis. FEMS Microbiol. Lett. 169, 165–170. Pleiss, J. A., Derrick, M. L., and Uhlenbeck, O. C. (1998). T7 RNA polymerase produces 50 end heterogeneity during in vitro transcription from certain templates. RNA 4, 1313–1317. Pley, H. W., Flaherty, K. M., and McKay, D. B. (1994a). Model for an RNA tertiary interaction from the structure of an intermolecular complex between a GAAA tetraloop and an RNA helix. Nature 372, 111–113. Pley, H. W., Flaherty, K. M., and McKay, D. B. (1994b). Three-dimensional structure of a hammerhead ribozyme. Nature 372, 68–74. Price, S. R., Ito, N., Oubridge, C., Avis, J. M., and Nagai, K. (1995). Crystallization of RNA-protein complexes. I. Methods for the large-scale preparation of RNA suitable for crystallographic studies. J. Mol. Biol. 249, 398–408. Pyle, A. M. (2002). Metal ions in the structure and function of RNA. J. Biol. Inorg. Chem. 7, 679–690. Ramakrishnan, B., and Sundaralingam, M. (1993). High resolution crystal structure of the A-DNA decamer d(CCCGGCCGGG). Novel intermolecular base-paired G*(G.C) triplets. J. Mol. Biol. 231, 431–444. Robertson, M. P., and Scott, W. G. (2008). A general method for phasing novel complex RNA crystal structures without heavy-atom derivatives. Acta Crystallogr. D Biol. Crystallogr. D64, 738–744. Romero, D. P., and Blackburn, E. H. (1991). A conserved secondary structure for telomerase RNA. Cell 67, 343–353.
Strategies in RNA Crystallography
139
Savchenko, A., Yee, A., Khachatryan, A., Skarina, T., Evdokimova, E., Pavlova, M., Semesi, A., Northey, J., Beasley, S., Lan, N., Das, R., Gerstein, M., et al. (2003). Strategies for structural proteomics of prokaryotes: Quantifying the advantages of studying orthologous proteins and of using both NMR and X-ray crystallography approaches. Proteins 50, 392–399. Sawaya, M. R., Pelletier, H., Kumar, A., Wilson, S. H., and Kraut, J. (1994). Crystal structure of rat DNA polymerase beta: Evidence for a common polymerase mechanism. Science 264, 1930–1935. Schnare, M. N., Damberger, S. H., Gray, M. W., and Gutell, R. R. (1996). Comprehensive comparison of structural characteristics in eukaryotic cytoplasmic large subunit (23 S-like) ribosomal RNA. J. Mol. Biol. 256, 701–719. Selmer, M., Dunham, C. M., Murphy, F. V., 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. Serganov, A., Huang, L., and Patel, D. J. (2009). Coenzyme recognition and gene regulation by a flavin mononucleotide riboswitch. Nature 458, 233–237. Sharp, P. A. (2009). The centrality of RNA. Cell 136, 577–580. Sheldrick, G. M. (2008). A short history of SHELX. Acta Crystallogr. A Found Crystallogr. 64, 112–122. Spitale, R. C., Torelli, A. T., Krucinska, J., Bandarian, V., and Wedekind, J. E. (2009). The structural basis for recognition of the PreQ0 metabolite by an unusually small riboswitch aptamer domain. J. Biol. Chem. 284, 11012–11016. Strobel, S. A., and Doudna, J. A. (1997). RNA seeing double: Close-packing of helices in RNA tertiary structure. Trends Biochem. Sci. 22, 262–266. Szewczak, A. A., Moore, P. B., Chang, Y. L., and Wool, I. G. (1993). The conformation of the sarcin/ricin loop from 28S ribosomal RNA. Proc. Natl Acad. Sci. USA 90, 9581–9585. Szymanski, M., Specht, T., Barciszewska, M. Z., Barciszewski, J., and Erdmann, V. A. (1998). 5S rRNA data bank. Nucleic Acids Res. 26, 156–159. Tereshko, V., Skripkin, E., and Patel, D. J. (2003). Encapsulating streptomycin within a small 40-mer RNA. Chem. Biol. 10, 175–187. Toor, N., Keating, K. S., Taylor, S. D., and Pyle, A. M. (2008). Crystal structure of a self-spliced group II intron. Science 320, 77–82. Torarinsson, E., Havgaard, J. H., and Gorodkin, J. (2007). Multiple structural alignment and clustering of RNA sequences. Bioinformatics 23, 926–932. Uhlenbeck, O. C. (1995). Keeping RNA happy. RNA 1, 4–6. Vicens, Q., Gooding, A. R., Laederach, A., and Cech, T. R. (2007). Local RNA structural changes induced by crystallization are revealed by SHAPE. RNA 13, 536–548. Wimberly, B. T., Brodersen, D. E., Clemons, W. M., Morgan-Warren, R. J., Carter, A. P., Vonrhein, C., Hartsch, T., and Ramakrishnan, V. (2000). Structure of the 30S ribosomal subunit. Nature 407, 327–339. Winkler, W. C., and Breaker, R. R. (2005). Regulation of bacterial gene expression by riboswitches. Annu. Rev. Microbiol. 59, 487–517.
C H A P T E R
S E V E N
Comparative Gel Electrophoresis Analysis of Helical Junctions in RNA David M. J. Lilley Contents 1. Principle and Theory of Electrophoresis of Bent or Kinked Nucleic Acids 2. Discrete Bending or Kinking of the Axis of a Duplex 3. The Analysis of Helical Junctions Using the Long–Short Arm Method 4. Experimental Strategies and Methods 5. Kink Turns in RNA 6. Four-Way RNA Junctions 7. Three-Way RNA Junctions 8. Complexes of Branched RNA with Proteins Acknowledgments References
144 145 146 147 149 149 152 155 155 155
Abstract Comparative gel electrophoresis provides information on the relative angles subtended between helical arms at a branchpoint in RNA. It is based upon the comparison of electrophoretic mobility in polyacrylamide gels of species containing two long arms, with the remaining one(s) being significantly shorter. Although the method currently lacks a really well-established basis of physical theory, it is very powerful, yet simple to apply. It has had a number of significant successes in RNA, DNA and DNA–protein complexes, and in all cases to date the results have stood the test of time and eventual comparison with crystallographic analysis.
Helical branchpoints, or junctions, abound in natural RNA molecules, where they connect helical segments (Lilley, 2000). Three- and fourway helical junctions are common elements in small, autonomously folding RNA species such as several of the nucleolytic ribozymes, and have also been selected by a number of riboswitches to create pockets that selectively Cancer Research UK Nucleic Acid Structure Research Group, MSI/WTB Complex, The University of Dundee, Dundee, United Kingdom Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69007-8
#
2009 Elsevier Inc. All rights reserved.
143
144
David M. J. Lilley
bind small molecule ligands. The local geometry of junctions exerts a major influence upon the long-range architecture of large RNA molecules. Electrophoresis in a gel matrix provides a powerful method for providing relative angular information about the disposition of helical arms about a branch point in nucleic acid junctions. Although mostly applied to DNA junctions initially (Cooper and Hagerman, 1987; Duckett et al., 1988; Gough and Lilley, 1985), it has been extensively used to give information on junctions in RNA molecules. The method is very simple and requires inexpensive apparatus. Despite this, the approach has not been widely adopted in many laboratories. One reason may be that electrophoresis is not underpinned by a good theoretical understanding compared to other structural methods, so that the conclusions that emerge are not quantitative in general. The symmetry of a given junction may be determined, but interaxial angles will not be measured in an absolute sense. However, determining which angles are larger or smaller than others is often enough to provide a good description of the structure. Comparative gel electrophoresis was first applied to the four-way (Holliday) DNA junction. This revealed the stacked X-structure of the junction, its dependence on the presence of metal ions, and the occurrence of two alternative conformers of the structure (Duckett et al., 1988). The stacked X-structure was confirmed by crystallography a decade after the electrophoretic experiments (Eichman et al., 2000; Ortiz-Lombardı´a et al., 1999). Since then, comparative gel electrophoresis has scored many more successes, with no failures to date. This has been reviewed recently (Lilley, 2008). The most recent of these is the four-way helical RNA junction found in the U1 snRNP; Nagai and coworkers (Pomeranz-Krummel et al., 2009) determined the structure crystallographically of the RNA in complex with a number of proteins, and confirmed the structure determined electrophoretically 14 years earlier. In addition to providing the global structure of junctions, in some circumstances comparative gel electrophoresis can provide information on multiple conformations and their interconversion, and on the structure of protein complexes. In combination with chemical synthesis, it can also probe detailed hydrogen bonding and ionic interactions within the structure of a junction or kink.
1. Principle and Theory of Electrophoresis of Bent or Kinked Nucleic Acids Most theoretical treatments of the gel electrophoresis of DNA molecules are based upon de Gennes’ (1971) concept of reptation. The migration of the nucleic acid is considered to occur in a ‘‘tube’’ formed by the polymer matrix of the gel, through which the DNA migrates rather like a
Comparative Gel Electrophoresis
145
snake. The force of the electric field pulling the DNA through the gel is balanced against the retardation due to friction. Lumpkin and Zimm (1982) derived the equation of motion of the nucleic acid along a gel tube as Q h2x m¼ ð7:1Þ x L2 where m is the mobility of the nucleic acid chain, Q is the effective charge, x is the frictional coefficient, hx is the end-to-end length in the direction of the electric field, and L is the contour length. This indicates that if a double-stranded DNA or RNA molecule becomes kinked (thus reducing the end-to-end distance), its electrophoretic mobility will be retarded. This relatively simple yet quite successful theory has been widely applied to analyze bent and kinked nucleic acid geometry. There have been a number of refinements to the theory (Barkema et al., 1994; Duke et al., 1992; Lerman and Frisch, 1982; Levene and Zimm, 1989; Lumpkin et al., 1985, 1989), to include nucleic acid elasticity for example (Deutsch, 1988). There is no well-developed theory at the present time for nucleic acids containing helical junctions, although there have been recent attempts to incorporate this (Heuer et al., 2005; Saha et al., 2006). However, most electrophoretic data on junctions are analyzed empirically. The general rule that increased kinking results in lower mobility seems to work well, unless bending is so severe that helices clash (see, e.g., Goody et al., 2003). The very basis of the comparative approach is that the conclusions result from the interpretation of relative mobilities of matched species.
2. Discrete Bending or Kinking of the Axis of a Duplex A sequence located in some specific point may cause a discrete bend or kink in the nucleic acid, either intrinsically or due to the interaction with another molecule such as a protein. This might be regarded as a two-way helical junction. An example is the kink turn (k-turn) commonly found in RNA (Klein et al., 2001), which introduces a sharp kink with an included angle of 60 . This can be investigated by placing the motif at the center of a long (e.g., 50 bp) duplex that is electrophoresed in a polyacrylamide gel under given conditions of salt composition and concentration, etc. In general kinked species exhibit lowered electrophoretic mobility, as the end-to-end distance is reduced. The species under study could be run alongside other species such as a perfect duplex, or a duplex with a central bulge kink for comparison. Sequence or functional group changes that
146
David M. J. Lilley
affect the stability of the kinked geometry will be manifested as altered electrophoretic mobility. A variant on the simple bend or kink is where two such bends are introduced into the same duplex molecule. Two kinks will generate a three-dimensional trajectory of the molecule, as first shown with A-tract bends in DNA (Drak and Crothers, 1991). The outer helical sections beyond the kinks will be related by a dihedral angle that will be a function of the local geometry of the kink, that is, the direction of the bend, the spacing between the two kinks, and the periodicity of the helix. Thus, the shape of the molecule, and hence its frictional properties and electrophoretic mobility, will vary systematically with the length of helix between the kinks. In these experiments (frequently called ‘‘phasing experiments’’), the overall length of the duplex is normally kept constant by removing basepairs from the arms as the spacer is lengthened. The end-to-end distance will vary sinusoidally as the central helix length between the kinks is varied, generating fastest mobility when the structure is trans-planar, and slowest for the cis-planar form. The variation in mobility will be a progressively damped sine function, with a period equivalent to that of the spacer helix and a phase determined by the structure of the kinks. Phasing experiments provided the first measurements of the periodicity of RNA (Bhattacharyya et al., 1990; Tang and Draper, 1990, 1994). The sinusoidal modulation of mobility proves that component kinks have a particular direction, and provides a sensitive way of revealing axial kinking when the extent of deviation is small. This was used to demonstrate kinking at the HIV TAR sequence (Riordan et al., 1992).
3. The Analysis of Helical Junctions Using the Long–Short Arm Method Helical junctions are structures in which a number of helical segments are connected by the covalent continuity of strands shared between them. There may or may not be additional unpaired nucleotides present at the positions where the strands exchange between the helices. These are named according to the IUB nomenclature (Lilley et al., 1995). The comparative gel electrophoresis method sets out to compare the relative angles included between pairs of helical arms radiating from a junction of three or more helices (Cooper and Hagerman, 1987; Duckett et al., 1988; Gough and Lilley, 1985). In a junction with n helical arms, there are nC2 combinations of the arms taken two at a time, where n
C2 ¼
n! 2!ðn 2Þ!
ð7:2Þ
Comparative Gel Electrophoresis
147
Thus, the basis of the long–short arm method is to compare the electrophoretic mobility of the nC2 species with two long arms and the remaining arms short. For a three-way junction there are three species to compare, each with one short arm. A four-way junction gives six species, each of which has two long and two short arms (Fig. 7.1). The symmetry and pattern of the mobilities of the different species can provide an excellent description of the global shape of the junction. In our experiments, we typically have long arms of 40 bp that are shortened to 15 bp; we find that this provides good discrimination between the different angles of junctions.
4. Experimental Strategies and Methods In general the electrophoretic methods require species in which the structure of interest (a kink or a junction) is embedded in the center of relatively long sections of duplex. In principle these can be constructed by hybridization of synthetic oligonucleotides, requiring species of 80 nt in length, but in practice this is too long for current methods for RNA synthesis. We have therefore generally adopted the approach of synthesizing hybrid DNA–RNA–DNA molecules, so that the functionally significant core of the molecule is made of RNA, but the long arms are made from DNA beyond one turn of helix. Hybrid species of 80 nt are easy to synthesize. The RNA–DNA junctions that are created in the long arms do not detectably affect the results. The component RNA chains may alternatively be generated by in vitro transcription, in which case the shorter helices may be replaced with short stem-loops—indeed, there is no reason why the individual species cannot be made as a single RNA transcript. The main drawback to this approach is that nonnatural nucleotides cannot be introduced, and for this chemical synthesis is essential. The species to be compared are loaded in 6% Ficoll (w/v) onto an 8% or 10% polyacrylamide gel with 29:1 monomer:bisacrylamide. We generally use 90 mM Tris–borate (pH 8.3) buffer containing either EDTA or added metal salts. Electrophoresis is typically performed for 1–2 days at 5 V/cm at room temperature. When salts are included in the electrophoresis buffer, this must be recirculated between the cathodic and anodic reservoirs at a rate of 1 L/h. This may require a little modification of commercial gel electrophoresis apparatus. Our RNA species generally include at least one strand that is radioactively [50 -32P]-labeled. When the electrophoresis is complete the gel is dried onto Whatman 3MM paper and exposed to storage phosphor plates at 4 C followed by phosphorimaging to provide a gel image.
148
David M. J. Lilley
B AC A
C
B
B
D
AB
BC
B A
A
C
C
D
D A
C
B
B
A
D
C
A
B
AD
C
CD
D
D A
C BD
D
Figure 7.1 The principle of the long–short arm method for the electrophoretic analysis of a four-way junction. The six species can be considered to be derived from the central junction comprising four long arms (e.g., 40 bp). These are named sequentially A, B, C, and D. The long–short arm species are generated by shortening two helical arms, to give the six species shown, named by their long arms.
Comparative Gel Electrophoresis
149
5. Kink Turns in RNA The k-turn provides an example of a discontinuity within an RNA duplex (Klein et al., 2001). This might be termed a two-way junction. It comprises a base bulge (usually three nucleotides) followed by AG and GA pairs and frequently a third non-Watson–Crick pair. K-turns generate a marked kinking of the axis with an included angle of 60 , stabilized by A-minor interactions. Unsurprisingly, this leads to a retardation of electrophoretic mobility in polyacrylamide gels (Goody et al., 2003; Matsumura et al., 2003) (Fig. 7.2). The kinking was found to be dependent on the presence of divalent cations, and also had rather precise sequence requirements. The kinking could also be demonstrated by fluorescence resonance energy transfer (FRET) between donor and acceptor fluorophores attached to the two ends of the helical arms (Goody et al., 2003). These two approaches were combined to analyze the contributions of specific hydrogen bonds to the folding into the kinked geometry (Liu and Lilley, 2007; Turner and Lilley, 2008).
6. Four-Way RNA Junctions Perfectly paired (4H) junctions exist in a number of important RNA species, including the hairpin ribozyme and U1 snRNA. Comparative gel electrophoresis showed that the junctions are relatively polymorphic (Duckett et al., 1995). The junctions undergo pairwise coaxial stacking of helices under all conditions. However, the angle subtended between the two axes varies with the ionic conditions. At low salt concentrations the conformation is parallel. Upon addition of moderate concentrations of divalent metal ions the electrophoretic pattern changed to one that could be interpreted in terms of a stacked structure with perpendicular axes. Recent single-molecule FRET studies have shown that this results to some degree from a rapid equilibrium between parallel and antiparallel conformations (Hohng et al., 2004). Increasing the concentration of metal ions further resulted in electrophoretic patterns indicative of a bias towards the antiparallel structure. Steady-state FRET experiments indicated that the transition between parallel and antiparallel forms is driven by the noncooperative binding of ions (Walter et al., 1998a). Comparative gel electrophoresis analysis of the 4H junction of U1 snRNA showed the junction adopted a coaxially stacked structure with almost perpendicular axes (Fig. 7.3). This result was very recently confirmed crystallographically (Pomeranz-Krummel et al., 2009). Perhaps the most extensively studied 4H junction in RNA is that of the hairpin ribozyme,
150
David M. J. Lilley
Kt-7
L1 L2 L3 −2b −1b
C helix
G A A
5⬘ G G C
GAAG
3⬘ C C G
AGGC
−2n −1n
dup
1n
NC helix
1n 2n
−1n
2n Kt-7
1b 2b
L1 −2n
L2
Figure 7.2 Electrophoretic retardation of RNA containing a kink-turn sequence. The sequence shown is that of Kt-7. A chemically synthesized 45 bp duplex comprised a 25-bp RNA section containing the Kt-7 sequence flanked by outer DNA sections of 20 bp. For comparison, a duplex species lacking the bulge and with the A G pairs changed to Watson–Crick basepairs was made. In addition, a series of variants in which the 20 -OH of selected nucleotides were substituted by H to test the effect of removing potential hydrogen bonds were made (Liu and Lilley, 2007). The radioactively labeled species were electrophoresed in 15% polyacrylamide in 90 mM Tris–borate (pH 8.3), 500 mM Mg2þ, and phosphorimaged. Note the pronounced retardation resulting from the presence of the K-turn motif in the RNA, and the reduced retardation resulting from preventing hydrogen bond formation in some cases. Tracks contained the unmodified Kt-7 sequence (Kt-7), the duplex (dup), or Kt-7 modified by removal of the 20 -OH from the indicated nucleotide.
both as an isolated junction (Walter et al., 1998b) and in situ in the ribozyme (Murchie et al., 1998; Walter et al., 1998c). Comparative gel electrophoresis revealed that the junction adopted an antiparallel stacked X-structure in the presence of divalent metal ions, and these conclusions were confirmed upon
151
Comparative Gel Electrophoresis
B
A
5⬘
AGG
3⬘
UCA
A U G C G C
CAG
G C C G U A
GUC
C
C
D
A D AB
AC
AD
BC
BD
CD
B
Figure 7.3 Analysis of the 4H four-way RNA junction of the human U1 snRNA by comparative gel electrophoresis (Duckett et al., 1995). The central sequence of the junction is shown. The A G pair at the center was retained in this analysis, although changing it to a Watson–Crick pair did not alter the global shape of the junction. The six long–short species can be considered to be derived from a junction with four arms of 40 bp. The central 20 bp comprises RNA, and the outer arms are DNA. The junction species were electrophoresed in an 8% polyacrylamide gel, in 90 mM Tris–borate (pH 8.3) and 1 mM Mg2þ. The mobility pattern of the six species is slow, slow, fast, fast, slow, slow. The simplest interpretation (shown on the right-hand side) is that of a stacked structure based on A on D and B on C coaxial stacking, with the axes nearly perpendicular. The pattern would also be consistent with a rapid exchange between nearly equal populations of parallel and antiparallel forms. However, a recent crystal structure has found a perpendicular stacked structure for this RNA junction (Pomeranz-Krummel et al., 2009).
the solution of the X-ray crystal structure of the ribozyme (Rupert and Ferre´-D’Amare´, 2001). The dynamic nature of this junction was later demonstrated using single-molecule FRET, showing transitions between antiparallel and parallel forms, and between alternative stacking conformers (Hohng et al., 2004). These studies showed that the folding of the ribozyme into its active form is greatly accelerated by the presence of the four-way
152
David M. J. Lilley
junction (Tan et al., 2003). The chirality of the isolated junction of the hairpin ribozyme was deduced to be opposite to that of the complete ribozyme using electrophoretic phasing experiments (Goody et al., 2004). Perfectly paired 4H RNA junctions are quite rare. Most natural fourway RNA junctions contain one or more formally unpaired nucleotides at the point of strand exchange. The structural and dynamic properties would be expected to be altered by the extra nucleotides. One example is the 2HS22HS1 four-way junction found in the hepatitis C virus internal ribosome entry site (IRES) element, which has been analyzed using a combination of comparative electrophoresis and FRET (Melcher et al., 2003). It was found that in contrast to the 4H RNA junctions, the IRES junction lost coaxial stacking in the absence of added metal ions, adopting an extended structure. On addition of divalent metal ions the junction folded by pairwise coaxial stacking of arms, adopting the stacking conformer that places the extra nucleotides onto the exchanging strands. The pattern of electrophoretic mobilities of the six long–short arm species was consistent with a perpendicular axis (Fig. 7.4). However, time-resolved FRET studies suggested an alternative interpretation of the pattern, with a rapid exchange between approximately equal populations of parallel and antiparallel conformations. Provided that the exchange is fast on the timescale of the gel electrophoresis, this should generate an average of the mobility patterns of the two forms, completely consistent with the pattern obtained. There is an important lesson here—we must be ready to consider more complex interpretations of electrophoretic results in some situations. An earlier crystal structure of the IRES junction had shown that it crystallized in a parallel conformation (Kieft et al., 2002), freezing out this component of the dynamic equilibrium.
7. Three-Way RNA Junctions Very few natural three-way junctions in RNA are perfectly paired 3H junctions. Most include one or more formally unpaired nucleotides in the regions connecting the helices. Such junctions are extremely common in natural RNA molecules (Lescoute and Westhof, 2006). The VS ribozyme provides a good example, where the basic core is organized by two threeway junctions (Fig. 7.5). These organize five helices (numbered II–VI) into an H-shape. Helices III, IV, and V create an HS1HS5HS2 junction, while helices II, III, and VI form a 2HS5HS2 junction. Both junctions undergo two-state folding induced by the noncooperative binding of divalent metal ions, and their structures have been analyzed as free junctions by comparative gel electrophoresis (Fig. 7.5). The II–III–VI junction folds by the
153
Comparative Gel Electrophoresis
Fast
B
A
5⬘ 3⬘
U A A U A U U CGG GGG GCC A CCC A G C G G C G
C
Parallel
Antiparallel
S
AB AC AS
BC
BS
CS
Dynamic averaging
Figure 7.4 Analysis of the 2HS12HS2 four-way junction of the HCV IRES by comparative gel electrophoresis. The sequence of the junction around the point of strand exchange is shown. Comparative gel electrophoresis in a 10% polyacrylamide gel was performed in the presence of 90 mM Tris–borate (pH 8.3), 1 mM Mg2þ, using the six long–short arm species, where arms were extended with DNA sections as before. The observed pattern of mobilities is interpreted in terms of a rapid exchange between approximately equal populations of parallel and antiparallel conformations as shown, with strand polarities indicated for clarity.
coaxial stacking of arms III and VI, with an acute angle between helices II and VI (Lafontaine et al., 2001). The III–IV–V junction folds by the coaxial stacking of arms IV and III, with an acute angle between helices V and III (Lafontaine et al., 2002). These structures are in good agreement with FRET measurements on the isolated junctions, and also with more recent small-angle X-ray scattering (SAXS) data from the isolated junctions and the complete ribozyme (Lipfert et al., 2008). The two junctions are joined through their common helix III, creating a continuous stack of helices IV, III, and VI. The dihedral angle subtended between helices II and V was determined by a variant of the helical phasing experiment discussed above, and from this a model of the complete trans-acting ribozyme was constructed (Lafontaine et al., 2002). This is broadly in agreement with the model derived from the SAXS data (Lipfert et al., 2008).
154
David M. J. Lilley
III–IV
III–V
IV–V
III–IV–V junction
IV 3 mM Mg2+
U
UUG
IV
CC U
V
A AC
GUA UG U C GU A U C G A U A III C G U A AC G A GCU GGG C
II
5⬘ 3⬘
V III
III
VI
CGA CCC GAACA II–III
II–VI
II III–VI
VI
II–III–VI junction
5 mM Mg2+
Figure 7.5 Analysis of the two three-way RNA junctions of the VS ribozyme by comparative gel electrophoresis. The secondary structure of the VS ribozyme is shown, with the sequences of the two component three-way junctions. Each was analyzed in isolation by comparative gel electrophoresis, comparing the mobilities of the three long–short arm species. As before, these species have a central core of RNA that is extended with DNA sections. The junction species were electrophoresed in 10% polyacrylamide gels in the presence of 90 mM Tris–borate (pH 8.3) with 3 ( junction III–IV–V) or 5 ( junction II–III–VI) mM Mg2þ. The structural interpretations of both sets of data are shown. Both junctions undergo coaxial stacking of two arms, with the third directed laterally.
Comparative Gel Electrophoresis
155
8. Complexes of Branched RNA with Proteins In principle the global structure of an RNA junction could be determined when it is complexed with bound protein. This has not been accomplished to date in RNA, but it has proven itself for DNA junctions. For example, the structure of DNA junctions bound by the junctionresolving enzymes T4 endonuclease VII (Po¨hler et al., 1996) and T7 endonuclease I (De´clais et al., 2003) have both been determined by comparative gel electrophoresis. It was found that both proteins substantially alter the global shape of the DNA junction in different ways. These structures were both recently confirmed by X-ray crystallography (Biertu¨mpfel et al., 2007; Hadden et al., 2007), showing that comparative gel electrophoresis functions reliably for protein complexes. There is no reason to expect that the method would not work equally well for RNA junction complexes. The binding of the ribosomal protein S15 to a threeway RNA junction has been studied by an electrophoretic approach that is related to comparative gel electrophoresis (Batey and Williamson, 1998).
ACKNOWLEDGMENTS I thank my coworkers who have contributed to the studies discussed in this chapter, including Derek Duckett, Alastair Murchie, Anamitra Bhattacharrya, Terry Goody, Jia Liu, Sonya Melcher, and Daniel Lafontaine. Cancer Research UK is acknowledged for financial support of the work of this laboratory.
REFERENCES Barkema, G. T., Marko, J. F., and Widom, B. (1994). Electrophoresis of charged polymers: Simulation and scaling in a lattice model of reptation. Phys. Rev. E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Top. 49(6), 5303–5309. Batey, R. T., and Williamson, J. R. (1998). Effects of polyvalent cations on the folding of an rRNA three-way junction and binding of ribosomal protein S15. RNA 4(8), 984–997. Bhattacharyya, A., Murchie, A. I. H., and Lilley, D. M. J. (1990). RNA bulges and the helical periodicity of double-stranded RNA. Nature 343, 484–487. Biertu¨mpfel, C., Yang, W., and Suck, D. (2007). Crystal structure of T4 endonuclease VII resolving a Holliday junction. Nature 449(7162), 616–620. Cooper, J. P., and Hagerman, P. J. (1987). Gel electrophoretic analysis of the geometry of a DNA four-way junction. J. Mol. Biol. 198, 711–719. De´clais, A.-C., Fogg, J. M., Freeman, A., Coste, F., Hadden, J. M., Phillips, S. E. V., and Lilley, D. M. J. (2003). The complex between a four-way DNA junction and T7 endonuclease I. EMBO J. 22, 1398–1409. de Gennes, P. G. (1971). Reptation of a polymer chain in the presence of fixed obstacles. J. Chem. Phys. 55, 572–578. Deutsch, J. M. (1988). Theoretical studies of DNA during gel electrophoresis. Science 240, 922–924.
156
David M. J. Lilley
Drak, J., and Crothers, D. M. (1991). Helical repeat and chirality effects on DNA gel electrophoretic mobility. Proc. Natl. Acad. Sci. USA 88, 3074–3078. Duckett, D. R., Murchie, A. I. H., Diekmann, S., von Kitzing, E., Kemper, B., and Lilley, D. M. J. (1988). The structure of the Holliday junction and its resolution. Cell 55, 79–89. Duckett, D. R., Murchie, A. I. H., and Lilley, D. M. J. (1995). The global folding of fourway helical junctions in RNA, including that in U1 snRNA. Cell 83, 1027–1036. Duke, T. A., Semenov, A. N., and Viovy, J. L. (1992). Mobility of a reptating polymer. Phys. Rev. Lett. 69(22), 3260–3263. Eichman, B. F., Vargason, J. M., Mooers, B. H. M., and Ho, P. S. (2000). The Holliday junction in an inverted repeat DNA sequence: Sequence effects on the structure of fourway junctions. Proc. Natl. Acad. Sci. USA 97(8), 3971–3976. Goody, T. A., Melcher, S. E., Norman, D. G., and Lilley, D. M. J. (2003). The kink-turn motif in RNA is dimorphic, and metal ion dependent. RNA 10, 254–264. Goody, T. A., Lilley, D. M. J., and Norman, D. G. (2004). The chirality of a four-way helical junction in RNA. J. Am. Chem. Soc. 126, 4126–4127. Gough, G. W., and Lilley, D. M. J. (1985). DNA bending induced by cruciform formation. Nature 313, 154–156. Hadden, J. M., De´clais, A.-C., Carr, S., Lilley, D. M. J., and Phillips, S. E. V. (2007). The structural basis of Holliday junction resolution by T7 endonuclease I. Nature 449 (7162), 621–624. Heuer, D. M., Yuan, C., Saha, S., and Archer, L. A. (2005). Effect of topological asymmetry on the electrophoretic mobility of branched DNA structures with and without singlebase mismatches. Electrophoresis 26, 64–70. Hohng, S., Wilson, T. J., Tan, E., Clegg, R. M., Lilley, D. M. J., and Ha, T. (2004). Conformational flexibility of four-way junctions in RNA. J. Mol. Biol. 336, 69–79. Kieft, J. S., Zhou, K., Grech, A., Jubin, R., and Doudna, J. A. (2002). Crystal structure of an RNA tertiary domain essential to HCV IRES- mediated translation initiation. Nat. Struct. Biol. 9(5), 370–374. Klein, D. J., Schmeing, T. M., Moore, P. B., and Steitz, T. A. (2001). The kink-turn: A new RNA secondary structure motif. EMBO J. 20(15), 4214–4221. Lafontaine, D. A., Norman, D. G., and Lilley, D. M. J. (2001). Structure, folding and activity of the VS ribozyme: Importance of the 2-3-6 helical junction. EMBO J. 20(15), 1415–1424. Lafontaine, D. A., Norman, D. G., and Lilley, D. M. J. (2002). The global structure of the VS ribozyme. EMBO J. 21, 2461–2471. Lerman, L. S., and Frisch, H. L. (1982). Why does the electrophoretic mobility of DNA in gels vary with the length of the molecule. Biopolymers 21, 995–997. Lescoute, A., and Westhof, E. (2006). Topology of three-way junctions in folded RNAs. RNA 12(1), 83–93. Levene, S. D., and Zimm, B. H. (1989). Understanding the anomalous electrophoresis of bent DNA molecules: A reptation model. Science 245, 396–399. Lilley, D. M. J. (2000). Structures of helical junctions in nucleic acids. Quart. Rev. Biophys. 33, 109–159. Lilley, D. M. J. (2008). Analysis of branched nucleic acid structure using comparative gel electrophoresis. Quart. Rev. Biophys. 41(1), 1–39. Lilley, D. M. J., Clegg, R. M., Diekmann, S., Seeman, N. C., von Kitzing, E., and Hagerman, P. (1995). Nomenclature Committee of the International Union of Biochemistry: A nomenclature of junctions and branchpoints in nucleic acids. Recommendations 1994. Eur. J. Biochem. 230, 1–2. Lipfert, J., Ouellet, J., Norman, D. G., Doniach, S., and Lilley, D. M. J. (2008). The complete VS ribozyme in solution studied by small-angle X-ray scattering. Structure 16, 1357–1367.
Comparative Gel Electrophoresis
157
Liu, J., and Lilley, D. M. J. (2007). The role of specific 20 -hydroxyl groups in the stabilization of the folded conformation of kink-turn RNA. RNA 13(2), 200–210. Lumpkin, O. J., and Zimm, B. H. (1982). Mobility of DNA in gel electrophoresis. Biopolymers 21, 2315–2316. Lumpkin, O. J., Dejardin, P., and Zimm, B. H. (1985). Theory of gel electrophoresis of DNA. Biopolymers 24(8), 1573–1593. Lumpkin, O., Levene, S. D., and Zimm, B. H. (1989). Exactly solvable reptation model. Phys. Rev. A 39(12), 6557–6566. Matsumura, S., Ikawa, Y., and Inoue, T. (2003). Biochemical characterization of the kinkturn RNA motif. Nucleic Acids Res. 31(19), 5544–5551. Melcher, S. E., Wilson, T. J., and Lilley, D. M. J. (2003). The dynamic nature of the fourway junction of the hepatitis C virus IRES. RNA 9(7), 809–820. Murchie, A. I. H., Thomson, J. B., Walter, F., and Lilley, D. M. J. (1998). Folding of the hairpin ribozyme in its natural conformation achieves close physical proximity of the loops. Mol. Cell 1, 873–881. Ortiz-Lombardı´a, M., Gonza´lez, A., Erijta, R., Aymamı´, J., Azorı´n, F., and Coll, M. (1999). Crystal structure of a DNA Holliday junction. Nat. Struct. Biol. 6, 913–917. Po¨hler, J. R. G., Giraud-Panis, M.-J. E., and Lilley, D. M. J. (1996). T4 endonuclease VII selects and alters the structure of the four-way DNA junction; binding of a resolutiondefective mutant enzyme. J. Mol. Biol. 260, 678–696. Pomeranz-Krummel, D. A., Oubridge, C., Leung, A. K. W., Li, J., and Nagai, K. (2009). Crystal structure of human spliceosomal U1 snRNP at 5.5 A˚ resolution. Nature 458, 475–480. Riordan, F. A., Bhattacharyya, A., McAteer, S., and Lilley, D. M. J. (1992). Kinking of RNA helices by bulged bases, and the structure of the human immunodeficiency virus transactivator response element. J. Mol. Biol. 226, 305–310. Rupert, P. B., and Ferre´-D’Amare´, A. R. (2001). Crystal structure of a hairpin ribozymeinhibitor complex with implications for catalysis. Nature 410, 780–786. Saha, S., Heuer, D. M., and Archer, L. A. (2006). Electrophoretic mobility of linear and starbranched DNA in semidilute polymer solutions. Electrophoresis 27, 3181–3194. 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. Tang, R. S., and Draper, D. E. (1990). Bulge loops used to measure the helical twist of RNA in solution. Biochemistry 29, 5232–5237. Tang, R. S., and Draper, D. E. (1994). On the use of phasing experiments to measure helical repeat and bulge loop-associated twist in RNA. Nucleic Acids Res. 22(5), 835–841. Turner, B., and Lilley, D. M. (2008). The importance of G.A hydrogen bonding in the metal ion- and protein-induced folding of a kink turn RNA. J. Mol. Biol. 381(2), 431–442. Walter, F., Murchie, A. I. H., Duckett, D. R., and Lilley, D. M. J. (1998a). Global structure of four-way RNA junctions studied using fluorescence resonance energy transfer. RNA 4, 719–728. Walter, F., Murchie, A. I. H., and Lilley, D. M. J. (1998b). The folding of the four-way RNA junction of the hairpin ribozyme. Biochemistry 37, 17629–17636. Walter, F., Murchie, A. I. H., Thomson, J. B., and Lilley, D. M. J. (1998c). Structure and activity of the hairpin ribozyme in its natural junction conformation; effect of metal ions. Biochemistry 37(40), 14195–14203.
C H A P T E R
E I G H T
The Structure and Folding of Branched RNA Analyzed by Fluorescence Resonance Energy Transfer David M. J. Lilley Contents 162 163 167 170 171 176 179 182 182
1. Theory of FRET 2. The Possible Effects of Fluorophore Orientation 3. Choice of Fluorophores 4. Construction of Fluorophore-Labeled RNA Species 5. Steady-State Measurements of FRET 6. Time-Resolved Measurements of FRET 7. Single-Molecule FRET Acknowledgments References
Abstract Fluorescence resonance energy transfer (FRET) is a spectroscopic means of obtaining distance information over a range up to 80 A˚ in solution. It is based on the dipolar coupling between the electronic transition moments of a donor and acceptor fluorophore attached at known positions on the RNA species of interest. It can be applied in ensembles of molecules, either by steady-state fluorescence or by lifetime measurements, but it is also very appropriate for single-molecule studies. In addition to the provision of distance information, recent studies have emphasized the orientation dependence of energy transfer.
Fluorescence (or Fo¨rster) resonance energy transfer (FRET) (Fo¨rster, 1948; Perrin, 1932) provides a spectroscopic way of estimating distances over a size range that is appropriate for biological macromolecules. It is based upon fluorescence, one of the most sensitive spectroscopic methods. Fluorescence is the emission of light from an excited molecule, having lost Cancer Research UK Nucleic Acid Structure Research Group, MSI/WTB Complex, The University of Dundee, Dundee, United Kingdom Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69008-X
#
2009 Elsevier Inc. All rights reserved.
159
160
David M. J. Lilley
vibrational energy in the excited state. Thus, the fluorescent photon is of longer wavelength than the photon that excited the molecule, the wellknown Stokes shift. Molecules interact with the electric component of light because of the change in electronic distribution between the ground state and an excited state, measured by the transition dipole moment vector. ^ can be written in Thus, the vector for a transition from states m to n (d) bra.ket notation as: d^ ¼ hCn jℜjCm i
ð8:1Þ
where Cm and Cn are the wavefunctions of the initial and final states, respectively, and ℜ is the dipole moment operator. The transition dipoles of two different fluorescent molecules may interact together in a dipolar process that leads to transfer of energy from one (the donor) to the other (the acceptor). This resonance energy transfer is strongly dependent upon the physical separation between the two. By tethering two smallmolecule fluorescent probes (fluorophores) to a biomolecule of interest at known positions, we can monitor the distance between these two points. For other reviews, see Clegg (1992, 2002), Lakowicz (1999), Walter and Burke (2000). Although FRET has been used in biochemical studies for half a century or more (Bennett, 1964; Cantor and Pechukas, 1971; Dale and Eisinger, 1976; Stryer and Haugland, 1967), its full exploitation in the study of nucleic acid structure and folding required the ability to synthesize oligonucleotides chemically and to attach fluorophores at known positions, and that has only been possible in the last 20 years. One of the first such studies of branched nucleic acids was the analysis of the structure of the four-way (Holliday) DNA junction, in which junctions with two arms terminally labeled with fluorescein and tetramethylrhodamine were studied in the steady state (Murchie et al., 1989). Since that time, FRET has been extensively used to study the structure and folding of branched nucleic acids, or junctions. Here, we define helical junctions as structures in which a number of helical segments are connected by the covalent continuity of strands shared between them (Lilley, 2000). There may or may not be additional unpaired nucleotides present at the positions where the strands exchange between the helices. These are named according to the IUB nomenclature (Lilley et al., 1995). The most common junctions found in natural RNA are three- and four-way junctions. Such junctions can act as important architectural elements in large RNA species, or as key folding motifs in small autonomously folding RNAs. We may broaden this definition to include a duplex containing a central structural feature that leads to a discontinuity in the axis; this could be regarded as a two-way helical junction (Gohlke et al., 1994). The kink-turn (k-turn) (Klein et al., 2001) is a good example of a two-way junction.
Fluorescence Resonance Energy Transfer
161
In a typical FRET experiment involving branched RNA or DNA, the nucleic acid would be labeled with donor and acceptor fluorophores, covalently attached at different locations. Analysis of FRET between the fluorophores then provides an estimate of the distance between them. Estimates of the absolute distance could provide input into molecular modeling to determine the structure, but the difficulties of getting accurate distances from FRET information are discussed below. Nevertheless, even relative information can be valuable. A series of vectors within the structure might be compared. For example, the three end-to-end distances of a threeway junction, or the six end-to-end distances of a four-way junction might be analyzed. Further information could be derived by systematic variation of the length of the helix to which a given fluorophore is attached. In some experiments the absolute value of FRET efficiency (or distance) may be less important than how it changes, for example, during folding induced by metal ion concentration or protein binding. FRET experiments on RNA are generally carried out in one of the three main ways. The simplest is by steady-state fluorimetry, where a solution of fluorophore-labeled RNA is excited at the wavelength of the donor, and then the intensity of the emitted light is scanned over the range of donor and acceptor emission. It requires relatively simple equipment, and unsophisticated data processing. FRET efficiency can be determined from the reduced fluorescence of the donor in the presence of the acceptor, or the enhanced emission of the acceptor. However, steady-state data are an average over the ensemble of molecules present, and no information is obtained about subpopulations of species within the ensemble. The second approach is to analyze the excitedstate lifetime of the donor. This becomes shortened in the presence of an acceptor due to the transfer of excitation energy by FRET, and so provides a measure of the efficiency of the process. Like the steady-state measurements, this is performed on a solution of donor–acceptor-labeled RNA. But the timedependence of the decay of donor molecules in the excited state can be fitted to models in which multiple populations of species have distinct fluorescent lifetimes, and the distributions of the different species can be estimated. Time-resolved fluorimetry requires a higher level of sophistication in terms of equipment and data processing. Lastly, the high sensitivity of fluorescence means that FRET experiments can be performed on single RNA molecules, either tethered to a glass slide in some manner, or freely diffusing in solution. Single-molecule FRET offers a number of advantages over ensemble methods. The ability to examine one molecule at a time allows us to divide the population into its constituent species, reveal heterogeneity within seemingly identical molecules, study molecules within impure samples or even cell extracts, and study kinetic processes that cannot be synchronized. There are a variety of ways to carry out single-molecule FRET experiments, all of which are based on microscopy and highly sensitive detection methods. Most laboratories have assembled their own equipment to perform this.
162
David M. J. Lilley
1. Theory of FRET FRET involves a resonance between singlet–singlet electronic transitions of the donor and acceptor fluorophores, arising from the dipolar coupling between the emission transition dipole of the donor and the absorption transition dipole of the acceptor. This leads to a transfer of excitation energy from the donor to the acceptor. The process occurs within a region much smaller than the wavelength of the light (the near field region), and therefore does not involve the transfer of real photons. FRET can be observed in a variety of ways, including a reduction in the fluorescent quantum yield of the donor, a corresponding shortening of the donor excited-state lifetime, and an increased fluorescent emission from the acceptor (if fluorescent). Either a classical or quantum mechanical analysis shows that the rate of the energy transfer process depends on the inverse sixth power of the distance between the two fluorophores (Fo¨rster, 1948). This is the basis of the use of the technique to provide structural information. In the laboratory the efficiency of the process (EFRET) is normally determined. This can be defined in a variety of ways, and is the proportion of donor excitation events that lead to excitation of the acceptor by dipolar coupling—a quantum yield of FRET: " 6 #1 1 tDA rDA kFRET EFRET ¼ ¼ 1þ ¼X ð8:2Þ tD R0 k i i where tDA and tD are the fluorescent lifetime in the presence and absence of the acceptor, respectively, rDA is the separation between the donor and acceptor. This results in the dependence on distance shown graphically in Fig. 8.1. EFRET can also be defined in terms of rates, where kFRET is the rate constant for energy transfer and ki are those for all the mechanisms of deactivating the donor excited state (including fluorescent emission from the donor, quenching, intersystems crossing, etc.). R0 is the characteristic Fo¨rster length for a given donor–acceptor pair of fluorophores, given by: 0:529 FD k2 JðlÞ ð8:3Þ Nn4 where the units of R0 and the wavelength l are in cm. FD is the fluorescent quantum yield of the donor in the absence of the acceptor and N is the Avogadro number. k is a parameter that depends on the relative orientation of the donor and acceptor transition moments, discussed in the following section. n is the refractive index of the medium in which the electric fields of the transition dipoles extend; a value of 1.33 is appropriate for the aqueous medium of biological macromolecules. J(l) is the normalized spectral overlap between donor emission and acceptor absorption, given by: R06 ¼
163
Fluorescence Resonance Energy Transfer
1
E FRET
0.8
0.6
0.4
0.2
0
0
0.5
1 r DA/R 0
1.5
2
Figure 8.1 The dependence of FRET efficiency (EFRET) as a function of the separation of donor and acceptor fluorophores (rDA) normalized to the Fo¨rster distance R0 (Eq. (8.2)), assuming a constant value of R0.
ð1 JðlÞ ¼
0
fD ðlÞeA ðlÞl4 dl ð1 fD ðlÞ dl
ð8:4Þ
0
where fD is the spectral shape function for the donor emission and eA is that for acceptor absorption (in M 1 cm 1). From equation (8.2) it can be seen that when rDA ¼ R0, the efficiency of FRET is 0.5. R0 values are frequently calculated on the basis of an assumption (stated or otherwise) that k2 ¼ 2/3. Indeed, this can be written as R0(2/3).
2. The Possible Effects of Fluorophore Orientation FRET arises from the dipolar coupling between the oscillating transition dipoles of the donor and acceptor fluorophores. The magnitude of the interaction depends on both the distance between them and their relative
164
David M. J. Lilley
orientation. The rate at which the acceptor is excited by FRET is proportional to the square of the scalar product of its transition dipole with the local electric field of the donor transition dipole. This is the origin of the k2 term in equation (8.3), given by: k2 ¼ ½ p^D p^A 3ð p^D^r DA Þð^r DA p^A Þ2 ¼ ðcos YT 3 cos YD cos YA Þ2 ð8:5Þ where p^D and p^A are the donor and acceptor transition dipole moment vectors, ^r DA is the vector between their centers, and the angles YT, YD, and YA are defined in Fig. 8.2A. k2 can take values between 0 and 4, as shown in Fig. 8.2B. If the value is not known, it becomes very hard to extract distances from FRET efficiencies. However, if at least one fluorophore is flexible, so that it experiences many orientations during the lifetime of the excited state, then k2 should average to 2/3. In the great majority of studies this assumption has been made. Where fluorescein is tethered to a phosphate group of RNA via a flexible linker, this is probably a good approximation. The negative charge is repelled by the backbone so that the fluorophore freely rotates in a cone, and the fluorescent anisotropy of fluorescein attached to DNA is typically low (usually 0.1). By contrast, the cyanine fluorophores interact strongly with DNA and RNA. We have shown that both Cy3 (Norman et al., 2000) and Cy5 (Iqbal
A D ΘD
f
ΘT
ΘA
A⬘
A
B
k 2= 0
k 2= 1
k 2= 4
Figure 8.2 Orientation of transition moments of cyanine fluorophores terminally attached to double-stranded DNA. (A) The orientation parameter k2. The transition dipole vectors for the coupled donor and acceptor fluorophore are indicated by the arrows, labeled D and A. Vector A0 is generated by the in-plane translation of vector A to share its origin with vector D. The definition of k2, given in Eq. (8.5), is based upon the angles shown. (B) If the fluorophores lie in parallel planes, the orientation parameter simplifies to k2 ¼ cos2 YT and varies between 0 and 1. The schematic shows the limiting cases, where the transition moments are parallel (k2 ¼ 1) and crossed (k2 ¼ 0). If the transition moments are colinear, k2 ¼ 4.
Fluorescence Resonance Energy Transfer
165
et al., 2008a) when attached to 50 -termini via C3 linkers (as often generated when coupled as phosphoramidites during synthesis) stack upon the ends of double-helical DNA very much in the manner of an additional basepair (Fig. 8.3). The relative immobilization of these fluorophores under these conditions suggests that orientation could be an important factor in FRET efficiency, and because this pair is commonly used in single-molecule studies, this could have significant practical consequences. If Cy3 and Cy5 are terminally attached to a duplex DNA or RNA molecule, the NMR structures suggest that their planes would be approximately parallel and coaxial, but the angle between their transitions moments (those are approximately directed along the polymethyne linkers between the two indole rings; Iqbal et al., 2008a) will depend upon the length of the helix and its periodicity. k2 should be maximal (k2 1) when the transition dipoles are parallel, and minimal (k2 0) when they are perpendicular. Thus, it would be expected that EFRET should be modulated with twice the periodicity of the helix. This experiment was carried out for both DNA and RNA–DNA hybrid helices, using both ensemble and single-molecule methods (Iqbal et al., 2008b). For both DNA and RNA–DNA duplexes, the FRET efficiency falls with length, but with a pronounced modulation (Fig. 8.4). EFRET is modulated with twice the helical periodicity; for example, for RNA there are clear maxima at 11 and 17 bp, with minima at 15 and 21 bp. The DNA helix exhibits the same behavior, but with a phase shift indicative of a more tightly wound helix. This is exactly the behavior expected for orientation dependence of FRET. However, the modulation is clearly
Figure 8.3 Molecular graphics image showing the structure of Cy3 stacked onto the terminal basepair of a DNA duplex, as determined by NMR (Norman et al., 2000).
166
David M. J. Lilley
1
0.8 RNA/DNA
E FRET
0.6
DNA
0.4
0.2
0
10
12
14
16 18 20 Duplex length/bp
22
24
Figure 8.4 Efficiency of energy transfer for Cy3 þ Cy5-labeled hybrid DNA and RNA/DNA duplexes as a function of duplex length (Iqbal et al., 2008b). EFRET was measured for each duplex species as phospholipid vesicle-encapsulated single molecules. The EFRET values are plotted for the DNA (squares) and DNA/RNA duplexes (circles) as a function of helix length, with the estimated errors. The lines show simulation of the data, using a model in which the major fraction of the fluorophores was stacked onto the helix undergoing lateral motion. For the DNA duplexes this was based on standard B geometry with a periodicity 10.5 bp/turn and a helical rise of ˚ /bp step; 31% of the fluorophore was allowed to be freely mobile (based on time3.6 A resolved analysis) with k2 ¼ 2/3, while the remaining fluorophore underwent lateral motion with a Gaussian half-width of 42 . For the DNA/RNA duplexes, the simulation was based on standard A geometry with a periodicity 12 bp/turn and a helical rise of ˚ /bp step; 12% of the fluorophore was allowed to be freely mobile (based on time3A resolved analysis) with k2 ¼ 2/3, while the remaining fluorophore underwent lateral motion with a Gaussian half-width of 42 . The fluorescent quantum yield for Cy3 was 0.30 attached to DNA and 0.35 attached to DNA/RNA. The refractive index was 1.33. The phase shift between the two helical forms is very clear. Single duplex molecules were studied trapped within phospholipid vesicles in 10 mM Tris–HCl (pH 8.0), 50 mM NaCl. Individual DNA or RNA/DNA duplex species were encapsulated in phospholipid vesicles comprising a 100:1 mixture of either L-a-phosphatidylcholine or 1,2dimyristoyl-sn-glycero-3-phosphocholine with 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(cap biotinyl). The vesicles were attached to quartz slides coated with biotin-functionalized polyethylene glycol, via NeutrAvidin. Encapsulated molecules were excited at 532 nm by prism-based total internal reflection.
incomplete (EFRET does not become zero at the minima), suggesting that the orientation is subject to averaging by a combination of lateral mobility of the fluorophores on the end of the helix, and torsional mobility of the helices. The data were well simulated by a model based on the geometric
Fluorescence Resonance Energy Transfer
167
properties of standard B- and A-form helices for DNA and RNA–DNA, and the positions of the fluorophores determined by NMR with a significant lateral averaging (Iqbal et al., 2008b), with or without a fraction of unstacked fluorophore as indicated by time-resolved studies (Iqbal et al., 2008b; Sanborn et al., 2007). This is not only a nice confirmation of the expected orientational dependence of FRET, but also a warning that such effects could significantly affect the interpretation of FRET data in terms of distance with a commonly used pair of fluorophores. Similar results have been obtained using more immobilized fluorophores (Bo¨rjesson et al., 2009; Lewis et al., 2005), but it is clear that the intrinsic immobility of the cyanine fluorophores is sufficient to result in significant variation of k2. Distances for ˚ if it is assumed that k2 ¼ 2/3, the duplex systems could be in error by 12 A where the largest discrepancy arises when the fluorophores are perpendicular (i.e., at the minima of EFRET). If the fluorophores are not constrained to lie in parallel planes, then k2 could reach a value of 4 in principle, so the potential error could be still greater. It is possible to imagine, for example, that states might be misassigned in a single-molecule time profile. The orientation dependence of FRET leads to something of a conundrum. If the fluorophores are mobile then k2 ¼ 2/3 may be a good approximation, and the distances between the fluorophores can be determined. But because of the flexibility, the position of the fluorophores relative to the RNA is not well known, so the interpretation of the distance in terms of the RNA structure is difficult. On the other hand, if the fluorophores are fixed so that their position on the RNA is known, this leads to a potential uncertainly with k2, so complicating the calculation of the distance. There are two ways to deal with this problem. One approach is to maximize the mobility—by selecting mobile fluorophores such as fluorescein and flexible linkers. In many cases, we are trying to distinguish between conformations such that the difference in donor–acceptor distance may be greater than the uncertainty in position. It may be possible to determine the average position of the mobile fluorophore, by reference to a less mobile fluorophore in a series of positions on a known structure (Norman et al., 2000), or using molecular dynamics calculations (Wozniak et al., 2008). Alternatively, if immobile fluorophores are chosen then the orientation dependence can potentially provide valuable angular information for structural modeling. It is likely that this will be exploited much more in the near future.
3. Choice of Fluorophores Small-molecule extrinsic fluorophores are generally used in structural and folding studies of RNA molecules (Fig. 8.5). There are a number of factors that will influence the choice of donor–acceptor fluorophores
168
David M. J. Lilley
HO
O
(H3C)2N
O
O
N(CH3)2
CO2
CO2
R
R
Tetramethylrhodamine
Fluorescein
N
N n
R⬘
R
Cyanine N rib
O
N
H N
N H
H
S N
N H O
N rib
1,3-diaza-2-oxophenothiazine
Figure 8.5 The chemical structures of some fluorophores commonly used in FRET experiments. 1,3-Diaza-2-oxophenothiazine is a cytosine analog (Wilhelmsson et al., 2001), and is shown basepaired to guanine (gray).
for FRET studies in RNA. One is the distance range to be explored. Given the dimensions of natural macromolecules it is good to choose a fluorophore pair that allows measurements up to 70 A˚ or more, requiring an R0 ˚ or greater. The spectral overlap integral for Cy3–Cy5 is J (2/3) of 55 A (l) ¼ 7.2 10 13 M 1 cm 3 (Iqbal et al., 2008b) (Fig. 8.6), giving R0 ˚ ; thus EFRET falls (2/3) ¼ 60.1 A˚. For fluorescein–Cy3, R0 (2/3) ¼ 56 A from 0.8 to 0.2 (a conservative range for the measurement of FRET) over ˚ . This is a very useful size range for the study of the distance rDA ¼ 44–70 A RNA folding. The orientation of the fluorophores is clearly another important factor. This may depend on the intrinsic properties of the fluorophores and the manner of their attachment to the nucleic acid. As discussed above,
169
Fluorescence Resonance Energy Transfer
3 ´ 105
0.6 Cy3 fluorescence
Cy5 absorption
2.5 ´ 105
0.4
2 ´ 105
0.3
1.5 ´ 105
0.2
1 ´ 105
0.1
0.5 ´ 104
0 500
550
600 650 Wavelength/nm
700
Cy5 absorption
Cy3 fluorescence
0.5
0 750
Figure 8.6 The fluorescence emission spectrum of Cy3 and the absorption spectrum of Cy5, showing the overlap between them.
fluorescein is quite mobile when it is attached to the 50 -termini of DNA or RNA. The cyanine fluorophores are predominantly stacked on the end, with a fraction that is unstacked at any moment (Iqbal et al., 2008b; Sanborn et al., 2007). This results in relatively complicated photophysical properties with multiple lifetimes. The dynamic and spectroscopic properties of tetramethylrhodamine are also quite complex (Neubauer et al., 2007; Va´mosi et al., 1996). If a fixed orientation of the fluorophores is required, then it may be advantageous to use fluorescent base analogs that are fixed in the helix by basepairing. The cytosine analog 1,3-diaza-2-oxophenothiazine is fluorescent and reasonably bright, but can basepair with guanine normally (Wilhelmsson et al., 2001) (Fig. 8.5). Wilhelmsson and colleagues (Bo¨rjesson et al., 2009) have recently synthesized two analogs 1,3-diaza-2oxophenoxazine (tCO) and 7-nitro-1,3-diaza-2-oxophenothiazine (tCnitro) that can act as a FRET donor–acceptor pair, with a calculated R0 (2/3) of ˚ . Although this is a relatively short Fo¨rster length, the base analogs can 27 A be placed quite close within the structure. It would be expected that energy transfer between tCO and tCnitro would be strongly affected by orientation, and this has been demonstrated for a series of DNA duplexes (Bo¨rjesson et al., 2009). These fluorophores have not been synthesized as ribonucleoside analogs to date, but this should be possible.
170
David M. J. Lilley
Fluorophores may be attached to RNA in a variety of ways. If the RNA is chemically synthesized, then some of the fluorophores (including fluorescein and Cy3) may be coupled to the 50 -termini as phosphoramidites. Alternatively, fluorophores may be coupled to a 20 -aminoribose incorporated into the RNA at a chosen point. A greater variety of fluorophores may be conjugated to aminolinkers as N-hydroxysuccinimide esters, either at the termini or at the 5 position of dU (assuming that a 20 -deoxyribose substitution can be tolerated at that position). In principle, fluorophores could be attached via the backbone phosphate groups, as demonstrated in DNA (Ozaki and Mclaughlin, 1992). At the present time the photophysical properties of fluorophores that have been conjugated postsynthetically to nucleic acids have not been well characterized in general.
4. Construction of Fluorophore-Labeled RNA Species In this laboratory RNA species are synthesized using ribonucleotide phosphoramidites with 20 -tert-butyldimethylsilyl protection. Fluorophores may be coupled to the 50 -terminus as phosphoramidites, with an average coupling efficiency of 97%. Alternatively, we conjugate fluorophores as N-hydroxysuccinimide esters to primary amine groups that have been incorporated as terminal or internal amino-linkers. After deprotection, the RNA is desalted by gel filtration followed by ethanol precipitation. All RNA species are purified by gel electrophoresis in polyacrylamide gels (usually 20%) containing 7 M urea. Fluorescently labeled species are significantly retarded, so that bands may be excised and the oligonucleotides electroeluted into 8 M ammonium acetate, and recovered by ethanol precipitation. Following the gel purification, it is very important to introduce a further stage of purification by reversed-phase HPLC. We use a C18 column, eluted with a linear gradient of 100 mM ammonium acetate/ acetonitrile, with a flow rate of 1 ml/min. The donor–acceptor-labeled RNA is then constructed by hybridization. The required stoichiometric combinations of fluorophore-labeled and unlabeled strands are mixed together in 90 mM Tris–borate (pH 8.3), 25 mM NaCl for 10 min. They are then placed at 80 C, followed by slow cooling over several hours. The doubly labeled species are purified by electrophoresis in a polyacrylamide gel under nondenaturing conditions at 4 C at 150 V in 90 mM Tris–borate (pH 8.3), 25 mM NaCl with recirculation at >1 L/h. The fluorescent species are recovered by band excision and electroelution.
171
Fluorescence Resonance Energy Transfer
5. Steady-State Measurements of FRET Steady-state fluorimetry has been extensively used to analyze folding in nucleic acids, including RNA. FRET efficiency can be measured from the quenching of the donor, or the enhanced emission from the acceptor. It can also be measured from the depolarization of the acceptor fluorescence. But of these the most sensitive method is to measure acceptor emission. For this we find Clegg’s normalized acceptor ratio (Clegg, 1992), the most straightforward means to extract EFRET, requiring a minimal number of samples. I have written a program in MATLAB to carry out this analysis. In this approach two emission spectra are recorded from a donor–acceptor labeled RNA sample, with excitation at two wavelengths (n1 and n2). In the general case (e.g., fluorescein–Cy3), the emission at a given wavelength of a sample excited primarily at the donor wavelength contains emission from the donor, emission from directly excited acceptor and emission from acceptor excited by energy transfer from the donor: Fðn1 n0 Þ / feD ðn0 ÞFA ðn1 ÞEFRET da þ eA ðn0 ÞFA ðn1 Þa þ eD ðn0 ÞFD ðn1 Þd½ð1 EFRET Þa þ ð1 aÞg 0
ð8:6Þ
0
¼ F ðn1 n Þ þ F ðn1 n Þ A
D
where d and a are the molar fraction of molecules labeled with donor and acceptor, respectively. Superscripts D and A refer to donor and acceptor, respectively. eD(n0 ) and eA(n0 ) are the molar absorption coefficients of donor and acceptor, respectively, and FD(n1) and FA(n1) are the fluorescent quantum yields of donor and acceptor, respectively. Thus, the spectrum contains components due to donor emission (FD(n1n0 ), i.e., the final term containing FD(n1)) and those due to acceptor emission (FA(n1n0 ), i.e., the first two terms containing FA(n1)). A spectrum of an RNA sample labeled only with donor is normalized to the donor peak and subtracted, leaving just the acceptor components, that is, FA(n1n0 ). The pure acceptor spectrum thus derived is then normalized to one from the same sample excited at a wavelength (n00 ) at which only the acceptor is excited, with emission at n2. This gives the normalized acceptor ratio: F A ðn1 n0 Þ EFRET deD ðn0 Þ eA ðn0 Þ FA ðn1 Þ ðratioÞA ¼ A ð8:7Þ ¼ þ A 00 F ðn1 n00 Þ eA ðn00 Þ e ðn Þ FA ðn2 Þ EFRET is directly proportional to (ratio)A, and can be easily calculated since eD(n0 )/eA(n00 ) and eA(n0 )/eA(n00 ) are measured from absorption spectra, and FA(n1)/FA(n2) is unity when n1 ¼ n2. An analogous normalization procedure for the measurement of efficiency from donor deactivation has been presented (Clegg, 1992) and used successfully for the analysis of the global
172
David M. J. Lilley
structure of a four-way DNA junction (Clegg et al., 1992). However, it is generally less sensitive than the acceptor ratio method. In this laboratory we use an SLM-Aminco 8100, equipped with Glan-Thompson polarizers. The electronics have been updated by the ISS Phoenix system. Measurements of FRET efficiency are performed under photon counting conditions, with the polarizers crossed at the magic angle (54.7 ) to remove polarization artifacts. Fluctuation of lamp intensity is corrected using a concentrated rhodamine B solution as a quantum counter. Ensemble steady-state measurements of FRET have been extensively used to study branched RNA molecules, including the three-way HS1HS7HS3 junction of the hammerhead ribozyme (Bassi et al., 1997; Penedo et al., 2004; Tuschl et al., 1994), the 2HS5HS3 (Lafontaine et al., 2001), and HS1HS5HS2 (Lafontaine et al., 2002) junctions of the VS ribozyme, a number of four-way 4H junctions (Walter et al., 1998a), including that of the hairpin ribozyme (Walter et al., 1998a,b) and its hinged equivalent lacking two arms (Walter et al., 1998c), and more complex junctions such as the 2HS12HS2 junction found in the HCV IRES (Melcher et al., 2003). The 4H junction derived from the IRES junction folds as an archetypical perfect four-way RNA junction (Fig. 8.7) (Melcher et al., 2003). The junction was constructed with four arms each of 12 bp, with fluorescein and Cy3 attached to the 50 -termini as phosphoramidites during synthesis. Six vectors were made, corresponding to the six possible end-to-end distances. In the presence of 10 mM Mg2þ ions, two of the vectors exhibited significantly more efficient energy transfer than the other four, consistent with a pronounced antiparallel structure based upon coaxial stacking of A on B and C on S arms (Fig. 8.7A). The structure of the junction is dependent on the concentration of divalent ions. The FRET efficiency of each vector changes with Mg2þ ion concentration (Fig. 8.7B); the data can be fitted to a two-state folding model: unfold EFRET ¼ EFRET þ DEFRET
KA ½Mg2þ n 1 þ KA ½Mg2þ n
ð8:8Þ
unfold is the FRET efficiency of the vector in the absence of Mg2þ where EFRET ions, DEFRET is the change in FRET efficiency with ion-induced folding, KA is the apparent association constant for Mg2þ ion binding, and n is the Hill coefficient. The data were fitted by nonlinear regression, giving ½Mg2þ 1=2 ¼ 150 mM and n ¼ 0.9, where ½Mg2þ 1=2 ¼ ð1=KA Þ1=n : Conformational changes in RNA structure can also be induced by protein binding, and FRET may be used to follow such a folding process. A good example of this is provided by the k-turn. The k-turn motif comprises a three-nucleotide bulge followed by trans sugar-Hoogsteen GA pairs (Klein et al., 2001), and introduces a very tight kink into the axis of helical RNA in the presence of Mg2þ ions (Goody et al., 2003).
173
Fluorescence Resonance Energy Transfer
B
U A A U A U CGG GCC
A
GGG CCC
C
G C G C C G
S
A 0.5
E FRET
0.4
A
C
S
B
0.3 0.2 0.1 BA
CA
AS
BC
BS
CS
Vector
B
0.5 BC
SB
0.24
SB
BC
0.22
0.42
0.38
0.2 10−7
E FRET
E FRET
0.46
10−6
10−5
10−4
10−3
10−2
10−1
[Mg2+]/M
Figure 8.7 Steady-state FRET analysis of a 4H four-way junction derived from the 2HS12HS2 junction of the HCV IRES (Melcher et al., 2003). The central sequence of the junction is shown. The four arms are sequentially named A, B, C and S. Donor– acceptor-labeled vectors for FRET analysis are constructed with 50 -terminally attached fluorescein (donor) and Cy3 (acceptor) on selected helical arms, named according to those arms in that order. Thus, BA is labeled with donor on the end of arm B, and acceptor on A. (A) Histogram of the FRET efficiencies of the six end-to-end vectors.
174
David M. J. Lilley
Folding into the kinked geometry requires the presence of metal ions, and this can be readily followed by FRET. This has been studied using a short RNA duplex, with a central k-turn motif and terminally attached fluorescein and Cy3 fluorophores (Goody et al., 2003). As the k-turn adopts its tightly kinked geometry, the fluorophores become closer and FRET efficiency increases. This can be combined with chemogenetic dissection of the RNA to determine interactions that are critical to the stability of the folded state (Liu and Lilley, 2007; Turner et al., 2005). K-turn folding may also be induced by the binding of L7Ae-related proteins (including the human 15.5 kDa protein)—a good example of induced fit (Turner et al., 2005). Titration of Archeoglobus L7Ae protein into the fluorescein–Cy3-labeled RNA solution brings about a marked increase in EFRET (Fig. 8.8A). Fitting to a stoichiometric binding model gives a dissociation constant of 65 pM, showing that the affinity is very high. However, to measure the affinity would require an undetectably low concentration of fluorescent RNA in order to be in equilibrium, so that this value is only an upper limit. FRET provided the means to measure this via the rates of association and dissociation. The association rate was measured by mixing 10 nM fluorescein–Cy3labeled RNA and 11 nM L7Ae protein in a stopped-flow mixer (Fig. 8.8B). The progress of the binding reaction was followed by the reduced fluorescence intensity of fluorescein at 515 nm, that is, as the fluorescein donor became increasingly quenched due to energy transfer to Cy3 as the RNA bound L7Ae and consequently folded into the kinked geometry. The data were fitted to two exponentials, with a faster rate of 0.59 s 1, corresponding to an association rate of kon ¼ 5.4 107 M 1 s 1. The dissociation rate was measured in the fluorimeter (this did not require fast kinetics methods), giving koff ¼ 0.002 s 1. From these rates we calculated a dissociation constant of KD ¼ 10 pM. It is possible to apply steady-state FRET measurements to larger constructs, the main limitation being the difficulty of preparing the RNA suitably labeled with donor and acceptor fluorophores. P-RNA (Smith et al., 2005) and ribosomes (Dorywalska et al., 2005) have been labeled by hybridization of fluorescent oligonucleotides. Clegg, Noller, and their collaborators have studied dynamic processes in the ribosome by steadystate FRET, using fluorophores attached to ribosomal proteins (Ermolenko et al., 2007a,b; Hickerson et al., 2005; Majumdar et al., 2005).
(B) FRET efficiency as a function of Mg2þ ion concentration for the SB and BC vectors. The data have been fitted to a two-state ion binding model. Fluorescence emission spectra were recorded at 4 C using an SLM-Aminco 8100 fluorimeter with modernized Phoenix electronics (ISS Inc., Champaign, IL, USA). Spectra were corrected for xenon lamp fluctuations and instrumental variations, and polarization artifacts were avoided by crossing excitation and emission polarizers at 54.7 .
175
Fluorescence Resonance Energy Transfer
F
GAA CCAGUCAGUGGC GAACCAUGUCAGG GGUCAGUCACCG
AGGGGUACAGUCC
Cy
A 0.5
E FRET
E FRET
0.4
0.15 0.3
r (fluorescein) r
0.1 0.05
0.2
10−11
10−10
10−9
10−8
10−7
10−6
[L7Ae]/M B
11
Fluorescence intensity
10 9 8 7 6 5 0
1
2
3
4
5
6
7
Time/s
Figure 8.8 Steady-state FRET analysis of the formation of a kinked geometry by a k-turn-containing RNA on the binding of L7Ae protein (Turner and Lilley, 2008; Turner et al., 2005). The sequence of the RNA is shown, with 50 -terminal fluorescein on the top strand and 50 -terminal Cy3 on the bottom strand. (A) FRET efficiency as a function of L7Ae concentration (closed circles). These data have been fitted to a stoichiometric binding model. In a separate experiment, singly fluorescein-labeled RNA of the same sequence was titrated with L7Ae and the fluorescence anisotropy r measured. Note that this changes very little over the range of protein concentration, indicating that the mobility of the fluorescein is not significantly altered. (B) Reaction progress curve for the association of L7Ae with Kt-7 RNA using a stopped-flow rapid mixer. Intensity of fluorescein emission at 515 nm is plotted as a function of time after mixing RNA and protein to final concentrations of 10 and 11 nM, respectively. The decrease in fluorescein fluorescence reflects increased FRET efficiency resulting from the kinking of the RNA on protein binding. The data are fitted to two exponential functions (line).
176
David M. J. Lilley
6. Time-Resolved Measurements of FRET The excited state of a fluorophore typically exhibits multiple lifetimes, so that the emission (I(t) at time t) decays with a rate: X t IðtÞ ¼ ð8:9Þ ai exp ti i where there are i lifetimes ti with amplitudes ai. In a FRET experiment, using an RNA species labeled with donor and acceptor, the lifetime of the donor excited state (tDA) will be shortened due to energy transfer to the acceptor, and this provides information on subpopulations of different conformational states within the ensemble: " # X ð1 X t t IDA ðtÞ ¼ drDA ð8:10Þ fn Pn ðRDA Þ ai exp ti ti ðR0 =rDA Þ6 0 n i where there are n distributions (each of amplitude fn) of donor–acceptor distance distributions Pn(rDA). The distributions are generally assumed to be Gaussian. Most of the fluorophores used in FRET measurements have excitedstate lifetimes in the nanosecond range, although the cyanine fluorophores have shorter lifetimes. There are broadly two ways to measure fluorescent lifetimes, working in either the time or the frequency domain. In the former, the ensemble of molecules in solution is excited by a very short pulse of light (ideally a d-pulse), usually from a titanium:saphire laser. The intensity of the emission as a function of time following the excitation pulse is measured by time-correlated photon counting, and a decay curve generated. The donor fluorescence decay is deconvoluted from the instrument response, and then fitted to multiple exponentials from which the lifetime distributions can be calculated. The decay curve for Cy3 50 terminally attached to a 16 bp DNA molecule is shown in Fig. 8.9. The data were fitted to three exponential rates, corresponding to lifetimes of 390 ps, 1.04, and 1.91 ns (Iqbal et al., 2008b). Levitus and coworkers (Sanborn et al., 2007) have made a detailed study of the photophysics of Cy3 terminally attached to DNA. They conclude that the short lifetime corresponds to Cy3 that is unstacked from the end of the helix, and thus able to relax by cis–trans photoisomerization in the polymethyne linker between indole rings. Working in the frequency domain, the sample is excited by light whose intensity is sinusoidally modulated at high frequency (typically MHz). Light from a continuous wave laser (e.g., the 488 nm line from
177
Fluorescence Resonance Energy Transfer
104
Counts
103
102 IRF 10
1 2
3
4
5
6 Time/ns
7
8
9
10
Figure 8.9 Time-resolved fluorescent lifetime analysis of Cy3 attached to doublestranded DNA (Iqbal et al., 2008b). Fluorescent decay curve for Cy3 attached to a 16 bp DNA duplex, showing the experimental data and the instrument response function (IRF), and the fit to three exponential functions (line). The decay curve was generated using time-correlated single-photon counting, after excitation by 200 fs pulses from a titanium:sapphire laser at 4.7 MHz.
an argon-ion laser to excite fluorescein) is modulated using an electrooptical modulator such as a Pockels cell. Because of the finite lifetime of the excited state, the emitted light is demodulated and phase shifted relative to the excitation. The data are plotted as the phase shift and modulation of the emitted light as a function of the modulation frequency. The donor–acceptor distance distributions were determined for k-turncontaining RNA as a function of Mg2þ concentration using this approach (Goody et al., 2003) (Fig. 8.10). The shift in both sets of curves with the increase in ionic concentration is apparent. The data required fitting to two Gaussian distributions of rDA distance to achieve an acceptable distribution of the residuals. These correspond to mean distances of ˚ , indicative of a tightly kinked geometry and 83 A˚ corresponding to 23 A an extended structure. At low of Mg2þ concentration, the extended structure was the major species, but the kinked geometry was dominant at high of Mg2þ concentration. Time-resolved FRET has been employed to study a number of RNA junctions, including the hammerhead ribozyme (Rueda et al., 2003), the hairpin ribozyme (Klostermeier and Millar, 2001; Walter et al., 1999), and the HCV IRES 2HS12HS2 junction (Melcher et al., 2003).
178
David M. J. Lilley
A 100
1 1 nM Mg2+
60
0.8 0.6
50 mM Mg2+
40
0.4
20
0.2
0
0 10
B
100 Frequency/MHz 1 nM Mg2+
0.016 P (R)
Modulation
Phase/deg
80
35%
65%
0.008
0 C 70%
P (R)
0.024
50 mM Mg2+
0.016
0.008
30%
0 0
20
40
60 80 100 120 140 Distance/Å
Figure 8.10 Two populations of conformation of k-turn-containing RNA using timeresolved FRET in the frequency domain by phase fluorimetry (Goody et al., 2003). Donor–acceptor distance distributions were analyzed using the same fluorescein–Cy3 end-labeled species studied by steady-state FRET (Fig. 8.8). (A) Plots of phase shift and demodulation of the fluorescein emission as a function of the presence of 1 nM (closed circles) and 50 mM (open squares) Mg2þ ions as a function of the modulation frequency. The best fits to the experimental data using one or two Gaussian distributions of rDA lengths are shown by the broken and solid lines, respectively. (B) Two-Gaussian rDA distributions P(R) calculated from the fit to the data obtained in 1 nM Mg2þ ions. The major species (65%) corresponds to the longer rDA distance. The integrated areas under the curves total to 100%. (C) Two-Gaussian rDA distributions calculated from the fit to the data obtained in 50 mM Mg2þ ions. The major species (70%) now corresponds to the shorter rDA distance. These measurements were performed at 4 C using a
Fluorescence Resonance Energy Transfer
179
7. Single-Molecule FRET The ability to detect single photons by devices of very high sensitivity and efficiency such as electron-multiplying charge-coupled device (EMCCD) chips and avalanche photodiode photodetectors has made it possible to carry out FRET experiments on single molecules (Ha et al., 1996, 1999a), and this has been applied to RNA in the last decade (Ha et al., 1999b; Zhuang et al., 2000). Very rapidly this approach has become probably the major way in which FRET is now applied to analysis of the folding and dynamics of RNA molecules. This is a very large subject in its own right, and a comprehensive coverage is not attempted here; there are other excellent reviews dedicated to this topic (Aleman et al., 2008; Ha, 2001; Joo et al., 2008; Pljevaljcic and Millar, 2008; Roy et al., 2008). Single-molecule FRET is performed on an inverted microscope, where the fluorescence is captured by the objective lens, and then split into donor and acceptor wavelengths using dichroic mirrors. There are two basic modes for doing this. One is to measure the burst of fluorescence from a free RNA molecule as it transiently diffuses through a confocal volume (Deniz et al., 1999). This uses essentially the same technology as fluorescence correlation spectroscopy. From the intensity at the donor and acceptor wavelengths, the efficiency of energy transfer can be measured, and a histogram constructed from many such events. However, because the molecule is typically observed for about a millisecond, no conformational dynamics are measured. This method has been used to study the conformation of the hairpin ribozyme (Pljevaljcic et al., 2004). Alternatively, the RNA can be immobilized on a slide, in which case single molecules can be studied for many seconds and conformational transitions observed. The time of observation is limited by photochemical effects on the fluorophores, but this can be extended with recently improved oxygen scavenging systems (Aitken et al., 2008; Rasnik et al., 2006). Again there are two choices. A wide-field approach can be used,
K2-Digital Phase Fluorimeter (ISS Inc., Champaign, IL, USA), with excitation at 488 nm from a vertically polarized, argon ion laser, intensity modulated at 39 frequencies between 4 and 300 MHz. Donor emission was measured using a 10 nm bandpass filter centered at 520 nm to exclude scattered incident light and acceptor fluorescence and a polarizer set at 54.7 to remove instrumental artifacts. Measurements were referenced to fluorescein in 10 mM NaOH, with a lifetime of 4.05 ns. Phase shift and modulation data were analyzed with the parameter estimation program CFS_LS ( Johnson and Faunt, 1992), according to the theory given in Melcher et al. (2003). Goodness of fit was evaluated by w2, the distribution of residuals and by a runs test. Confidence intervals (single standard deviation) were determined by the Bootstrap method (Efron and Tibshirani, 1993).
180
David M. J. Lilley
where fluorescence from many single molecules is collected simultaneously using an EMCCD camera, or light from one molecule at a time can be analyzed, and the surface scanned using a nanopositioning stage. The latter allows higher time resolution, but as the data acquisition rates of cameras become faster this is less important. For the former, the excitation can be carried out using prism-based total internal reflection microscopy, where the surface-bound molecules are excited by the evanescent wave. This requires an objective lens with a high numerical aperture. Or epi-illumination can be used, where the excitation passes through the objective lens (Sase et al., 1995). There are a number of ways in which RNA can be tethered to the surface for observation. Most require that one strand of the RNA carries a biotin covalently attached to its 50 -terminus, which can be accomplished during synthesis. This may then be bound via a streptavidin molecule to biotinylated BSA coating the surface of the slide, or alternatively to PEG. A way of avoiding a direct connection with the surface while localizing the RNA on the slide is to use phospholipid vesicle encapsulation (Boukobza et al., 2001; Okumus et al., 2004). In this approach the molecule can diffuse freely within the confines of a 200 nm diameter vessel (a volume of 4 10 18 L). The data shown in Fig. 8.4 were collected in this manner. A relatively early application of single-molecule FRET was to a ribosomal three-way RNA junction (Ha et al., 1999). The adenine riboswitch is a more complex three-way junction, an HS2HS3HS8 junction (Mandal and Breaker, 2004). Structural transitions were observed by measurement of FRET between fluorophores tethered to 5-amino-allyluridine nucleotides incorporated into the terminal loops (Lemay et al., 2006). The RNA exhibits repeated transitions between two states of high (EFRET ¼ 0.9) and low (EFRET ¼ 0.25) FRET efficiency (Fig. 8.11). The high FRET state would be consistent with the loop–loop interaction observed in the crystal (Serganov et al., 2004). In the absence of added metal ions, the riboswitch remains largely in the open (low-FRET) state, but the folded (high-FRET) state becomes stabilized as the ionic concentration is raised. The dynamics were altered upon binding its adenine ligand. Singlemolecule FRET has also been applied to four-way RNA junctions (Hohng et al., 2004), the junction-containing hairpin ribozyme (Nahas et al., 2004; Okumus et al., 2004; Tan et al., 2003), and its minimal hinged form (Zhuang et al., 2002). The data for the full form of the hairpin ribozyme were interpreted in terms of a three-state folding process, in which the dynamics of the four-way junction juxtaposes the loops, which may then undergo a docking process. In a subsequent study, the conformational dynamics of the ribozyme were exploited to use FRET to follow cycles of cleavage and ligation in the active ribozyme (Nahas et al., 2004). This might be regarded as the first example of single-molecule ribozyme enzymology.
181
Fluorescence Resonance Energy Transfer
Cy5
Cy3
0.8 2 mM Mg2+
0.4 0
E FRET
0.8 0.5 mM Mg2+
0.4 0 0.8
0.05 mM Mg2+
0.4 0 0.8
0 mM Mg2+
0.4 0 0
5
10
15
20 Time/s
25
30
35
Figure 8.11 Conformational transitions in the pbuE adenine riboswitch nucleotide binding domain (Mandal and Breaker, 2004) observed by single-molecule FRET (Lemay et al., 2006). The riboswitch was fluorescently labeled in the terminal loops, and tethered to the surface of a quartz slide via a biotin terminally attached to the third, open helix. Time records of FRET efficiency (50 ms integration time) as a function of elapsed time for single riboswitch molecules are shown at various Mg2þ ion concentrations. These data were recorded using a total internal reflection fluorescence microscope with 532 nm laser excitation and a back-illuminated electron-multiplying CCD camera (Andor iXon). The imaging buffer was 50 mM Tris–HCl (pH 8.1), 6% (w/w) glucose, 1% 2-mercaptoethanol, 0.1 mg/mL glucose oxidase, 0.02 mg/mL glucose catalase, and the indicated concentration of MgCl2. Measurements were performed at room temperature (22 C). Single-molecule FRET efficiency after background correction was approximated by IA/(IA þ ID), where IA and ID are the fluorescence intensities of the Cy5 acceptor and Cy3 donor, respectively.
Single-molecule studies have revealed that some branched RNA species exhibit surprisingly heterogeneous dynamics (Lemay et al., 2006; Tan et al., 2003; Zhuang et al., 2002), while others (such as a simple 4H RNA junction) do not. Rates of conformational transitions can vary a hundred
182
David M. J. Lilley
fold from molecule to molecule; these properties are very persistent, with very infrequent interconversion of kinetic properties by a given molecule. Suspicion initially focused upon possible inhomogeneity of the surface to which the molecules were bound. However, this possibility is excluded because the same heterogeneity in hairpin ribozyme molecules was found when they were encapsulated in phospholipid vesicles (Okumus et al., 2004). So, it seems that the heterogeneity may be somehow built into the structure of more complex RNA molecules. This is not well understood at the present time. Single-molecule FRET has been applied to the folding of the group I intron ribozyme (Lee et al., 2007a; Russell et al., 2002; Zhuang et al., 2000), the group II intron ribozyme (Steiner et al., 2008), the VS ribozyme (Pereira et al., 2008), and the interaction of a tetraloop and its receptor (Hodak et al., 2005). It has been used to study folding of telomerase RNA (Stone et al., 2007) and dynamic processes in the ribosome during translation (Blanchard et al., 2004a,b; Cornish et al., 2008, 2009; Lee et al., 2007b; Wang et al., 2007). As the systems become more complex, the profiles of FRET efficiency with time can become more complicated, with many states interconverting. Hidden Markov modeling algorithms have been used to uncover the states within such profiles (McKinney et al., 2006). In a recent development, FRET has been used to study the dynamics of single DNA junctions while under the application of stretching force (Hohng et al., 2007); this experiment should be directly applicable to branched RNA species. Single-molecule FRET has become an extremely powerful way to study the dynamics of RNA in complex assemblies. The main limitation to this is now probably the ease with which fluorescent labels can be incorporated where needed.
ACKNOWLEDGMENTS I thank my collaborators and coworkers who have contributed to the studies discussed in this review, including Bob Clegg, Tim Wilson, Taekjip Ha, Carlos Penedo, Jo Ouellet, Sonya Melcher, Terry Goody, Ben Turner, David Norman, and Daniel Lafontaine. Cancer Research UK is acknowledged for financial support of the work of this laboratory.
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. Aleman, E. A., Lamichhane, R., and Rueda, D. (2008). Exploring RNA folding one molecule at a time. Curr. Opin. Chem. Biol. 12, 647–654.
Fluorescence Resonance Energy Transfer
183
Bassi, G. S., Murchie, A. I. H., Walter, F., Clegg, R. M., and Lilley, D. M. J. (1997). Ioninduced folding of the hammerhead ribozyme: A fluorescence resonance energy transfer study. EMBO J. 16, 7481–7489. Bennett, R. (1964). Radiationless intermolecular energy transfer. I. Singlet-singlet transfer. J. Chem. Phys. 41, 3037–3040. 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. Bo¨rjesson, K., Preus, S., El-Sagheer, A. H., Brown, T., Albinsson, B., and Wilhelmsson, L. M. (2009). Nucleic acid base analog FRET-pair facilitating detailed structural measurements in nucleic acid containing systems. J. Am. Chem. Soc. 131, 4288–4293. 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. Cantor, C. R., and Pechukas, P. (1971). Determination of distance distribution functions by singlet-singlet energy transfer. Proc. Natl. Acad. Sci. USA 68, 2099–2101. Clegg, R. M. (1992). Fluorescence resonance energy transfer and nucleic acids. Methods Enzymol. 211, 353–388. Clegg, R. M. (2002). FRET tells us about proximities, distances, orientations and dynamic properties. J. Biotechnol. 82, 177–179. Clegg, R. M., Murchie, A. I. H., Zechel, A., Carlberg, C., Diekmann, S., and Lilley, D. M. J. (1992). Fluorescence resonance energy transfer analysis of the structure of the four-way DNA junction. Biochemistry 31, 4846–4856. Cornish, P. V., Ermolenko, D. N., Noller, H. F., and Ha, T. (2008). Spontaneous intersubunit rotation in single ribosomes. Mol. Cell 30, 578–588. Cornish, P. V., Ermolenko, D. N., Staple, D. W., Hoang, L., Hickerson, R. P., Noller, H. F., and Ha, T. (2009). Following movement of the L1 stalk between three functional states in single ribosomes. Proc. Natl. Acad. Sci. USA 106, 2571–2576. Dale, R. E., and Eisinger, J. (1976). Intramolecular energy transfer and molecular conformation. Proc. Natl. Acad. Sci. USA 73, 271–273. 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. 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. Efron, B., and Tibshirani, R. J. (1993). An introduction to the Bootstrap. Chapman Hall, New York. Ermolenko, D. N., Spiegel, P. C., Majumdar, Z. K., Hickerson, R. P., Clegg, R. M., and Noller, H. F. (2007a). The antibiotic viomycin traps the ribosome in an intermediate state of translocation. Nat. Struct. Mol. Biol. 14, 493–497. Ermolenko, D. N., Majumdar, Z. K., Hickerson, R. P., Spiegel, P. C., Clegg, R. M., and Noller, H. F. (2007b). Observation of intersubunit movement of the ribosome in solution using FRET. J. Mol. Biol. 370, 530–540. Fo¨rster, T. (1948). Zwischenmolekulare energiewanderung und fluoreszenz. Ann. Phys. 2, 55–75. Gohlke, C., Murchie, A. I. H., Lilley, D. M. J., and Clegg, R. M. (1994). The kinking of DNA and RNA helices by bulged nucleotides observed by fluorescence resonance energy transfer. Proc. Natl. Acad. Sci. USA 91, 11660–11664.
184
David M. J. Lilley
Goody, T. A., Melcher, S. E., Norman, D. G., and Lilley, D. M. J. (2003). The kink-turn motif in RNA is dimorphic, and metal ion dependent. RNA 10, 254–264. Ha, T. (2001). Single-molecule fluorescence resonance energy transfer. Methods 25, 78–86. 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. Ha, T., Ting, A. Y., Liang, J., Caldwell, W. B., Deniz, A. A., Chemla, D. S., Schultz, P. G., and Weiss, S (1999a). Single-molecule fluorescence spectroscopy of enzyme conformational dynamics and cleavage mechanism. Proc. Natl. Acad. Sci. USA 96, 893–898. Ha, T., Zhuang, X., Kim, H. D., Orr, J. W., Williamson, J. R., and Chu, S. (1999b). Ligand-induced conformational changes observed in single RNA molecules. Proc. Natl. Acad. Sci. USA 96, 9077–9082. Hickerson, R., Majumdar, Z. K., Baucom, A., Clegg, R. M., and Noller, H. F. (2005). Measurement of internal movements within the 30 S ribosomal subunit using Forster resonance energy transfer. J. Mol. Biol. 354, 459–472. Hodak, J. H., Fiore, J. L., Nesbitt, D. J., Downey, C. D., and Pardi, A. (2005). Docking kinetics and equilibrium of a GAAA tetraloop-receptor motif probed by single-molecule FRET. Proc. Natl. Acad. Sci. USA 102, 10505–10510. Hohng, S., Wilson, T. J., Tan, E., Clegg, R. M., Lilley, D. M. J., and Ha, T. (2004). Conformational flexibility of four-way junctions in RNA. J. Mol. Biol. 336, 69–79. Hohng, S., Zhou, R., Nahas, M. K., Yu, J., Schulten, K., Lilley, D. M. J., and Ha, T. (2007). Fluorescence-force spectroscopy maps two-dimensional reaction landscape of the Holliday junction. Science 318, 279–283. Iqbal, A., Wang, L., Thompson, K. C., Lilley, D. M. J., and Norman, D. G. (2008a). The structure of cyanine 5 terminally attached to double-stranded DNA: Implications for FRET studies. Biochemistry 47, 7857–7862. Iqbal, A., Arslan, S., Okumus, B., Wilson, T. J., Giraud, G., Norman, D. G., Ha, T., and Lilley, D. M. J. (2008b). Orientation dependence in fluorescent energy transfer between Cy3 and Cy5 terminally-attached to double-stranded nucleic acids. Proc. Natl. Acad. Sci. USA 105, 11176–11181. Johnson, M. L., and Faunt, L. M. (1992). Parameter estimation by least-squares methods. Methods Enzymol. 210, 1–37. Joo, C., Balci, H., Ishitsuka, Y., Buranachai, C., and Ha, T. (2008). Advances in singlemolecule fluorescence methods for molecular biology. Ann. Rev. Biochem. 77, 51–76. Klein, D. J., Schmeing, T. M., Moore, P. B., and Steitz, T. A. (2001). The kink-turn: A new RNA secondary structure motif. EMBO J. 20, 4214–4221. Klostermeier, D., and Millar, D. P. (2001). Tertiary structure stability of the hairpin ribozyme in its natural and minimal forms: Different energetic contributions from a ribose zipper motif. Biochemistry 40, 11211–11218. Lafontaine, D. A., Norman, D. G., and Lilley, D. M. J. (2001). Structure, folding and activity of the VS ribozyme: Importance of the 2-3-6 helical junction. EMBO J. 20, 1415–1424. Lafontaine, D. A., Norman, D. G., and Lilley, D. M. J. (2002). The global structure of the VS ribozyme. EMBO J. 21, 2461–2471. Lakowicz, J. R. (1999). Principles of Fluorescence Spectroscopy. 2nd edn. Plenum Press, New York. Lee, T. H., Lapidus, L. J., Zhao, W., Travers, K. J., Herschlag, D., and Chu, S. (2007a). Measuring the folding transition time of single RNA molecules. Biophys. J. 92, 3275–3283. Lee, T. H., Blanchard, S. C., Kim, H. D., Puglisi, J. D., and Chu, S. (2007b). The role of fluctuations in tRNA selection by the ribosome. Proc. Natl. Acad. Sci. USA 104, 13661–13665.
Fluorescence Resonance Energy Transfer
185
Lemay, J.-F., Penedo, J. C., Tremblay, R., Lilley, D. M. J., and Lafontaine, D. A. (2006). Folding of the adenine riboswitch. Chem. Biol. 13, 857–868. Lewis, F. D., Zhang, L., and Zuo, X. (2005). Orientation control of fluorescence resonance energy transfer using DNA as a helical scaffold. J. Am. Chem. Soc. 127, 10002–10003. Lilley, D. M. J. (2000). Structures of helical junctions in nucleic acids. Q. Rev. Biophys. 33, 109–159. Lilley, D. M. J., Clegg, R. M., Diekmann, S., Seeman, N. C., von Kitzing, E., and Hagerman, P. (1995). Nomenclature Committee of the International Union of Biochemistry: A nomenclature of junctions and branchpoints in nucleic acids. Recommendations 1994. Eur. J. Biochem. 230, 1–2. Liu, J., and Lilley, D. M. J. (2007). The role of specific 20 -hydroxyl groups in the stabilization of the folded conformation of kink-turn RNA. RNA 13, 200–210. Majumdar, Z. K., Hickerson, R., Noller, H. F., and Clegg, R. M. (2005). Measurements of internal distance changes of the 30S ribosome using FRET with multiple donor-acceptor pairs: Quantitative spectroscopic methods. J. Mol. Biol. 351, 1123–1145. Mandal, M., and Breaker, R. R. (2004). Adenine riboswitches and gene activation by disruption of a transcription terminator. Nat. Struct. Mol. Biol. 11, 29–35. McKinney, S. A., Joo, C., and Ha, T. (2006). Analysis of single-molecule FRET trajectories using hidden Markov modeling. Biophys. J. 91, 1941–1951. Melcher, S. E., Wilson, T. J., and Lilley, D. M. J. (2003). The dynamic nature of the fourway junction of the hepatitis C virus IRES. RNA 9, 809–820. Murchie, A. I. H., Clegg, R. M., von Kitzing, E., Duckett, D. R., Diekmann, S., and Lilley, D. M. J. (1989). Fluorescence energy transfer shows that the four-way DNA junction is a right-handed cross of antiparallel molecules. Nature 341, 763–766. Nahas, M. K., Wilson, T. J., Hohng, S., Jarvie, K., Lilley, D. M. J., and Ha, T. (2004). Observation of internal cleavage and ligation reactions of a ribozyme. Nat. Struct. Mol. Biol. 11, 1107–1113. Neubauer, H., Gaiko, N., Berger, S., Schaffer, J., Eggeling, C., Tuma, J., Verdier, L., Seidel, C. A., Griesinger, C., and Volkmer, A. (2007). Orientational and dynamical heterogeneity of rhodamine 6G terminally attached to a DNA helix revealed by NMR and single-molecule fluorescence spectroscopy. J. Am. Chem. Soc. 129, 12746–12755. Norman, D. G., Grainger, R. J., Uhrin, D., and Lilley, D. M. J. (2000). The location of Cyanine-3 on double-stranded DNA; importance for fluorescence resonance energy transfer studies. Biochemistry 39, 6317–6324. 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. Ozaki, H., and Mclaughlin, L. W. (1992). The estimation of distances between specific backbone-labeled sites in DNA using fluorescence resonance energy transfer. Nucleic Acids Res. 20, 5205–5214. Penedo, J. C., Wilson, T. J., Jayasena, S. D., Khvorova, A., and Lilley, D. M. J. (2004). Folding of the natural hammerhead ribozyme is enhanced by interaction of auxiliary elements. RNA 10, 880–888. Pereira, M. J., 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. Perrin, F. (1932). The´orie quantique des transferts d’activation entre mole´cules de me´me espe`ce. Cas des solutions fluorescentes. Ann. Chim. Phys. 17, 283–314. Pljevaljcic, G., and Millar, D. P. (2008). Single-molecule fluorescence methods for the analysis of RNA folding and ribonucleoprotein assembly. Methods Enzymol. 450, 233–252.
186
David M. J. Lilley
Pljevaljcic, G., Millar, D. P., and Deniz, A. A. (2004). Freely diffusing single hairpin ribozymes provide insights into the role of secondary structure and partially folded states in RNA folding. Biophys. J. 87, 457–467. Rasnik, I., McKinney, S. A., and Ha, T. (2006). Nonblinking and long-lasting singlemolecule fluorescence imaging. Nat. Methods 3, 891–893. Roy, R., Hohng, S., and Ha, T. (2008). A practical guide to single-molecule FRET. Nat. Methods 5, 507–516. Rueda, D., Wick, K., McDowell, S. E., and Walter, N. G. (2003). Diffusely bound Mg2+ ions slightly reorient stems I and II of the hammerhead ribozyme to increase the probability of formation of the catalytic core. Biochemistry 42, 9924–9936. Russell, R., Zhuang, X., Babcock, H. P., Millett, I. S., Doniach, S., Chu, S., and Herschlag, D. (2002). Exploring the folding landscape of a structured RNA. Proc. Natl. Acad. Sci. USA 99, 155–160. Sanborn, M. E., Connolly, B. K., Gurunathan, K., and Levitus, M. (2007). Fluorescence properties and photophysics of the sulfoindocyanine Cy3 linked covalently to DNA. J. Phys. Chem. B 111, 11064–11074. Sase, I., Miyata, H., Corrie, J. E., Craik, J. S., and Kinosita, K. Jr. (1995). Real time imaging of single fluorophores on moving actin with an epifluorescence microscope. Biophys. J. 69, 323–328. Serganov, A., Yuan, Y. R., Pikovskaya, O., Polonskaia, A., Malinina, L., Phan, A. T., Hobartner, C., Micura, R., Breaker, R. R., and Patel, D. J. (2004). Structural basis for discriminative regulation of gene expression by adenine- and guanine-sensing mRNAs. Chem. Biol. 11, 1729–1741. 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. Steiner, M., Karunatilaka, K. S., Sigel, R. K., and Rueda, D. (2008). Single-molecule studies of group II intron ribozymes. Proc. Natl. Acad. Sci. USA 105, 13853–13858. Stone, M. D., Mihalusova, M., O’Connor, C. M., Prathapam, R., Collins, K., and Zhuang, X. (2007). Stepwise protein-mediated RNA folding directs assembly of telomerase ribonucleoprotein. Nature 446, 458–461. Stryer, L., and Haugland, R. P. (1967). Energy transfer: A spectroscopic ruler. Proc. Natl. Acad. Sci. USA 58, 719–726. 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. Turner, B., and Lilley, D. M. (2008). The importance of G.A hydrogen bonding in the metal ion- and protein-induced folding of a kink turn RNA. J. Mol. Biol. 381, 431–442. Turner, B., Melcher, S. E., Wilson, T. J., Norman, D. G., and Lilley, D. M. J. (2005). Induced fit of RNA on binding the L7Ae protein to the kink-turn motif. RNA 11, 1192–1200. Tuschl, T., Gohlke, C., Jovin, T. M., Westhof, E., and Eckstein, F. (1994). A threedimensional model for the hammerhead ribozyme based on fluorescence measurements. Science 266, 785–789. Va´mosi, G., Gohlke, C., and Clegg, R. M. (1996). Fluorescence characteristics of 5-carboxytetramethylrhodamine linked covalently to the 50 end of oligonucleotides: Multiple conformers of single-stranded and double-stranded dye-DNA complexes. Biophys. J. 71, 972–994. Walter, N. G., and Burke, J. M. (2000). Fluorescence assays to study structure, dynamics, and function of RNA and RNA-ligand complexes. Methods Enzymol. 317, 409–440. Walter, F., Murchie, A. I. H., Duckett, D. R., and Lilley, D. M. J. (1998a). Global structure of four-way RNA junctions studied using fluorescence resonance energy transfer. RNA 4, 719–728.
Fluorescence Resonance Energy Transfer
187
Walter, F., Murchie, A. I. H., and Lilley, D. M. J. (1998b). The folding of the four-way RNA junction of the hairpin ribozyme. Biochemistry 37, 17629–17636. Walter, N. G., Hampel, K. J., Brown, K. M., and Burke, J. M. (1998c). Tertiary structure formation in the hairpin ribozyme monitored by fluorescence resonance energy transfer. EMBO J. 17, 2378–2391. Walter, N. G., Burke, J. M., and Millar, D. P. (1999). Stability of hairpin ribozyme tertiary structure is governed by the interdomain junction. Nat. Struct. Biol. 6, 544–549. Wang, Y., Qin, H., Kudaravalli, R. D., Kirillov, S. V., Dempsey, G. T., Pan, D., Cooperman, B. S., and Goldman, Y. E. (2007). Single-molecule structural dynamics of EF-G–ribosome interaction during translocation. Biochemistry 46, 10767–10775. Wilhelmsson, L. M., Holmen, A., Lincoln, P., Nielsen, P. E., and Norden, B. (2001). A highly fluorescent DNA base analogue that forms Watson-Crick base pairs with guanine. J. Am. Chem. Soc. 123, 2434–2435. Wozniak, A. K., Schroder, G. F., Grubmuller, H., Seidel, C. A., and Oesterhelt, F. (2008). Single-molecule FRET measures bends and kinks in DNA. Proc. Natl. Acad. Sci. USA 105, 18337–18342. Zhuang, X., Bartley, L. E., Babcock, H. P., Russel, R., Ha, T., Herschlag, D., and Chu, S. (2000). A single molecule study of RNA catalysis and folding. Science 288, 2048–2051. Zhuang, X. W., Kim, H. D., 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
N I N E
Analysis of RNA Folding by Native Polyacrylamide Gel Electrophoresis Sarah A. Woodson* and Eda Koculi† Contents 1. Introduction 2. Theory of Gel Electrophoresis 2.1. Mobility of macromolecules 2.2. Chemical exchange 3. Electrophoresis Equipment 4. Stability of Folded RNA Measured by Native PAGE 4.1. Casting and prerunning gels 4.2. Sample preparation 4.3. Running the gel 4.4. Data analysis 5. RNA Folding Kinetics 6. RNA Compactness and Native PAGE Mobility 7. Probing the Function of Conformers Resolved by Native PAGE 7.1. Measuring RNA activity in situ with two-dimensional PAGE 7.2. Ligand-induced conformational change 8. Controls and Further Considerations 9. Summary Acknowledgments References
190 191 191 193 194 196 196 197 198 198 199 200 201 201 203 204 205 205 205
Abstract Polyacrylamide gel electrophoresis under native conditions (native PAGE) is a well-established and versatile method for probing nucleic acid conformation and nucleic acid–protein interactions. Native PAGE has been used to measure RNA folding equilibria and kinetics under a wide variety of conditions. Advantages of this method are its adaptability, absolute determination of reaction endpoints, and direct analysis of conformational hetereogeneity within a sample. Native PAGE is also useful for resolving ligand-induced structural changes. * {
T.C. Jenkins Department of Biophysics, Johns Hopkins University, Baltimore, Maryland, USA Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, Evanston, Illinois, USA
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69009-1
#
2009 Elsevier Inc. All rights reserved.
189
190
Sarah A. Woodson and Eda Koculi
1. Introduction Polyacrylamide electrophoresis under nondenaturing conditions (native PAGE) has become a popular method for analyzing nucleic acid–protein complexes(Fried andCrothers,1981; Garner andRevzin,1981),DNAbendingand flexibility (Koo et al., 1986; Wu and Crothers, 1984), and conformational changes in RNA (Emerick and Woodson, 1994; Friederich and Hagerman, 1997; Pyle et al., 1990; Stahl et al., 1979). One source of its popularity is its adaptability: native PAGE can be used over a wide range of conditions to measure folding reactions, ligand binding, and even to select populations for in vitro evolution (Bevilacqua and Bevilacqua, 1998; Kim et al., 2003; Ryder et al., 2008; Tuerk and Gold, 1990). The permutations on this method are limited only by the imagination of the experimenter. Another advantage is that native PAGE requires small amounts of material that can be produced with standard molecular biology techniques and very little specialized equipment. Native PAGE offers two other advantages for quantitative studies of macromolecule interactions. First, different conformations of the macromolecule can be visualized and enumerated, as long as they have different electrophoretic mobilities. This is also an advantage when measuring RNA– protein binding, as complexes with different stoichiometries can be distinguished (Fried and Daugherty, 1998). Second, the fraction of the population in each conformational state can be quantified in absolute terms. The latter is enormously helpful in obtaining reliable endpoints for folding or binding reactions. In contrast, footprinting data and spectroscopic data such as FRET must often be interpreted relative to some saturation point, with the assumption that all of the components in the system are active. Native PAGE has several disadvantages. First, it is pseudoequilibrium method, like nitrocellulose filtration and footprinting. The separation of folded or bound polynucleotide requires that macromolecules are ‘‘caged’’ within the gel matrix during several hours of electrophoresis (Fried and Liu, 1994). As discussed below, the quality of the separation and interpretation of the results depends on the dynamics of the system and the rate of transport through the gel. Second, the electrophoretic mobility cannot be easily predicted from theory and thus interpreted in terms of a physical property such as hydrodynamic radius. For these reasons, it is desirable to complement native PAGE studies with other methods such as activity assays, footprinting, UV, and fluorescence spectroscopy or small angle scattering. This chapter will focus on the use of native PAGE to investigate folding of the Tetrahymena group I ribozyme. However, these protocols are easily adapted to other ribozymes and structured RNAs (e.g., Adilakshmi et al., 2005; Lafontaine et al., 2002; Pinard et al., 2001; Severcan et al., 2009). Detailed discussions of gel mobility shift methods for measuring protein–
191
RNA Folding by Native PAGE
nucleic acid interactions are available elsewhere (Fried and Daugherty, 1998; Ryder et al., 2008). In the following section we discuss the parameters that govern the design of a native PAGE experiment and present protocols for the analysis of RNA folding reactions.
2. Theory of Gel Electrophoresis 2.1. Mobility of macromolecules One of the most successful models for gel electrophoresis is the reptation theory of Lumpkin and Zimm for the migration of double-stranded DNA (Lumpkin, 1982). An in-depth discussion can be found in Zimm and Levene (1992); for a synopsis see Bloomfield et al. (2000). The velocity n of a charged particle in a solution with an electric field E depends on the electrical force Fel ¼ ZqE, in which Z is the number of charges and q is the charge of a proton, and the frictional force Ffr ¼ fv, in which f is the frictional coefficient. At steady state, these forces balance and the velocity is v ¼ ZqE=f . The electrophoretic mobility m is the velocity relative to the field strength, m ¼ vE ¼ Zq=f . During gel electrophoresis, the migration of macromolecules is obstructed by the polymer matrix, and thus depends on the molecular weight as well as the frictional coefficient with the matrix. In the reptation model, DNA is assumed to move in a worm-like fashion through virtual tubes in the gel polymer matrix (Fig. 9.1A). The central result is that the electrophoretic mobility depends on the mean-square end-to-end distance of the macromolecule, and inversely on length (L) or molecular weight. The electrophoretic mobility is expressed as m¼
hh2x iZq L 2 fgel
ð9:1Þ
in which L is the contour length (related to the number of residues) and hx is the component of the end-to-end vector of the polymer that is aligned with the electric field (Zimm and Levene, 1992). Thus, DNA molecules that are bent have a shorter end-to-end distance and migrate more slowly than those that are straight (Wu and Crothers, 1984). Another prediction of reptation theory is that molecules move fastest when the entire chain is in the same tube. Partial unfolding or branching of the helix makes this less likely, and consequently impede migration, resulting in anomalous migration patterns that can be used to model helical junctions and bend angles (Lilley, 2008; Zinkel and Crothers, 1990).
192
Sarah A. Woodson and Eda Koculi
A x,E hx B
+
+ R − − − + +
E
C
150 nm
Figure 9.1 Models for macromolecular electrophoresis. (A) Reptation of long DNA fragments through a polyacrylamide gel. Redrawn from Bloomfield et al. (2000). (B) Ogston sieve model, which applies when Rg of the macromolecule is smaller than the diameter of the pore. (C) Scanning electron micrograph of the interior of a 7.5% (w/v) polyacrylamide gel. Reprinted from Yuan et al. (2006) with permission.
In practice, the migration of a nucleic acid through the gel is complicated by its interactions with the ions in the gel running buffer (Mohanty and McLaughlin, 2001; Stellwagen et al., 2003), field-dependent bias in the orientation of the leading edge of the DNA, and by the elasticity of the gel matrix (Zimm and Levene, 1992). Folded RNAs are similar in size to the pores of a typical 4–20% polyacrylamide gel (1–8 nm) (Chrambach and Rodbard, 1971) and thus may be more appropriately described by a sieving model in which the macromolecule must avoid collisions with the gel matrix (Sartori et al., 2003) (Fig. 9.1B). In this model, proposed by Ogston (Ogston, 1958), the gel mobility m relative to the mobility in free solution mo decreases with gel concentration c and a retardation coefficient Kr, or log m ¼ log mo Krc (Ferguson, 1964). Therefore, tightly folded RNAs travel more rapidly than unfolded RNAs of the same molecular weight. The absolute electrophoretic mobility is difficult to predict, however, because of dynamic fluctuations in the macromolecule and in the gel matrix itself (Locke and Trinh, 1999; Stellwagen et al., 2003; Yuan et al., 2006). Happily, the observation of a relative change in gel mobility is sufficient for many applications.
RNA Folding by Native PAGE
193
2.2. Chemical exchange The dynamics of the system are an important consideration when native PAGE is used to analyze RNA folding or protein–nucleic acid interactions. If chemical exchange between conformers (or bound and free RNA) is rapid relative to the rate of transport through the gel, a single band of average mobility will be observed (Cann, 1996). If exchange is slow relative to the rate of transport, then bands with different mobilities will be observed, corresponding to each conformational state. If the exchange rate is comparable to the rate of transport, then an additional band or zone of intermediate mobility will also be observed, representing material that has undergone exchange during electrophoresis. Even a small amount of exchange can result in fuzzy bands or a faint track of labeled material between the two principle bands. Examples of RNA conformers in fast exchange can be found in folding studies on the P4–P6 domain of the Tetrahymena ribozyme and the yeast bI5 mitochondrial ribozyme (Buchmueller et al., 2000; Szewczak and Cech, 1997). In these studies, the average electrophoretic mobility increased as more of the RNA is folded, but only a single band was observed (Fig. 9.2A). The fraction of folded RNA is extracted from the mobility of the RNA, relative to an external standard such as a DNA restriction fragment or relative to the maximum mobility of the RNA when it assumed to be completely folded. The mobility of the P4–P6 RNA decreased with the gel percentage, further confirming that mobility reflects the average compactness of the RNA (Fig. 9.2B) (Szewczak and Cech, 1997). Studies on the Tetrahymena ribozyme and pre-RNA provide an example of RNA conformers in slow exchange (Emerick and Woodson, 1994; Pan et al., 2000). In this case, each conformer migrates at a characteristic rate, producing several distinct bands, or if there are many structures, a broad smear of material (Fig. 9.3A and B). The native RNA migrates fastest, while unfolded or misfolded forms migrate more slowly. Each band is quantified separately, and the fraction of folded RNA is obtained from the amount of material in the native band. In general, we have found that even small changes in electrophoretic mobility reliably signal a change in structure that ought not to be ignored. The ability to separate liganded forms of an RNA by native PAGE depends not only on the rate of chemical exchange but also on the size of the molecule and its rate of diffusion within the pores of the gel during electrophoresis. In general, complexes with small molecules are more difficult to detect by native PAGE than complexes with large molecules such as proteins. Moreover, positively charged ligands move opposite to the RNA in an electric field. Thus, ions such as Mg2þ must be added to the running buffer so that the RNA remains associated with Mg2þ ions as it travels through the gel. Interactions between lambda N peptides and box B RNA
194
BP5/5a
U1 + U2
U2
P4-P6
A
U1
Sarah A. Woodson and Eda Koculi
Extended Compact
B −3.8
10 mM MgCl2
Log M
−4
−4.2
P4P6 U1 U2 U1U2 BP55a
−4.4
−4.6
0
2
4
6
8
10
Gel percentage Figure 9.2 Native gel electrophoresis of the Tetrahymena P4–P6 RNA. (A) The folded and extended forms equilibrate rapidly, producing a single band whose mobility reflects the average structure of the RNA. U1, U2, BP5/5a refer to mutations in a hinge region. RNAs were run on native 6% (19:1 mono:bis) polyacrylamide gel in TBE þ 10 mM MgCl2 at 25 C. (B) Ferguson plot shows that the relative mobility (M) depends linearly on gel concentration, as predicted by the Ogston model. Reprinted from Szewczak and Cech (1997).
hairpins were successfully measured by adding peptide to the gel before polymerization (Cilley and Williamson, 1997).
3. Electrophoresis Equipment A variety of commercial apparatuses or ‘‘gel boxes’’ for vertical PAGE are suitable for native PAGE experiments. Temperature control is critical to the separation of RNA conformers (slow exchange), and thus the box must be designed for use with a circulating water bath. We have obtained good
195
RNA Folding by Native PAGE
A
B
wt L2L5cP3 0 0.5 1 10 0 0.5 1 10 min
{I}
{INS} Mg2+
{U} IN
N <30 s Native gel
C
I I N
N Na+ K+
[Co(NH3)6]3+ spd3+ Mg2+ Ba2+ 1 0.8
fN
0.6 0.4 0.2 0 0.001
0.01
0.1
1 10 [Cation] (mM)
100
1000
104
Figure 9.3 Native gel electrophoresis of the Tetrahymena ribozyme. (A) Folded (N) and misfolded (I) forms of the RNA are in slow exchange and are trapped within the matrix of the gel, migrating at different rates. Folded RNA is stabilized by Mg2þ in the running buffer. (B) Folding kinetics. Ribozyme was incubated in folding buffer containing MgCl2 for 0–10 min before samples were loaded on a native 8% (29:1) polyacrylamide gel in THEM3 at 4 C. wt, wild type; L2P5cP3 is a mutant that increases the folding rate. The fraction of native RNA is determined from the volume of counts in the native band relative to the total counts in the lane. (C) Folding equilibrium of the ribozyme in different cations measured by native PAGE. The fraction of N was fit to equation 9.2. Redrawn from Pan et al. (2000) and Heilman-Miller et al. (2001).
results with an Owl Penguin P10DS gel box (ThermoScientific) connected to a refrigerated recirculating water bath. Cold water circulates from the water bath through a central core that lies between two vertical polyacrylamide gel sandwiches. The surfaces that contact the gel sandwich are made of alumina, in order to improve heat transfer from the gel. A 20 20 cm format supplies the most area for critical separations, but smaller formats may be satisfactory for some applications. In addition to casting materials (plates, spacers, combs), an electrophoresis power supply capable of delivering
196
Sarah A. Woodson and Eda Koculi
1000 V and maintaining constant power is also needed. It is helpful to place a small water bath or temperature block nearby for incubating samples prior to loading. In general, we find that the internal temperature of the gel must be 10 C for good separation of RNA conformers. Separation of the Tetrahymena ribozyme is lost when the temperature rises above 15 C. With our apparatus, an internal temperature of 10 C can be achieved by setting the water bath to 0–4 C and by applying no more than 15 W electrical power to each gel (or 30 W for two gels). Although the running buffer is 2–4 C, the gel interior is warmer owing its electrical resistance. The water bath must recirculate vigorously and have adequate capacity. We use an old Neslab RTE-110 (5 L) with cooling capacity of 300 W at 0 C and 12 L/min pumping rate. It should be filled with 1:1 (v/v) water:ethylene glycol to prevent icing. One water bath can be used to cool two doublesided gel boxes. We insulate the tubing from the water bath with foam to help maintain a low temperature. If the temperature of the water bath is set below 2 C, care should be taken that ice or precipitates do not obstruct the current and produce uneven results. Running gels in a cold room does not provide sufficient heat transfer and is unsuitable for the analysis of RNA conformers (although it may work for protein gel mobility shifts).
4. Stability of Folded RNA Measured by Native PAGE 4.1. Casting and prerunning gels Solutions for native polyacrylamide gels are prepared and cast in glass frames according to standard methods. We use 0.5-mm-thick Teflon spacers and combs with 24–30 wells in a 20-cm-wide gel. Short and long glass plates are assembled with side spacers and clamped on each side with large black binder clips. Polymerization of native gels is particularly sensitive to contaminants. Grease or silicone lubricants are to be avoided and plates should be cleaned meticulously with soap, water, and finally ethanol before and after each use. The bottom of the casting frame may be sealed with electrical tape if desired, although this is not necessary. To resolve folded and unfolded conformations of the 387 nt Tetrahymena ribozyme, we use 8% acrylamide (29:1 mono:bisacrylamide) in 34 mM Tris, 66 mM Hepes (pH 7.5), 0.1 mM EDTA, and 3 mM MgCl2 (THEM3). Hepes is used instead of borate to maintain the native structure of the ribozyme (Buchmueller and Weeks, 2004; Pyle et al., 1990), whereas the MgCl2 concentration is chosen to be just sufficient to maintain the RNA in its folded state. Twenty-five milliliters of acrylamide solution per gel is prepared and degassed using RNase-free water. 200 mL 10% ammonium
RNA Folding by Native PAGE
197
persulfate and 50 mL TEMED is added to begin polymerization, and the solution is poured immediately into the prepared frame. The comb is inserted and the gel frame is laid flat until the polyacrylamide has polymerized completely (about 45 min). When the gel has polymerized, the comb is removed and wells are flushed thoroughly with deionized (RNase-free) water. At this stage, the gel can be covered with plastic wrap and stored overnight at 4 C, if desired. The binder clamps (and tape if used) are removed and the gel is placed on the gel box. The upper and lower reservoirs are filled with 1 THEM and prerun for 30–45 min with the water bath at the desired temperature, making sure that the current is not blocked by ice, precipitate, or air bubbles. The gel temperature can be monitored with a contact thermometer or small thermal probe inside the gel.
4.2. Sample preparation RNA samples for native PAGE can be prepared in many different buffers or salts and at different temperatures, depending on the scientific question to be addressed. We provide a protocol for measuring the folding of the Tetrahymena ribozyme that can be adapted to other RNAs as needed. Since native gels are easily overloaded, best results are obtained with low (0.1–10 mg/mL) RNA concentrations, although we have used up to 1 mg/ mL RNA. Sample volumes should be 2–4 mL per lane to produce tight bands. To detect small quantities of RNA, the RNA must be labeled with a fluorescent dye or a radioisotope. 32P-labeling is most sensitive and easily quantified with storage phosphorescence scanners. To measure the folding equilibrium of the Tetrahymena L-21Sca ribozyme in various counterions, 32P-labeled ribozyme is prepared by T7 in vitro transcription with a-[32P]-ATP, according to established protocols (Zaug et al., 1988). Free label is removed with a size-exclusion spin column (e.g., TE-10, BD Sciences), and the RNA used without further purification (Emerick and Woodson, 1993). If desired, the RNA can be end-labeled with 32P and purified by denaturing PAGE. As for all biochemical experiments on RNA, care must be taken to use water, reagents, and plastic consumables that are free of RNase. For the folding reactions (5–10 mL), labeled ribozyme (1000–2000 cpm/ mL) is added to HE buffer (50 mM Hepes adjusted to pH 7.5 with sodium hydroxide, 1 mM EDTA, pH 8), 10% (v/v) glycerol, 0.01% (w/v) xylene cyanol, plus the desired concentration of MgCl2 or other salt (HeilmanMiller et al., 2001; Koculi et al., 2004). At least one sample should contain no MgCl2, representing the ‘‘unfolded’’ RNA, and one sample should contain enough MgCl2 to fold the RNA completely. The reactions are incubated at the desired temperature for sufficient time for the reaction to reach equilibrium. We incubate the Tetrahymena ribozyme 2–4 h in a water
198
Sarah A. Woodson and Eda Koculi
bath at 30 C. Shorter incubation times suffice at higher temperatures, or for RNAs that fold rapidly (Rangan et al., 2004).
4.3. Running the gel When the folding reactions have reached equilibrium, 2 mL of each sample is loaded into separate lanes of a native 8% polyacrylamide gel that was prepared and prerun as described above (<10 C). It is helpful to use narrow tips that allow the sample to be placed near the bottom of the well. The current should be applied to the gel as soon as possible after the samples are loaded. Gels must be run long enough to resolve the conformational species of interest, but not so long that smaller RNAs run off the bottom of the gel. To resolve the folded and unfolded forms of the Tetrahymena ribozyme, gels are run 4 h at 15 W. In this protocol, glycerol and tracking dyes were added to the RNA folding reaction, so the samples can be loaded directly without further manipulation. However, glycerol slightly perturbs the folding free energy of the ribozyme (Pan et al., 1999), and prevents the formation of some RNA–protein complexes (Bellur and Woodson, 2009). An alternative approach is to prepare samples without glycerol, then rapidly mix it with a tenth volume of glycerol and dyes just before loading aliquots on the gel. At the end of the run, the buffer reservoirs are drained and the gel frame is disassembled, taking appropriate precautions if using radioactive samples. One glass plate is carefully removed and the gel is transferred to a piece of Whatman 3 mm filter paper. The other side of the gel is covered with plastic wrap and completely dried in a vacuum gel dryer at low heat (to avoid cracking). The dried gel is directly exposed to an imager or to X-ray film.
4.4. Data analysis Although a qualitative analysis is sometimes sufficient, in most cases it is desirable to quantify the amount of labeled material in each conformational state. This can be done using commercial or free image analysis software to integrate pixels within a certain area. One approach is to obtain an intensity profile of each lane and integrate the area under each peak (usually from one row of pixels, although some programs can average several rows; Das et al., 2005). Area integration can distinguish broad and sharp peaks, and overlapping peaks can be sometimes deconvoluted. A disadvantage of this method is that the results are sensitive to distortions in the gel because only a thin strip of each lane is used. Alternatively, peaks areas can be defined manually and the entire volume of the peak integrated. This method is more tolerant of imperfect gels. However, because the bands in native gels are often broad, it is critical to set a uniform criterion (such as pixel intensity) for defining peak boundaries and
RNA Folding by Native PAGE
199
to carefully subtract the background. One way to estimate the background is to integrate a similar number of pixels from a region of the gel in which no sample was loaded. Ideally, at least the folded RNA will produce a sharp band that can be clearly distinguished from other bands on the gel (Fig. 9.3). If this is not the case, it may be worth optimizing the electrophoresis conditions to improve the separation or changing the folding buffer to increase the stability of the RNA. Once the peaks have been integrated, the amount of each labeled species relative to the total is easily calculated for each lane in the gel P ( fi ¼ countsi =countstotal countsi = i countsi ). The results of the experiment can be fit to an appropriate model. For folding experiments, the fraction of native RNA versus Mg2þ concentration C was fit to the Hill equation: ðC=Cm Þn ð9:2Þ fN ¼ fN ð0Þ þ ½ fN ðmaxÞ fN ð0Þ 1 þ ðC=Cm Þn where ƒN(0) is the fraction of folded RNA without Mg2þ, ƒN(max) is the maximum fraction of folded RNA in saturating Mg2þ, Cm is the midpoint of the folding transition, and n represents the cooperativity of the folding equilibrium with respect to Mg2þ (Fig. 9.3C). Alternatively, the free energy of the folding transition can be calculated from O¼
1 DGUN 2 1 C 2 ðdfN =dCÞmax pffiffiffi ¼ max DC 8 RT lnð3 þ 2 2Þ
ð9:3Þ
where R is the gas constant, T is temperature, and DGUN is the free energy change associated with the Mg2þ-dependent folding transition (Pan et al., 1999). O is a dimensionless quantity that is related to the Hill coefficient and the midpoint of folding transition, Cmax is the concentration of Mg2þ at which the derivative of fraction native with respect to C reaches its maximum, (dfN/dC)max is the maximum value of the derivative, DC is the width of the curve at 1/2(dfN/dC)max. This method has the advantage of being less sensitive to variations in the upper and lower baselines of the curve. In both cases, DGUN is assumed to vary with ln C, so that DGUN ¼ DGref n ln C (Fang et al., 1999; Pan et al., 1999).
5. RNA Folding Kinetics Native PAGE can also be used to measure the kinetics of RNA folding. The time resolution of the method is limited by the time needed to mix the samples and to load them into the well of the gel itself, which typically require 15–30 s. Thus, this method is only suitable for reactions
200
Sarah A. Woodson and Eda Koculi
with half-lives greater than 1 min (Fried and Crothers, 1981). As discussed above, the folding reaction effectively stops once the samples enter the gel, trapping the molecules in whatever conformation they held when they entered the gel. Refolding of the Tetrahymena ribozyme is very slow at the low temperatures of the running buffer and the gel. Because samples must be loaded on the gel immediately, the native gel is cast and prerun before the start of the reaction. Folding reactions are set up in a sufficient volume so that aliquots can be withdrawn at various times (20–40 mL). We obtain the most reproducible results by first mixing all of the reaction components except the RNA. This mixture is warmed to the desired reaction temperature (e.g., 30–50 C) in a water bath or heating block placed near the native gel apparatus. The folding reaction is then started by adding a 5 or 10 stock of unfolded RNA to the folding buffer. To follow the reaction over time, a 2 mL aliquot is removed as soon as possible (e.g., 15 s) and loaded in the first well of the gel. With the current running (do not touch the gel), reaction aliquots are carefully loaded in adjacent wells at different times, with intervals chosen to span the expected half-life of the reaction. It is recommended that one also prepare samples of unfolded and fully renatured RNA as controls. The gels are run and analyzed as described above. The increase in the fraction folded RNA over time is fit to a rate equation, such as fN ðtÞ ¼ Amax ½1 expðktÞ
ð9:4Þ
in which k is the observed rate constant and Amax ƒN(1) is the maximum amplitude of the reaction. One difference between time courses and equilibrium experiments is that samples loaded at the beginning of the reaction travel further through the gel than samples loaded at the end of the reaction (Fig. 9.3B). The resulting curvature in the banding pattern does not interfere with the analysis as long as all lanes run long enough to achieve the desired resolution of the sample. Control samples can be loaded at the beginning and the end of the time course to facilitate band assignment.
6. RNA Compactness and Native PAGE Mobility As discussed above, the ability of folded RNAs to migrate through the gel depends on their size relative to the pores of the gel and on structural fluctuations that ‘‘snag’’ the RNA on obstacles within the matrix during electrophoresis. This principle can be used to estimate the compactness of the folded RNA, as more tightly folded RNAs should sieve through the gel more easily. Buchmueller et al. (2000) measured the mobility of the bI5 ribozyme in native gels containing 0–20 mM MgCl2. Double-stranded 174 bp RNA,
RNA Folding by Native PAGE
201
which is not expected to change structure upon the addition of Mg2þ, was loaded on the same gel as a control. The velocity of the ribozyme through the gel relative to the dsRNA control increased with Mg2þ, indicating the RNA was more tightly folded. This change in mobility correlated with a decrease in the Stokes’ radii (RH) measured by gel permeation chromatography (Buchmueller et al., 2000). We measured the mobility of the Tetrahymena ribozyme relative to DNA restriction fragments in native polyacrylamide gels containing MgCl2, CaCl2, and SrCl2 (Koculi et al., 2007). In these studies, the folded RNA migrated fastest in MgCl2 and slowest in SrCl2, consistent with a less tightly packed and more dynamic structure in the larger cations that was also reflected in the breadth of the RNA peak in gel permeation chromatography (Koculi et al., 2007) (Fig. 9.4). We also measured the radius of gyration (Rg) of the Tetrahymena ribozyme by small angle X-ray scattering, and found that it differs very little in MgCl2, CaCl2, and SrCl2 (Moghaddam et al., 2009). Thus, the data from native PAGE must be interpreted cautiously, as both size and dynamic fluctuations can lower the mobility of an RNA (Olson et al., 1993).
7. Probing the Function of Conformers Resolved by Native PAGE One of the most important questions is how the conformational states resolved by native PAGE differ in their structure and function. Sometimes the bands can be assigned to functional states by comparing the results with those from other assays. For example, the proportion of the fast migrating band of the Tetrahymena pre-RNA at different MgCl2 concentrations correlated with its self-splicing activity under the same conditions, allowing this band to be assigned to the native state (Pan and Woodson, 1998). This correlation between solution activity and native PAGE results also provided reassurance that the amount of folded RNA trapped by the native gel corresponded faithfully with the proportion of active RNA.
7.1. Measuring RNA activity in situ with two-dimensional PAGE An important control is to determine whether the species with different electrophoretic mobility can exchange. This is done by eluting RNA from each band in the native gel, incubating it for an appropriate time under folding conditions, and repeating the native PAGE analysis. Assuming the RNA conformations are in exchange, it is not possible to determine the activity of each species in solution. However, the reactivity or conformation
202
Sarah A. Woodson and Eda Koculi
A
MgCl2
CaCl2
SrCl2
ΦX U F
ΦX U F
ΦX U F Mg
Origin 1353 1078 872 603
Origin
Origin
1353 1078 872
1353
310 281 271
603
234 194
310 281
118
271
1078 872
603
310 281
234 B
271
Mg2+
Relative movement of RNA
1.2 Ca2+ 310 bp 1.1
Sr2+
1 271 bp 0.9
0.8 0.025
0.03 0.035 0.04 0.045 0.05 Counterion charge density (C/Å3)
0.055
Figure 9.4 Mobility of the Tetrahymena ribozyme in different divalent metal ions. (A) The unfolded (U) and folded (F) ribozyme was run next to FX DNA size markers on native 8% PAGE in THE buffer with 3 mM MgCl2, CaCl2, or SrCl2. (B) The relative RNA mobility decreased with the size and charge density of the metal ion. Reprinted from Koculi et al. (2007).
of the RNA can be probed in situ by diffusing substrates or modifying reagents into the gel. Detailed protocols for ‘‘fingerprinting’’ RNA conformers can be found elsewhere (Woodson, 2001).
RNA Folding by Native PAGE
203
For the Tetrahymena pre-RNA, native and partly misfolded conformers were first separated on a 6% native gel as described above (Emerick and Woodson, 1994). After the first round of electrophoresis, gel slices corresponding to each lane were excised and laid on the bottom of a glass plate for a second round of denaturing electrophoresis. A solution of GTP, which is a cofactor for self-splicing of the pre-RNA, was pipetted onto the surface of the gel slices. After a few minutes, self-splicing was quenched by applying a solution of 8 M urea and EDTA. A second plate and spacers were added to the first plates, and the denaturing polyacrylamide gel cast in place over the gel slices from the first dimension. The second dimension of denaturing PAGE resolved the products of the self-splicing reaction, which are lower in molecular weight. Only the fast migrating band gave rise to fully spliced RNAs, demonstrating that this band represented the native (N) form of the pre-RNA (Emerick and Woodson, 1994). An alternative approach is to probe the conformation of the RNA in the native gel by chemical modification (Emerick and Woodson, 1994). Rather than using two-dimensional electrophoresis, each band was excised from the native gel and placed in a 1.5 mL tube on ice. A solution of dimethylsulfate (DMS) was pipetted over the gel slice, followed by a quench of 1 M beta-mercaptoethanol. The modified RNA was then eluted from the gel slice and analyzed by primer extension (Woodson, 2001). This approach has the advantage of probing the RNA structure in situ. However, the extent of modification within the gel slice is difficult to control, and many chemical modification reagents are not reactive enough to work within the gel matrix. An alternative is chemical modification interference or nucleotide analog interference (NAIM) (Christian and Yarus, 1992; OrtolevaDonnelly et al., 1998; Pan and Woodson, 1998; see chapter by Strobel). In these methods, the RNA is modified in solution before native PAGE, to identify modifications that interfere with folding or ligand binding.
7.2. Ligand-induced conformational change Another method of probing the functions of an RNA’s conformational states is to determine whether each conformational species can interact specifically with ligands. If the ligand changes the mobility of the RNA, then ligand binding can be assessed simultaneously by native PAGE. For example, Draper and coworkers separated two isoforms of a pseudoknot within the E. coli alpha operon mRNA by native PAGE (Gluick et al., 1997). Ribosomal protein S4, which represses translation of the alpha operon, selectively retarded the migration of the fast isomer, showing that the fast form of the RNA is bound by the protein (Schlax et al., 2001). We observed that GTP shifts the Tetrahymena pre-RNA to a conformation that migrates just a bit more slowly than the native pre-RNA (Emerick et al., 1996; Pan et al., 1999). Only the native form of the RNA was affected;
204
Sarah A. Woodson and Eda Koculi
the mobility of bands containing unfolded or misfolded RNA did not change in the presence of GTP. The shifted band contains a complex of spliced RNAs (Emerick et al., 1996), and was thus a useful diagnostic for the RNA’s catalytic activity (Pan et al., 1999).
8. Controls and Further Considerations Because conformational changes in RNA or short DNAs typically cause small changes in electrophoretic mobility, analysis of nucleic acid folding requires careful optimization of electrophoresis conditions. By contrast, protein–nucleic acid interactions are typically easier to analyze by native PAGE because the molecular weight and positive charge of the protein produces a relatively large shift in gel mobility. In designing the experiments, the native conformation of the RNA (or the RNA–protein complex in a gel shift assay) must be trapped in the matrix of the gel during loading of the sample and remain stable during the electrophoresis run. First, the concentration of MgCl2 in the running buffer, as well as the buffer itself, can be varied, depending on the stability of the RNA to be studied. Second, the polyacrylamide concentration and crosslinker ratio should be optimized for each system. We have used 6% polyacrylamide (29:1 mono:bis) for a 500–700 nt RNAs, 8% for 200–400 nt ribozymes, and 8–12% for oligonucleotides. Evidence that native PAGE results reflect solution conditions rather than conditions in the gel come from experiments on the Tetrahymena pre-RNA in different ions. While the Tetrahymena ribozyme can fold in Ca2þ, splicing requires Mg2þ ions in the active site (Grosshans and Cech, 1989; Streicher et al., 1996). As discussed above, a GTP-dependent mobility shift is diagnostic for the catalytic competence of the pre-RNA (Pan et al., 1999). The GTP-dependent shift was observed when Mg2þ was added to the samples and Ca2þ was added to the gel running buffer, but not when the samples contained Ca2þ and the gel contained Mg2þ. Thus, the native PAGE results reflect the conformational state of the RNA before it was loaded on the gel, rather than any changes that might have occurred in the gel itself. A final concern is the extent to which the proportion of each conformer is faithfully captured by the native gel. Although very small structural differences, such as isomerization within an active site, may not be resolved unless they alter the hydrodynamic profile of the RNA, native PAGE results generally correlate well with other measures of RNA folding in solution. Because 10–30 s are needed for samples to enter the gel, native PAGE is most successful at resolving conformational states that do not exchange within this time (Fig. 9.3A). The entrapment of different conformers is
RNA Folding by Native PAGE
205
aided by the low temperature of the running buffer, which slows conformational exchange in the sample well and in the gel itself. For example, misfolded forms of the Tetrahymena ribozyme refold very slowly at 4 C, and are easily separated from the native form (Pan and Woodson, 1998). However, if the ribozyme is first incubated in another ion such as Naþ that allows the RNA to come close to the native structure, these native-like intermediates are captured as the native form when the RNA encounters Mg2þ in the gel running buffer (Figure 9.3A) (HeilmanMiller et al., 2001). Similarly, the Azoarcus ribozyme rapidly forms nativelike, compact intermediates in Mg2þ concentrations below that required for catalytic activity (Rangan et al., 2003). These intermediates also appear in the folded state when assayed by native PAGE.
9. Summary Native PAGE is a versatile method for probing the equilibria and kinetics of RNA folding reactions, and the interactions between RNAs and their ligands. Its principal advantage is the ability to resolve and quantify conformational heterogeneity within a system. Native PAGE is best suited for resolving large scale structural changes and those that are in slow exchange. The mobility of individual macromolecules is also sensitive to conformational fluctuations during electrophoresis, and future developments may lead to further use of electrophoresis through gels and soluble polymers for the study of molecular dynamics (Sartori et al., 2003).
ACKNOWLEDGMENTS The authors thank the many members of the Woodson laboratory who have contributed to these methods over the years, and the NIH (GM46686) for support.
REFERENCES Adilakshmi, T., Ramaswamy, P., and Woodson, S. A. (2005). Protein-independent folding pathway of the 16S rRNA 50 domain. J. Mol. Biol. 351, 508–519. Bellur, D. L., and Woodson, S. A. (2009). A minimized rRNA-binding site for ribosomal protein S4 and its implications for 30S assembly. Nucleic Acids Res. 37, 1886–1896. Bevilacqua, J. M., and Bevilacqua, P. C. (1998). Thermodynamic analysis of an RNA combinatorial library contained in a short hairpin. Biochemistry 37, 15877–15884. Bloomfield, V. A., Crothers, D. M., and Tinoco, I. J. (2000). Nucleic Acids: Structures, Properties, and Functions. University Science Books, Sausalito, CA. Buchmueller, K. L., and Weeks, K. M. (2004). Tris–borate is a poor counterion for RNA: A cautionary tale for RNA folding studies. Nucleic Acids Res. 32, e184.
206
Sarah A. Woodson and Eda Koculi
Buchmueller, K. L., Webb, A. E., Richardson, D. A., and Weeks, K. M. (2000). A collapsed, non-native RNA folding state. Nat. Struct. Biol. 7, 362–366. Cann, J. R. (1996). Theory and practice of gel electrophoresis of interacting macromolecules. Anal. Biochem. 237, 1–16. Chrambach, A., and Rodbard, D. (1971). Polyacrylamide gel electrophoresis. Science 172, 440–451. Christian, E. L., and Yarus, M. (1992). Analysis of the role of phosphate oxygens in the group I intron from Tetrahymena. J. Mol. Biol. 228, 743–758. Cilley, C. D., and Williamson, J. R. (1997). Analysis of bacteriophage N protein and peptide binding to boxB RNA using polyacrylamide gel coelectrophoresis (PACE). RNA 3, 57–67. Das, R., Laederach, A., Pearlman, S. M., Herschlag, D., and Altman, R. B. (2005). SAFA: Semi-automated footprinting analysis software for high-throughput quantification of nucleic acid footprinting experiments. RNA 11, 344–354. Emerick, V. L., and Woodson, S. A. (1993). Self-splicing of the Tetrahymena pre-rRNA is decreased by misfolding during transcription. Biochemistry 32, 14062–14067. Emerick, V. L., and Woodson, S. A. (1994). Fingerprinting the folding of a group I precursor RNA. Proc. Natl. Acad. Sci. USA 91, 9675–9679. Emerick, V. L., Pan, J., and Woodson, S. A. (1996). Analysis of rate-determining conformational changes during self- splicing of the Tetrahymena intron. Biochemistry 35, 13469–13477. Fang, X., Pan, T., and Sosnick, T. R. (1999). A thermodynamic framework and cooperativity in the tertiary folding of a Mg(2þ)-dependent ribozyme. Biochemistry 38, 16840–16846. Ferguson, K. A. (1964). Starch-gel electrophoresis—Application to the classification of pituitary proteins and polypeptides. Metabolism 13(Suppl.), 985–1002. Fried, M., and Crothers, D. M. (1981). Equilibria and kinetics of lac repressor–operator interactions by polyacrylamide gel electrophoresis. Nucleic Acids Res. 9, 6505–6525. Fried, M. G., and Daugherty, M. A. (1998). Electrophoretic analysis of multiple protein–DNA interactions. Electrophoresis 19, 1247–1253. Fried, M. G., and Liu, G. (1994). Molecular sequestration stabilizes CAP–DNA complexes during polyacrylamide gel electrophoresis. Nucleic Acids Res. 22, 5054–5059. Friederich, M. W., and Hagerman, P. J. (1997). The angle between the anticodon and aminoacyl acceptor stems of yeast tRNA(Phe) is strongly modulated by magnesium ions. Biochemistry 36, 6090–6099. Garner, M. M., and Revzin, A. (1981). A gel electrophoresis method for quantifying the binding of proteins to specific DNA regions: Application to components of the Escherichia coli lactose operon regulatory system. Nucleic Acids Res. 9, 3047–3060. Gluick, T. C., Gerstner, R. B., and Draper, D. E. (1997). Effects of Mg2þ, Kþ, and Hþ on an equilibrium between alternative conformations of an RNA pseudoknot. J. Mol. Biol. 270, 451–463. Grosshans, C. A., and Cech, T. R. (1989). Metal ion requirements for sequence-specific endoribonuclease activity of the Tetrahymena ribozyme. Biochemistry 28, 6888–6894. Heilman-Miller, S. L., Thirumalai, D., and Woodson, S. A. (2001). Role of counterion condensation in folding of the Tetrahymena ribozyme. I. Equilibrium stabilization by cations. J. Mol. Biol. 306, 1157–1166. Kim, S., Shi, H., Lee, D. K., and Lis, J. T. (2003). Specific SR protein-dependent splicing substrates identified through genomic SELEX. Nucleic Acids Res. 31, 1955–1961. Koculi, E., Lee, N. K., Thirumalai, D., and Woodson, S. A. (2004). Folding of the Tetrahymena ribozyme by polyamines: Importance of counterion valence and size. J. Mol. Biol. 341, 27–36.
RNA Folding by Native PAGE
207
Koculi, E., Hyeon, C., Thirumalai, D., and Woodson, S. A. (2007). Charge density of divalent metal cations determines RNA stability. J. Am. Chem. Soc. 129, 2676–2682. Koo, H. S., Wu, H. M., and Crothers, D. M. (1986). DNA bending at adenine. Thymine tracts. Nature 320, 501–506. Lafontaine, D. A., Norman, D. G., and Lilley, D. M. (2002). The global structure of the VS ribozyme. EMBO J. 21, 2461–2471. Lilley, D. M. (2008). Analysis of branched nucleic acid structure using comparative gel electrophoresis. Q. Rev. Biophys. 41, 1–39. Locke, B. R., and Trinh, S. H. (1999). When can the Ogston-Morris-Rodbard-Chrambach model be applied to gel electrophoresis? Electrophoresis 20, 3331–3334. Lumpkin, O. J. (1982). Mobility of DNA in gel electrophoresis. Biopolymers 21, 2315–2316. Mohanty, U., and McLaughlin, L. (2001). On the characteristics of migration of oligomeric DNA in polyacrylamide gels and in free solution. Annu. Rev. Phys. Chem. 52, 93–106. Moghaddam, S., Caliskan, G., Chauhan, S., Hyeon, C., Briber, R. M., Thirumalai, D., and Woodson, S. A. (2009). Metal ion dependence of cooperative collapse transitions in RNA. J. Mol. Biol. 393, 753–764. Ogston, A. G. (1958). The spaces in a uniform random suspension of fibres. Trans. Faraday Soc. 54, 1754–1757. Olson, W. K., Marky, N. L., Jernigan, R. L., and Zhurkin, V. B. (1993). Influence of fluctuations on DNA curvature. A comparison of flexible and static wedge models of intrinsically bent DNA. J. Mol. Biol. 232, 530–554. Ortoleva-Donnelly, L., Szewczak, A. A., Gutell, R. R., and Strobel, S. A. (1998). The chemical basis of adenosine conservation throughout the Tetrahymena ribozyme. RNA 4, 498–519. Pan, J., and Woodson, S. A. (1998). Folding intermediates of a self-splicing RNA: Mispairing of the catalytic core. J. Mol. Biol. 280, 597–609. Pan, J., Thirumalai, D., and Woodson, S. A. (1999). Magnesium-dependent folding of selfsplicing RNA: Exploring the link between cooperativity, thermodynamics, and kinetics. Proc. Natl. Acad. Sci. USA 96, 6149–6154. Pan, J., Deras, M. L., and Woodson, S. A. (2000). Fast folding of a ribozyme by stabilizing core interactions: Evidence for multiple folding pathways in RNA. J. Mol. Biol. 296, 133–144. Pinard, R., Hampel, K. J., Heckman, J. E., Lambert, D., Chan, P. A., Major, F., and Burke, J. M. (2001). Functional involvement of G8 in the hairpin ribozyme cleavage mechanism. EMBO J. 20, 6434–6442. Pyle, A. M., McSwiggen, J. A., and Cech, T. R. (1990). Direct measurement of oligonucleotide substrate binding to wild-type and mutant ribozymes from Tetrahymena. Proc. Natl. Acad. Sci. USA 87, 8187–8191. Rangan, P., Masquida, B., Westhof, E., and Woodson, S. A. (2003). Assembly of core helices and rapid tertiary folding of a small bacterial group I ribozyme. Proc. Natl. Acad. Sci. USA 100, 1574–1579. Rangan, P., Masquida, B., Westhof, E., and Woodson, S. A. (2004). Architecture and folding mechanism of the Azoarcus Group I Pre-tRNA. J. Mol. Biol. 339, 41–51. Ryder, S. P., Recht, M. I., and Williamson, J. R. (2008). Quantitative analysis of protein–RNA interactions by gel mobility shift. Methods Mol. Biol. 488, 99–115. Sartori, A., Barbier, V., and Viovy, J. L. (2003). Sieving mechanisms in polymeric matrices. Electrophoresis 24, 421–440. Schlax, P. J., Xavier, K. A., Gluick, T. C., and Draper, D. E. (2001). Translational repression of the Escherichia coli alpha operon mRNA: Importance of an mRNA conformational switch and a ternary entrapment complex. J. Biol. Chem. 276, 38494–38501. Severcan, I., Geary, C., Verzemnieks, E., Chworos, A., and Jaeger, L. (2009). Square-shaped RNA particles from different RNA folds. Nano Lett. 9, 1270–1277.
208
Sarah A. Woodson and Eda Koculi
Stahl, D. A., Walker, T. A., Meyhack, B., and Pace, N. R. (1979). Precursor-specific nucleotide sequences can govern RNA folding. Cell 18, 1133–1143. Stellwagen, E., Lu, Y., and Stellwagen, N. C. (2003). Unified description of electrophoresis and diffusion for DNA and other polyions. Biochemistry 42, 11745–11750. Streicher, B., Westhof, E., and Schroeder, R. (1996). The environment of two metal ions surrounding the splice site of a group I intron. EMBO J. 15, 2556–2564. Szewczak, A. A., and Cech, T. R. (1997). An RNA internal loop acts as a hinge to facilitate ribozyme folding and catalysis. RNA 3, 838–849. Tuerk, C., and Gold, L. (1990). Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase. Science 249, 505–510. Woodson, S. A. (2001). Probing RNA folding pathways by RNA fingerprinting. Curr. Protoc. Nucleic Acid. Chem. 14, Chapter 11: Unit 11.4. Wu, H. M., and Crothers, D. M. (1984). The locus of sequence-directed and proteininduced DNA bending. Nature 308, 509–513. Yuan, C., Rhoades, E., Heuer, D. M., Saha, S., Lou, X. W., and Archer, L. A. (2006). Comprehensive interpretation of gel electrophoresis data. Anal. Chem. 78, 6179–6186. Zaug, A. J., Grosshans, C. A., and Cech, T. R. (1988). Sequence-specific endoribonuclease activity of the Tetrahymena ribozyme: enhanced cleavage of certain oligonucleotide substrates that form mismatched ribozyme-substrate complexes. Biochemistry 27, 8924–8931. Zimm, B. H., and Levene, S. D. (1992). Problems and prospects in the theory of gel electrophoresis of DNA. Q. Rev. Biophys. 25, 171–204. Zinkel, S. S., and Crothers, D. M. (1990). Comparative gel electrophoresis measurement of the DNA bend angle induced by the catabolite activator protein. Biopolymers 29, 29–38.
C H A P T E R
T E N
Using Analytical Ultracentrifugation (AUC) to Measure Global Conformational Changes Accompanying Equilibrium Tertiary Folding of RNA Molecules Somdeb Mitra Contents 210 211 216 216 217 217 218 218 218 219 220 222 222 224 224 225 226 230 230 231 232 232
1. Introduction 2. Theoretical Background 3. Materials and Instrumentation 3.1. Reagents 3.2. Instrument setup 3.3. Analytical cell assembly 4. Performing an RNA Folding Experiment 4.1. Experimental premise 4.2. Checking for an active self-assembly reaction 4.3. Choosing folding conditions 4.4. Sample preparation 4.5. Sample loading and rotor assembly 4.6. Data acquisition 5. Data Analysis 5.1. Premise of the data analysis 5.2. Information content of the g(s*) distributions 5.3. Steps in data analysis 6. Case Studies 6.1. tRNA 6.2. Group I intron ribozyme 6.3. IRES RNA 6.4. M-box riboswitch
Department of Biochemistry, Albert Einstein College of Medicine, Bronx, New York, USA Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69010-8
#
2009 Elsevier Inc. All rights reserved.
209
210
Somdeb Mitra
7. Conclusions Acknowledgments References
232 233 233
Abstract Analytical ultracentrifugation (AUC) is a powerful technique to determine the global conformational changes in RNA molecules mediated by cations or small molecule ligands. Although most of the developments in the field of AUC have been centered on studies involving protein molecules, the experimental methods as well as the analytical approaches have been successfully adapted and applied to the study of a variety of RNA molecules ranging from small riboswitches to large ribozymes. Most often AUC studies are performed in conjunction with other structural probing techniques that provide complementary information on local changes in the solvent accessibilities at specific regions within RNA molecules. This chapter provides a brief theoretical background, working knowledge of instrumentation, practical considerations for experimental setup, and guidelines for data analysis procedures to enable the design, execution, and interpretation of sedimentation velocity experiments that detect changes in the global dimensions of an RNA molecule during its equilibrium folding.
1. Introduction The transition from a newly synthesized elongated chain-like state to an ordered functional conformation for almost any biological polymer is usually accompanied by a decrease in its hydrodynamic radius. This phenomenon is typically referred to as ‘‘compaction.’’ For RNA molecules this global structural collapse from the elongated unfolded state to their native fold usually occurs in two distinct phases. The first phase results from base pairing among nucleotides to generate an ensemble of independently stable entities possessing the native or native-like secondary structure (Brion and Westhof, 1997). The second phase is due to both local and long-range tertiary contact formation (Thirumalai et al., 2001). The collapse accompanying secondary structure formation generally occurs on the ‘‘microsecond’’ time scale (Cole and Crothers, 1972; Cole et al., 1972). This transition is difficult to detect with a conventional mixing apparatus. The slower second phase spanning the ‘‘millisecond’’ to ‘‘second’’ temporal regime has been a subject of intense investigation (Das et al., 2003; Kwok et al., 2006; Russell et al., 2002a,b). Information acquired on the changes in global dimension, when melded with information on local structural changes at single nucleotide spatial resolution, aid in the generation of significantly detailed structural models of both equilibrium
Studying RNA Folding by Analytical Ultracentrifugation
211
and kinetic folding intermediates (Baird et al., 2005; Jonikas et al., 2009; Laederach et al., 2006, 2007). Such structural models, in turn, provide deep insights into the physical principles governing the stability of RNA structures as well as their conformational dynamics (Laederach et al., 2007; Shcherbakova et al., 2008). The techniques that have been used to study global conformational changes of RNA molecules include analytical ultracentrifugation (AUC) (Costantino and Kieft, 2005; Dann et al., 2007; Takamoto et al., 2002), small-angle X-ray scattering (SAXS) (Baird et al., 2005; Russell et al., 2000; Takamoto et al., 2004), native gel electrophoresis (Buchmueller et al., 2000; Koculi et al., 2006, 2007; Lilley et al., 1992; Perez-Salas et al., 2004), size exclusion chromatography (Dann et al., 2007), and fluorescence resonance energy transfer (FRET) (Bartley et al., 2003; Bokinsky et al., 2003; Hohng et al., 2004; Kim et al., 2002; Pljevaljcic´ et al., 2005; Russell et al., 2002a,b; Sattin et al., 2008; Silverman and Cech, 1999, 2001). Among these techniques, AUC deserves a special mention because it is a very sensitive method that requires relatively smaller amounts of sample and reports precisely, based on fundamental hydrodynamic principles, the size and shape of RNA molecules in solution. Since this technique rests on optical absorption properties of RNA molecules in the UV range (260 nm), it evades the necessity of labeling the RNA with a separate radioactive or fluorescent reporter. AUC has been used to determine equilibrium folding transitions as well as ligand-induced equilibrium conformational changes of RNA molecules ranging from large ribozymes to small riboswitches (Costantino and Kieft, 2005; Dann et al., 2007; Takamoto et al., 2002). In this chapter, we design AUC-based experiments for measuring the changes in the hydrodynamic radii during cation-induced tertiary folding of RNA molecules under equilibrium conditions.
2. Theoretical Background We first summarize the theoretical background that should allow a researcher to design, conduct, and interpret the results of an AUC experiment aimed at studying equilibrium RNA folding. For further details regarding the physical principles of mass transport involved in the AUC technique, we recommend the readers to consult elegant and elaborate discourses on the topic including the following book chapters (Cantor and Schimmel, 1980; Cole et al., 2008; Fujita, 1975; Ignacio Tinoco et al., 2003; Scott and Schuck, 2005; Tanford, 1961). AUC experiments can be conducted in two modes: sedimentation velocity (SV) (Cole et al., 2008; Howlett et al., 2006; Laue and Stafford, 1999; Lebowitz et al., 2002) and sedimentation equilibrium (SE)
212
Somdeb Mitra
(Howlett et al., 2006; Laue, 1995). Each technique utilizes different aspects of the principles of mass transport under two opposing physical forces, sedimentation and diffusion. When a homogenous solution of macromolecules contained in a cell (a closed compartment) is subjected to a centrifugal force, the solute particles tend to sediment toward the bottom of the cell, establishing a solvent–solution boundary. Diffusion forces tend to redistribute the concentration gradient formed at boundaries. In SV experiments, the samples are spun at high angular velocities so that sedimentation overcomes diffusion and the bulk solute is transported toward the bottom of the cell. By monitoring the rate of boundary movement and hence the rate of sedimentation, one can calculate the size, shape, and conformational homogeneity of the macromolecular solute under a given set of solution conditions. In SE experiments, the solution is subjected to smaller angular velocities so that the molecular flux due to diffusion forces is balanced by the flux due to sedimentation, resulting in the formation of a smooth concentration gradient along the length of the cell. By monitoring this equilibrium distribution of concentrations, one can precisely calculate the stoichiometries of interactions and association constants as well as the average molecular weights of the solute species. We shall limit our discussion exclusively to SV experiments which provide all the necessary information about the global conformational changes during RNA folding. In an SV experiment, the sedimentation and the diffusion forces that determine the net rate of movement of the solvent–solution boundary, stem from two intrinsic properties of the solute molecules, their sedimentation coefficient (s) and their diffusion coefficient (D). Whereas D depends predominantly on the shape of the solute particles, s depends both on its shape and on its mass. The diffusion coefficient, D, is defined as the ratio of the flux of molecules ( Jx, moving in the direction x under diffusive forces) to the concentration gradient of the molecules ð@c=@xÞ. The dependence of D on the molecular shape stems from its relation to the frictional coefficient f: D¼
kT ; f
ð10:1Þ
where k is Boltzmann constant and T is the absolute temperature. If the solute particles are nonhydrated and have a perfect spherical shape with a radius r, f determines the forces that the solute particles feel (in a direction opposite to their motion) when they move in a solution of viscosity because f ¼ 6pr. However, macromolecular solutes are seldom perfectly spherical; their shapes are more often best approximated as ellipsoids of revolution, for example, prolate (elongated spindle-shaped) and oblate (flattened disc-shaped) ellipsoids. The shape factor
Studying RNA Folding by Analytical Ultracentrifugation
F¼
f fo
213
ð10:2Þ
is defined as the ratio of the observed f to that calculated for a sphere of the same volume ( fo). The shape factor provides a clue to the shape that the macromolecule assumes in solution under a given set of conditions. Macromolecules are almost always solvated thereby increasing their hydrodynamic volume. Thus, calculation of absolute hydrodynamic parameters rather than relative changes requires correction for solvation. The sedimentation coefficient, s, depends both on the shape and the effective buoyant mass (Mb) of the solute particle vrÞg Mb ¼ mð1
ð10:3Þ
where m and v stand for the mass and the partial specific volume of the solute particle, respectively, r is the solvent density, and g is the acceleration due to gravity. When a solution contained in a sample cell is spun in a centrifuge rotor at an angular velocity, o, the linear velocity (u) of a solute particle of mass m located at a distance x from the center of the rotor is given by: u¼
mð1 vrÞg 2 ox f
ð10:4Þ
Combined with the definition of buoyant mass (Eq. (10.3)), this equation can be written as s¼
u Mb ¼ f o2 x
ð10:5Þ
This ratio of the linear velocity of a particle to its angular acceleration, in a centrifugal field, is the sedimentation coefficient (s). As evident from the equation, the term Mb results in the dependence of s on the mass, whereas the term f results in its dependence on the size and shape of the macromolecule. The goal of a SV experiment is to monitor the rate of movement of the solvent–solution boundary (or, the evolution of the radial concentration gradient @c=@x as a function of time), described by the Lamm equation which takes into account the combined effects of sedimentation and diffusion, as follows: 2 @c @ c 1 @c @c 2 þ ¼D so x þ 2c ð10:6Þ @t @x2 x @x @x The instrument routinely used in our lab for AUC experiments is the Beckman XL-I analytical ultracentrifuge. We exploit the UV absorption property of RNA molecules at 260 nm to employ the absorbance optics built in the centrifuge. The optical system of a Beckman analytical ultracentrifuge is shown schematically in Fig. 10.1A. The instrument is
B
Sample meniscus Sample solution
A Reference
Top view
Toroidal diffraction grating
Sample
Incident light detector
Solvent meniscus Reference (blank)
Reflector
C
Sample meniscus 1.00
Sample/reference cell assembly TVS Rotor
Boundary Absorbance (O.D.)
Imaging system for radial scanning
Slit (2 nm) Aperture Xenon flash lamp
Plateau
0.75 0.50
Baseline
0.25 0.00 −0.25
Photomultiplier tube
−0.50 6.1 6.2 6.3 Solvent meniscus
Figure 10.1 (Continued )
6.4
6.5 6.6 6.7 Radius (cm)
6.8
6.9
7.0
7.1
Studying RNA Folding by Analytical Ultracentrifugation
215
designed such that a light beam is directed, through a monochromator, to the cells (sample holders) in the rotor. The solutions are contained within sector-shaped cavities in the centerpiece of the cells (Fig. 10.1B), tightly fit between two transparent quartz windows, housed within a cylindrical metal compartment. The UV beam at 260 nm enters through one quartz window, passes through the solution, exits through the quartz window at the opposite end and is collected by a detector. As the absorbance of the solute RNA molecules differ from that of the solvent, the solvent–solute boundary is clearly detected in the absorbance scans (Fig. 10.1C). Successive scans at regular time intervals, during centrifugation, depict the migration of the boundary toward the bottom of the cell and provide a direct measurement of the radial distribution of the concentration gradient as a function of time [c(x, t)] (Fig. 10.1C). As described in further details in Section 5, we analyze the scans using the software DCDTþ (Philo, 2006), which converts the raw concentration profiles into time derivatives ð@c=@tÞ and fits these values to approximate unbounded solutions of the Lamm equation (Philo, 2000; Stafford, 1994). As the rotor speed (o) and the concentration of the macromolecules (c) are known, and the time (t) and the radial concentration distribution [c(x, t)] are obtained from the scans of absorbance profiles, the fitting yields values of s and D. As both parameters are dependent on the solvent viscosity and temperature, they are transformed to standard values with reference to a standard temperature (20 C) and a standard solvent (water) and reported as s20,w and D20,w. This standardization allows analysis of the changes in the intrinsic properties of solute molecules with changes in solution condition and is a prerequisite in cationmediated folding studies of RNA molecules. The s and the D values, due to their dependence on f, reflect the shape of an RNA molecule. For example, an extended unfolded RNA structure formed in the absence of divalent cation sediments and diffuses slowly, producing smaller s and D values. Compact-folded RNA structure formed Figure 10.1 (A) Overview of the optical system and rotor arrangement in the Beckman Optima analytical ultracentrifuge. One cell holder is zoomed in, to show the positioning of the cells within the rotor. Courtesy of Beckman Coulter, Inc. [Adapted from Greg Ralston; Introduction to Analytical Ultracentrifugation; http://www.beckman. com/literature/Bioresearch/361847.pdf.] (B) Schematic diagram of a centerpiece inside an assembled cell. The height of the solvent column is slightly greater than that of the sample column, due to slight excess volume of the solvent inside, which helps in separation of the solvent and the sample menisci in the optical scans. (C) Examples of optical scans obtained from an AUC run, showing progressive movement of the boundary in a radial direction toward the bottom of the cell. Eleven scans are shown, spaced at regular time intervals. The scans are recorded as absorbance (OD) versus radius (cm). The first set of (downwards) spikes represent the air–solvent meniscus while the second set of (upwards) spikes represent the air–sample solution meniscus.
216
Somdeb Mitra
in the presence of divalent cations would show the opposite trend. We calculate the Stoke’s radius (RH) of the RNA molecule from D using the program SEDNTERP (http://www.jphilo.mailway.com/download. htm), wherein the hydration of the molecule is assumed to be 0.59 and the partial specific volume ( v) to be 0.53 cm3/g. The RH value, assuming that the RNA molecule adopts a nearly spherical geometry, is directly proportional to its degree of compaction under a given set of solution conditions. As mentioned before, if the RNA molecules tend to adopt nonspherical global shapes in solution, the shape factor (Eq. (10.2)) calculated using SEDNTERP indicates the extent of divergence from a spherical shape. The shape factor has been shown to decrease as the molecules fold, indicating that the folded state better resembles a sphere than the unfolded state (Costantino and Kieft, 2005). Alternatively, since molecules often adopt ellipsoidal geometries both in the unfolded and the folded states, their shapes may be more accurately represented by the axial ratio (a/b) parameter (estimated from f using SEDNTERP), where a and b are the major and minor axes of the ellipsoids, respectively. However, in AUC studies on cation-mediated RNA conformational changes, both the RH and the axial ratio (a/b) have been found to decrease in an identical manner, as the RNA molecules become compactly folded with increasing concentration of divalent or monovalent ions (Takamoto et al., 2002). Therefore, SV experiments report, in multiple ways, on the equilibrium changes in the global dimensions of RNA molecules as they fold from the ensemble of unfolded states to the native state.
3. Materials and Instrumentation 3.1. Reagents Our folding experiments are typically performed in CE buffer (pH 7.3) which contains 10 mM sodium cacodylate and 0.1 mM EDTA. When titrations are performed with divalent cations, potassium chloride (KCl) is added to a final concentration of 100 mM and the buffer is referred to as Cacodylate-EDTA-K (CEK) buffer. Due to high susceptibility of RNA molecules to ribonuclease (RNase) degradation, special care should be taken to avoid nuclease contaminations. We purchase high quality RNase free reagents from Ambion (DEPC-treated water and stock solutions of 2 M KCl, 1 M MgCl2, 0.5 M EDTA). Ultra pure (98%) sodium cacodylate is purchased in its trihydrated form (C2H6AsNaO23H2O; mol. wt. 214.03 g/mol) from Sigma. The RNA under investigation is in vitro transcribed by run-off transcription using the MEGAscriptÒ T7 Kit (Ambion), usually from
Studying RNA Folding by Analytical Ultracentrifugation
217
PCR-generated linear DNA template. The template contains a T7 promoter sequence immediately upstream to the RNA sequence, the first nucleotide after the T7 promoter (þ1) is changed to G to improve transcription yield and the last two nucleotides of the template are methylated at 20 -O position to reduce 30 -end heterogeneity of the transcribed RNA. The transcript is purified either on a 5% denaturing PAGE or by the MEGAclearTM kit (Ambion) and dissolved in CE buffer at stock concentrations of 10–50 mM.
3.2. Instrument setup The Beckman XL-I analytical ultracentrifuge is run with the An-60 Ti rotor that can hold up to three sample cells. Before starting an experiment, the rotor and the monochromator are placed inside the centrifuge chamber and allowed to equilibrate at the desired temperature (25 C for the experiments we describe herein) under vacuum inside the chamber. The xenon lamp produces wavelengths between 190 and 800 nm; a monochromator allows selection of a given wavelength. A wavelength scan should be performed before initiating an experiment to check if the intensity of light is sufficient. If not, the lamp should be cleaned as described in the Beckman XL-x manual. We perform scans at 260 nm for RNA folding experiments. An absorbance spectrophotometer collects the light after it passes through the solution. The concentration of the RNA samples should be between 0.1 and 1 O.D. (given the optical path length of 1.2 cm, for an RNA molecule of about 250 nucleotides with mol. wt. of approximately 75 kDa, this absorbance range spans a concentration range of approximately 50–500 nM.
3.3. Analytical cell assembly We maintain a set of AUC cells dedicated to RNA experiments. The cells are regularly cleaned with RNaseZap (Ambion) which removes RNase contaminations that might occur during handling of the cells in shared AUC facilities. We use aluminum-epon/charcoal centerpieces with quartz windows. As shown in Fig. 10.1B, each centerpiece has two sector-shaped cavities with a 1.2 cm optical path length, one for the sample solution and one for the reference solution (blank). The cells are assembled according to the instructions provided in the Beckman manual. It is important to note that after assembling the cells, the screw ring that closes them should be tightened to 120 psi using a torque wrench to prevent leakage. With the screw rings facing the user, the sample is loaded into the right sector and the blank solution is loaded into the left sector of the centerpiece, through the hole at the top. The reference solution is always prepared in slight excess amounts compared to the sample solution
218
Somdeb Mitra
(350 vs. 300 ml in our experiments) so that after loading the solution–air interfaces (meniscus) of the two solutions are at slightly different heights (Fig. 10.1B). This helps in distinguishing the sample meniscus from the solvent meniscus (Fig. 10.1C).
4. Performing an RNA Folding Experiment In this section, we indicate the general factors that an experimentalist should consider while setting up a cation-mediated equilibrium RNA folding experiment, provide a stepwise protocol for sample preparation and outline the data collection procedure using the absorption optics of the analytical ultracentrifuge.
4.1. Experimental premise The premise of a cation-mediated equilibrium RNA folding experiment is that, in a low ionic strength solution (corresponding to about 10 mM monovalent ions), an RNA sample exists as an ensemble of molecules with stable secondary structure but devoid of tertiary structure (Doty et al., 1959; Draper et al., 2005; Freier et al., 1986; Fresco et al., 1960; Manning, 1978). Due to the negative charges on the phosphodiester backbones, the base-paired secondary structure modules repel each other (Bai et al., 2005, 2007; Murthy and Rose, 2000; Tan and Chen, 2006), resulting in an extended global conformation (detected by AUC as a sample with small s20,w and D20,w values and high shape factor value). Addition of positively charged counterions predominantly exerts two effects. First, the screening of the electrostatic repulsion generates a relaxed ensemble in which the secondary structure modules can approach each other (Das et al., 2003; Takamoto et al., 2004), thereby facilitating tertiary contact formation. Second, site-specific binding of ions stabilizes the tertiary structure by neutralizing pockets of high negative charge densities. Both of these effects lead to a compact, folded native state (Draper, 2004; Draper et al., 2005; Grilley et al., 2006; Misra et al., 2003) that is detected by AUC as a sample with high s20,w and D20,w values and low shape factor value.
4.2. Checking for an active self-assembly reaction Very low RNA concentrations (<0.1 O.D.) produce low signal to noise while at high RNA concentrations (1 O.D.) the detector saturates and does not yield accurate data. At higher concentrations, RNA molecules
Studying RNA Folding by Analytical Ultracentrifugation
219
may aggregate and, therefore, deviate from ideal nonassociating solution behavior that is essential to a reliable hydrodynamic analysis. In order to check for self-assembly and determine an appropriate RNA concentration for a hydrodynamic analysis, we first analyze a series of dilutions of the RNA sample, ranging between 0.08 and 1.2 O.D. Two concentration series are determined, the first in CEK buffer (10 mM potassium cacodylate, 0.1 mM EDTA at pH 7.3, 100 mM KCl) and the second in CEK buffer with 10 mM Mg2þ. These series correspond to the unfolded and folded states of the RNA molecule being analyzed, respectively. These absorption data are collected and analyzed to determine the standard sedimentation coefficient (s20,w) of the samples. The s20,w values of both the unfolded and the folded samples are plotted as a function of RNA concentration (in absorption units). If s20,w is constant or decreases linearly with concentration, the sample is behaving as monodisperse particles. In such cases, linear fits to the data are extrapolated to zero RNA concentration to determine the s∘20;w , which indicates the sample’s hydrodynamic behavior at infinite dilution. A significant downward slope, usually observed in the absence of Mg2þ, indicates nonideal behavior at high concentrations due to electrostatic repulsion between negatively charged RNA backbones (Costantino and Kieft, 2005). An upward slope reflects RNA aggregation with increasing concentration. We have more frequently observed this phenomenon in samples containing higher RNA concentrations in the presence of Mg2þ. Samples that aggregate may show multiple components in the sedimentation boundaries. The presence of aggregation precludes the conduct of a hydrodynamic analysis such as described in this chapter.
4.3. Choosing folding conditions We generally begin our studies on equilibrium folding of an RNA molecule by characterizing the changes in its global dimensions as a function of increasing concentrations of either a monovalent ion like potassium (Kþ) or a divalent ion like magnesium (Mg2þ). The choice of Kþ is based on its physiological relevance (Cayley et al., 1991) as well as due to the observation of specific Kþ binding sites in crystal structures of RNA molecules (Basu et al., 1998; Conn et al., 2002; Stahley et al., 2007). Site-specific binding of Mg2þ has been observed in numerous RNA crystal structures (Adams et al., 2004; Cate et al., 1996, 1997; Draper et al., 2005; Golden et al., 2005; Guo et al., 2004; Klein et al., 2004) and thermodynamic stabilization of RNA tertiary structures by Mg2þ ions is a well-studied phenomenon in the RNA folding literature (Draper et al., 2005; Grilley et al., 2006, 2007; Misra and Draper, 2000; Sattin et al., 2008; Soto et al., 2007).
220
Somdeb Mitra
Sodium (Naþ) is often used as the monovalent ion of choice to exclusively elucidate the effect of electrostatic charge neutralization on RNA folding, by avoiding the thermodynamic effects of site-specific ion binding (Takamoto et al., 2002, 2004). It has been observed that when divalent ion-mediated RNA folding experiments are initiated with low monovalent ion concentrations in the buffer (10 mM Naþ or Kþ) misfolded species are often populated, most likely due to nonnative docking of secondary structure modules in a rigid ensemble (Sclavi et al., 1998). We therefore choose to have 100 mM monovalent ions in our buffers, during Mg2þ-mediated folding experiments, to generate an electrostatically relaxed environment with greater conformational flexibility (Kwok et al., 2006; Laederach et al., 2007). It is recommended to carry out a set of SV measurements of the unfolded and the folded state ensembles as a function of RNA concentration as described earlier before collecting the data points for the entire equilibrium transition. The shape of the apparent sedimentation coefficient distribution curves (discussed in Section 5) derived from such measurements, indicate whether the initial and the final states are conformationally homogeneous. Appearance of more than one peak in such distributions would indicate the existence of distinct structural states with different hydrodynamic properties.
4.4. Sample preparation 1. Dilute the RNA solution (10–50 mM) in ‘‘CEK buffer’’ (10 mM potassium cacodylate, 0.1 mM EDTA at pH 7.3, 100 mM KCl) for Mg2þ titration or in 1 CE buffer (for monovalent ion titration). We consistently get good results with RNA concentrations corresponding to an absorbance of 0.5–0.6 O.D. Prepare RNA in appropriate amounts for up to three folding data points for each centrifuge run. 2. Denature the RNA solution by heating at 95 C for 2 min. Cool the samples to room temperature for 15 min and give a quick spin to bring down the condensates on the lids of the tubes back into solution. 3. Prepare the titration reaction mix in each reaction tube as shown in the following tables. Add all the components except the main counterion that would mediate folding (Mg2þ for Mg titration and Kþ for K titrations). For each reaction mix, prepare a reference solution of slightly excess volume but with identical reagent concentrations, for ‘‘blank’’ absorbance measurements. [Note: For Kþ titrations, prepare two tubes with 10 mM MgCl2 for comparison of the final states of the RNA in highest Kþ concentration and in 10 mM Mg2þ.]
221
Studying RNA Folding by Analytical Ultracentrifugation
Mg2+ titrations For RNA samples RNA CE Mg2þ conc. (10–50 mM ) buffer (10) X mM
Y ml for (0.5 O.D.)
30 ml
KCl (2 M )
MgCl2 (1 M )
H2O
15 ml
0.3X ml
(300 (Y þ 300 ml 45 þ 0.3X)) ml
Total volume
For blanks RNA Mg2þ conc. (60 mM) (ml)
CE buffer (10) (ml)
KCl MgCl2 (2 M) (ml) (ml)
H2O
Total volume
X mM
35 ml
17.5 ml
(350 – 52.5 þ 0.35X) ml
350 ml
0
0.35X ml
Kþ titrations For RNA samples Kþ conc.
RNA (5–10 mM) (ml)
CE buffer KCl (2 M) H2O (10) (ml)
Total volume
X mM
Y ml for (0.5 O.D.)
30 ml
(300 (Y þ 30 þ 0.15X) ml
300 ml
0.15X ml
For blanks Kþ conc.
RNA (60 mM) (ml)
CE buffer KCl (ml) (10) (ml)
H2O
Total volume
X mM
0
35 ml
(350 (35 þ 0.175X) ml
350 ml
0.175 ml
4. Aliquot appropriate volume of RNA to each reaction mix (but not to the corresponding reference solutions) and incubate both the reaction mix and the reference solutions at 50 C for 5 min. 5. Add appropriate concentrations of the folding cation (Mg2þ or Kþ) and continue incubation at 50 C for 30 min. 6. Take the tubes out and place them at 25 C (folding temperature) for 1 h.
222
Somdeb Mitra
4.5. Sample loading and rotor assembly 1. Load each reaction mix and its corresponding blank solution into the appropriate cavities of the assembled centerpieces as discussed above. Label each cell on its outer surface with a marker and record which sample has been loaded into which cell; it is a common error to mix them up when loading the rotor. 2. Seal the holes of the centerpieces with the small circular red gaskets and close them by gently tightening the brass screws through the threading above the holes. Blow off any dust particles sticking to the outer walls of the quartz windows with compressed air. 3. Weigh the cells carefully to make sure that they are balanced (the allowed tolerance is 0.5 g). Set aside the cell that differs the most in terms of its weight to balance it against the reference cell of the rotor. The weight of the reference cell is adjusted by screwing in the standard weights supplied by Beckman. 4. Take the temperature equilibrated rotor out of the centrifuge chamber. Place the reference cell at position 4 and the cell balanced against it at position 2. Place the other two balanced cells opposite to each other. Align the markings at the bottom edges of the cells with the markings on the peripheries of the holes at the bottom of the rotor. Follow the Beckman XL-I instruction manual for correct placing of cells in the rotor. The side of the cells with the screw rings must face upwards with the fill holes of the centerpieces facing the center of the rotor. 5. Place the loaded rotor into the centrifuge chamber, attach the monochromator to its holder and close the sliding door. Start the vacuum pump and wait for the pressure to come down to 100 mm.
4.6. Data acquisition 1. On the screen of the computer attached to the instrument, open the ‘‘Beckman XL-x’’ software that controls the centrifuge run and data collection. Wait until the system initializes. 2. In the File menu, select and open New File. This brings up the window where the run parameters can be specified. Select the 4 hole in the rotor option. Set Speed to 3000 rpm and Temperature to the desired experimental temperature (25 C in our experiments). 3. There are four boxes, one for each AUC cell. The fourth one is for the reference cell. You can either fill up the boxes individually or check the box ‘‘All settings identical to cell 1.’’ However, if you do check this option, do not forget to change the name of the samples in the ‘‘Comment’’ box. 4. In the ‘‘Cell #’’ box, from the three scanning options on the right side, select Velocity. From the options for choosing the optics, select Absorbance. The Rmin and the Rmax options allow the user to choose the desired radial
Studying RNA Folding by Analytical Ultracentrifugation
223
distance range to be scanned. We generally start by collecting scans between 5.9 and 7.1 cm; this range may be reduced according to the number and quality of the scans at the beginning and at the end. Make sure you change the scanning wavelength in the W1 box to 260 nm. In the ‘‘Comment’’ box assign a name that would allow you to recognize the sample contained in the cell, later during the data analysis stage. 5. Click and open the Options window to make sure all the boxes are unchecked. Do not specify the Method at this stage. Make sure that the pressure inside the centrifuge chamber has come down to at least 30–40 mm and then start the rotor by clicking on ‘‘Start single scan.’’ This brings up the windows that display the absorption patterns of the solutions in each cell in real time. 6. The rotor speed goes up to 3000 rpm before starting the absorbance scan. Collect a set of scans at 3000 rpm (by clicking on the ‘‘Start single scan’’ option each time) to check if the samples are leaking from the cells. A leak is usually indicated by a systematic drop in the intensity of absorption and by an increase in the pressure inside the centrifuge chamber. 7. If no leakage is detected, change the rotor speed by typing the desired value in the Speed box and click ‘‘Start single scan.’’ This brings up the rotor to the desired speed before collecting a set of absorbance scans. Let the rotor spin at the desired velocity for atleast 5–10 min before initializing actual data acquisition. [Note: The choice of rotor speed in a SV experiment depends inversely on the size of the RNA molecule being studied. The larger the molecule is, the faster it sediments. The velocity should be chosen appropriately so as to allow collection of at least 50–60 absorbance scans before the solvent– solution boundary hits the base of the cell. For a molecule of about 400 nucleotides (mol. wt.130 kDa), we choose velocities around 25,000 rpm (30,000 g), whereas for a molecule of about 250 nucleotides (mol. wt. 75 kDa), we choose velocities around 30,000 rpm ( 85,000 g). When initiating folding studies on a new RNA molecule, it is desirable to run the samples at different velocities to find the appropriate range. However, once a velocity is decided upon, all the experiments with a given RNA molecule should be performed at the same velocity to allow comparison between data sets.] 8. If the initial scans at the desired velocity show no signs of leakage, click to open the Method window. Specify the number of scans (usually no more than 60) and close the box. Open the Options window and check the ‘‘No delay calibration’’ option. If you want the rotor to stop automatically after data collection, check the ‘‘Stop XL after last scan’’ option. Save the file by clicking on ‘‘Save file as’’ from the drop down list in the File section.
224
Somdeb Mitra
9. Start experimental data acquisition by clicking on ‘‘Start method scan.’’ About 2–3 h is required to complete a set of 60 absorbance scans. 10. Once data collection is over, stop the centrifuge, wait until the rotor stops, release the vacuum pump and wait for the pressure to come up that would allow opening of the chamber door. Remove the monochromator, take out the rotor and remove the cells from the rotor. Remove the samples from the cells and disassemble them as described in the Beckman XL instruction manual. Wash the windows and centerpieces with RNaseZap solution and deionized water before reassembling them for the next set of runs. [Note: Sometimes it might be hard to push the assembly out of the cell. Do not apply too much force which can damage the windows and the centerpiece, but instead, place the cell inside a refrigerator at 4 C for about 15 min. Take out the cells and gently push out the assembly; it should come out easily.]
5. Data Analysis Several softwares are available for analysis of SV data, like SVEDBERG (Philo, 1997), SEDFIT & SEDPHAT (Schuck, 2000, 2004), and DCDTþ (Philo, 2000, 2006). We analyze the SV data using the DCDTþ software (Philo, 2000, 2006) based on the time-derivative ð@c=@tÞ analysis method (Stafford, 1992, 1994). The software, details about its release versions and operational instructions are made available at http://www.jphilo. mailway.com/dcdtþ.htm. We briefly summarize the theory behind the method and guidelines for data analysis in the context of an RNA folding experiment.
5.1. Premise of the data analysis The radial concentration scans obtained from the UV spectrophotometer of the analytical ultracentrifuge can be either converted to a radial derivative of the concentrations at a given instant of time ð@c=@rÞt or to the time derivative of the concentrations at fixed radial position ð@c=@tÞr (Stafford, 1992). The @c=@t method, as the name implies, uses the temporal derivative which results in elimination of time independent (random) sources of noise in the data, thereby greatly increasing the precision of sedimentation boundary analysis (Stafford, 1992). Numerically, this process is implemented by subtracting pairs of radial concentration scans obtained at uniformly and closely spaced time intervals ½ðc2 c1 Þ=ðt2 t1 Þ. The values are then plotted as a function of radius to obtain ð@c=@tÞr versus r curves (Stafford, 1994). It can be shown that the apparent sedimentation coefficient s*
Studying RNA Folding by Analytical Ultracentrifugation
225
(i.e., s calculated by completely ignoring the effects of diffusion) of a molecule is related to the radial position r by 1 r s∗ ¼ ð10:7Þ ln o2 t rm where rm is the fixed position of the sample meniscus. Therefore, the abscissa of the ð@c=@tÞr versus r plot can be converted to s*. Once this is done, a further improvement in signal-to-noise ratio is achieved by averaging the ½ðc2 c1 Þ=ðt2 t1 Þ values at a constant value of s*. The average ð@c=@tÞr values are then converted to the apparent sedimentation coefficient distribution function g(s∗), expressed as 2 @c 1 o2 t 2 r ∗ ð10:8Þ gðs Þ ¼ @t corr c0 lnðrm =rÞ rm where ð@c=@tÞcorr is the corrected value of ð@c=@tÞr taking into account contributions from the plateau region and c0 is the initial loading concentration (Stafford, 1997). In this way, the ð@c=@tÞr versus r plots are converted to gðs∗ Þ versus s∗ plots that form the basis of our analysis of the SV data of RNA folding experiments using the software DCDTþ.
5.2. Information content of the g(s*) distributions The gðs∗ Þ versus s∗ plots can be closely approximated by Gaussian distributions for noninteracting systems. The number of peaks observed in this distribution indicates the degree of homogeneity of the sample. For an ensemble of RNA molecules that possess a unique or almost identical global conformations, the gðs∗ Þ distribution would display a single peak centered around the mean weight averaged sedimentation coefficient, sw ð s∗p s∗ gðs∗ Þ ds∗ sw ¼ 0ð ð10:9Þ ∗ ∗ gðs Þ ds The maxima of the gðs∗ Þ versus s∗ curve, therefore, denotes the sedimentation coefficient of the molecule, s. The presence of multiple peaks could indicate either the existence of conformationally distinct ensembles or self-association of the RNA molecules or both. The gðs∗ Þ distribution function is calculated by completely ignoring the effect of diffusion; an assumption that has been shown to yield nearly accurate results for molecules with molar weight >50 kDa. However, the diffusion coefficients can still be obtained from the standard deviation (s)
226
Somdeb Mitra
of the fits of the gðs∗ Þ distribution to a Gaussian function, according to the following relation: ðsrm o2 tÞ2 ð10:10Þ 2t where the symbols have their usual meanings (Stafford, 1997). Once s and D have been determined from the Gaussian fits, the molecular weight can be estimated from the Svedberg equation: M ¼ ðs=DÞ½RT =ð1 vrÞ, where v is the specific volume of the RNA molecule and r is the solvent density. The concentration of the sample is obtained by integrating the area under the gðs∗ Þ distribution. D¼
5.3. Steps in data analysis 1. The Beckman XL software is installed in the computer controlling the instrument such that all the information associated with a set of scans from a run are stored in a folder designated by the date and time of the run. Export the folders containing the digitized information of the experimental absorption scans out of the controlling computer to a separate computer in which the latest version of DCDTþ (we use version 2.2.1) has been downloaded and installed. 2. We use SEDNTERP for estimating the values of several hydrodynamic parameters and, therefore, have the software downloaded and installed in the same computer used for analysis. Use SEDNTERP to estimate the density (r) and viscosity () of your experimental solution (buffer and ions) using the list of standard reagents and buffer components. Wherever needed, use 0.53 cm3/g as the partial specific volume (v) and 0.59 as the hydration of an RNA molecule. [Note: While analyzing the data of an RNA folding experiment in which folding was mediated by titrating in increasing concentrations of the folding cation, make sure to adjust the buffer density and viscosity for each cation concentration.] 3. The data from each AUC cell is individually opened and analyzed in DCDTþ. Open DCDTþ, browse to the folder containing the data and load the scans from a cell. 4. Once the scans from a cell are loaded and displayed, specify the meniscus position and the beginning and end positions of data scans. At this stage the raw scans are displayed as absorbance versus radial position curves (Fig. 10.2A). Click Next to allow the software to calculate the dc/dt versus s∗ distribution curves. 5. Click Auto adjust to allow the software to automatically select a subset of good data scans that would generate a good Gaussian distribution
A
B 0.8
0.00022 0.00020 0.00018 0.00016 0.00014 0.00012
0.7
0.5
−dc/dt (OD/s)
Absorbance
0.6
0.4
0.00010 0.00008 0.00006
0.3 0.2 0.1
0.00004 0.00002
0.0 −0.1 −0.2 6.1
0.00000 −0.00002 6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
7.0
7.1
7.2
Radius (cm) s(w) = 7.127 +/− 0.04, c = 0.737 +/− 0.004
C
0
2
3
4
5
6
7
8
9
s* (20, w) (Svedberg)
10
11
12
13
14
15
D 0.20
0.20 0.15
g (s*) (OD/Svedberg)
g (s*) (OD/Svedberg)
1
0.10 0.05 0.00 −0.05
0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 −0.02 0.05 0.00
−0.10 0
1
2
3
4
5
6
7
8
9
s* (20, w) (Svedberg)
10
11
12
13
14
−0.05 0
1
2 3 Residuals
4
5
6
7
8
9
10
11
12
13
14
s* (20, w) (Svedberg)
Figure 10.2 Steps in the DCDTþ analysis of sedimentation velocity data collected on an RNA molecule of 250 nucleotides ( 75 kDa) in CEK buffer at 35,000 rpm. (A) Only 16 raw scans (out of the 50) are depicted for clarity. The red line indicates the sample meniscus, while the
228
6.
7.
8.
9.
Somdeb Mitra
(Fig. 10.2B). At this stage, you may also convert the apparent sedimentation coefficient s∗ to the standard apparent sedimentation coefficient s∗ 20;w . Proceed to the Next step to specify the range of s∗ 20;w that would be used to generate the gðs∗ Þ versus s∗ distribution. We generally choose the 20;w default range selected by the software and proceed to the Next step to produce the distribution. At this stage, the software also computes the weight averaged s∗ 20;w value (Fig. 10.2C). The Next step entails fitting the gðs∗ Þ versus s∗ 20;w distribution to Gaussian function. Several fitting option are made available in DCDTþ. Depending on the number of peaks/shoulders observed, we treat the distribution as composed of a single/multiple species, and fit the gðs* Þ versus s*20;w curve accordingly. We usually select the fitting options as s/ D for molecular mass units and fit D for each species, so that after the fitting session, we obtain as our results, the initial loading concentration C0 (from the area under the plot), s20,w (from the center point), and D20,w (from the standard deviation of the fit) (Fig. 10.2D). Analyze the data from each cell for all the runs, in exactly the same way so as to obtain the s20,w and the D20,w of the RNA molecules at each concentration of the folding cation. The s20,w values, when plotted as a function of cation concentration, show how the RNA molecules assume an increasingly compact shape (higher s20,w) with increasing cation concentration. Also calculate the s20,w/D20,w value at each cation concentration. Since, s/D is proportional to the molecular weight, the plot of s20,w/D20,w versus cation concentration should produce a straight line, provided the molecules do not aggregate at high ionic strengths. Use the s20,w and the D20,w and C0 values obtained from DCDTþ, as input parameters in the program SEDNTERP to calculate the Stoke’s radii (RH) of the RNA molecules. A plot of RH versus cation concentration should show the opposite trend as compared to the s20,w versus cation concentration plot, that is, the hydrodynamic radius of the RNA molecules decrease with increasing cation concentration. The axial ratio (a/b) is also calculated using SEDNTERP and when plotted as a function of cation concentration, should show the same trend as the RH plot (an example is provided in Fig. 10.3B).
two green lines indicate the radial range of data scans used in the analysis (specified by the user). (B) The dc/dt versus s∗ 20;w curves for the same scans, derived by transformation of the X-axis, are shown. (C) The dc/dt versus s∗ distribution is converted to the gðs∗ Þ versus s∗ 20;w distribution. As the distribution shows predominantly one peak, it is fit to a single Gaussian in the next step. (D) Fitting gðs∗ Þ versus s∗ 20;w distribution for a single species using a Gaussian function. The slight deviation from the fit at the left end does not improve if the distribution is fit with two Gaussians. The green line at the bottom shows the residuals of the fit.
229
Studying RNA Folding by Analytical Ultracentrifugation
Unfolded (ensemble average); high RH
B
0
Energy
RH (Å)
A
2
[MgCl2] (mM) 4 6 8
80 75 70 65 60 55 50 45
MgCl2 NaCl
MgCl2 NaCl
a/b
20
Folded; low RH
15 10 5 0.0
C
P5
P2.1
P9.1 Tetraloop
P9.1 P2.1 Non native P13
Core Tetraloop
P9.1 Native P2.1 P13
2.0
10 mM Mg2+
P5 P9
P9 P5
P8 J8/7 P7 P2 P6
P9.2
200 mM Na+
L5c
P9 P5abc
80 mM Na+
0.5 1.0 1.5 [Nacl] (M)
A Bulge
Conformational space
10 mM Na+
10
P5 P9 J8/7
(P14) L2 P2.1
P9.1 (P13)
Tetraloop
Figure 10.3 Analysis of the equilibrium folding of the group I intron ribozyme derived from Tetrahymena thermophila. (A) A schematic of an RNA folding funnel, depicting the progressive reduction of conformational entropy as the molecule approaches the lowest free energy state. An ensemble of unfolded structures, with a much larger average Stoke’s radius, populate the rim of the funnel, while, the folded native state populating the tip of the funnel represents a compact structure, with a smaller Stoke’s radius. [Reprinted from Shcherbakova et al. (2008), Copyright (2008), with permission from Elsevier.] (B) Results of sedimentation velocity analysis of Mg2þ- and Naþ-mediated compaction of the Tetrahymena ribozyme. Both the RH and the a/b ratio show the same trend, that is, decrease with increasing ionic strength. The Mg2þ-mediated compaction isotherm is much more cooperative and occurs predominantly in one step (Kd 0.2 mM), with a very small second phase (Kd 7.5–8.9 mM). The Naþmediated compaction isotherm is distinctly biphasic. The first phase is highly cooperative with a Kd of 40 mM, while the second phase with a Kd 590 mM has a much lower cooperativity. The native state of the molecule is slightly less compact in Naþ ions as compared to Mg2þ ions. [Reprinted by permission from Macmillan Publishers Ltd, Nature Structural Biology (Takamoto et al., 2002), Copyright (2002).] (C) Schematic representation of the equilibrium folding of the Tetrahymena ribozyme, combining the results of local structure probing by hydroxyl radical footprinting and global conformational analysis by AUC. The ribozyme adopts an extended structure, devoid of tertiary contacts, in 10 mM Naþ ions. In the first phase of monovalent ion-mediated compac˚ (complete by 80 mM Naþ ions), a distinct set of native and some tion from 80 to 60 A nonnative tertiary interactions are formed that comprise the native L9-P5 tetraloop– receptor interaction, the native L5b-J6a/6b tetraloop–receptor interaction a nonnative
230
Somdeb Mitra
10. Titration curves obtained by either plotting s20,w or RH values as a function of cation concentration [M], are scaled to fractional saturation by nonlinear leastsquares fitting of the data against the coupled equations pi ¼ pi;lower þ ðpi;upper pi;lower Þ Y i
ð10:11Þ
and Y i ¼
KdinH ½MnH 1 þ KdinH ½MnH
ð10:12Þ
using the program OriginÒ (OriginLabsÒ ) where i denotes an AUC condition (e.g., X mM of ion Mþ), pi is the signal quantitated at a given value of i, pi,lower and pi,upper are the lower and upper transition endpoints, respectively. Eq. (10.12) is the Hill equation that is used to describe cooperative processes, where Kdi is the equilibrium dissociation constant, [M] is the ligand concentration, and nH is the Hill coefficient. The fitting of folding isotherms to the Hill equation is purely phenomenological and the fitted curves do not represent cation binding but simply a transition between two states, mediated by the cation. Therefore, for the purpose of analyzing cation-mediated RNA folding isotherms, Kd and nH represent, respectively, the midpoint and the cooperativity of the folding transition, whereas [M] denotes the concentration of the monovalent or divalent cation (Mþ or M2þ). The best-fit values and 65% confidence intervals are reported.
6. Case Studies We briefly discuss here the major conclusions of a few studies that have employed AUC to characterize the equilibrium conformational changes of a variety of RNA molecules.
6.1. tRNA In one of the earliest examples of the application of AUC to study RNA folding, Henley et al. (1966) demonstrated that the s20,w values of tRNA molecules (unfractionated tRNAs from yeast) decreased with P13-like contact between P9.1 and P2.1. The molecule further compacts in the presence of 200 mM Naþ ions, with a concomitant formation of more native tertiary contacts and reorganization of the P13 contact to its native form. Finally, only divalent ions (10 mM Mg2þ) drive the formation of all the native tertiary contacts within the catalytic core of ˚ ). [Reprinted from the ribozyme, thereby generating its fully compact form (45 A Uchida et al. (2003), Copyright (2003), with permission from Elsevier.]
Studying RNA Folding by Analytical Ultracentrifugation
231
increasing temperature, indicating that the molecules unfolded to larger extended structures upon heating. The thermal denaturation profiles indicated that conformational changes in tRNA occurred in two stages, first, the loss of tertiary structure (between 20 and 40 C) and second, progressive loss of secondary structure with increase in temperature beyond 40 C. Moreover, the changes in hydrodynamic properties due to thermal denaturation were very similar in the presence of divalent (Mg2þ) and monovalent (Naþ) ions, albeit, the RNA was much more stable in the presence of Mg2þ than Naþ. A practical difficulty, encountered in this study was performing the AUC runs at temperatures higher than 40 C; at such high temperatures, the oil vapors from the diffusion pump interfere with the UV absorption optics. They circumvented this problem by using a different type of optical scanning known as the Schlieren optics, which generates the data as profiles of radial derivative of concentration distributions as a function of radius (as opposed to the concentration versus radius scans obtained from UV optics).
6.2. Group I intron ribozyme The ribozyme derived from the self-splicing group I intron of the protozoa Tetrahymena thermophila remains one of the most well-studied large RNA molecule (388 nucleotides). Decades of investigations on the folding of this ribozyme have established the classic paradigm of cation-mediated RNA structure formation; stabilization of a disordered ensemble of extended RNA structures into a compact-folded conformation (Fig. 10.3A). A study aimed at characterizing the equilibrium folding behavior of the Tetrahymena ribozyme (Takamoto et al., 2002) (Fig. 10.3B) demonstrated ˚ occurred that while Mg2þ mediated compaction to a RH of 45 A predominantly in a highly concerted single step (with Kd 0.2 mM), compaction mediated by Naþ ions occurred in a much less concerted manner in two distinct phases. In the first phase, the RNA compacted ˚ with a transition midpoint of 40 mM, whereas in from 80 to 60 A ˚ with a transition midpoint the second phase it further compacted to 50 A of 587 mM. As shown in Fig. 10.3C, comparison with hydroxyl radical footprinting data on the same RNA, under identical experimental condition, revealed that while both native and nonnative tertiary contacts are formed during the first compaction phase, only native tertiary contact formation drives the ribozyme to its folded structure in the second phase. The slightly larger global dimension observed in 1.5 M Naþ as compared to that in 10 mM Mg2þ indicates the inability of monovalent ions, even at sufficiently high concentrations, to fully compact this RNA to its native shape.
232
Somdeb Mitra
6.3. IRES RNA An internal ribosomal entry site (IRES) RNA in the single-stranded RNA genome of the cricket paralysis virus (CrPV) mediates 50 -cap independent ribosome recruitment and translation initiation. Costantino and Kieft (2005), using a combination of hydroxyl radical footprinting and AUC, demonstrated that the CrPV IRES can independently fold into a compact conformation in the presence of Mg2þ, with a solvent excluded and highly structured core. This compact IRES domain is essential for its ribosome binding activity and direct insertion into the mRNA decoding groove of the ribosome. Upon Mg2þ titration, the s20,w of CrPV IRES RNA, like the Tetrahymena ribozyme, undergoes a highly cooperative compaction transition; the process involves a change in the f/fo ratio from 2.29 to 1.95, indicating that the compact form assumes an increasingly spherical shape upon folding.
6.4. M-box riboswitch AUC was used (Brautigam et al., 2009; Dann et al., 2007) along with other techniques to demonstrate that a structured region (termed M-box) in the 50 -untranslated region of the mRNA, encoding a magnesium transporter in Bacillus subtilis, folded into a compact structured domain in response to varying Mg2þ concentrations. The compact conformation, upon Mg2þ binding, induces structural changes in the upstream regions of the mRNA transcript which results in premature transcription termination. Therefore, the M-box region serves as a ‘‘riboswitch’’ which monitors intracellular Mg2þ concentration and maintains Mg2þ homeostasis by regulating the expression of Mg2þ transporters. This compact conformation, characterized by a high s20,w, is not formed at lower concentration of Mg2þ where a much more extended structure with low s20,w, prevails.
7. Conclusions SV experiments using analytical ultracentrifuges have proved to be an extremely useful method for detecting equilibrium conformational changes in RNA molecules, mediated by cations or small ligand molecules. It directly measures the intrinsic properties of RNA molecules in solution, thereby circumventing the need for comparison with calibrated values of know standards as in the case of indirect measurements (like electrophoresis with known mol. wt. markers). Information about global hydrodynamic features of an RNA molecule, derived from AUC experiments is generally used in conjunction with information about local changes at specific sites in
Studying RNA Folding by Analytical Ultracentrifugation
233
the same RNA molecule. Such local structural information can be acquired from complementary techniques like hydroxyl radical footprinting (refer Chapter 2, Part I, Vol. 468, Methods in Enzymology), DMS footprinting, or partial RNase digestion. AUC data can be used as structural filters in computational techniques, like NAST (Nucleic Acid Simulation Toolkit; Jonikas et al., 2009), which use knowledge-based potentials to generate coarse-grained models of folding intermediates of RNA molecules.
ACKNOWLEDGMENTS The author is a member of the laboratory of Michael Brenowitz within which the studies described in this chapter were conducted. This chapter was supported by grant 1RO1GM085130 from the National Institute of General Medical Sciences of the National Institutes of Health.
REFERENCES Adams, P., Stahley, M., et al. (2004). Crystal structure of a group I intron splicing intermediate. RNA 10(12), 1867–1887. Bai, Y., Das, R., et al. (2005). Probing counterion modulated repulsion and attraction between nucleic acid duplexes in solution. Proc. Natl. Acad. Sci. USA 102(4), 1035–1040. Bai, Y., Greenfeld, M., et al. (2007). Quantitative and comprehensive decomposition of the ion atmosphere around nucleic acids. J. Am. Chem. Soc. 129(48), 14981–14988. Baird, N. J., Westhof, E., et al. (2005). Structure of a folding intermediate reveals the interplay between core and peripheral elements in RNA folding. J. Mol. Biol. 352(3), 712–722. Bartley, L. E., Zhuang, X., et al. (2003). Exploration of the transition state for tertiary structure formation between an RNA helix and a large structured RNA. J. Mol. Biol. 328 (5), 1011–1026. Basu, S., Rambo, R., et al. (1998). A specific monovalent metal ion integral to the AA platform of the RNA tetraloop receptor. Nat. Struct. Biol. 5(11), 986–992. Bokinsky, G., Rueda, D., et al. (2003). Single-molecule transition-state analysis of RNA folding. Proc. Natl. Acad. Sci. USA 100(16), 9302–9307. Brautigam, C. A., Wakeman, C. A., et al. (2009). Methods for analysis of ligand-induced RNA conformational changes. Methods Mol. Biol. 540, 77–95. Brion, P., and Westhof, E. (1997). Hierarchy and dynamics of RNA folding. Annu. Rev. Biophys. Biomol. Struct. 26, 113–137. Buchmueller, K. L., Webb, A. E., et al. (2000). A collapsed non-native RNA folding state. Nat. Struct. Biol. 7(5), 362–366. Cantor, C. R., and Schimmel, P. R. (1980). Biophysical Chemistry, Part II: Techniques for the study of biological structure and function. W. H. Freeman and Company. II, San Francisco. Cate, J., Gooding, A., et al. (1996). Crystal structure of a group I ribozyme domain: Principles of RNA packing. Science 273(5282), 1678–1685. Cate, J., Hanna, R., et al. (1997). A magnesium ion core at the heart of a ribozyme domain. Nat. Struct. Biol. 4(7), 553–558.
234
Somdeb Mitra
Cayley, S., Lewis, B. A., et al. (1991). Characterization of the cytoplasm of Escherichia coli K-12 as a function of external osmolarity. Implications for protein–DNA interactions in vivo. J. Mol. Biol. 222(2), 281–300. Cole, P., and Crothers, D. (1972). Conformational changes of transfer ribonucleic acid. Relaxation kinetics of the early melting transition of methionine transfer ribonucleic acid (Escherichia coli ). Biochemistry 11(23), 4368–4374. Cole, P., Yang, S., et al. (1972). Conformational changes of transfer ribonucleic acid. Equilibrium phase diagrams. Biochemistry 11(23), 4358–4368. Cole, J. L., Larry, J. W., et al. (2008). Analytical ultracentrifugation: Sedimentation velocity and sedimentation equilibrium. In ‘‘Biophysical Tools for Biologists, Volume One: In Vitro Techniques,’’ ( J. J. Correira and I. H. William Detrich, eds.), Vol. 84. Burlington, MA, USA. Conn, G., Gittis, A., et al. (2002). A compact RNA tertiary structure contains a buried backbone-Kþ complex. J. Mol. Biol. 318(4), 963–973. Costantino, D., and Kieft, J. S. (2005). A preformed compact ribosome-binding domain in the cricket paralysis-like virus IRES RNAs. RNA 11(3), 332–343. Dann, C. E. III., Wakeman, C. A., et al. (2007). Structure and mechanism of a metal-sensing regulatory RNA. Cell 130(5), 878–892. Das, R., Kwok, L., et al. (2003). The fastest global events in RNA folding: Electrostatic relaxation and tertiary collapse of the Tetrahymena ribozyme. J. Mol. Biol. 332(2), 311–319. Doty, P., Boedtker, H., et al. (1959). Configurational studies of polynucleotides and ribonucleic acid. Ann. N. Y. Acad. Sci. 81, 693–708. Draper, D. E. (2004). A guide to ions and RNA structure. RNA 10(3), 335–343. Draper, D., Grilley, D., et al. (2005). Ions and RNA folding. Annu. Rev. Biophys. Biomol. Struct. 34, 221–243. Freier, S., Kierzek, R., et al. (1986). Improved free-energy parameters for predictions of RNA duplex stability. Proc. Natl. Acad. Sci. USA 83(24), 9373–9377. Fresco, J., Alberts, B., et al. (1960). Some molecular details of the secondary structure of ribonucleic acid. Nature 188, 98–101. Fujita, H. (1975). Foundations of Ultracentrifugal Analysis. Wiley, New York. Golden, B. L., Kim, H., et al. (2005). Crystal structure of a phage Twort group I ribozymeproduct complex. Nat. Struct. Mol. Biol. 12(1), 82–89. Grilley, D., Soto, A., et al. (2006). Mg2þ –RNA interaction free energies and their relationship to the folding of RNA tertiary structures. Proc. Natl. Acad. Sci. USA 103 (38), 14003–14008. Grilley, D., Misra, V., et al. (2007). Importance of partially unfolded conformations for Mg (2þ)-induced folding of RNA tertiary structure: Structural models and free energies of Mg2þ interactions. Biochemistry 46(36), 10266–10278. Guo, F., Gooding, A. R., et al. (2004). Structure of the Tetrahymena ribozyme: Base triple sandwich and metal ion at the active site. Mol. Cell 16(3), 351–362. Henley, D. D., Lindahl, T., et al. (1966). Hydrodynamic changes accompanying the thermal denaturation of transfer ribonucleic acid. Proc. Natl. Acad. Sci. USA 55(1), 191–198. Hohng, S., Wilson, T. J., et al. (2004). Conformational flexibility of four-way junctions in RNA. J. Mol. Biol. 336(1), 69–79. Howlett, G. J., Minton, A. P., et al. (2006). Analytical ultracentrifugation for the study of protein association and assembly. Curr. Opin. Chem. Biol. 10(5), 430–436. Ignacio Tinoco, J., Sauer, K., et al. (2003). Molecular motion and transport properties. In ‘‘Physical Chemistry: Principles and Applications in Biological Sciences,’’ ( J. Challice, ed.). Upper Saddle River, New Jersey, USA. Jonikas, M. A., Radmer, R. J., et al. (2009). Coarse-grained modeling of large RNA molecules with knowledge-based potentials and structural filters. RNA 15(2), 189–199.
Studying RNA Folding by Analytical Ultracentrifugation
235
Kim, H., Nienhaus, G., et al. (2002). Mg2þ-dependent conformational change of RNA studied by fluorescence correlation and FRET on immobilized single molecules. Proc. Natl. Acad. Sci. USA 99(7), 4284–4289. Klein, D., Moore, P., et al. (2004). The contribution of metal ions to the structural stability of the large ribosomal subunit. RNA 10(9), 1366–1379. Koculi, E., Thirumalai, D., et al. (2006). Counterion charge density determines the position and plasticity of RNA folding transition states. J. Mol. Biol. 359(2), 446–454. Koculi, E., Hyeon, C., et al. (2007). Charge density of divalent metal cations determines RNA stability. J. Am. Chem. Soc. 129(9), 2676–2682. Kwok, L., Shcherbakova, I., et al. (2006). Concordant exploration of the kinetics of RNA folding from global and local perspectives. J. Mol. Biol. 355(2), 282–293. Laederach, A., Shcherbakova, I., et al. (2006). Local kinetic measures of macromolecular structure reveal partitioning among multiple parallel pathways from the earliest steps in the folding of a large RNA molecule. J. Mol. Biol. 358(4), 1179–1190. Laederach, A., Shcherbakova, I., et al. (2007). Distinct contribution of electrostatics, initial conformational ensemble, and macromolecular stability in RNA folding. Proc. Natl. Acad. Sci. USA . Laue, T. M. (1995). Sedimentation equilibrium as thermodynamic tool. Methods Enzymol. 259, 427–452. Laue, T. M., and Stafford, W. F. III. (1999). Modern applications of analytical ultracentrifugation. Annu. Rev. Biophys. Biomol. Struct. 28, 75–100. Lebowitz, J., Lewis, M. S., et al. (2002). Modern analytical ultracentrifugation in protein science: A tutorial review. Protein Sci. 11(9), 2067–2079. Lilley, D. M., Bhattacharyya, A., et al. (1992). Gel electrophoresis and the structure of RNA molecules. Biotechnol. Genet. Eng. Rev. 10, 379–401. Manning, G. (1978). The molecular theory of polyelectrolyte solutions with applications to the electrostatic properties of polynucleotides. Q. Rev. Biophys. 11(2), 179–246. Misra, V. K., and Draper, D. E. (2000). Mg(2þ) binding to tRNA revisited: The nonlinear Poisson–Boltzmann model. J. Mol. Biol. 299(3), 813–825. Misra, V., Shiman, R., et al. (2003). A thermodynamic framework for the magnesiumdependent folding of RNA. Biopolymers 69(1), 118–136. Murthy, V. L., and Rose, G. D. (2000). Is counterion delocalization responsible for collapse in RNA folding? Biochemistry 39(47), 14365–14370. Perez-Salas, U. A., Rangan, P., et al. (2004). Compaction of a bacterial group I ribozyme coincides with the assembly of core helices. Biochemistry 43(6), 1746–1753. Philo, J. S. (1997). An improved function for fitting sedimentation velocity data for lowmolecular-weight solutes. Biophys. J. 72(1), 435–444. Philo, J. S. (2000). A method for directly fitting the time derivative of sedimentation velocity data and an alternative algorithm for calculating sedimentation coefficient distribution functions. Anal. Biochem. 279(2), 151–163. Philo, J. S. (2006). Improved methods for fitting sedimentation coefficient distributions derived by time-derivative techniques. Anal. Biochem. 354(2), 238–246. Pljevaljcic´, G., Klostermeier, D., et al. (2005). The tertiary structure of the hairpin ribozyme is formed through a slow conformational search. Biochemistry 44(12), 4870–4876. Russell, R., Millett, I. S., et al. (2000). Small angle X-ray scattering reveals a compact intermediate in RNA folding [see comments]. Nat. Struct. Biol. 7(5), 367–370. Russell, R., Millett, I., et al. (2002a). Rapid compaction during RNA folding. Proc. Natl. Acad. Sci. USA 99(7), 4266–4271. Russell, R., Zhuang, X., et al. (2002b). Exploring the folding landscape of a structured RNA. Proc. Natl. Acad. Sci. USA 99(1), 155–160. Sattin, B. D., Zhao, W., et al. (2008). Direct measurement of tertiary contact cooperativity in RNA folding. J. Am. Chem. Soc. 130(19), 6085–6087.
236
Somdeb Mitra
Schuck, P. (2000). Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and Lamm equation modeling. Biophys. J. 78(3), 1606–1619. Schuck, P. (2004). A model for sedimentation in inhomogeneous media. II. Compressibility of aqueous and organic solvents. Biophys. Chem. 108(1–3), 201–214. Sclavi, B., Sullivan, M., et al. (1998). RNA folding at millisecond intervals by synchrotron hydroxyl radical footprinting. Science 279(5358), 1940–1943. Scott, D. J., and Schuck, P. (2005). A Brief Introduction to the Analytical Ultracentrifugation of Proteins for Beginners. Royal Society of Chemistry, Cambridge, UK. Shcherbakova, I., Mitra, S., et al. (2008). Energy barriers, pathways, and dynamics during folding of large, multidomain RNAs. Curr. Opin. Chem. Biol. 12(6), 655–666. Silverman, S. K., and Cech, T. R. (1999). RNA tertiary folding monitored by fluorescence of covalently attached pyrene. Biochemistry 38(43), 14224–14237. Silverman, S., and Cech, T. (2001). An early transition state for folding of the P4–P6 RNA domain. RNA 7(2), 161–166. Soto, A., Misra, V., et al. (2007). Tertiary structure of an RNA pseudoknot is stabilized by ‘‘diffuse’’ mg(2þ) ions. Biochemistry 46(11), 2973–2983. Stafford, W. F. III. (1992). Boundary analysis in sedimentation transport experiments: A procedure for obtaining sedimentation coefficient distributions using the time derivative of the concentration profile. Anal. Biochem. 203(2), 295–301. Stafford, W. F. III. (1994). Boundary analysis in sedimentation velocity experiments. Methods Enzymol. 240, 478–501. Stafford, W. F. (1997). Sedimentation velocity spins a new weave for an old fabric. Curr. Opin. Biotechnol. 8(1), 14–24. Stahley, M. R., Adams, P. L., et al. (2007). Structural metals in the group I intron: A ribozyme with a multiple metal ion core. J. Mol. Biol. 372(1), 89–102. Takamoto, K., He, Q., et al. (2002). Monovalent cations mediate formation of native tertiary structure of the Tetrahymena thermophila ribozyme. Nat. Struct. Biol. 9(12), 928–933. Takamoto, K., Das, R., et al. (2004). Principles of RNA compaction: Insights from the equilibrium folding pathway of the P4–P6 RNA domain in monovalent cations. J. Mol. Biol. 343(5), 1195–1206. Tan, Z., and Chen, S. (2006). Nucleic acid helix stability: Effects of salt concentration, cation valence and size, and chain length. Biophys. J. 90(4), 1175–1190. Tanford, C. (1961). Physical Chemistry of Macromolecules. Wiley, New York. Thirumalai, D., Lee, N., et al. (2001). Early events in RNA folding. Annu. Rev. Phys. Chem. 52, 751–762. Uchida, T., Takamoto, K., et al. (2003). Multiple monovalent ion-dependent pathways for the folding of the L-21 Tetrahymena thermophila ribozyme. J. Mol. Biol. 328(2), 463–478.
C H A P T E R
E L E V E N
Use of Small Angle X-ray Scattering (SAXS) to Characterize Conformational States of Functional RNAs1 Sebastian Doniach*,† and Jan Lipfert‡ Contents 1. Small Angle X-ray Scattering (SAXS) as a Tool for Global Structure Determination at Low Resolution 2. SAXS and Conformational Changes in Small Functional RNAs 3. SAXS Data Acquisition and Effects of Radiation Damage 4. SAXS Data Analysis 4.1. Model free analysis 4.2. Ab initio three-dimensional shape reconstructions 5. Low-Resolution Atomic Scale Models of RNA: Fitting Secondary Structure RNA Models to Three-Dimensional Shape Models 6. Determining the Thermodynamics of RNA Folding Using Bead Models 7. Concluding Remarks Acknowledgments References
238 238 239 239 239 241 243 245 248 249 249
Abstract Small-angle X-ray scattering (SAXS) is emerging as an important technique to characterize the structure of RNA molecules. While lower in resolution than X-ray crystallography or NMR spectroscopy, SAXS has the great advantage to have virtually no molecular weight limitations and does not require crystallization. In addition, SAXS can be readily applied under a large range of solution * { { 1
Department of Applied Physics, Stanford University, Stanford, California, USA Department of Physics, Stanford University, Stanford, California, USA Kavli Institute of Nanoscience, Delft University of Technology, Delft, The Netherlands Contribution to book edited by D. Herschlag
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69011-X
#
2009 Elsevier Inc. All rights reserved.
237
238
Sebastian Doniach and Jan Lipfert
conditions, allowing to monitor RNA folding, ligand binding, and to characterize partially folded intermediates. Here, we review how the development of SAXS as a structural technique is driven by advances in computer algorithms that allow to reconstruct low-resolution electron density maps ab initio from scattering profiles. In addition, we delineate how these low-resolution models can be used in free energy electrostatics calculations. Finally, we discuss how one can exploit the hierarchical nature of RNA folding by combining the low resolution, global information provided by SAXS with local information on RNA structure, from either experiments or state-of-the-art RNA structure prediction algorithms, to further increase the resolution and quality of models obtained from SAXS.
1. Small Angle X-ray Scattering (SAXS) as a Tool for Global Structure Determination at Low Resolution X-ray scattering from molecules in dilute solution is a classic technique dating back to the use of static X-ray tube sources in the 1960s and earlier. However, the relatively low X-ray flux from such sources makes the acquisition of an X-ray scattering profile a matter of hours of exposure. More recently, the use of synchrotron radiation X-rays has made the acquisition of SAXS data much faster, down to a fraction of second exposure time on third generation electron storage ring sources. When coupled with revolutionary advances in data analysis, whereby a low resolution three-dimensional electron density map may be recovered from the one-dimensional X-ray scattering profile, SAXS has now become a routine technique for characterizing conformational changes in biomolecules (Lipfert and Doniach, 2007; Petoukhov and Svergun, 2007; Putnam et al., 2007). To date such methods have been used to study proteins in solution. Only recently have these methods been applied to the study of RNA molecules in solution (Lipfert et al., 2007a).
2. SAXS and Conformational Changes in Small Functional RNAs The discovery of catalytic activity in RNA molecules by Cech, Altman, and coworkers in 1982 (Zaug et al., 1983) has led to considerable interest in the structure–function relationships of RNA molecules in biology. Furthermore, it is becoming clear that previously unknown RNA machinery, specified in the noncoding regions of the genome, is essential for controlling expression of genes (Amaral and Mattick, 2008; Mattick, 2007).
Use of Small Angle X-ray Scattering
239
As for proteins, X-ray crystallography is the method of choice for determining the structure of RNA molecules when they are in a state that allows for their crystallization. However, even in a folded conformation, RNA molecules present problems for crystallization, partly as a result of their high negative charge. Changes in conformation involved in the function of RNA molecules often lead to large changes in molecular size and shape and involve only partially folded conformations, which preclude crystallization. In such cases SAXS has proved to be a useful technique which complements more local structural techniques such as FRET, hydroxy radical footprinting, and others, as it gives a measure of the global changes in size and shape of the molecule.
3. SAXS Data Acquisition and Effects of Radiation Damage An additional practical consideration for measurement of RNA using SAXS is the relatively high scattering contrast of RNA molecules relative to that of proteins as a result of the high density of phosphate moieties in the polynucleotide backbone. This means that shorter exposure times or more dilute samples may be used compared to SAXS measurements for proteins. In addition, it turns out that RNA is also more resistant to radiation damage than are protein samples, thereby allowing for longer exposure of dilute RNA samples than is possible in the case of protein samples. A heuristic reason for the extra ‘‘radiation hardness’’ of RNAs turns out to be that the more negatively charged RNA or DNA molecules rather ‘‘fall apart’’ into smaller pieces when hit by X-rays (cf. also the synchrotron footprinting experiments of Mike Brenowitz and coworkers (Sclavi et al., 1998)). As smaller pieces scatter much less, the signal will be influenced less by molecules falling apart. In contrast, damage tends to aggregate proteins. This effect becomes rapidly manifested as the aggregates quickly start to dominate the SAXS signal, due to the (MW)2 dependence of the scattering signal. Detailed protocols and additional practical considerations for SAXS measurements of RNA have been published recently (Lipfert et al., 2009).
4. SAXS Data Analysis 4.1. Model free analysis Scattering intensities are isotropic about the direction of the incoming X-ray beam as all molecular orientations are averaged for a dilute solution of molecules. The scattering intensity is obtained from the two-dimensional
240
Sebastian Doniach and Jan Lipfert
CCD data by circular averaging and expressed as a function of scattering angle in terms of the wave vector of momentum transfer q, with q ¼ 4p sin (y)/l, where 2y is the total scattering angle and l the X-ray wavelength. The scattering profiles may be represented as a plot of I(q) versus q. However, other plots are also useful. An example is the use of the Kratky plot in which q2 I(q) is plotted versus q. This plot is useful on account of Porod’s law (Porod, 1951), which states that for large q the scattering from an object with a well-defined surface falls approximately as q 4, which leads to a decrease q 2 in the Kratky representation for large q. Folded proteins or nucleic acids therefore tend to have a characteristic peak in the Kratky plot (Kratky and Porod, 1949). For unfolded or random coil polymer states, Kratky and Porod (1949) showed that the scattering from a random polymer falls like q 1, which leads to a linear rise at high q in the Kratky plot for completely denatured proteins (Doniach, 2001) or nucleic acids. For RNA molecules in low salt buffer conditions (200 mM monovalent salt, often referred to as ‘‘unfolded’’ conditions), the Kratky plot typically exhibits a broad and featureless peak at intermediate q values, due to the presence of secondary structure and partial ordering due to electrostatic repulsion. Only upon addition of high concentrations of a denaturant (e.g., 7 M urea) does the RNA completely unfold, giving rise to the typical q 1 behavior in the scattering profiles at large q. The onset of RNA folding upon addition of Mg2þ to the low salt buffer can typically be monitored by the appearance of a sharper peak, at lower q values than the initial broad peak, in the Kratky plot (Das et al., 2003; Lipfert et al., 2007a, 2008). The most straightforward application of SAXS is in the determination of the radius of gyration, Rg, of the molecule of interest. This may be estimated from Guinier analysis of the low q data and is a model-free result obtained from the Guinier approximation, valid at low q: ! q2 Rg2 IðqÞ Ið0Þ exp ð11:1Þ 3 dependent only on the X-ray wavelength. Here I(0) is the forward scattered intensity and is proportional to the square of the number of electrons in the molecule, relative to the background density of the solvent, and hence scales as the square of the molecular weight. As a result of this q-dependence, the Guinier plot (Guinier, 1939) of the low q region is represented by plotting log(I ) versus q2. Looking at the slope of this curve gives a practical way of visualizing Rg. This Guinier analysis, which is available numerically in a number of software packages, then also allows for a measure of the molecular weight of the molecule through the extrapolation to q ¼ 0. The forward scattering intensity, I(0), may be calibrated by measurements of molecular weight standards, often DNA or proteins samples of known molecular weight and concentration (Lipfert et al., 2009). After calibration of I(0),
241
Use of Small Angle X-ray Scattering
Guinier analysis provides a straightforward way to measure the molecular weight of the molecule or molecular complex under study. The radius of gyration Rg may be expressed in terms of the pair distribution of electron density in the molecule, P(r): ð Dmax r 2 PðrÞ dr Rg ¼ 0ð Dmax ð11:2Þ 2 PðrÞ dr 0
The pair distribution, P(r), and associated maximum atom–atom distance in the molecule, Dmax, may be estimated via the computer program GNOM (Svergun, 1992). This program performs a regularized indirect transform of the scattering data to obtain P(r), which may be thought of as a histogram of interatomic distances. The P(r) function has a maximum at the most probable intermolecular distance and goes to zero at Dmax, the maximum intramolecular distance. In order to estimate Dmax, regularized transforms are performed with the input parameter Dmax varied in steps of 2 A˚. Values of Dmax are chosen that yielded solutions that (i) fit the experimental data well and (ii) have a smooth and strictly positive P(r) function. While P(r) provides a useful guide to the general size and internal distribution of electron density in the molecule, it does not give much insight into the three-dimensional form of the molecule, which is not surprising as it only deals with the one-dimensional aspect of the SAXS profile. Nonetheless, it should be noted that the one-dimensional SAXS profiles or, equivalently, the P(r) distributions have the potential to guide ab initio structural prediction algorithms which start from the nucleotide sequence (cf. Zheng and Doniach (2002) and Wu et al. for the protein case (Wu et al., 2005; Zheng and Doniach, 2002)).
4.2. Ab initio three-dimensional shape reconstructions Although the idea of three-dimensional shape reconstruction from SAXS data dates back to Svergun and Stuhrmann (1991) and Svergun et al. (1996), the method became much more practical after the discovery by Chacon et al. (1998) that the placement of scattering point ‘‘beads,’’ or ‘‘dummy atoms,’’ in a spatial array in such a way that the resulting scattering profile fits the experimental data, could lead to low-resolution three-dimensional representations of the electron density in the molecule. This idea was further developed by Walther et al. (2000), who tested it out on a variety of protein structures, and by Svergun et al. who developed the user friendly programs DAMMIN (Svergun, 1999) and GASBOR (Svergun et al., 2001) which are freely available.
242
Sebastian Doniach and Jan Lipfert
The development of such algorithms which produce candidate threedimensional electron density maps from one-dimensional scattering profiles has proven extremely useful for understanding changes in conformational size and shape and has let to a boom in the use of SAXS as a structural technique, in particular for proteins and protein complexes. Although the software was developed for protein scattering, it has been subsequently shown by Lipfert et al. (2007b) that it also works for RNA molecules. Test cases with known high-resolution structure from X-ray crystallography include the P4–P6 fragment of the Tetrahymena group I intron ribozyme, tRNA, and short DNA duplexes. The advantage of SAXS for RNA structure determination is that it also may be used for partially unfolded structures for which crystallization is impossible or even in the case of folded states of molecules for which crystallization has not yet become available. It is important to realize that the reconstruction of a three-dimensional density map from a one-dimensional scattering profile is not unique. Alternative solutions for a given three-dimensional reconstruction cannot in general be avoided, owing to the lack of phase and angular orientation information inherent in SAXS measurements. Such alternative solutions are most prominent for geometrical bodies of high symmetry such as cylinders. However, for real molecules, as shown by Walther et al. (2000), the lack of intrinsic molecular symmetry is generally found to result in a reasonably unique shape reconstruction, usually obtained after averaging over a number of runs of the reconstruction algorithm. After a number of three-dimensional reconstructions are performed they may be averaged using the program SUPCOMB (Kozin and Svergun, 2001), which performs an initial alignment of structures based on their axes of inertia followed by measurement of their overlap by minimization of the normalized spatial discrepancy (NSD). For two sets of points S1 ¼ 1, . . ., N1 and S2 ¼ 1, . . ., N2 the NSD is defined as "
!#1=2 N1 N2 1 1 X 1 X 2 2 r ðs1;i ; S2 Þ þ r ðs2;i ; S1 Þ NSDðS1 ; S2 Þ ¼ 2 N1 d22 i ¼ 1 N2 d12 i ¼ 1 ð11:3Þ where r(s1,i, S2) is minimum value among the distances between si,1 and all points from S2 and di is the average distance between neighboring points in Si. The NSD has the property that it is zero for identical objects and larger than one for objects that systematically differ from one another (Kozin and Svergun, 2001). The aligned structures can be averaged using the program DAMAVER (Volkov and Svergun, 2002), giving an effective occupancy of
Use of Small Angle X-ray Scattering
243
Figure 11.1 Atomic resolution structure of yeast tRNAPhe (PDB accession code 1TRA), rendered as black sticks and reconstructed density (red transparent surface). The reconstructed density was generated from the filtered consensus bead model by smoothing with a Gaussian kernel. Figure adapted from Lipfert et al. (2007b).
each voxel. Keeping all occupied voxels generates a convex hull of all models filtering at half maximal occupancy provides ‘‘filtered’’ models. In order to better visualize the results, the reconstructed bead models may be converted to electron density maps using real space convolution with a Gaussian kernel with the program SITUS (Wriggers and Chacon, 2001; Wriggers et al., 1999). An example is shown in Fig. 11.1. A kernel ˚ and voxel spacing of 2 A˚ were employed. Molecular width of 6 A graphics were prepared using the program PyMOL (DeLano, 2002).
5. Low-Resolution Atomic Scale Models of RNA: Fitting Secondary Structure RNA Models to Three-Dimensional Shape Models Although the low resolution density maps obtained from the threedimensional reconstruction algorithms give insight into the general shape of a molecular conformation, they intrinsically lack information about how the nucleotide sequence fills the space delineated by the map. Progress on solving this problem of fitting the known (or putative) secondary structure
244
Sebastian Doniach and Jan Lipfert
of an RNA molecule into a three-dimensional shape has recently been made by Lipfert et al. (2008) for the case of the VS ribozyme, which is the largest of the known nucleolytic ribozymes, and the only one for which there is no crystal structure. The principle adopted here results from the hierarchical structural organization of RNA molecules (Brion and Westhof, 1997) which generally consist of Watson–Crick double helices joined by junctions which may be two-way, three-way, or four-way (Lilley, 2000). Here, we adopted a global approach in which helical components are assembled as cylindrical bodies into the three-dimensional shape of the molecule. The three-way helical junctions all play key roles in directing the architecture of the ribozyme. The low-resolution electron density envelope is good enough to fit cylindrical representations of helical segments and has provided a starting point for more detailed structural modeling of the VS ribozyme to include the junctions and tertiary interactions. The VS ribozyme can be synthesized in either its cis or trans form. A lowresolution electron density map calculated from SAXS data for the cis-acting form of the ribozyme is shown in Fig. 11.2. The map was calculated using the scattering profiles obtained in the presence of 10 mM Mg2þ. The electron density maps of the cis and trans forms reveal extended structures with maximum lengths of 110 and 120 A˚, respectively, consistent with the Dmax values. However, there are some significant differences between the two envelopes. The most obvious difference between the cis and trans
PT
PL
PT
PR
PL Concave face
PT
PR
PL
Figure 11.2 Three views of the cis VS ribozyme (Lipfert et al., 2008). The central view shows the concave face. The three protuberances extend from the central flat box, labeled PL, PR, and PT. PR is the most prominent. The two side views are shown in the left and right. These clearly reveal the thinness of the envelope.
Use of Small Angle X-ray Scattering
245
structures is the presence of a protuberance in the cis-ribozyme, strongly suggesting the position of an additional helix. A ridge runs diagonally across the concave face, suggesting the path of a helix. The irregular envelope of the electron density calculated for the cis form of the ribozyme and its subcomponents places considerable limitations on how helices can be fitted, particularly in view of the overall length and a thickness that is close to that of a single helix. A fitting process was performed using cylinders of appropriate length and diameter to represent individual helical sections. In addition, SAXS measurements were made on RNA fragments of the full molecule. Information obtained from fitting to the SAXS data for several of these constructs also helped guide the placement of the helices. A model of the complete cis-acting form of the VS ribozyme was constructed using the cylinder model as a starting point. This exercise was performed to see whether a stereochemically acceptable model based upon the SAXS-derived structure could be constructed that would be consistent with all of the available data. A number of aspects of the structure are presently unknown, and the resulting model should not be confused with an atomic resolution structure. However, it has heuristic value upon which future experiments on the folding and function of the ribozyme may be based. In the final atomic scale model the cylinders with RNA helices were replaced by RNA helices manually located to fit the electron density envelope. In most cases, these were canonical A-form helices and were treated as rigid bodies. Finally, the structure of this model was refined in an iterative manner involving manual adjustment and energy minimization using XPLOR (see Fig. 11.3). More recently, the use of the RNA structure prediction algorithm MC-SYM (Parisien and Major, 2008) is being used to model the structure of the TPP riboswitch (M. Ali, J. Lipfert, S. Seifert, D. Herschlag, and S. Doniach, in preparation).
6. Determining the Thermodynamics of RNA Folding Using Bead Models As RNA molecules are highly negatively charged due to their sugar– phosphate backbone, the conformational state of these molecules is strongly dependent on the concentration of positively charged counterions. In the absence of Mg2þ or other polyvalent counterions, and in low monovalent salt concentrations, RNA becomes denatured as a result of ineffective Debye screening of the repulsive forces between different parts of the polymer. As a result, quantitative representation of the counterion induced free energy changes involved in RNA folding and function is of central
246
Sebastian Doniach and Jan Lipfert
IV
IV
V
V III
III II
II
G638
G638
I
I A756
A756
VI
VII
VI
VII
Figure 11.3 Parallel-eye stereoscopic image of a model of the complete VS ribozyme (Lipfert et al., 2008). The model was constructed by connecting previously defined helical sections of a low-resolution model fitting the density map shown in Fig 11.2. Energy-minimization refinement against the standard stereochemical restraints was used to regularize and refine the structure. The scissile phosphate is shown as a sphere, and the probable active site components A756 and G638 are annotated.
importance in understanding RNA folding and the structure–function relationships for RNA molecules. The ability to bind small-molecule ligands or to form protein complexes adds another dimension to this question for certain functional RNAs. Poisson–Boltzmann theory is widely used to estimate free energy landscapes for counterion induced conformational changes (Chu et al., 2008; Draper, 2008; Draper et al., 2005). Here, one estimates the free energy associated with a cloud of counterions attracted to an RNA (or other charged molecule) by solving self consistently Poisson’s equation for the electrostatic potential on a spatial grid embedding the molecule coupled to the Boltzmann factors determining the local concentration of counterions as a function of the electrostatic potential. To do this (numerically) requires a detailed model of the distribution of fixed charges for the RNA molecule involved. As a consequence, PB analysis has mostly been limited to cases for which high-resolution structures are available from crystallography. However, a complete understanding of
Use of Small Angle X-ray Scattering
247
RNA folding requires quantitative models for the free energy of both the folded and unfolded conformations, for which no crystal structures can be obtained. In a recent work by Lipfert et al. (2007b), it has been shown that the bead models obtained in three-dimensional reconstruction of the SAXS data can provide a substitute for a detailed atomic model. It was shown that the composition of the associated ions could be adequately calculated using PB theory in combination with SAXS-derived low-resolution bead models. In addition, work in progress suggests that it is possible to use the same strategy to compute the electrostatic contribution to the free energy of RNA folding ( J. Lipfert, A. Sim, D. Herschlag, and S. Doniach, in preparation). To illustrate these ideas we show the results of Lipfert et al. on ion binding to the P4–P6 fragment from the Tetrahymena ribozyme (Lipfert et al., 2007b). Das et al. (2005) measured the number of bound excess Mg2þ ions as a function of MgCl2 concentration in a 2 M NaCl background for wild-type P4–P6 and for a mutant that does not bind Mg2þ to the ‘‘metal ion core.’’ As the two specific ‘‘metal ion core’’ Mg2þ binding sites are in the interior of the molecules, our PB calculations (which exclude ions from the interior of the molecule) are not expected to capture their contribution to the overall ion binding. We therefore compare the PB simulations with the ion binding data obtained for the P4–P6 mutant that does not exhibit specific ion binding to the ‘‘metal ion core.’’ This approach neglects small, but measurable, differences between the mutant and wild-type P4–P6 solution structures in high salt concentrations (Takamoto et al., 2002). The theoretical predictions using the reconstructed bead model (Fig. 11.4, thick, dashed line) agree reasonably well with the PB calculations for the atomic resolution coordinates (Fig. 11.4, thin, solid line). They slightly under-predict the excess number of Mg2þ ions determined experimentally, in particular those obtained using the fluorescence indicator HQS (Fig. 11.4, circles). Overall the agreement with the experiment is remarkable, however, given the fact that the data were obtained in 2 M NaCl background and that PB theory is generally expected to be valid only in the low concentration limit. Once the electrostatic potential, F(x) is known on the grid, the Coulomb contribution of the cloud of counterions to the free energy, DG, may be estimated for each new conformation represented by the bead model for partially folded RNAs. However, it should also be mentioned that another important contribution to the total DG is the site-specific binding of Mg2þ ions. This is a general feature of many crystal structures for folded RNAs. Theoretical estimates of the free energy of specific Mg2þ binding would need to take into account possible covalent contributions and so requires very detailed positional information about the bound Mg2þ and its binding site environment (mostly oxygen). Thus, it is really only possible for states in which high-resolution crystallographic data is available.
248
Sebastian Doniach and Jan Lipfert
4
Mg2+ bound
3
2
1
0 –4.5
–4
–3 –3.5 log[Mg2+] (mM)
–2.5
–2
Figure 11.4 Ion binding to a P4–P6 mutant that does not exhibit specific Mg2þ binding in the ‘‘ion core.’’ The number of excess Mg2þ ions was measured using a fluorescence indicator (circles) and atomic emission spectroscopy (squares) by Das et al. (2005) in 2 M NaCl background. Theoretical predictions were obtained from PB calculations using the PDB coordinates (thin, solid lines) or the reconstructed bead model with uniformly assigned charges (Lipfert et al., 2007b) (thick, dashed lines). Figure adapted from reference Bai et al. (2007).
However, as a practical matter, once the counterion cloud contributions have been worked out as discussed above then DG values for specific Mg2þ binding can be deduced as parameters defining their contributions to thermodynamic models of measured folding curves for the transition between different conformational states of the RNA.
7. Concluding Remarks The biophysics of functional RNAs is still at an early stage of development. Many techniques are needed to help relate structure to function for this rapidly growing and highly significant frontier of biology. In particular, ab initio structure prediction methods are being developed (Das and Baker, 2007, 2008; Das et al., 2008; Parisien and Major, 2008) that can give reasonable structural models on a local scale. It is at this stage that SAXS, as a global structural technique, is probably the only way available to distinguish predicted structures that provide global structural electron density compatible with the SAXS data. In addition, in the context of structure–function relationships for functional RNA molecules, SAXS measurements contribute in a fundamental way to sorting out some of the needed details of the conformational changes that are central to the function of these molecules.
Use of Small Angle X-ray Scattering
249
ACKNOWLEDGMENTS We acknowledge our collaborators, in particular, Mona Ali, Yu Bai, Vincent B. Chu, Rhiju Das, Daniel Herschlag, David M. J. Lilley, Jonathan Ouellet, So¨nke Seifert, and Adelene Y. Sim, for help and discussions. This research was generously supported by NIH grant PO1 GM066275. Computing resources were provided by the Bio-X2 computer cluster at Stanford University (NSF award CNS-0619926). Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Office of Science, and Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357.
REFERENCES Amaral, P. P., and Mattick, J. S. (2008). Noncoding RNA in development. Mamm. Genome 19, 454–492. 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. Brion, P., and Westhof, E. (1997). Hierarchy and dynamics of RNA folding. Annu. Rev. Biophys. Biomol. Struct. 26, 113–137. Chacon, P., Moran, F., Diaz, J. F., Pantos, E., and Andreu, J. M. (1998). Low-resolution structures of proteins in solution retrieved from X-ray scattering with a genetic algorithm. Biophys. J. 74, 2760–2775. Chu, V. B., Bai, Y., Lipfert, J., Herschlag, D., and Doniach, S. (2008). A repulsive field: Advances in the electrostatics of the ion atmosphere. Curr. Opin. Chem. Biol. 12, 619–625. Das, R., and Baker, D. (2007). Automated de novo prediction of native-like RNA tertiary structures. Proc. Natl. Acad. Sci. USA 104, 14664–14669. Das, R., and Baker, D. (2008). Macromolecular modeling with rosetta. Annu. Rev. Biochem. 77, 363–382. Das, R., Kwok, L. W., Millett, I. S., Bai, Y., Mills, T. T., Jacob, J., Maskel, G. S., Seifert, S., Mochrie, S. G., Thiyagarajan, P., Doniach, S., Pollack, L., et al. (2003). The fastest global events in RNA folding: Electrostatic relaxation and tertiary collapse of the Tetrahymena ribozyme. J. Mol. Biol. 332, 311–319. 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. Das, R., Kudaravalli, M., Jonikas, M., Laederach, A., Fong, R., Schwans, J. P., Baker, D., Piccirilli, J. A., Altman, R. B., and Herschlag, D. (2008). Structural inference of native and partially folded RNA by high-throughput contact mapping. Proc. Natl. Acad. Sci. USA 105, 4144–4149. DeLano (2002). http://www.delanoscientific.com Doniach, S. (2001). Changes in biomolecular conformation seen by small angle X-ray scattering. Chem. Rev. 101, 1763–1778. Draper, D. E. (2008). RNA folding: Thermodynamic and molecular descriptions of the roles of ions. Biophys. J. 95, 5489–5495. Draper, D. E., Grilley, D., and Soto, A. M. (2005). Ions and RNA folding. Annu. Rev. Biophys. Biomol. Struct. 34, 221–243. Guinier, A. (1939). La diffraction des rayons X aux tre`s petits angles: Application a` l’e´tude de phe´nome`nes ultramicroscopiques. Ann. Phys. (Paris) 12, 161–237. Kozin, M. B., and Svergun, D. I. (2001). Automated matching of high- and low-resolution structural models. J. Appl. Crystallogr. 34, 33–41.
250
Sebastian Doniach and Jan Lipfert
Kratky, O., and Porod, G. (1949). Ro¨ntgenuntersuchung gelo¨ster Fadenmoleku¨le. Recl. Trav. Chim. Pays Bas 68, 1106–1122. Lilley, D. M. (2000). Structures of helical junctions in nucleic acids. Q. Rev. Biophys. 33, 109–159. Lipfert, J., and Doniach, S. (2007). Small-angle X-ray scattering from RNA, proteins, and protein complexes. Annu. Rev. Biophys. Biomol. Struct. 36, 307–327. Lipfert, J., Das, R., Chu, V. B., Kudaravalli, M., Boyd, N., Herschlag, D., and Doniach, S. (2007a). Structural transitions and thermodynamics of a glycine-dependent riboswitch from Vibrio cholerae. J. Mol. Biol. 365, 1393–1406. Lipfert, J., Chu, V. B., Bai, Y., Herschlag, D., and Doniach, S. (2007b). Low-resolution models for nucleic acids from small-angle X-ray scattering with applications to electrostatic modeling. J. Appl. Crystallogr. 40, S229–S234. Lipfert, J., Ouellet, J., Norman, D. G., Doniach, S., and Lilley, D. M. (2008). The complete VS ribozyme in solution studied by small-angle X-ray scattering. Structure 16, 1357–1367. Lipfert, J., Herschlag, D., and Doniach, S. (2009). Riboswitch conformations revealed by small-angle X-ray scattering. Methods Mol. Biol. 540, 141–159. Mattick, J. S. (2007). A new paradigm for developmental biology. J. Exp. Biol. 210, 1526–1547. Parisien, M., and Major, F. (2008). The MC-Fold and MC-Sym pipeline infers RNA structure from sequence data. Nature 452, 51–55. Petoukhov, M. V., and Svergun, D. I. (2007). Analysis of X-ray and neutron scattering from biomacromolecular solutions. Curr. Opin. Struct. Biol. 17, 562–571. Porod, G. (1951). Die Ro¨ntgenkleinwinkelstreuung von dichtgepackten kolloiden systemen. Kolloid-Zeitschrift & Zeitschrift Fur Polymere 124, 83–114. Putnam, C. D., Hammel, M., Hura, G. L., and Tainer, J. A. (2007). X-ray solution scattering (SAXS) combined with crystallography and computation: Defining accurate macromolecular structures, conformations and assemblies in solution. Q. Rev. Biophys. 40, 191–285. Sclavi, B., Sullivan, M., Chance, M. R., Brenowitz, M., and Woodson, S. A. (1998). RNA folding at millisecond intervals by synchrotron hydroxyl radical footprinting. Science 279, 1940–1943. Svergun, D. I. (1992). Determination of the regularization parameter in indirect-transform methods using perceptual criteria. J. Appl. Crystallogr. 25, 495–503. Svergun, D. I. (1999). Restoring low resolution structure of biological macromolecules from solution scattering using simulated annealing. Biophys. J. 76, 2879–2886. Svergun, D. I., and Stuhrmann, H. B. (1991). New developments in direct shape determination from small-angle scattering: 1. Theory and model-calculations. Acta Crystallogr. A 47, 736–744. Svergun, D. I., Volkov, V. V., Kozin, M. B., and Stuhrmann, H. B. (1996). New developments in direct shape determination from small-angle scattering: 2. Uniqueness. Acta Crystallogr. A 52, 419–426. Svergun, D. I., Petoukhov, M. V., and Koch, M. H. (2001). Determination of domain structure of proteins from X-ray solution scattering. Biophys. J. 80, 2946–2953. Takamoto, K., He, Q., Morris, S., Chance, M. R., and Brenowitz, M. (2002). Monovalent cations mediate formation of native tertiary structure of the Tetrahymena thermophila ribozyme. Nat. Struct. Biol. 9, 928–933. Volkov, V. V., and Svergun, D. I. (2002). 12th International Conference on Small-Angle Scattering. Venice, Italy. Walther, D., Cohen, F. E., and Doniach, S. (2000). Reconstruction of low-resolution threedimensional density maps from one-dimensional small-angle X-ray solution scattering data for biomolecules. J. Appl. Crystallogr. 33, 350–363.
Use of Small Angle X-ray Scattering
251
Wriggers, W., and Chacon, P. (2001). Using Situs for the registration of protein structures with low-resolution bead models from X-ray solution scattering. J. Appl. Crystallogr. 34, 773–776. Wriggers, W., Milligan, R. A., and McCammon, J. A. (1999). Situs: A package for docking crystal structures into low-resolution maps from electron microscopy. J. Struct. Biol. 125, 185–195. Wu, Y. H., Tian, X., Lu, M. Y., Chen, M. Z., Wang, Q. H., and Ma, J. P. (2005). Folding of small helical proteins assisted by small-angle X-ray scattering profiles. Structure 13, 1587–1597. Zaug, A. J., Grabowski, P. J., and Cech, T. R. (1983). Autocatalytic cyclization of an excised intervening sequence RNA is a cleavage-ligation reaction. Nature 301, 578–583. Zheng, W., and Doniach, S. (2002). Protein structure prediction constrained by solution X-ray scattering data and structural homology identification. J. Mol. Biol. 316, 173–187.
C H A P T E R
T W E LV E
Time-Resolved X-ray Scattering and RNA Folding Lois Pollack* and Sebastian Doniach†,‡ Contents 254 254 255 256 256 257 259 260 261 261 262 262 262
1. 2. 3. 4.
Introduction SAXS Studies of RNA Data Acquisition Time-Resolved SAXS Methods 4.1. Stopped flow mixers and RNA folding 4.2. Continuous flow mixers and RNA folding 5. Mixer Fabrication 5.1. X-ray beam parameters 5.2. Sample preparation 5.3. Samples and radiation damage 6. Data Analysis of SAXS Measurements 6.1. Radius of gyration 6.2. Principal component analysis 6.3. Three-dimensional reconstruction from one-dimensional SAXS data 7. Concluding Remarks Acknowledgments References
264 266 266 267
Abstract Time-resolved small-angle X-ray scattering (SAXS) reports changes in the global conformation of macromolecules and is thus a valuable probe of structural transitions like RNA folding. Time-resolved SAXS has been applied to study folding of the Tetrahymena ribozyme. This chapter describes the methods that enable acquisition and analysis of time-resolved SAXS data and insights into RNA folding gained from these studies.
* School of Applied and Engineering Physics, Cornell University, Ithaca, New York, USA Department of Applied Physics, Stanford University, Stanford, California, USA Department of Physics, Stanford University, Stanford, California, USA
{ {
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69012-1
#
2009 Elsevier Inc. All rights reserved.
253
254
Lois Pollack and Sebastian Doniach
1. Introduction Conformational changes in biomolecules, including folding, can be induced by solvent exchange through mixing. For proteins, this change is usually accomplished by diluting a solution of a protein–denaturant mixture (urea or guanidine hydrochloride) with buffer so that the resulting solvent has a denaturant concentration well below a threshold for denaturation. For polynucleotides such as RNA, denaturation results from the repulsive Coulomb forces between phosphate groups in the phosphate–sugar backbone which are imperfectly screened in a buffer containing minimal monovalent salt (NaCl or KCl). In these systems folding is induced by the addition of buffer containing higher concentrations of cations, typically 10 mM Mg2þ (as MgCl2). Small-angle X-ray scattering (SAXS) provides valuable information about the global structure of macromolecules, such as the size and shape of RNA (Lipfert and Doniach, 2007). Time-resolved SAXS reports the changes in the global conformation that accompany a structural transition like folding. The earliest time-resolved SAXS studies of RNA folding (Russell et al., 2000) relied on manual mixing techniques and reported significant compaction of unfolded RNA within 30 s of the addition of Mg2þ. Access to shorter measurement times was enabled by application of rapid mixing devices. Stopped flow mixers were used to measure folding events on time scales as short as tens of milliseconds after initiation of folding (Russell et al., 2002); more specialized continuous flow mixers accessed times beginning within about 1 ms of the initiation of folding (Akiyama et al., 2002; Pollack et al., 1999). Application of these techniques to the RNA folding problem has revealed discrete and temporally well-resolved phases of molecular compaction triggered by the addition of cations (Das et al., 2003; Russell et al., 2002). These transitions occur on time scales as short as single millisecond and as long as minutes. In conjunction with studies of mutants (Das et al., 2003; Kwok et al., 2006) or when coupled with other techniques that report local structure formation on comparable time scales (Kwok et al., 2006; Schlatterer et al., 2008), time-resolved SAXS experiments have helped to provide insight into the mechanisms that drive large-scale conformational transitions. This chapter describes the methods involved in acquiring and interpreting time-resolved SAXS data of RNA folding.
2. SAXS Studies of RNA We begin with a brief review of static SAXS (see Chapter 11 of this volume). X-rays incident on a sample are deflected or scattered by objects with electron densities that differ from that of the surrounding buffer. For
255
Time-Resolved SAXS and RNA Folding
˚ , the objects with dimensions larger than the X-ray wavelength, typically 1 A angular dependence of the scattering intensity reflects phase differences between X-rays scattered from electrons at different positions within the object. The intensity falls off more rapidly with angle for a larger (compared to smaller) object, as more substantial phase differences result from the larger distances present. Because SAXS reports molecular conformation in solution, it is ideally suited to measure conformational changes, for example, those due to folding; however, the molecules in solution are randomly oriented with respect to the beam, thus the spatially averaged structure is measured. This latter consideration is important in interpreting SAXS data.
3. Data Acquisition Small-angle scattering profiles report the intensity of X-rays scattered as a function of angle, relative to the y ¼ 0 direction of the incident beam. Typically, intensity is plotted as a function of q ¼ 4p sin y/l or s ¼ 2 sin y/l ¼ q/2p. A cartoon representing a SAXS beamline is shown in Fig. 12.1. Two profiles must be acquired to extract the scattering originating from the macromolecule. The first measures the scattering of RNA plus buffer in the cell. This signal also includes parasitic photons. The second measures scattering from buffer alone in the cell along with the parasitic signal. All profiles are scaled to compensate for variations in X-ray beam intensity, monitored by a diode that is integrated into the beam stop (see Fig. 12.1) and by ionization chambers placed before the sample. The ‘‘RNA absent’’ or background profile is subtracted from the ‘‘RNA
Scattering image CCD Incident X-rays
Sample
Converted to I(q) via radial integration
2q Beam stop
Figure 12.1 A schematic representation of a SAXS experiment. The X-ray beam is incident from the left and scatters from the sample. A detector, located to the right of the sample, records the angular variation of intensity of scattered X-rays. The shape of this scattering profile contains information about the global structural features of the molecules in the sample. More details about the beamline components, as well as the process for converting CCD images into one dimensional curves of intensity versus angle, can be found elsewhere in this volume (Chapter 19).
256
Lois Pollack and Sebastian Doniach
present’’ profile, yielding the scattering profile of the RNA alone. Note that the background signal contains contributions not only from the buffer but also from all windows and other components along the beamline, including notorious slit scatter that should be minimized but cannot always be eliminated. It is important to note that this so-called ‘‘parasitic scattering’’ tends to be greatest in the forward direction, surrounding the beam stop. Standard experimental protocol suggests that ‘‘background’’ images be acquired both before and after an RNA-containing image, using the same sample cell whenever possible. The ‘‘before’’ and ‘‘after’’ scattering profiles should match identically and can then be subtracted from the RNA scattering profile. Signal averaging is almost always required for time-resolved measurements. Signals are small because of the relatively low number of scattered photons resulting from the short exposure times employed for manual mixing or stopped flow measurements or the small sample volumes associated with continuous flow mixers.
4. Time-Resolved SAXS Methods Time-resolved SAXS studies report the time course of conformational changes following the addition of cations that trigger RNA folding. The ultimate time resolution of these experimental methods is determined both by the type of mixer used and the measurement time.
4.1. Stopped flow mixers and RNA folding Mixing times of several milliseconds are typical for stopped flow, though newer commercial mixers boast dead times approaching the single millisecond. After mixing, solutions are ejected into a SAXS compatible cell, described in more detail in Section 4.1.1. The X-ray beam may be gated to collect snapshots of the scattering profile at a predetermined set of intervals in order to minimize the total radiation exposure and hence the radiation damage to the sample. Short exposures are acquired after mixing is complete, at carefully calibrated delay times. In this case the time resolution is determined by the shortest opening time of the shutter and by the sensitivity of the detector. Alternatively, the sample can be exposed continuously to the X-ray beam and the detector read periodically. As the readout time of typical CCD detectors may be quite slow (hundreds of milliseconds), this also limits the time resolution of the measurement. Again, total radiation dose needs to be minimized to avoid excessive radiation damage to the sample. In these experiments, timing is controlled by computer coordination between the mixer and the beamline shutter or detector. Stopped
Time-Resolved SAXS and RNA Folding
257
flow places no restrictions on the longest time delays, in principle the sample can be left to age for hours; however, repeated exposure of the same sample should be avoided as it can result in radiation damage. 4.1.1. Stopped flow instrumentation Commercially available stopped flow mixers (from BioLogic as used at the Advanced Photon Source (APS) and the Cornell High Energy Synchrotron Source (CHESS) or Unisoku, Osaka, Japan, as used at Stanford Synchrotron Radiation Laboratory (SSRL)) can be readily modified for compatibility with synchrotron X-ray beams. The mixed sample is pushed either into customized cuvettes, machined to fit into the mixer head, or into X-ray compatible capillaries. The path length of the cuvettes/capillary can be adjusted for any X-ray energy; for example, a 1-mm path length works well at X-ray energies of 8 keV. Most importantly, the ‘‘windows’’ on the cuvette, through which X-rays pass, must not contribute significantly to the overall background scattering. Both mica and silicon nitride windows have been successfully employed for SAXS experiments (Andresen et al., 2004; Lipfert et al., 2006).
4.2. Continuous flow mixers and RNA folding Continuous flow mixers are employed to access folding events on timescales shorter than 10 ms. In addition to sharp time resolution, the use of continuous flow remediates radiation damage because the (constantly replenished) sample is only briefly exposed to the X-ray beam. As indicated above, the mixing time of RNA with Mg2þ is a critical factor in determining the time resolution of the experiment. Mixing times as short as tens of microseconds have been demonstrated using microfluidic mixers with hydrodynamic focusing (Park et al., 2006), although the mixing time of the present generation of SAXS compatible mixers (shown schematically in Fig. 12.2) is of order 1 ms. RNA is unfolded in a low salt buffer and flows into one port of the continuous flow mixer. An equivalent buffer containing an additional 10 mM MgCl2 flows into two orthogonal ports. The flow of these two channels focuses the RNA-containing solution into a thin jet. Diffusion of small Mg2þ ions across this stream occurs rapidly, triggering folding. The RNA folds as it flows down the channel; the device can be translated relative to the X-ray beam to probe the RNA conformation at any position. In this way measurement of time delay is converted into distance of the point of measurement from the point of mixing, divided by the velocity of the stream. Hence, time dependence is measured within this mixer by sampling numerous positions along the outlet channel. In the simplest case, the ‘‘time resolution’’ of the measurement is determined by the passage time of flowing sample across the beam-defined window. For
258
Lois Pollack and Sebastian Doniach
C Mg2+ X-ray probe beam Unfolded RNA no Mg2+ A
B
Mg2+
Figure 12.2 A schematic of continuous flow mixers used for time-resolved SAXS studies of RNA folding. The molecular shapes depicted in the figure represent coarse grain models of RNA conformations, described in Russell et al. (2002).
example, if the stream flows at 10 cm/s and the X-ray beam width is 10 mm, the time resolution of the measurement equals 100 ms. The time resolution of the measurement also depends on several important mixer parameters: the diffusive mixing time (determined by the jet width and the diffusion coefficient of the ions (Knight et al., 1998)), the beam size, uncertainty in the mixing time resulting from the hydrodynamic focusing process, and depth-dependent variations in the flow speed of the jet. Each of these issues has been addressed independently, but will be briefly reviewed here for completeness. The first uncertainty arises from premature mixing of RNA with Mg2þ ions as the jet is being focused. The RNA that flows along the edges of the jet contacts Mg2þ at an earlier time than the RNA that flows along the midplane of the device. Premixing times depend both on channel dimension and flow speed, and introduce uncertainty on the order of 1 ms into the mixing times. These effects can be eliminated using a more complicated mixer, employing sheath flow (Park et al., 2006). Time resolution can also be limited by the parabolic flow profile of a confined fluid in the low Reynolds number (laminar flow) regime. The fluid velocity at the walls approaches zero. If the probe beam sample molecules spread over the entire width of a channel, their differing velocities must be considered. Those in close proximity to the walls travel very slowly, whereas those at the center of the channel flow most rapidly. To compensate for this effect, we flow an extra layer of buffer against the walls
259
Time-Resolved SAXS and RNA Folding
Silicon nitride window
PDMS
Silicon
Silicon nitride window
PDMS
Figure 12.3 A more realistic depiction of the hydrodynamic focusing mixer used for time-resolved SAXS studies. This figure, adapted from Lisa Kwok’s thesis (Ph.D. Cornell), illustrates the multilayer construction of the mixing device.
where the velocity gradient is largest. Figure 12.3 illustrates a more realistic schematic of the mixer. Its fabrication is described in Section 5. An additional benefit of this more complex geometry is the lack of contact between RNA and the windows, which eliminates concerns about sample sticking.
5. Mixer Fabrication The rapid mixer is composed of layers that are fabricated separately and then assembled together. The main four channel device, represented by the cartoon of Fig. 12.2, is etched through a 1-mm-thick silicon wafer using an anisotropic Bosch process RIE (Unaxis 770, Unaxis). The depth of this etch requires a thick mask. We use a 7 mm layer of PECVD silicon dioxide (GCI PECVD; Group Sciences Incorporated, San Jose, CA). This mixer is sandwiched between two 100 mm thick poly(dimethylsiloxane) (PDMS) layers (Duffy et al., 1998), which contain channels in a T configuration.
260
Lois Pollack and Sebastian Doniach
These three channels overlay the side and outlet channels of the mixer and provide the buffer flow that separates the RNA-containing jet from the windows. These layers were fabricated by curing PDMS onto masters created by selectively exposing SU-8 photoresist. Finally, the PDMS–silicon device is sandwiched between two silicon nitride windows. These windows are fabricated by coating a wafer with a thin silicon nitride film, then through etching from the bottom to expose rectangular silicon nitride membranes. The silicon is cut to size, and these windows are attached to the PDMS layers, forming a complete device. Additional liquid PDMS is used to seal the device, if necessary. Once assembled, the device is mounted in a stainless steel holder. A large ( mm2) pad, fabricated at the end of each channel, serves as the interface between external plumbing and the microfluidic channels. The custommachined holder contains a series of channels that interface on one end with the fabricated pad and on the other end with standard connectors for Luer lock tubing. O-rings positioned between the stainless steel holder and the chip reservoirs provide the required fluid seal. The Luer lock fittings interface the chips to standard tubing and eventually to syringes that contain the samples and buffers. Finally, Harvard syringe pumps are used to drive the flow through the channels. Precision syringe pumps (Harvard PHD 2000 Harvard Apparatus) are used to drive the RNA-containing solution, while more robust pumps (Harvard 22) are used to drive the substantial buffer flow. By varying the settings on the pumps, flow speeds ranging from mm/s through tens of cm/s can be achieved. Both short and long times can be easily accessed in a single device. These mixers enable studies on time scales ranging from submillisecond up to 0.2 s.
5.1. X-ray beam parameters Because of the short measurement times of stopped flow, or the small sample volumes associated with continuous flow, a synchrotron source is required for time-resolved SAXS studies of RNA folding. To date studies have been carried out at the APS, at CHESS, and at SSRL. The optics vary by beamline, but all involve intensity enhancement by an insertion device. Monochromatic undulator beam was employed for stopped flow experiments at the 12-ID station at APS (Seifert et al., 2000). Multilayer beam was employed at SSRL beamline 4-2 (Tsuruta et al., 1998), and focused multilayer beam was employed at the CHESS G1 station (Kazimirov et al., 2006). For the continuous flow cell, ‘‘pink’’ or 3% bandwidth undulator beam was employed at the 8-ID beamline at APS (Sandy et al., 1999). All experiments employed a CCD detector to record a 2D image of the scattering, as illustrated in Fig. 12.1.
Time-Resolved SAXS and RNA Folding
261
5.2. Sample preparation Since RNA secondary structure is relatively stable, RNA folding experiments which start with a sample that has been prepared in the absence of Mg2þ generally need to anneal the sample in a suitable buffer at relatively high temperatures between 50 and 90 C. This observation is a result of experiments on folding of the Tetrahymena ribozyme, in which a kinetically stable misfolded state was found to form when folding was initiated by addition of Mg2þ for a sample prepared at low temperature (Russell et al., 2002). This state was shown to give an Rg value of 51 A˚, identical to that for the folding intermediate. A subsequent 50 C incubation was required to achieve the reduced Rg of the native state. It is important to note that the pathways that lead to the native and misfolded states diverge from each other late in folding, concomitant with or ˚ intermedisubsequent to the rate-limiting step. This indicates that the 51 A ate that was observed prior to the rate-limiting step is populated by molecules that fold correctly as well as incorrectly. This kind of observation has led to the idea that functional RNA molecules tend to fold an order of magnitude more slowly than proteins, as kinetic traps may form on the folding pathway which then take time to unfold again so that the molecule can finally reach its functional folded state. However, the lack of time-resolved RNA folding measurements to date on a range of different functional RNAs means that this idea has not yet been substantiated in depth.
5.3. Samples and radiation damage Acquisition of a complete time series typically requires between 1 and 10 mg of RNA, dissolved to concentrations between 1 and 4 mg/ml. As a control for aggregation due to the relative high RNA concentration, data are acquired at two or three different concentrations. Ample quantities of RNA were prepared by in vitro transcription and purified as described in Russell and Herschlag (1999) and Schlatterer et al. (2008). Molecules were unfolded in 50 mM Na-MOPS or K-MOPS buffer, pH 7.0. In most cases folding was initiated by the addition of buffer containing sufficient MgCl2 to raise the free concentration of Mg2þ to 10 mM. The high flux of synchrotron X-rays generates large numbers of hydroxyl radicals that can radiation damage RNA samples (e.g., Sclavi et al., 1997). Radiation damage is manifested by time-dependent changes in scattering profiles, most notably at the lowest angle, which signify aggregation. These changes can be small, but accumulate with time. As mentioned above, flow cells remediate this damage, but increase the rate of RNA consumption. Further measures that limit radiation damage include the use of Tris–HCl buffer that rapidly scavenged hydroxyl radicals (Fang et al., 2000).
262
Lois Pollack and Sebastian Doniach
6. Data Analysis of SAXS Measurements 6.1. Radius of gyration Early time-resolved data on protein folding emphasized the time dependence of the forward scattering intensity I(0) and the radius of gyration Rg (Chen et al., 1996, 1998; Eliezer et al., 1993, 1995). Radius of gyration values were calculated using the Guinier approximation. The first time-resolved SAXS studies of Tetrahymena ribozyme folding (Russell et al., 2000) were achieved by manual mixing. Russell et al. reported changes in the radius of gyration of this ribozyme upon initiation of tertiary folding by the addition of Mg2þ to ˚ RNA under low-salt conditions. The Rg value decreased from 74 to 51 A within the experimental dead time of 1 min, yielding a minimum rate constant for the compaction of 3 min 1, at least 20-fold faster than the overall rate constant for folding. Under these experimental conditions (15 C), 95% of the ribozyme population misfolds to a state that is stable for hours, but ˚ at elevated temperature rapidly converts to the native state, with an Rg of 47 A (Russell and Herschlag, 1999). These experiments indicate that a compact set of intermediates forms early in the folding process. Despite the straightforward nature of these measurements, it took several years before the radius of gyration of a full folding course could be measured, largely due to experimental difficulties in accessing very low angle data due to parasitic scattering. Kwok et al. (2006) reported time-dependent Rg measured over the full course of folding. Representative curves are shown in Fig. 12.4 for the wild-type ribozyme initiated from a standard (20 mM) low-salt initial condition as well as from a higher salt (100 mM) initial condition. Both continuous and stopped flow techniques were employed. Time-resolved SAXS studies of the full-length ribozyme display a number of discrete transitions, with amplitudes that depend on the presence of specific tertiary contacts as well as on salt conditions. The parallel application of a local structural probe, in this case time-dependent hydroxyl radical footprinting, was essential in interpreting the Rg data. Application of these coupled techniques clearly indicated that a rapid compaction occurs prior to significant tertiary contact formation in the full-length ribozyme (Kwok et al., 2006). Comparable studies on folding of the P4–P6 domain display compaction that is more coincident with tertiary contact formation and suggest that folding of this domain is limited by stiff hinges in addition to electrostatics (Schlatterer et al., 2008).
6.2. Principal component analysis Although the radius of gyration provides the overall size of the molecule, the full scattering profile contains much additional information. Segel et al. (1999) introduced the use of principal component analysis of the kinetics (see also Doniach, 2001) to separate scattering profiles of intermediate states
263
Time-Resolved SAXS and RNA Folding
80
Radius of gyration (A)
75 70 65 60 55 50 45 40 −6
−4
–2 0 log10 time (s)
2
4
Figure 12.4 The time-dependent radius of gyration for the Tetrahymena ribozyme following the addition of 10 mM Mg2þ to RNA in solutions containing either low (20 mM, brown) or moderate (100 mM, blue) concentrations of monovalent ions.
occupied along the kinetic folding pathway. This approach was successfully applied in the first full time-resolved studies of RNA folding, combining both continuous and stopped flow data (Russell et al., 2002). Here, data acquired for folding times ranging from 5 ms to 1000 s were fit according to singular value decomposition (SVD) analysis (Doniach, 2001; Henry and Hofrichter, 1992). This analysis identifies the smallest number of independent curves that, in different combination, can recreate a dataset of related experimental curves. A matrix representing the experimental data is created with columns corresponding to scattering profiles acquired at different times. SVD (MATLAB) is used to transform this matrix of scattering profiles in a product of three matrices: [U W V ]T. The leftmost matrix, U, contains the set of column vectors that form an orthonormal basis for the data matrix. The center matrix, W, is diagonal and contains the so-called singular values for each basis curve in U. These coefficients dictate the weight of each basis curve that is needed to accurately reconstruct each experimental curve. Finally, V provides a measure of the weights of each basis curve required to reconstruct the data. V contains the amplitudes of each basis vector for the initial data profiles being analyzed. The number of significant singular values (i.e., with a magnitude significantly larger than noise) indicates the number of independent basis curves required to completely reconstruct the data. Application of SVD analysis to the time-resolved SAXS data of Russell et al. (2002) clearly indicated two significant singular values, and therefore enabled data analysis by projection of each independent scattering profile onto two states. The folded and unfolded states were selected because of
264
Lois Pollack and Sebastian Doniach
their physical significance. A least-squares fit was used to optimize this decomposition. In most cases, analysis of the residuals of the fit was consistent with random noise, confirming the applicability of this method. In some cases small systematic deviation was observed, suggesting the need for a third SVD component; however, inclusion of an additional component did not significantly alter the conclusions of the analysis based on two components. This decomposition was applied to curves displaying the product of Is2 versus s; plots of SAXS data in this form are called Kratky plots. This format readily distinguished extended from compact conformations and emphasizes changes in scattering profiles at larger angle, reflecting subtle yet significant changes in molecular conformation. Kratky plots of unfolded RNA typically have a small shoulder or peak at low angle, followed by a gentle rise. Folded RNA, which are much more compact, display strong peaks in the low to mid angle region. This analysis was applied in the first millisecond scale studies of RNA folding (Russell et al., 2002) in addition to follow-up studies (Das et al., 2003) that probed folding of a mutant RNA in which the five long range tertiary contacts were knocked out. Figure 12.5 shows Kratky plots in addition to the coefficients of the linear projection of each time-dependent scattering profile on the unfolded and folded states. Each set of fractional weights can be fit to a double (for the wild-type) or single (for the knockout mutant) exponential to extract kinetic parameters. The data shown in this figure were acquired via both continuous (open circles) and stopped (closed circles) flow. For all datasets, SVD analysis reveals two dominant components. However, the presence of more than two states cannot be ruled out (Das et al., 2003).
6.3. Three-dimensional reconstruction from one-dimensional SAXS data Analysis of the data in terms of low-resolution three-dimensional electron density maps of the scattering molecules in different states dates back to Svergun and Stuhrmann (1991). This approach received a qualitative advance by the work of Chacon et al. (1998), who showed how density maps could be generated by placing point scatters on a grid to reproduce the observed scattering profile. The key step here is the breaking of isotropic symmetry of the data by placing scattering centers (‘‘beads’’) on a lattice. The uniqueness of the resulting reconstructions was investigated for several proteins by Walther et al. (2000), who also developed an algorithm (SAXS 3d) for generating the bead models. At the same time, Svergun and collaborators (Petoukhov and Svergun, 2003; Svergun, 1999) developed algorithms DAMMIN and GASBOR for building bead models which are now widely used.
265
Time-Resolved SAXS and RNA Folding
A
B s2 I(s), quintuple mutant
s2 I(s), wild type
100
Time (s)
10
1
0.1
0.01
0.001 0
C
0
0.01 s (A−1)
0.02
0
0.01 s (A−1)
0.02
1
PF
0.8 0.6 0.4 0.2 0
Partial collapse
0
0.01
0.02 Time (s)
0.2
0.4
0.6
0.8
Figure 12.5 SAXS data representing a full time series following the addition of Mg2þ to trigger folding of the Tetrahymena ribozyme. Scattering profiles of two constructs are shown as Kratky plots. Panel (A) illustrates folding of the wild-type ribozyme. Panel (B) illustrates folding of a construct in which five key long range contacts have been disrupted by mutation. Panel (C) illustrates the two-state projections described in the text. Here, PF represents the percentage of the ‘‘folded’’ scattering profile that provides the best fit to each time-dependent curve. Data are shown for both the wild-type (squares) and mutant (circle) ribozymes (Figure reprinted with permission from Das et al., 2003).
Application of these algorithms to SAXS data on small RNA molecules was initiated by Lipfert and Doniach (2007), who showed that threedimensional density maps giving the shape outline of small RNAs and
266
Lois Pollack and Sebastian Doniach
DNAs (such as tRNA and short DNA duplexes) generated from SAXS data fitted the X-ray crystallography—derived structures at low resolution on the ˚. scale of 10–15 A In a recent study, Pollack and collaborators (Lamb et al., 2008) applied shape generation algorithms to RNA folding measurements. They investigated the limits of applicability of reconstruction methods to time-resolved data by providing examples and discussing the pros and cons of this approach, in particular addressing two primary concerns. First, signal-tonoise ratio associated with time-resolved data is much lower than that of equilibrium data. Time-resolved data are more subject to noise from parasitic scattering that can only be subtracted out in an average way than are static data for which longer exposure times allow better averaging over fluctuations in the parasitic scattering. Each individual time-resolved measurement will thus have a variable parasitic background making the analysis of multiple components contributing to the overall scattering profile more difficult. Thus, more detailed study of time-resolved intermediates in RNA conformation changes will require averaging over many time-resolved scattering runs. Second, during folding, the ensemble of states present may be more heterogeneous than in a static experiment, especially if multiple parallel pathways are present. In an attempt to reconstruct a reaction intermediate, this approach was only applied to ‘‘long lived’’ intermediates, for example, those states occurring during a plateau in the kinetic curve.
7. Concluding Remarks SAXS data have the advantage of providing information of global changes in molecular conformation. In this sense it is complementary to local probes of RNA structure such as NMR, FRET, fluorescent probes, single-molecule measurements, and hydroxyl radical footprinting. It also has the advantage of time resolution over time scales ranging from tens of microseconds to hours. The disadvantages are the fairly large quantities of sample needed to overcome radiation damage limitations and the need to use specialized synchrotron radiation facilities.
ACKNOWLEDGMENTS We acknowledge our collaborators for help and discussions, in particular: K. Andresen, Y. Bai, M. Brenowitz, R. Das, S. M. Gruner, D. Herschlag, J. Jacob, L. W. Kwok, J. S. Lamb, G .S. Maskel, T. T. Mills, I. S. Millett, S. G. J. Mochrie, B. Nakatani, V. Pande, H. Y. Park, R. Russell, J. Schlatterer, S. Seifert, I. Shcherbakova, H. Smith, M. W. Tate and P. Thiyagarajan.
Time-Resolved SAXS and RNA Folding
267
This research was generously supported by NIH grant PO1 GM066275, NASA through (NAG8-1178) and NSF through Cornell’s Nanobiotechnology Center and through MCB0347220. Computing resources were provided by the Bio-X2 computer cluster at Stanford University (NSF award CNS-0619926). Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Office of Science, and Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. CHESS is supported by the NSF and the NIH/National Institute of General Medical Sciences under award DMR-0225180.
REFERENCES Akiyama, S., Takahashi, S., Kimura, T., Ishimori, K., Morishima, I., Nishikawa, Y., and Fujisawa, T. (2002). Conformational landscape of cytochrome c folding studied by microsecond-resolved small-angle X-ray scattering. Proc. Natl. Acad. Sci. USA 99, 1329–1334. Andresen, K., Das, R., Park, H. Y., Smith, H., Kwok, L. W., Lamb, J. S., Kirkland, E. J., Herschlag, D., Finkelstein, K. D., and Pollack, L. (2004). Spatial distribution of competing ions around DNA in solution. Phys. Rev. Lett. 93, 248103. Chacon, P., Moran, F., Diaz, J. F., Pantos, E., and Andreu, J. M. (1998). Low-resolution structures of proteins in solution retrieved from X-ray scattering with a genetic algorithm. Biophys. J. 74, 2760–2775. Chen, L. L., Hodgson, K. O., and Doniach, S. (1996). A lysozyme folding intermediate revealed by solution X-ray scattering. J. Mol. Biol. 261, 658–671. Chen, L. L., Wildegger, G., Kiefhaber, T., Hodgson, K. O., and Doniach, S. (1998). Kinetics of lysozyme refolding: Structural characterization of a non-specifically collapsed state using time-resolved X-ray scattering. J. Mol. Biol. 276, 225–237. Das, R., Kwok, L. W., Millett, I. S., Bai, Y., Mills, T. T., Jacob, J., Maskel, G. S., Seifert, S., Mochrie, S. G. J., Thiyagarajan, P., Doniach, S., Pollack, L., et al. (2003). The fastest global events in RNA folding: Electrostatic relaxation and tertiary collapse of the Tetrahymena ribozyme. J. Mol. Biol. 332, 311–319. Doniach, S. (2001). Changes in biomolecular conformation seen by small angle X-ray scattering. Chem. Rev. 101, 1763–1778. Duffy, D. C., McDonald, J. C., Schueller, O. J. A., and Whitesides, G. M. (1998). Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Anal. Chem. 70, 4974–4984. Eliezer, D., Chiba, K., Tsuruta, H., Doniach, S., Hodgson, K. O., and Kihara, H. (1993). Evidence of an associative intermediate on the myoglobin refolding pathway. Biophys. J. 65, 912–917. Eliezer, D., Jennings, P. A., Wright, P. E., Doniach, S., Hodgson, K. O., and Tsuruta, H. (1995). The radius of gyration of an apomyoglobin folding intermediate. Science 270, 487–488. Fang, X. W., Littrell, K., Yang, X., Henderson, S. J., Siefert, S., Thiyagarajan, P., Pan, T., and Sosnick, T. R. (2000). Mg2þ-dependent compaction and folding of yeast tRNA (Phe) and the catalytic domain of the B-subtilis RNase P RNA determined by smallangle X-ray scattering. Biochemistry 39, 11107–11113. Henry, E. R., and Hofrichter, J. (1992). Singular value decomposition—Application to analysis of experimental-data. Methods Enzymol. 210, 129–192. Kazimirov, A., Smilgies, D. M., Shen, Q., Xiao, X. H., Hao, Q., Fontes, E., Bilderback, D. H., Gruner, S. M., Platonov, Y., and Martynov, V. V. (2006). Multilayer X-ray optics at CHESS. J. Synchrotron Radiat. 13, 204–210. Knight, J. B., Vishwanath, A., Brody, J. P., and Austin, R. H. (1998). Hydrodynamic focusing on a silicon chip: Mixing nanoliters in microseconds. Phys. Rev. Lett. 80, 3863–3866.
268
Lois Pollack and Sebastian Doniach
Kwok, L. W., Shcherbakova, I., Lamb, J. S., Park, H. Y., Andresen, K., Smith, H., Brenowitz, M., and Pollack, L. (2006). Concordant exploration of the kinetics of RNA folding from global and local perspectives. J. Mol. Biol. 355, 282–293. Lamb, J., Kwok, L., Qiu, X. Y., Andresen, K., Park, H. Y., and Pollack, L. (2008). Reconstructing three-dimensional shape envelopes from time-resolved small-angle X-ray scattering data. J. Appl. Crystallogr. 41, 1046–1052. Lipfert, J., and Doniach, S. (2007). Small-angle X-ray scattering from RNA, proteins, and protein complexes. Annu. Rev. Biophys. Biomol. Struct. 36, 307–327. Lipfert, J., Millett, I. S., Seifert, S., and Doniach, S. (2006). Sample holder for small-angle X-ray scattering static and flow cell measurements. Rev. Sci. Instrum. 77, 046108. Park, H. Y., Qiu, X. Y., Rhoades, E., Korlach, J., Kwok, L. W., Zipfel, W. R., Webb, W. W., and Pollack, L. (2006). Achieving uniform mixing in a microfluidic device: Hydrodynamic focusing prior to mixing. Anal. Chem. 78, 4465–4473. Petoukhov, M. V., and Svergun, D. I. (2003). New methods for domain structure determination of proteins from solution scattering data. J. Appl. Crystallogr. 36, 540–544. Pollack, L., Tate, M. W., Darnton, N. C., Knight, J. B., Gruner, S. M., Eaton, W. A., and Austin, R. H. (1999). Compactness of the denatured state of a fast-folding protein measured by submillisecond small-angle X-ray scattering. Proc. Natl. Acad. Sci. USA 96, 10115–10117. Russell, R., and Herschlag, D. (1999). New pathways in folding of the Tetrahymena group I RNA enzyme. J. Mol. Biol. 291, 1155–1167. Russell, R., Millett, I. S., Doniach, S., and Herschlag, D. (2000). Small angle X-ray scattering reveals a compact intermediate in RNA folding. Nat. Struct. Biol. 7, 367–370. Russell, R., Millettt, I. S., Tate, M. W., Kwok, L. W., Nakatani, B., Gruner, S. M., Mochrie, S. G. J., Pande, V., Doniach, S., Herschlag, D., and Pollack, L. (2002). Rapid compaction during RNA folding. Proc. Natl. Acad. Sci. USA 99, 4266–4271. Sandy, A. R., Lurio, L. B., Mochrie, S. G. J., Malik, A., Stephenson, G. B., Pelletier, J. F., and Sutton, M. (1999). Design and characterization of an undulator beamline optimized for small-angle coherent X-ray scattering at the advanced photon source. J. Synchrotron Radiat. 6, 1174–1184. Schlatterer, J. C., Kwok, L. W., Lamb, J. S., Park, H. Y., Andresen, K., Brenowitz, M., and Pollack, L. (2008). Hinge bending: A barrier to RNA folding. J. Mol. Biol. 379, 859–870. Sclavi, B., Woodson, S., Sullivan, M., Chance, M. R., and Brenowitz, M. (1997). Timeresolved synchrotron X-ray ‘‘footprinting’’, a new approach to the study of nucleic acid structure and function: Application to protein-DNA interactions and RNA folding. J. Mol. Biol. 266, 144–159. Segel, D. J., Bachmann, A., Hofrichter, J., Hodgson, K. O., Doniach, S., and Kiefhaber, T. (1999). Characterization of transient intermediates in lysozyme folding with timeresolved small-angle X-ray scattering. J. Mol. Biol. 288, 489–499. Seifert, S., Winans, R. E., Tiede, D. M., and Thiyagarajan, P. (2000). Design and performance of a ASAXS instrument at the advanced photon source. J. Appl. Crystallogr. 33, 782–784. Svergun, D. I. (1999). Restoring low resolution structure of biological macromolecules from solution scattering using simulated annealing. Biophys. J. 76, 2879–2886. Svergun, D. I., and Stuhrmann, H. B. (1991). New developments in direct shape determination from small-angle scattering. I. Theory and model-calculations. Acta Crystallogr. A 47, 736–744. Tsuruta, H., Brennan, S., Rek, Z. U., Irving, T. C., Tompkins, W. H., and Hodgson, K. O. (1998). A wide-bandpass multilayer monochromator for biological small-angle scattering and fiber diffraction studies. J. Appl. Crystallogr. 31, 672–682. Walther, D., Cohen, F. E., and Doniach, S. (2000). Reconstruction of low-resolution threedimensional density maps from one-dimensional small-angle X-ray solution scattering data for biomolecules. J. Appl. Crystallogr. 33, 350–363.
C H A P T E R
T H I R T E E N
2-Aminopurine as a Probe of RNA Conformational Transitions Kathleen B. Hall Contents 269 270 273 273 274 276 278 278 284 284
1. 2. 3. 4.
Introduction 2AP Structure and Photophysics RNA Oligonucleotides with 2AP Steady-State Fluorescence and RNA Folding 4.1. Group I ribozyme folding 4.2. TPP riboswitch folding 5. Time-Resolved Fluorescence Intensity Decay 5.1. The IRE hairpin RNA loop Acknowledgments References
Abstract 2-aminopurine (2AP) is a fluorescent nucleobase that provides the means to probe structure and dynamics of RNA molecules. Because 2AP can base pair with Uridine, it can replace normal A:U pairs without substantial deformation of duplexes. It is best used as a probe of ostensibly single-stranded regions: its fluorescence intensity reports on base stacking and its fluorescence decay lifetimes report on its conformational dynamics. Three examples of its use are described here, illustrating how 2AP fluorescence has been used to probe RNA folding and hairpin loop dynamics.
1. Introduction Among fluorescence probes of nucleic acid structure, dynamics, and folding, 2-aminopurine (2AP) is unique in that its introduction into an RNA can often be done with little perturbation of the RNA structure. Whereas other fluorescent nucleotide analogs have higher quantum yields than 2AP, their modifications have the potential to disrupt secondary and Department of Biochemistry and Molecular Biophysics, Washington University School of Medicine, St. Louis, Missouri, USA Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69013-3
#
2009 Elsevier Inc. All rights reserved.
269
270
Kathleen B. Hall
tertiary interactions in an RNA, making them problematic for probing RNA properties. Thus, 2AP remains the substitution of choice as a fluorescent reporter. What follows is a brief summary of the structural and spectroscopic properties of 2AP to provide a context for a discussion of how it has been and could be used to monitor RNA folding and dynamics. Several examples of its incorporation into RNA molecules are then discussed, noting how its fluorescence has been interpreted in the context of specific RNA properties. Since data analysis is a large part of fluorescence measurements, steady-state and time-resolved data are presented, including problems and pitfalls in their interpretation.
2. 2AP Structure and Photophysics The structure of 2AP allows it to base pair with both cytosine and uridine (Fig. 13.1A). While its base pairing with uridine is a wobble pair similar to a G:U pair, its pairing with cytosine is noncanonical, but both 2AP:U and 2AP:C can be accommodated in an RNA duplex. A more typical use of 2AP is as a replacement for a purine in an ‘‘unstructured’’ region of RNA such as a hairpin loop or internal bulge, where its fluorescence changes can report on conformational changes or ligand binding. This use of 2AP reflects its photophysical properties, which are completely controlled by its structural context. The fluorescence properties of free 2AP are simple. A Jablonski diagram of 2AP (Fig. 13.1B) computed with time-dependent density functional theory (TDDFT) finds a dominant singlet excited state transition from S0 to S1 at 292 nm ( Jean and Hall, 2001). In solution, the free nucleobase has a fluorescence excitation maximum of 305 nm and an emission maximum of 360 nm at pH 7. Its quantum yield is not high: 0.68 at pH 7.0 in 100 mM NaCl, 25 C. Its fluorescence lifetime in aqueous solution is 10 ns at 22 C and is described by a single exponential decay. When 2AP is incorporated into an RNA molecule, its photophysics becomes extremely complex ( Jean and Hall, 2001; Rachofsky et al., 2001). Two different environments give rise to two different fluorescent states: (1) when 2AP stacks with another nucleobase, its fluorescence is severely quenched; and (2) when 2AP moves independently of flanking bases, its fluorescence decay becomes multiexponential. The physical basis of these different phenomena relates to the interaction of 2AP with other nucleobases. Quenching of 2AP fluorescence is most dramatic when 2AP is within a duplex. In a series of computational calculations of 2AP electronic properties, we have arrived at a picture of stacked bases that explains, at least in part, the origin of 2AP fluorescence quenching. The extent of quenching
A
N N N
N N H H
N N N
2-Aminopurine
N N H H
B
H H N N O
S2
2-Aminopurine : Cytosine S2
S1
296 nm S1 Dark (0.094)
N
O
N N N
N N H H
H N O
2-Aminopurine : Uracil
N
O O
N N N
N H
H N O
S2
S1
331 nm (0.008)
S0 Guanine : Uracil
2APpT B-form
275 nm (0.002)
292 nm (0.12)
2AP
288 nm (0.023) 301 nm (0.034)
2APpT A-form
N
N H H
Figure 13.1 (A) Structure of 2-aminopurine and its base pairing partners. For reference, a G:U wobble pair is shown. (B) Jablonski diagrams indicating the electronic transitions of 2AP alone and in an A-form and B-form stack with thymine.
272
Kathleen B. Hall
depends on the identity of the stacking nucleobase(s) on the 50 and/or 30 side of 2AP, the relative geometry of the stacked bases, and the stability of ˚ apart the stacking. In a standard RNA A-form duplex, the bases are 2.8 A (the helical rise/residue) and the spatial overlap of their aromatic rings is extensive; far more so than the overlap of stacked bases in a B-form duplex. The close stacking of these conjugated p-systems leads to mixing of their molecular orbitals, creating what we termed a ‘‘supermolecule’’ composed of 2AP and its nearest neighbors. These supermolecules have unique electronic states that are neither the sum of the monomers nor perturbations of the monomers. When 2AP quenching results from supermolecule formation, the mechanisms of quenching depend on the geometry and identity of the flanking bases. An example of the electronic properties of 2AP stacked with Thymine in an A-form and B-form geometry illustrates these differences (Fig. 13.1B). Jablonski diagrams derived from TDDFT calculations for 50 -2APpT show how geometry controls excited state transitions. In the A-form dimer, the S0 (ground state) to S1 (first excited state) and S0 to S2 (second excited state) transitions have nearly equal and weak intensities (oscillator strengths of 0.034 and 0.023; compare to the 2AP S0 to S1 transition with an oscillator strength of 0.12) with an energy separation of only 13 nm. In the B-form dimer, there is a strong S0 to S2 transition which is 2AP-like, but the presence of a lower S1 state means that internal conversion will rapidly occur (Kasha’s rule). In the A-form dimer, fluorescence emission intensity will be reduced by the weak transition intensities; in the B-form dimer, quenching results from nonradiative energy transfer to the (dark) S1 state. Quenching occurs in both cases, but if the excited state P lifetimes t ¼ 1/(krad þ knrad) were measured, they would differ: B-form geometry will lead to a shorter lifetime than the A-form. The measured lifetime(s) of 2AP fluorescence are an important source of information on the structure and dynamics of its environment. Although the free 2AP has a single excited state decay time (lifetime), 2AP in an RNA can have several. In the above dimers, two different static structures give rise to two different fluorescence lifetimes, but in the context of an RNA in solution, the bases can sample several conformations. Even in a duplex, the bases flip in and out at some frequency determined by the duplex structure, their hydrogen bonding partner and their flanking bases (Gue´ron et al., 1987). The shortest decay time measured is limited by the resolution of the instrument (the instrument response function, IRF); using time-correlated single photon counting instruments, measured lifetimes have ranged from 0.3 to 10 ns, but the number of components is also finite: typically no more than four decay times are realistically fit to the data. There is often not a unique interpretation of fluorescence lifetimes in terms of RNA structure, and to develop an accurate description of structure and dynamics, other experimental variables must be included.
2-Aminopurine as a Probe of RNA Conformational Transitions
273
2AP photophysics is determined by its environment, which makes 2AP a useful probe of RNA structure and dynamics but because its properties depend on its environment, the incorporation of 2AP into an RNA must be done cautiously, preferably with prior knowledge of the site—duplex or single strand, tight turn or floppy loop, etc. With that caveat in mind, 2AP has been used in several ways to provide unique information on the folding and dynamics of both small and large RNAs.
3. RNA Oligonucleotides with 2AP All RNAs that contain 2AP must be chemically synthesized. The phosphoramidite form of 2AP is commercially available from Glen Research (Sterling, VA, USA) for in-house synthesis. Dharmacon (Boulder, CO, USA) and IBA (Go¨ttingen, Germany) do chemical synthesis of RNAs with 2AP. Both suppliers offer protected and deprotected forms of the oligos; if money is no object, these suppliers will purify the RNA by denaturing gel electrophoresis or HPLC. RNAs with 2AP require no special handling beyond normal RNA care. One potential limitation of 2AP fluorescence studies is the requirement for a UV-light source for excitation. For steady-state fluorescence measurements, this is no problem since all fluorometers can excite at 305 nm. For lifetime measurements, the light source is more challenging since diode lasers currently cannot provide this wavelength and tunable titanium– sapphire (Ti:S) lasers must be tripled. Our in-house time-correlated single photon counting (TCSPC) instrument uses a 76-MHz mode-locked Ti:S oscillator (Coherent Mira 900F) with a Coherent Verdi Nd:YVO4 pump laser. The output frequency is tripled to 300 nm with a UOplaz tripler for 2AP excitation. The pulse rate is controlled with a NEOS pulse picker, since the 76 MHz pulses from the Ti:S laser are too rapid for 2AP excitation/decay. If 2AP in an RNA has an isotropic decay lifetime near the 10 ns lifetime of the free 2AP, the 13 ns time interval between laser pulses will not allow complete decay of the signal; we typically reduce the frequency to 7.6 MHz.
4. Steady-State Fluorescence and RNA Folding Among studies of RNA folding using 2AP fluorescence, only two are used here as examples of what can be learned by steady-state fluorescence. These are large RNAs (150–200 nt) that have many duplex regions, loops, and internal bulges that adopt a unique active tertiary fold in the presence of ligand.
274
Kathleen B. Hall
4.1. Group I ribozyme folding The small Azoarcus ribozyme from its pre-tRNA(Ile) has many tertiary interactions in common with larger members of the Group I intron family, including a G1A2A3A4 tetraloop (Ikawa et al., 1999) that docks with the corresponding tetraloop receptor (Fig. 13.2). The GAAA tetraloop structure has been solved by NMR ( Jucker et al., 1996) where it was found to have an extensive hydrogen bonding network among bases and riboses in
B
3.2
0.00 0.25 0.50 1.00 2.00 5.00 20.0
(Mg2+) mM
2.8 2.4
Fraw
2
A
1.6 1.2
A3 A2
0.8 0.4 0
340
360
400 380 Wavelength (nm)
420
1 0.8
Frel
0.6 0.4 0.2 0 0.1
1
10
(Mg2+) in mM
Figure 13.2 (A) Structure of the GAAA tetraloop (pdb 1ZIF). (B) Steady-state 0 fluorescence emission spectra (Fraw) of 5 GA#2APA as a function of Mg2þ concentration, and a plot of the change in relative intensity (Frel) with the fit to a three-state model of folding (courtesy of S. Woodson).
2-Aminopurine as a Probe of RNA Conformational Transitions
275
the loop. A2 is stacked over A3 which is stacked over the sheared G1:A4 base pair. A3 is hydrogen bonded to the 20 -OH ribose of G1 through its N6/O6 while A2 is unconstrained by any hydrogen bonds. In the GAAA loop in the Azoarcus ribozyme, A3 was replaced by 2AP. To introduce 2AP into the GAAA site, the Woodson lab (Chauhan et al., 2009) made an unusual construct of the ribozyme that uses two RNAs: a 187-nt body that contains the tetraloop receptor and a 15-nt strand comprising the 30 end of the ribozyme. The short RNA sequence 50 2AP 291Agccacacaaaccg is hybridized to the large RNA (lower case indicates hybridizing sequences) and leads to formation of a nicked 0 0 cGA#2APAg tetraloop from the cGA1903 . . .5 2AP191Ag juxtapositions at the ends of the stem. Although the tetraloop is nicked [GA#2APA], the ribozyme is active and its folding thermodynamics are normal. However, using 2AP fluorescence as a sensitive monitor of formation of the tertiary interaction, the data showed that the kinetics of folding was slowed (Chauhan et al., 2009). Tertiary folding of the Azoarcus ribozyme requires Mg2þ ions (Tanner and Cech, 1996) and includes formation of a GAAA:receptor interaction (Davis et al., 2005). The introduction of 2AP provided a probe of the folding as a function of Mg2þ, and the data are quite intriguing (Chauhan et al., 2009). As shown in Fig. 13.2, the fluorescence intensity of 2AP first increases at Mg2þ concentrations up to 0.5 mM, then progressively decreases until the final 20 mM Mg2þ point where the RNA has acquired its stable tertiary structure. 2AP fluorescence intensity thus identifies two transitions, which are described by a three-state model: U ! IC ! N. The ‘‘Unfolded’’ state has secondary structure but no tertiary contacts; the ‘‘Intermediate’’ state(s) have core helices that are collapsed into a compact form but lack ‘‘Native’’ contacts. The initial increase in 2AP fluorescence corresponds to the U ! IC transition, but the subsequent decrease in fluorescence shows that formation of the final interaction between the GAAA loop and its docking site occurs later in the folding process. An imaginative interpretation of the 2AP environment at equilibrium could reflect its structure in the RNA. For example, when the 2AP-oligo is first hybridized to the long strand prior to the start of the Mg2þ titration, 2AP stacks at the terminus of the duplex which quenches its fluorescence. As Mg2þ is added, the GAAA loop structure forms, the tertiary structure collapses (IC state), and 2AP fluorescence increases which suggests that its stacking is reduced. If 2AP were securely stacked over the G1A4 base pair as it is in the typical GAAA tetraloop, then its fluorescence should be quenched, so these data suggest that the tetraloop structure is not static. As more Mg2þ is added, there is a cooperative conformational change to the Native state, which involves formation of the GAAA:receptor interaction that again quenches 2AP fluorescence. In this interaction, 2AP is confined to a stacked conformation that leads to efficient quenching through ‘‘supermolecule’’ formation.
276
Kathleen B. Hall
The interpretation of 2AP fluorescence in terms of the folding pathway of the ribozyme indicates that there are at least three states in the process (Fig. 13.2). Stopped-flow fluorescence experiments over the folding timecourse of 10 s show that the initial increase in 2AP fluorescence occurs rapidly (kobs ¼ 106 s 1) and is followed by three slower phases that greatly decrease its fluorescence. It seems clear that the nicked state of the GAAA tetraloop has slowed the kinetics of folding to separate the stages of the path and reveal intermediates (Chauhan et al., 2009). Some of these intermediates are misfolded, but others are on the normal folding pathway. The broken GAAA tetraloop allows discrimination of these steps in the pathway, through a mechanism that could involve excessive nucleobase flexibility (the stacking is not sufficient to rigidify their orientation) that requires more time to make contacts within the receptor site. The conformational heterogeneity of the [GA#2APA] loop could be assessed as a function of Mg2þ conformation using time-resolved fluorescence measurements to determine the number of 2AP decay lifetimes. This dynamic information on the loop motions could contribute to the model for the general mechanism of this particular tertiary interaction.
4.2. TPP riboswitch folding When the Escherichia coli thiamine pyrophosphate (thiM TPP) riboswitch binds to TPP, it undergoes a conformational change to occlude the Shine– Delgarno ribosome binding site, effectively blocking translation of the downstream mRNA (Winkler et al., 2002). In a technical tour-de-force, the Micura lab (Lang et al., 2007) introduced 2AP at seven positions in the aptamer domain where the ligand is bound and three positions in the complete 151 nt aptamer/expression platform (Fig. 13.3), with the goal of using 2AP fluorescence to monitor binding of TPP and folding of the RNA. Working from crystal structures of the aptamer domain (Edwards and Ferre´-D’Amare´, 2006; Serganov et al., 2006; Thore et al., 2006) RNAs were synthesized chemically with a single 2AP located at sites predicted to perturb neither the tertiary fold nor TPP binding. All RNAs were constructed by enzymatic splint ligation (Lang et al., 2007; Moore, 1999) of chemically synthesized RNA strands. In a splint ligation, a DNA template complementary to both the 30 -OH and 50 -P sequences of the RNAs to be joined is hybridized to the RNAs. The junction is closed by T4 DNA ligase and the ligated RNA is purified. The TPP riboswitch RNAs were purified by HPLC in 6 M urea at 80 C and the products verified by mass spectroscopy. Steady-state 2AP fluorescence was used to monitor TPP binding and riboswitch folding. All of the 2AP-containing aptamer domains underwent a change in fluorescence intensity (either an increase or decrease) when TPP bound; in the full-length riboswitch, only one site was responsive to
277
2-Aminopurine as a Probe of RNA Conformational Transitions
thiM 151
A
B
U62AP 1.2
C
3
0.6 0.4
3⬘
0
TPP
2 1 0
0.2
P1 5⬘ A12AP
4
0.8 F/a.u.
U62AP
1 (F−F0)/(Ff −F0) (a.u.)
A128AP
P8
0
350
400 450 l/nm
5 10 15 Concentration TPP (µM)
20
U62AP 151 nt ‘full length riboswitch’ 25 ⬚C 0.8
22 18 14 10 6
5 eq TPP 0
10 20 30 40 50 60 Time (s)
Observed rate k⬘ (s −1)
Fluorescence (a.u.)
26
0.6 0.4 0.2 0.0
k25⬚ C = 8.13 ± 0.16 ⫻ 10−4 M−1 s−1
0 1 2 3 4 5 6 7 Concentration TPP cTPP (µM)
Figure 13.3 (A) Schematic diagram of the thiM ribozyme, showing the positions of 2AP insertions. (B) Fluorescence emission intensity of 2AP62 as a function of added TPP. (C) Stopped-flow fluorescence of 2AP62 riboswitch upon addition of 5 TPP. [RNA] ¼ 0.3 mM in 50 mM KMOPS, 100 mM KCl, 2 mM MgCl2, 25 C, pH 7.5. Data are described by a first-order exponential, and the apparent rate constants as a function of added TPP are shown to the right (Lang et al., 2007).
binding/folding; the other two sites showed virtually no change in fluorescence. Binding isotherms were fit by a 1:1 stoichiometry and KD,app ranging from 250 to 400 nM at 25 C. The U62-to-2AP62 substitution was effective in both the aptamer and the full-length transcript, so it is discussed here in more detail. As shown in Fig. 13.3, U62 is within an internal bulge in the aptamer. In a cocrystal structure, it is extruded from the body of the RNA and is
278
Kathleen B. Hall
completely solvent-exposed. In the free aptamer, 2AP62 has low fluorescence intensity, consistent with a predominantly stacked conformation. Upon TPP addition, 2AP62 fluorescence increases significantly, consistent with the cocrystal structure. The steady-state fluorescence intensity was measured as a function of time after addition of TPP in a stopped-flow instrument: 2AP62 response allows calculation of a rate constant for binding/ conformational rearrangement of 1 105 M 1 s 1 (Fig. 13.3). Analogous experiments were done for each of the riboswitch constructs to construct a map of the rate constants as a function of position. From these data, the authors constructed a model of the folding/binding process of the RNA. The introduction of 2AP in so many different structural contexts in the TPP riboswitch provides an opportunity to use it as a probe of the folding at both local and global levels. From the perspective of each local structure and dynamics, what is lacking is time-resolved data that would show how the 2AP is contained within the structure, and by implication how flexible that particular region of the RNA is in the free and bound states. The intrinsic flexibility of an RNA can be a critical part of its function, but is a parameter that is often difficult to assess. With the introduction of 2AP, at least part of that description could be obtained.
5. Time-Resolved Fluorescence Intensity Decay This chapter might be viewed as a plea for the measurement of timeresolved fluorescence of 2AP when it is introduced into an RNA. The resulting picture of the flexibility of the nucleobase in different chemical and physical contexts would be extremely useful as we build our understanding of the dynamics of RNAs. Measurements of fluorescence intensity can be excellent reporters of RNA folding, as the above examples show, but especially when fluorescence is quenched, the most popular interpretation is that the base is stacked with the attendant implication that its position is static. However, if 2AP remains flexible but spends part of its time stacked with another nucleobase, the solution average fluorescence intensity will still be reduced. Only by measurement of the decay lifetimes of 2AP will its true dynamics be discerned.
5.1. The IRE hairpin RNA loop This simple RNA stemloop serves as an illustration of how both steady-state and time-resolved fluorescence together provide a detailed description of the properties of an RNA (Hall and Williams, 2004). The sequence of the iron response element (IRE) RNA hairpin loop [C6A7G8U9G10C11] is phylogenetically conserved (numbered from our construct). Cytidine 6 and
279
2-Aminopurine as a Probe of RNA Conformational Transitions
Guanosine 10 make a hydrogen-bonded base pair, but other nucleobase positions are poorly constrained by NMR data (Addess et al., 1997; Laing and Hall, 1996; McCallum and Pardi, 2003). To better characterize the loop structure and nucleotide motions, 2AP replaced A7 and G8 in the IRE (Fig. 13.4). Although neither A7 nor G8 is constrained by hydrogen bonding in the loop, their structure and dynamics are quite different, as the 2AP fluorescence clearly showed. Steady-state fluorescence measurements showed that 2AP7 and 2AP8 spectra have identical emission and excitation maxima, indicating their environments were similar, but the fluorescence intensity of 2AP7 is only a quarter that of 2AP8 at 20 C (Fig. 13.4). NMR structural data indicated that A7 was stacked over the C6:G10 base pair at 20 C which could account for 2AP7 quenching. Indeed, 2AP7 fluorescence
2AP8
Fluorescence intensity (a.u.)
2.5⫻105
2AP7
2⫻105
IRE P8
1.5⫻105
1⫻105
5⫻104 IRE P7 0 200
250
300
350
400
450
500
Wavelength
Fluorescence intensity
3.5⫻105 4C 15C 25C 35C
3⫻105
2.5⫻105 2⫻105
1.5⫻105 1⫻105 5⫻104 0
340
360
380
400
Wavelength (nm)
420
440
Fluorescence intensity (a.u.)
1.2⫻105
4⫻105
1⫻105 8⫻104 6⫻104 4⫻104 2⫻104
0
340
360
380
400
420
440
Wavelength (nm)
Figure 13.4 Structure of the IRE RNA with A7 and G8 indicated (Hall and Williams, 2004). Top right: fluorescence excitation and emission spectra of IRE 2AP7 and IRE 2AP8. Bottom: fluorescence emission intensity of each IRE as a function of temperature, 4 mM RNAs, 30 mM NaCl, 10 mM potassium phosphate, pH 7.0, 20 C. The buffer baseline was subtracted from each spectrum (the Raman line is at 350 nm).
280
Kathleen B. Hall
intensity increases with temperature from 4 to 35 C, consistent with a disruption of its stacking interactions. Since the hairpin melting temperature is 60 C in these conditions, the 2AP fluorescence is reporting on changes in the loop conformation. The data are consistent with stacking of 2AP7 that leads to quenching of its fluorescence due to formation of an electronic ‘‘supermolecule’’ state which is disrupted at higher temperatures. In contrast, the fluorescence emission intensity of 2AP8 decreases from 4 to 35 C. This behavior is similar to that of free 2AP, which undergoes quenching of its fluorescence due to collisions with solvent; higher temperatures increase the efficiency of collisional quenching. However, another interpretation of the data is that as the loop conformation becomes more flexible, the 2AP8 nucleobase spends a portion of its time stacked with the A7 nucleobase, which leads to transient formation of the electronic supermolecule and so to quenching. To further characterize the mobility of the IRE loop, time-resolved isotropic fluorescence emission decay components of the IRE RNAs were determined as a function of temperature. Some details of the measurements and data assessment will be necessary here to appreciate both the utility of the information and caveats about its literal interpretation. Considering first the TCSPC instrument itself, some uncertainty in the measurements arise from its intrinsic parameters. With 300 nm incident light, the IRF of the photomultiplier tube ranged from 190 to 276 ps full-width at half-height (FWHH). The width of the IRF and the time resolution (32.5 ps/channel) limit the short components that can be reliably extracted from the fit, and certainly those <200 ps will have large errors on their amplitudes and lifetimes. Fluorescence emission decay components as short as 9–20 ps (Larsen et al., 2001) and 30–70 ps (Guest et al., 1991) (and much shorter by Wan et al., 2000) have been measured for 2AP in a stacked conformation, but in our instrument, a fit to such a short lifetime would be inaccurate. Fluorescence decay data were processed with FluoFit (PicoQuant) (Globals is another popular analysis package). This software package provides many options for data acquisition and processing, including decay lifetime distributions and amplitudes, calculated using nonlinear least squares fitting with parameter variation using Marquardt–Levenberg or Monte Carlo sampling. Error testing can be done using several algorithms, including assessment of reduced w2, distribution and autocorrelation of weighted residuals, and our preferred method: support plane analysis and bootstrap. It also allows for global analysis. Rigorous and critical evaluation of fitting parameters is essential for these experiments, since there is seldom a unique fit to the data when several lifetimes are present. An example of a fit of 2AP8-IRE data using FluoFit is given in Fig. 13.5, and several features need to be noted. The top curve is the 2AP8-IRE
281
2-Aminopurine as a Probe of RNA Conformational Transitions
923
308
1539
2154 No. 1 2 3
2770 Decay (ns) 0.670 3.365 6.879
Ampl./% 46.1 32.0 22.0
4
Intensity (counts)
10
103
2AP-IRE
102
101
IRF
Resids.
3.0 0.0
−3.0
0.01
0.02
0.03
0.04 0.05 0.06 Time (ns ⫻ 1⫻103)
Fitting function: 3-Exponential, reconvolution No. Decay (ns) Error decay 0.041 1 0.670 0.317 2 3.365 0.093 3 6.879 Av. 4.911 X2(reduced):
0.07
Amplitude 48143.241 33398.510 22932.099
0.08
0.09
0.10
rel. amplitude 46.1% 32.0% 22.0%
1.936
Figure 13.5 Time-resolved fluorescence decay spectrum of 2AP8-IRE at 27 C in 30 mM NaCl, 10 mM potassium phosphate, pH 7.0.
isotropic decay, which has returned to baseline by 70 ns. For these experiments, the laser repetition rate has been divided by 10 so that every tenth pulse enters the sample (the other pulses are ‘‘dumped’’ by the pulse picker). In these experiments, the pulse discrimination was not perfect, and the spike observed at 18 ns results from the N þ 1 pulse. Also note the bump in the spectra at 66 ns; it is an instrument artifact that we cannot eliminate but does not affect the data. There are two decays shown: the top intensity trace is the decay of 2AP8 in the IRE; the bottom is the IRF. The IRF is collected from scattered light
282
Kathleen B. Hall
(a solution of dilute Ludox in water) and measures the effective deadtime of the instrument. In this experiment, FWHH of the IRF is 238 ps. The FluoFit software convolutes the IRF (in the frequency domain as a fast Fourier transform) with a (multi)exponential model containing guesses of both amplitude(s) and lifetime(s). Deviations of the resultant curve inform the next input guesses, and the iterative process is repeated until the fitting (hopefully) converges. The goodness of fit is then evaluated. In our experiments, the best fit was judged by the shape of the support planes and the reduced w2. As this example illustrates, the reduced w2 is 1.9, which is reasonable. As the residuals show, there is a spike at the initial value and a bump at 66 ns, and some oscillations near the beginning of the curve. The fitting is therefore not perfect, and most of the error comes from the initial points. In fitting the data, the coincidence of the IRF and the beginning of the decay are extremely sensitive; fitting should include the initial rise and decay of the intensity. The wider the IRF, the more difficult this becomes and the less accurate the fit. In practice, we have found that the support plane analysis is the most useful metric for judging goodness of fit. The support planes will show how the errors in amplitude and decay time are distributed; they are not necessarily symmetric errors. A well-defined value will have a smooth parabolic fit around the minima, indicating it is well-determined. A poor value could have unbounded errors indicating that the values are not unique. A fit can have a reasonable w2 but have poorly defined limits on the values of amplitude or decay time if the data are not of sufficient quality (perhaps due to instrument parameters such as too rapid excitation pulses or low signal intensity). The issue of a unique fit to a multiexponential decay is not a trivial one, and the mathematics of fitting are complex. The experimentalist has the responsibility for judging the goodness of fit and for describing the reliability of the reported data. Spending a few hours fitting the data should be sufficient to inspire caution and a rueful perspective on what is fact and what is fiction. Despite the caveats, time-resolved fluorescence made a significant contribution to the description of the loop motions in the two 2AP-IRE RNAs (Hall and Williams, 2004). Using two labeled RNAs and comparing steadystate and time-resolved data as a function of temperature shows clear differences between the two purine positions in the loop. Properties of 2AP8-IRE will be used as an example here. Figure 13.5 shows the decay of 2AP8-IRE at 27 C, fit to a three-exponential decay function. The longest lifetime of 6.9 0.1 ns suggests that the base spends time (20%) unstacked (free 2AP has a decay lifetime of 10–11 ns at 20 C). A second decay component has an amplitude of 32% and a lifetime of 3.3 0.3 ns. This lifetime should be viewed as an average of unresolved lifetimes that arise from different structural contexts of the 2AP in the RNA, keeping in mind that lifetimes differing by less than a factor of two will not be resolved in these experiments (Lakowicz, 2006). Possible 2AP conformations that
2-Aminopurine as a Probe of RNA Conformational Transitions
283
result in this timescale of motion include exchange between a stacked and unstacked conformation occurring on the ns timescale (dynamic motion), or the opposite situation: a weak interaction between two bases at a static distance from each other, or with a relative orientation that reduces their shared p electron density. Finally, the shortest lifetime of 2AP8, 0.6 ns, has about 50% amplitude, indicating that (at this temperature) the base is sampling conformations that lead to nonradiative charge transfer. This 600 ps lifetime is close to the resolving power of the instrument and should have the largest error. Indeed, the support plane of the decay time shows that the lower bound is poorly determined, despite the deceptively small error from the fit. 2AP lifetimes ranging from 200 to 400 ps have been reported by Nordland et al. (1989) in DNAs and seem to be characteristic of transient stacking. 5.1.1. Time-resolved fluorescence of the GA#2APA tetraloop Were these data available, what would they show? If the loop is not formed in the absence of Mg2þ and 2AP is stacked on the end of the duplex, then 2AP should exhibit one predominant very short lifetime as a result of stable supermolecule formation. In low concentrations of Mg2þ, its lifetime should increase as it becomes part of the broken loop structure and loses its stacking. If its orientation is not tightly constrained, it will have more than one lifetime (perhaps similar to that of 2AP8 in the IRE). At higher concentrations of Mg2þ, it will become more constrained as the global structure collapses and the tetraloop docks with the receptor. Its decay could be reduced from three or four components to two or even only one; that information would reveal how flexible it remains as part of the tertiary interaction. 5.1.2. Time-resolved fluorescence of the 2AP62 in the riboswitch What could these data reveal about the folding of the riboswitch? Although the fluorescence of 2AP62 is quenched before TPP is bound, it is unlikely to be rigidly stacked in the internal bulge. Measurements of its lifetime(s) would provide a description of its mobility and by implication, the flexibility of the bulge. 2AP62 becomes much more fluorescent when TPP binds, indicating that it has been extruded from the duplex. Time-resolved fluorescence data (and especially time-resolved anisotropy, which is another useful measurement) would show if 2AP62 simply sticks out of the structure or whether it undergoes conformational exchange. Again, these data provide information on the dynamics of the larger RNA to show how much flexibility it retains upon formation of its tertiary structure. In summary, 2AP incorporation into an RNA can be used to map its folding pathway, its kinetics of local and global folding, and its intrinsic dynamics. It is uniquely able to report on RNA properties since it fits tidily into secondary and tertiary structures without the structural perturbations
284
Kathleen B. Hall
that are an inevitable consequence of other fluorescent base analogs with their additional chemical substituents. Collecting steady-state and timeresolved fluorescence measurements together provides a more accurate description of the RNA that helps to interpret its behavior.
ACKNOWLEDGMENTS This work was supported in part by the NIH (R01 GM077231). Thanks to Professor Sarah Woodson for Fig. 13.2.
REFERENCES Addess, K. J., Basilion, J. P., Klausner, R. D., Rouault, T. A., and Pardi, A. (1997). Structure and dynamics of the iron responsive element RNA: Implications for binding of the RNA by Iron regulatory binding proteins. J. Mol. Biol. 274, 72–83. Chauhan, S., Behrouzi, R., Rangan, P., and Woodson, S. A. (2009). Structural rearrangements linked to global folding pathways of the Azoarcus Group I ribozyme. J. Mol. Biol. 386, 1167–1178. Davis, J. H., Tonelli, M., Scott, L. G., Jaeger, L., Williamson, J. R., and Butcher, S. E. (2005). RNA helical packing in solution: NMR structure of a 30 kDa GAAA tetraloopreceptor complex. J. Mol. Biol. 351, 371–382. Edwards, T. E., and Ferre´-D’Amare´, A. R. (2006). Crystal structures of the thi-box riboswitch bound to thiamine pyrophosphate analogs reveal adaptive RNA-small molecule recognition. Structure 14, 1459–1468. Gue´ron, M., Kochoyan, M., and Leroy, J. L. (1987). A single mode of DNA base pair opening drives imino proton exchange. Nature 328(6125), 89–92. Guest, C. R., Hochstrasser, R. A., Sowers, L. C., and Millar, D. P. (1991). Dynamics of mismatched base pairs in DNA. Biochemistry 30, 3271–3279. Hall, K. B., and Williams, D. J. (2004). Dynamics of the IRE hairpin loop probed by 2-aminopurine fluorescence and stochastic dynamics simulations. RNA 10(1), 34–47. Ikawa, Y., Naito, D., Aono, N., Shiraishi, H., and Inoue, T. (1999). A conserved motif in group IC3 introns is a new class of GNRA receptor. Nucleic Acids Res. 27, 1859–1865. Jean, J. M., and Hall, K. B. (2001). 2-Aminopurine fluorescence quenching and lifetimes: Role of base stacking. Proc. Natl. Acad. Sci. USA 98, 37–41. Jucker, F. M., Heus, H. A., Yip, P. F., Moors, E. H. M., and Pardi, A. (1996). A network of heterogeneous hydrogen bonds in GNRA tetraloops. J. Mol. Biol. 264, 968–980. Laing, L. G., and Hall, K. B. (1996). A model of the iron responsive element RNA hairpin loop structure determined from NMR and thermodynamic data. Biochemistry 35, 13586–13594. Lakowicz, J. R. (2006). Principles of Fluorescence Spectroscopy. 3rd edn. Kluwer/Plenum Academic Press. Lang, K., Rieder, R., and Micura, R. (2007). Ligand-induced folding of the thiM TPP riboswitch investigated by a structure-based fluorescence spectroscopic approach. Nucleic Acids Res. 35, 5370–5378. Larsen, O. F. A., van Stokkum, I. H. M., Gobets, B., van Grondelle, R., and van Amerongen, H. (2001). Probing the structure and dynamics of a DNA hairpin by ultrafast quenching and fluorescence depolarization. Biophys. J. 81, 1115–1126.
2-Aminopurine as a Probe of RNA Conformational Transitions
285
McCallum, S. A., and Pardi, A. (2003). Refined solution structure of the iron-responsive element RNA using residual dipolar couplings. J. Mol. Biol. 326, 1037–1050. Moore, M. J. (1999). Joining RNA molecules with T4 DNA ligase. Methods Mol. Biol. 118, 11–19. Nordland, T. M., Andersson, S., Nilsson, L., Rigler, R., Grasland, A., and McLaughlin, L. W. (1989). Structure and dynamics of a fluorescent DNA oligomer containing the EcoR1 recognition sequence: Fluorescence, molecular dynamics, and NMR studies. Biochemistry 28, 9095–9103. Rachofsky, E. L., Osman, R., and Ross, J. B. A. (2001). Probing structure and dynamics of DNA with 2-aminopurine: Effects of local environment on fluorescence. Biochemistry 40, 946–956. Serganov, A., Polonskaia, A., Phan, A. T., Breaker, R. R., and Patel, D. J. (2006). Structural basis for gene regulation by thiamine phyrophosphate-sensing riboswitch. Nature 441, 1167–1171. Tanner, M., and Cech, T. (1996). Activity and thermostability of the small self-splicing group I intron in the pre-tRNA(Ile) of the purple bacterium Azoarcus. RNA 2, 74–83. Thore, S., Leibindgut, M., and Ban, N. (2006). Structure of the eukaryotic thiamine pyrophosphate riboswitch with its regulatory ligand. Science 312, 1208–1211. Wan, C., Fiebig, T., Schiemann, O., Barton, J. K., and Zewail, A. H. (2000). Femptosecond direct observation of charge transfer between bases in RNA. Proc. Natl. Acad. Sci. 97, 14052–14055. Winkler, W., Nahvi, A., and Breaker, R. R. (2002). Thiamine derivatives bind messenger RNAs directory to regulate bacterial gene expression. Nature 419, 952–956.
C H A P T E R
F O U R T E E N
Fluorescence Polarization Anisotropy to Measure RNA Dynamics Xuesong Shi* and Daniel Herschlag† Contents 1. Introduction 2. General Information of FPA Measurement 2.1. L-format and T-format 2.2. Determining anisotropy and minimizing background signal 2.3. Choosing a wavelength for anisotropy measurement 2.4. Testing polarizer alignment 2.5. Factors that influence anisotropy 3. Choice of FPA Probes 4. Measuring FPA in a Simple Duplex, using an 11mer Control RNA Duplex as an Example 4.1. Sequence design 4.2. Sample preparation 4.3. FPA measurement 5. Use of FPA to Study Helical Dynamics of RNA, with a Junction Model Construct as an Example 5.1. Assembling the constructs without LacI 5.2. Lac Repressor (LacI) 5.3. Assembling model constructs with LacI and carrying out FPA measurements 6. Use of FPA to Study Helical Dynamics in a Complex RNA, with the Tetrahymena Group I Intron Ribozyme as an Example 7. Salt Dependence and Normalization of FPA with a Short Control Duplex Acknowledgments References
* {
288 289 289 289 289 290 290 291 292 292 292 293 294 294 296 298 299 301 301 301
Department of Biochemistry, Stanford University, Stanford, California, USA Departments of Biochemistry and Chemistry, Stanford University, Stanford, California, USA
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69014-5
#
2009 Elsevier Inc. All rights reserved.
287
288
Xuesong Shi and Daniel Herschlag
Abstract RNA requires helical motion to fold and carry out its function. As RNA helical motion occurs on the nanosecond timescale, the timescale probed by fluorescence dyes, fluorescence polarization anisotropy (FPA) is a simple, yet powerful, technique to study helical dynamics in RNA. With the recent development of several fluorescent base analogs that have a nanosecond timescale lifetime in a duplex, FPA has begun to be used for characterizing RNA dynamics. Using the probe 6-methylisoxanthopterin (6-MI) as an example, we describe the procedure for carrying out FPA experiments on model oligonucleotide systems and in a complex RNA, the Tetrahymena group I intron. For smaller RNA systems, isolating the motion of the target helix from the overall tumbling of the whole RNA system is necessary, and nucleic acids binding proteins can be incorporated into the RNA system to increase the overall size of the system, slow the overall tumbling, and thereby reduce the anisotropy contribution from the overall tumbling to negligible. The procedure for incorporating one such protein, the Lac Repressor, is given as an example.
1. Introduction Like other macromolecules, RNA is dynamic. The process of folding into a structured RNA involves dynamic rearrangement of RNA helices, and many RNAs function via a series of conformational transitions. Thus, measurement of the dynamics of individual helices will be required to fully understand RNA folding and function. Additionally, dynamic information can also be used to provide information about local structural features. As the movements of individual helices are often on the nanosecond timescale, fluorescence polarization anisotropy (FPA) (Bucci and Steiner, 1988; Duhamel et al., 1996; LiCata and Wowor, 2008; Thomas et al., 1980) is a natural choice for studying RNA dynamics; FPA measures the rate of depolarization of a fluorophore during its lifetime, which is often in the low nanosecond regime. Despite its potential, FPA has been rarely used for studying RNA dynamics, largely due to a lack of suitable probes. With the recent development of several new fluorescent base analogs, wider application of the FPA technique for measuring RNA dynamics has become possible (Shi et al., 2009). Besides FPA, NMR and EPR have both been successfully used in characterizing RNA dynamics on the nanosecond timescale (Edwards and Sigurdsson, 2007; Grant et al., 2009; Zhang et al., 2006, 2007). The fluorescence-based FPA methods complement NMR and EPR well in that FPA is not limited by the size of the RNA being studied and does not require large amount of samples as NMR, is a more direct readout of dynamics, involves relatively inexpensive equipment, and is potentially applicable to transient intermediates due to its high time resolution.
Fluorescence Polarization Anisotropy to Measure RNA Dynamics
289
2. General Information of FPA Measurement 2.1. L-format and T-format Fluorescence anisotropy is normally measured as r¼
IVV ðIHV =IHH ÞIVH : IVV þ 2ðIHV =IHH ÞIVH
ð14:1Þ
In Eq. (14.1), I is fluorescent intensity; the subscript letters, V for vertical and H for horizontal, represent the polarization direction of the two polarizers on the excitation and emission light path, respectively; and the ratio, IHV/IHH, calibrates for the difference in the emission channel’s sensitivity towards vertical and horizontal polarized components. Anisotropy, r, can be measured by either L-format or T-format. In the L-format, all four fluorescence intensities, IVV, IVH, IHV, and IHH, are measured using a single channel of a photodetector so that each intensity needs to be measured separately. If the fluorimeter has two emission channels then anisotropy can also be measured in a T-format, which allows fluorescence intensities pairs, IVV/IVH or IHV/IHH, to be measured simultaneously via the two emission channels. Thus, measurements in the T-format are faster than in the L-format.
2.2. Determining anisotropy and minimizing background signal To obtain the correct anisotropy of the dye, background signals from the buffer solution and macromolecule solutes must be factored out. The anisotropy signal from the dye is determined using background corrected values of IVV, IVH, IHV, and IHH, which are obtained by subtracting the intensities of reference samples that contain all of the components of the real sample, except the dye, from measured sample intensities. This procedure minimizes the background signal’s contribution to the uncertainty in anisotropy values. It is desirable to optimize the buffer condition to minimize background signal. All sample components should preferably produce very small fluorescence signal (via fluorescence or scattering) relative to the fluorophore signal. If the background signal is strong, each component in the buffer solution should be tested for fluorescence in water at the experimental wavelength. The high-signal components should be replaced or used at lower concentration.
2.3. Choosing a wavelength for anisotropy measurement To find the optimum excitation wavelength for an anisotropy measurement, the excitation wavelength dependence of anisotropy (r) is plotted to determine the range of excitation wavelengths, lex, that give a constant
290
Xuesong Shi and Daniel Herschlag
value of r. The excitation wavelength that is used should be well within this range and give the strongest emission intensity, unless there are additional complicating features of the system, such as much greater background contributions at the maximum emission wavelength.
2.4. Testing polarizer alignment The alignment of polarizers should be tested periodically to maintain data accuracy and consistency. We dilute glycogen (Sigma; cat. #G8751) solution until it reaches a stable maximum anisotropy. The scattered light from diluted glycogen solution is 100% polarized with r equal to 1 (Lakowicz, 2006). We consider the alignment to be satisfactory if the maximum anisotropy is larger than 0.98. A stock glycogen solution of the final diluted concentration is kept at 20 C for future testing. The glycogen solution should not be overdiluted, otherwise there can be effects from background signal.
2.5. Factors that influence anisotropy Consider the simplest case in which a dye molecule is rigidly attached to a spherical macromolecule. If the dye molecule has a single exponential decay in its lifetime then the expected steady anisotropy is r¼
r0 1 þ ðt=yÞ
ð14:2Þ
where r0 is the fundamental anisotropy, which is related to the angle between the absorption and emission dipoles of the dye, and y ¼ V/RT, where is viscosity, V is the hydrodynamic volume, R is the gas constant, T is temperature, and y is the rotational correlation time. Given knowledge of r0 and t, the fluorescence lifetime of the dye, information about the rotational dynamics of the macromolecule, y, can be obtained from the observed anisotropy. In practice, the attachment of the dye to the macromolecule is often not completely rigid. When the segmental motion of the fluorophore is much more rapid than the tumbling of the macromolecule, such motion can simply app be represented by using apparent fundamental anisotropy, r0 , instead of r0, app app where r0 < r0 . The value of r0 can be obtained by measuring anisotropy with increasing viscosity of the solution, such that at high enough viscosity, the term r/y ¼ tRT/V approaches zero and app
robs ¼ app
r0 1 þ ðt=yÞ
ð14:3Þ
approaches r0 . The observed anisotropy, robs is influenced by the properties app of the dye, r0 and t, as well as solution properties, and T. All these parameters are needed to relate robs to macromolecule motion.
291
Fluorescence Polarization Anisotropy to Measure RNA Dynamics
3. Choice of FPA Probes One of the main advantages of using a fluorescent base analog over other fluorescent dyes is that the movement of fluorescent base analog can be strongly coupled with the labeled RNA helices through base pairing with the complementary base in the helix and base stacking. Such strong couplings allow the dynamics of the RNA helices to be relatively directly reported by the FPA of the base analogs. Specifically, labeling RNA with regular fluorescent dyes often requires a covalent linker. This linker reduces the motional coupling between the RNA and dye and convolutes the interpretation of FPA data with potential linker–RNA interactions. Although dyes that are intercalators, such as ethidium bromide, can achieve strong coupling with RNA in its motion, binding of intercalator dyes is not sequence specific and introduces significant structural changes in RNA. In contrast, fluorescent base analogs introduce little perturbation to RNA structure. To be used as a FPA probe for RNA helical dynamics, a fluorescent base analog also needs to maintain a significant level of fluorescence intensity and a lifetime on the nanosecond timescale once incorporated into a duplex. Only a small fraction of the reported fluorescence base analogs (Asseline, 2006; Rist and Marino, 2002; Wilson and Kool, 2006) satisfy this criteria; notable examples include 6-methylisoxanthopterin (6-MI), 1,3-diaza-2oxophenothiazine (tC), and 1,3-diaza-2-oxophenoxazine (tC ) (Hawkins et al., 1997; Sandin et al., 2005, 2008; Wilhelmsson et al., 2001) (Fig. 14.1). The fluorescent properties of these three dyes are summarized in Table 14.1. As more accurate measurements of anisotropy can be achieved with a higher ratio of dye intensity to background signal, it is desirable to use a brighter dye and a dye that emit at a wavelength far from the absorption maximum of RNA, thereby minimizing the background signal from the natural bases. All three dyes noted above are reasonably bright at excitation wavelengths that are well distanced from the absorption maximum of the natural bases. The maximum excitation wavelength for 6-MI, tC , and tC monomers are 340, 360, and 375 nm, respectively. Although tC has the most red-shifted excitation wavelength, it is about three times less bright in H N
O
NH2 N
N N
H3C O
6MI
HN
HN N
N O
S
N tC
O
N tC⬚
Figure 14.1 Structures of 6-MI, tC, and tCo.
O
292
Xuesong Shi and Daniel Herschlag
Table 14.1 Fluorescence properties of 6-MI, tC, and tCo incorporated in an RNA (6-MI) or DNA duplex (tC and tCo)
Quantum yield
Lifetime (ns)
6-MIa
0.21
tC
0.18–0.21 (Sandin et al., 2005) 0.17–0.27 (Sandin et al., 2008)
4.4 (Bucci and Steiner, 1988) 5.9–6.9 (Sandin et al., 2005)
tCo
a
3.5–4.8 (Sandin et al., 2008)
Extinction coefficient of free nucleoside (M1 cm1)
e340 4350 (Hawkins et al., 1997) e375 ¼ 4000 (Wilhelmsson et al., 2003) e360 ¼ 9000 (Sandin et al., 2008)
In the sequence 50 -CUFUC-30 , where F ¼ 6-MI, base paired to its complement.
a duplex than tC (Sandin et al., 2008). 6-MI can be as bright as tC within specially designed sequences. One advantage of tC and tC over 6-MI is that fluorescent properties of tC and especially tC are less sensitive to the neighboring base sequences than 6-MI. This reduced sensitivity allows tC and tC to be incorporated into RNA with a minimum degree of sequence design and alteration. The main advantage of 6-MI over tC and tC is that 6-MI-modified oligos are commercially available (Fidelity Systems Inc., Gaithersburg, MD). Procedures for synthesizing and incorporating of all three dyes are available in the literature (Hawkins, 2007; Sandin et al., 2007, 2008). The protocols given in the following section are based on using 6-MI as the labeling dye.
4. Measuring FPA in a Simple Duplex, using an 11mer Control RNA Duplex as an Example 4.1. Sequence design The ideal RNA sequence surrounding 6-MI in an oligonucleotide is YUFUY, where F is 6-MI and Y is a pyrimidine. Exchanging the neighboring U to other nucleotides (except T) or exchanging Y to G can shorten the lifetime of 6-MI to the subnanosecond timescale (Hawkins et al., 1997).
4.2. Sample preparation 6-MI-containing 11mer RNA oligonucleotides were obtained from Fidelity Systems (Gaithersburg, MD). Oligonucleotides were purified by ion exchange HPLC on a DNAPAC PA100 column (Dionex, Sunnyvale, CA)
Fluorescence Polarization Anisotropy to Measure RNA Dynamics
293
using a linear gradient of 10 mM to 2 M ammonium acetate, pH 5.5, in 10% acetonitrile. The collected fractions were dried in a speed-vac overnight to remove residual salt and then resuspended in 50 mL TE buffer (25 mM Tris–HCl, 0.1 mM EDTA, pH 8.0) for long-term storage at 20 C. Note: 6-MI-containing oligonucleotides can also be made following a literature protocol (Hawkins, 2007). Note: If the experiments are highly sensitive to salt, then it is recommended to desalt the oligonucleotides by a reverse phase cartridge, such as a C-18 sep-Pak cartridge from Waters (Milford, MA), before speed-vac concentrating. Alternatively, the HPLC-purified oligonucleotides can be desalted and concentrated by ethanol precipitation. The 6-MI-labeled 11mer oligonucleotide is annealed with an excess amount of the complementary strand at a molar ratio of 1:5 at 95 C for 2 min before gradually cooling to room temperature over 30 min. The completion of hybridization can be checked by native 15% polyacrylamide gel electrophoresis (15% for 10–25 bp in a duplex). The mixture is diluted to a final concentration of 100–200 nM of the 6-MI strand in 50 mM NaMOPS, pH 7.0, with the desired amount of added salt before FPA measurements. A minimum concentration of 100 nM 6-MI is suggested to maintain a ratio of dye intensity to background that is sufficiently high.
4.3. FPA measurement We performed FPA measurements on a Fluorolog-3 spectrometer from Horiba Jobin Yvon (Edison, NJ) using L-format. For 6-MI studies, excitation and emission wavelengths were set to 350 and 425 nm, respectively. As anisotropy is relatively sensitive to temperature, a water bath is connected to the sample compartment for temperature control. The first measurement is made after incubation in the sample compartment for 10 min to allow for temperature equilibration. For each measurement, the average of three consecutive readouts of anisotropy and fluorescent intensities is recorded, and three to four of such measurements are made over 10 min to ensure that the solution is well equilibrated and that there is no time dependence in the anisotropy. Each anisotropy value obtained is corrected for the background by calibrating the four intensities, IVV, IVH, IHV, and IHH, with background intensities separately; for example, IVV ¼ IVV,observed IVV,background. Then the anisotropy is calculated using the corrected intensities as Eq. (14.1). The three to four corrected anisotropy values are averaged to give a value with a standard deviation that is generally smaller than 0.002. Results from two to four independent samples measured on at least 2 different days are averaged to give the final anisotropy value. With all anisotropy values pooled, the standard deviation is generally smaller than 0.0025.
294
Xuesong Shi and Daniel Herschlag
5. Use of FPA to Study Helical Dynamics of RNA, with a Junction Model Construct as an Example RNA structure is largely composed of helical segments connected by various kinds of junctions. Here, we broadly define a junction as any nonWatson Crick base-paired region. With the properly base-paired helical segments being relatively rigid, the flexibility in RNA predominately comes from the junction regions. The effects of different types and sequences of junctions on helical dynamics can be studied individually by using a minimal construct consisting of a dye-labeled helix connected to a nonlabeled helix through the target junction. Thus, the impact of junctions on the motion of the labeled helix can be assessed by FPA. However, the observed anisotropy is determined from depolarization of the dye, which can arise from not only the junction-modulated movement of dye-labeled helix but also from the overall tumbling of the RNA molecule. One way to separate the desired junction-modulated motion from overall tumbling is to increase the overall size of the construct, thereby rendering overall tumbling too slow to cause any significant depolarization over the lifetime of the dye so that the contribution of tumbling to the observed anisotropy is negligible. To increase the construct size, we extended the nonlabeled helix to include a 20-bp LacI operon DNA sequence, which allowed us to bind the 154 kDa Lac Repressor (LacI). (This protein is equivalent to about a 320-bp DNA duplex in volume (Chalikian and Breslauer, 1998; Fischer et al., 2004).) Construction and FPA measurement of model systems consisting of single-stranded AAA (A3) and UUU (U3) junctions are described in Scheme 14.1.
5.1. Assembling the constructs without LacI The A3 and U3 constructs were assembled from two single-stranded oligos, an 11mer short strand and a 50mer long strand (Scheme 14.1). The short strand contains 6-MI and the long strand contains the LacI operon sequence. The long strand was obtained from Integrated DNA Technologies (Coralville, IA) and purified by denaturing polyacrylamide gel electrophoresis. The constructs (Scheme 14.1) were assembled in two steps. In the first step, the purified partially self-complementary long strand was annealed at 95 C for 5 min, gradually cooled to 50 C over 40 min and then quickly cooled on ice to form a duplex with short hangover regions on both sides. Note: The long strand is partially self-complementary (The 36 nt sequence at the 30 -end of the 50mer is self-complementary, Scheme 14.1), and this property reduces the number of oligos that need to be purchased. However,
Duplex constructs* 5⬘-r(GGACAGGAGGG -X - AGUUA)d(GCGAATTGTGAGCGCTCACAATTCGC)r(UAACU) r(CCCUCCUFUCC)-3⬘ 3⬘-r(CCUFUCCUCCC) (UCAAU)d(CGCTTAACACTCGCGAGTGTTAAGCG)r(AUUGA) - X - GGGAGGACAGG-5⬘ Construct A3 Construct U3 Construct segments Short strand Long strand
: X = AAA : X = UUU (11 mer): 5⬘-r(CCCUCCUFUCC)-3⬘ F = 6-MI (50 mer): 5⬘-r(GGACAGGAGGG-AAA-AGUUA) d(GCGAATTGTGAGCGCTCACAATTCGC)r(UAACU) 5⬘-r(GGACAGGAGGG-UUU-AGUUA) d(GCGAATTGTGAGCGCTCACAATTCGC)r(UAACU)
The Lacl operon sequence is underlined. *r(...) represents RNA d(...) represents DNA
Scheme 14.1 Example constructs with single-stranded AAA and UUU junction.
296
Xuesong Shi and Daniel Herschlag
the disadvantage of having a partially self-complementary sequence is the possibility of forming alternative stable hairpin structures. After proper annealing, the purity of the annealed complex should be assessed by nondenaturing gel electrophoresis. In our case there were negligible amounts (2–4%) of hairpin. High oligonucleotide concentration in the annealing step can help favor intermolecular complex formation. In the second step, 5 mM of the annealed long strand duplexes were hybridized with 2 mM short strand at 38 C for 30 min. Complete hybridization (>98%) was confirmed by nondenaturing gel electrophoresis. The mixture was diluted to a final concentration of 100 nM short strand and 250 nM long strand duplexes for FPA measurement. Excess annealed long strand duplexes can be added to ensure that there is no further increase in anisotropy, as this species does not contain the fluorescence dye.
5.2. Lac Repressor (LacI) 5.2.1. Transformation Add 1 mL Plasmid pMDB1 (a gift from Michael Brenowitz, Albert Einstein College of Medicine) into 100 mL competent BL21-DE3-pLysS cells (Promega, Madison, WI). Incubate the mixture on ice for 30 min, then heat shock the sample at 42 C for 90 s to allow plasmid uptake. After the heat shock, immediately ice the sample for another 2 min. Add 1 mL LB media to the cells and grow for 1 h at 37 C with shaking. Use microbeads to spread cells evenly onto a LB agarose plate prewarmed at 37 C containing Carbenicillin and Chloramphenicol, which will select for cells that have taken up the plasmid. Grow at 37 C overnight. On the second day, there should be identifiable colonies growing on the plate. Cover the plate with parafilm and store at 4 C to avoid overgrowth of colonies. Before the end of the day, pick a colony from the plate and add the colony into 100 mL LB together with 100 mL of 50 mg/mL Carbenicillin and 100 mL of 10 mg/mL Chloramphenicol, then grow overnight at 37 C with shaking. 5.2.2. Cell growth and induction Pellet cells for 5 min in two 50 mL Falcon tubes (e.g., 2000 rpm in a Beckman J-6M swinging bucket centrifuge with Beckman JS-4.2 rotor). (We use sterile Falcon tubes (BD Biosciences, San Jose, CA, cat. #352070) to minimize the chance of nuclease contamination.) The supernatant is removed. Optionally, one pellet can be resuspended in equal volume of LB (1% Bacto Tryptone, 0.5% yeast extract and 0.5% NaCl in water) and 60% glycerol, divided into multiple fractions and stored at 80 C for future use. The other pellet is resuspended in 25 mL at 37 C LB, then add into 2 L LB containing 2 mL of 50 mg/mL Carbenicillin and 2 mL of 10 mg/mL Chloramphenicol. Grow at 37 C until the OD600 is close to 1,
Fluorescence Polarization Anisotropy to Measure RNA Dynamics
297
which normally takes about 4–6 h. Remove 1 mL for SDS–PAGE. Add 0.2 mL 1 M IPTG to a final concentration of 0.1 mM to induce gene expression, continue to grow for another 2 h at 37 C. Remove 1 mL for SDS–PAGE analysis of the amount of LacI products. The result of IPTG induction is tested by 12% SDS–PAGE. An aliquot (100 mL) from each of the two 1 mL pre-IPTG and after-IPTG samples are centrifuged for 5 min at 18,500g. The pellets are resuspended in 14 mL of gel loading buffer and incubated at 90 C for 6 min before loading onto the gel. If the IPTG induction is successful, the after-IPTG lane should have a strong band at around 38.5 kDa, the molecular weight of LacI monomer, and this band should be much weaker in the pre-IPTG lane. Centrifuge the after-IPTG cells in two 1 L centrifuge bottles at 4500 g for 15 min, remove supernatant, resuspend in 0.01 M Tris–HCl, pH 8.0, transfer to 50 mL Falcon tubes, centrifuge for 7 min at 4500 g, remove supernatant and freeze the pellets at 80 C. 5.2.3. Cell lysis The frozen pellets are thawed on ice, then resuspended in buffer A (25 mM Bis–Tris, pH 6.0, 1 mM MgCl2, 1 mM DTT, 15% glycerol, and 150 mM KCl). The cell suspension is lysised by French press at 1100 psi two to three times. The lysed samples are centrifuged at 38,000g for 20 min at 4 C. The supernatant is transferred to a fresh tube, and the DNA is precipitated by adding protamine sulfate (6 mg/mL) while stirring to a final concentration of 0.3 mg/mL. The sample is centrifuged at 38,000g for 20 min at 4 C. The supernatant is decanted and buffer exchanged with buffer A using Amicon Ultra-15 concentrator (Millipore, Billerica, MA) with a molecular weight cutoff of 10 kDa. 5.2.4. FPLC purification The sample solution is first purified using a HiTrap SP HP (GE Healthcare Bio-Sciences, Piscataway, NJ) cation exchange column using buffer A as the starting buffer and buffer B (25 mM Bis–Tris, pH 6.0, 1 mM MgCl2, 1 mM DTT, 15% glycerol, and 2 M KCl) as the elution buffer. The samples are eluted with a linear gradient of 150–650 mM KCl over 6 min at 5 mL/min. The LacI peak elutes at 380–420 mM KCl. The amount and purity of LacI in each collected fraction can be tested by 12% SDS–PAGE. The main LacI containing fractions are pooled and then further purified over a HiPrep 16/60 Sephacryl 200 (Pharmacia Biotech, now GE Healthcare Bio-Sciences) gel filtration column preequilibrated with the running buffer C (25 mM Tris–HCl, pH 8.0, 1 mM EDTA, 1 mM DTT, 15% glycerol, and 600 mM KCl). Load no more than 5 mL of sample per run for a HiPrep 16/60 Sephacryl 200 column (5 mL is 1/24 of the column bed volume of 120 mL). The column is run at 1 mL/min with buffer C. The elution profile contains a smaller peak at around 26–34 mL and a larger peak
298
Xuesong Shi and Daniel Herschlag
around 40–56 mL. The latter peak is identified as LacI by SDS–PAGE. The fractions collected of the latter peak are concentrated to about 200 mL with Amicon Ultra-15 concentrator (Millipore) with a molecular weight cutoff of 10 kDa. The concentrated solution is diluted three times with 25 mM Tris–HCl, pH 8.0, then further diluted with 10 mL of storage buffer (25 mM Tris–HCl, pH 8.0, 0.1 mM EDTA, 5% glycerol, and 200 mM NaCl). The solution is concentrated to about 200 mL and stored in 20 mL aliquots at 80 C after measuring its absorbance. The concentration of LacI tetramer is calculated with an extinction coefficient of 0.10 mM 1 cm 1 at 280 nm. Note: With 5% glycerol, the stock LacI solution is viscous. It is desirable to control the final glycerol concentration to be less than 0.5% or, in the other words, to concentrate the stock LacI solution to be at least 10 times more than the experimental LacI concentration. We concentrate LacI to 30 mM (OD280 ¼ 3) or above. Introduction of 0.5% glycerol increases the viscosity of the solution by slightly less than 1% and gives a negligible increase in the anisotropy of <1%.
5.3. Assembling model constructs with LacI and carrying out FPA measurements Concentrated LacI solution is thawed on ice. After adjusting to the experimental salt concentration, the solution is vortexed briefly followed by centrifugation for 1 min at 16,000g. The clear solution is transferred to a separate tube and kept on ice. The annealed sample mixture of 100 nM short strand and 250 nM long strand duplex is titrated with the concentrated LacI until the anisotropy saturates (Fig. 14.2). The buffer condition is 50 mM NaMOPS, pH 7.0, and 280 mM added NaCl, for a total Naþ concentration of 300 mM. Measure anisotropy as described in Section 4. Note: Experiments with LacI were carried out with a salt concentration of 300 mM Naþ or more to weaken the nonspecific LacI DNA binding, as nonspecific binding can affect FPA values. As nonspecific LacI DNA binding is associated with the release of several counterions, the Kd of nonspecific LacI DNA binding is strongly salt dependent. From literature data (Frank et al., 1997), the dissociation constant for nonspecific LacI binding weakens from about 0.6(0.4) mM at 150 mM Kþ to about 41 ( 16) mM at 300 mM Kþ at 15 C. Oversaturating the solution with LacI will increase the free LacI concentration, which results in more nonspecific LacI binding. Thus, it is desirable to limit the LacI concentration to be no more than 1 mM over the saturation value. Note: During and after the LacI measurement, check the cuvette to make sure the solution remains clear over the entire experiments. Clouding is often an indication of LacI aggregation. This observation also typically gives
299
Fluorescence Polarization Anisotropy to Measure RNA Dynamics
Anisotropy
0.29
0.28
0.27
0.26
0.0
0.5
1.0
1.5
LacI (mM)
Figure 14.2 Titration of oligonucleotide model construct with increasing LacI. Conditions: 50 mM NaMOPS, pH 7.0, 280 mM NaCl, 15 C with 100 nM of 6-MI-containing short strands and 250 nM of long strand duplex.
time-dependent fluorescence anisotropy values. If there is LacI aggregation, redo the experiments with freshly made buffer solutions.
6. Use of FPA to Study Helical Dynamics in a Complex RNA, with the Tetrahymena Group I Intron Ribozyme as an Example Dynamics of individual helices of complex RNAs can be directly characterized by FPA by incorporating 6-MI into the target helix, provided that the complex RNA is large enough that its overall tumbling is much slower than the lifetime of 6-MI. The Tetrahymena group I ribozyme, roughly 400 nt in size, is one such RNA. The following protocol is for characterizing the dynamics of the P1 duplex in this ribozyme. The P1 duplex consists of an internal guidance strand (IGS) and a complementary substrate strand. The IGS strand connects to the rest of the ribozyme through a single-stranded junction, and the P1 duplex is formed by hybridizing the substrate strand, which is labeled with 6-MI, with the IGS (Fig. 14.3). 1. L-16 ScaI ribozyme was prepared by in vitro transcription with T7 RNA polymerase at 30 C for 40 min in the presence of 4 mM MgCl2, NTPs (0.5 mM each), 40 mM DTT, 2 mM spermidine, 40 mM Tris–HCl, pH 8.0, and 0.01% Triton X-100 (Karbstein et al., 2007). The transcription product was then purified by 8% denaturing polyacrylamide gel electrophoresis.
300
Xuesong Shi and Daniel Herschlag
6-MI
Substrate IGS
Junction P1 duplex
Figure 14.3 The Tetrahymena group I ribozyme and its P1 duplex. The P1 duplex consists of an internal guide sequence (IGS) and an oligonucleotide substrate, which can be conveniently labeled with 6-MI.
2. After purification, the ribozyme was folded at 50 C for 30 min in 50 mM NaMOPS, pH 7.0, and 10 mM MgCl2 (Bartley et al., 2003). The folded solution was then buffer exchanged into the experimental buffer condition using the YM30 (30 kDa cutoff) Microcon centrifugal filter (Millipore) by concentrating from about 500 to 10–20 mL three times. The concentration of the folded ribozyme is determined by absorbance at 260 nM (using a NanoDrop ND-1000 Spectrophotometer, Thermo Fisher Scientific, Waltham, MA). 3. The concentrated ribozyme solution was annealed with concentrated 6-MI labeled substrate strands under the experimental buffer condition in a 4:1 molar ratio of ribozyme to substrate for 30 min at 38 C. The annealed solution was then diluted to 200 nM of 6-MI and 0.8 mM ribozyme for FPA measurement. To ensure the sample has properly hybridized, the annealed solution was titrated with the concentrated ribozyme solution to a final ribozyme concentration, typically 2 mM, to ensure the saturation of FPA signal. FPA was measured as in Section 4. Note: Alternatively, the folded ribozyme can be mixed with the 6-MI labeled substrate strands in a 3:1 ratio and annealed at 38 C for 30 min. The annealed mixture can then be buffer exchanged with the experimental buffer using YM30 (30 kDa cutoff) Microcon centrifugal filter (Millipore) by concentrating from about 500 to 10–20 mL three times. This buffer exchange step also removes any substrate strand not associated with the ribozyme. After measuring concentration, the concentrated mixture can be used for FPA measurement by dilute to a final ribozyme concentration of 0.5 mM or higher. The obtained anisotropy is the same, within error (0.005), as that obtained by the above protocol. Structured RNAs that are significantly smaller than the Tetrahymena ribozyme can be studied in the same way as a model oligonucleotide
Fluorescence Polarization Anisotropy to Measure RNA Dynamics
301
construct by extending the structured RNA to include a LacI operon sequence or an RNA binding protein recognition sequence and increasing the size of the RNA by binding the corresponding protein to it.
7. Salt Dependence and Normalization of FPA with a Short Control Duplex FPA results obtained at different salt conditions may not be directly comparable because the fluorescence properties of 6-MI, including the lifetime (t), are salt dependent. The salt dependence of the FPA of a helix in a complex construct should thereby be normalized relative to the FPA of a short control duplex of the same sequence of the targeted helix to account for salt effects on the local environment of the 6-MI fluorophore. The normalization ratio, rnorm, can be calculated as the ratio between the apparent rotational correlation time, y, of the constructs and the control duplex only, rnorm ¼ yconstruct/ycontrol. y is related to the rate of anisotropy decay, with larger y associated with higher anisotropy. If the basic Perrin equation for a sphere (Eq. (14.3)) is used to simplify calculation, then app
rnorm ¼
ðr1 =rcontrol Þ 1 : app ðr1 =rconstruct Þ 1
ACKNOWLEDGMENTS We thank Emilia Mollova for her work in this area and Tara Benz-Moy for helpful discussions and comments. This work was supported by NIH Grants PO1 GM066275 and GM49243.
REFERENCES Asseline, U. (2006). Development and applications of fluorescent oligonucleotides. Curr. Org. Chem. 10, 491–518. 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. Bucci, E., and Steiner, R. F. (1988). Anisotropy decay of fluorescence as an experimental approach to protein dynamics. Biophys. Chem. 30, 199–224. Chalikian, T. V., and Breslauer, K. J. (1998). Volumetric properties of nucleic acids. Biopolymers 48, 264–280. Duhamel, J., Kanyo, J., DinterGottlieb, G., and Lu, P. (1996). Fluorescence emission of ethidium bromide intercalated in defined DNA duplexes: evaluation of hydrodynamic components. Biochemistry 35, 16687–16697.
302
Xuesong Shi and Daniel Herschlag
Edwards, T. E., and Sigurdsson, S. T. (2007). Site-specific incorporation of nitroxide spinlabeles into 20 -position of nucleic acids. Nat. Protoc. 2, 1954–1962. Fischer, H., Polikarpov, I., and Craievich, A. F. (2004). Average protein density is a molecular-weight-dependent function. Protein Sci. 13, 2825–2828. Frank, D. E., Saecker, R. M., Bond, J. P., Capp, M. W., Tsodikov, O. V., Melcher, S. E., Levandoski, M. M., and Record, M. T. (1997). Thermodynamics of the interactions of lac repressor with variants of the symmetric lac operator: Effects of converting a consensus site to a non-specific site. J. Mol. Biol. 267, 1186–1206. Grant, G. P. G., Boyd, N., Herschlag, D., and Qin, P. Z. (2009). Motions of the substrate recognition duplex in a group I intron assessed by site-directed spin-labeling. J. Am. Chem. Soc. 131, 3136–3137. Hawkins, M. E. (2007). Synthesis, purification and sample experiments for fluorescent pteridine-containing DNA: Tools for studying DNA interactive systems. Nat. Protoc. 2, 1013–1021. Hawkins, M. E., Pfleiderer, W., Balis, F. M., Porter, D., and Knutson, J. R. (1997). Fluorescent properties of pteridine nucleoside analogs as monomers and incorporated into oligonucleotides. Anal. Biochem. 244, 86–95. Karbstein, K., Lee, J., and Herschlag, D. (2007). Probing the role of a secondary structure element at the 50 - and 30 -splice sites in group I intron self-splicing: The Tetrahymena L-16ScaI ribozyme reveals a new role for the G*U pair in self-splicing. Biochemistry 46, 4861–4875. Lakowicz, J. R. (2006). Principles of Fluorescence Spectroscopy. 3rd edn. Springer, New York. LiCata, V. J., and Wowor, A. J. (2008). Applications of fluorescence anisotropy to the study of protein-DNA interactions. Methods Cell Biol. 84, 243–262. Rist, M. J., and Marino, J. P. (2002). Fluorescent nucleotide base analoges as probes of nucleic acid structure, dynamics and interactions. Curr. Org. Chem. 6, 775–793. Sandin, P., Wilhelmsson, L. M., Lincoln, P., Powers, V. E. C., Brown, T., and Albinsson, B. (2005). Fluorescent properties of DNA base analogue tC upon incorporation into DNA — negligible influence of neighbouring bases on fluorescence quantum yield. Nucleic Acids Res. 33, 5019–5025. Sandin, P., Lincoln, P., Brown, T., and Wilhelmsson, L. M. (2007). Synthesis and oligonucleotide incorporation of fluorescent cytosine analogue tC: A promising nucleic acid probe. Nat. Protoc. 2, 615–623. Sandin, P., Borjesson, K., Li, H., Martensson, J., Brown, T., Wilhelmsson, L. M., and Albinsson, B. (2008). Characterization and use of an unprecedently bright and structurally non-perturbing fluorescent DNA base analog. Nucleic Acids Res. 36, 157–167. Shi, X., Mollova, E. T., Pljevaljcic, G., Millar, D. P., and Herschlag, D. (2009). Probing the dynamics of the P1 helix within the Tetrahymena group I intron. J. Am. Chem. Soc. 131, 9571–9578. Thomas, J. C., Allison, S. A., Appellof, C. J., and Schurr, J. M. (1980). Torsion dynamics and depolarization of fluorescence of linear macromolecules. II. Fluorescence polarization anisotropy measurements of a clean viral phi 29 DNA. Biophys. Chem. 12, 177–188. Wilhelmsson, L. M., Holmen, A., Lincoln, P., Nielson, P. E., and Norden, B. (2001). A highly fluorescent DNA base analog that forms Watson-Crick base pairs with guanine. J. Am. Chem. Soc. 123, 2434–2435. Wilhelmsson, L. M., Sandin, P., Holmen, A., Albinsson, B., Lincoln, P., Norden, B., and Phys, J. (2003). Photophysical characterization of fluorescent DNA base analog, tC. J. Phys. Chem. B 107, 9094–9101. Wilson, J. N., and Kool, E. T. (2006). Fluorescent DNA base replacements: reporters and sensors for biological systems. Org. Biomol. Chem. 4, 4265–4274. Zhang, Q., Sun, X. Y., Watt, E. D., and Al-Hashimi, H. M. (2006). Resolving the motional modes that code for RNA adaptation. Science 311, 653–656. Zhang, Q., Stelzer, A. C., Fisher, C. K., and Al-Hashimi, H. M. (2007). Visualizing spatially correlated dynamics that directs RNA conformational transitions. Nature 450, 1263–1267.
C H A P T E R
F I F T E E N
Studying RNA Using Site-Directed Spin-Labeling and Continuous-Wave Electron Paramagnetic Resonance Spectroscopy Xiaojun Zhang,* Pavol Cekan,† Snorri Th. Sigurdsson,† and Peter Z. Qin* Contents 1. Site-Directed Spin-Labeling 1.1. Spin-labeling during chemical synthesis of the nucleic acid 1.2. Postsynthetic spin-labeling 2. Acquisition and Processing of cw-EPR Spectrum 2.1. Sample preparation and insertion 2.2. Critical coupling of cavity 2.3. Spectral acquisition 2.4. Spectral processing 3. Spectral Analysis 4. Examples of Application 4.1. Example 1: cw-EPR studies of the HIV TAR RNA 4.2. Example 2: Studying RNA/RNA interactions via cw-EPR reported tR effects Acknowledgments References
304 305 307 308 310 310 311 315 318 320 320 324 326 326
Abstract In site-directed spin-labeling (SDSL), a stable nitroxide radical is attached to a specific location within a macromolecule and electron paramagnetic resonance (EPR) spectroscopy is used to interrogate the local environment surrounding the nitroxide. The SDSL strategy enables probing site-specific structural and dynamic features of RNA in solution without being limited by the size of the molecule, thus serving as a unique tool in biophysical studies of RNA. This chapter describes the use of continuous-wave (cw)-EPR to study dynamic features of RNAs as well as to * {
Department of Chemistry, University of Southern California, Los Angeles, California, USA Department of Chemistry, Science Institute, University of Iceland, Reykjavik, Iceland
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69015-7
#
2009 Elsevier Inc. All rights reserved.
303
304
Xiaojun Zhang et al.
monitor interactions between them. Various approaches for attaching nitroxide spin labels to nucleic acids are described, followed by detailed descriptions of cwEPR spectral acquisition and processing procedures. Specific examples are subsequently used to illustrate analysis of EPR spectra, showing how information regarding the parent RNA can be extracted.
Information about RNA structure and movement is critical for our understanding of how RNA is able to carry out its multifaceted functions. One spectroscopic technique that has shown great promise to study RNA, as well as other biopolymers, is electron paramagnetic resonance (EPR) spectroscopy, also named electron spin resonance (ESR) spectroscopy. EPR is a magnetic resonance technique that monitors the behaviors of unpaired electrons, and has long been used to study structure and dynamics of biomolecules (see recent reviews by Klug and Feix, 2008; Sowa and Qin, 2008). Structural information can be obtained by distance measurements, that is, by determination of distances between two spin-centers, and is the topic of another chapter in this volume (see Chapter 16 in this volume). This chapter focuses on continuous-wave (cw)-EPR, which has been commonly used for studying dynamic features and interactions between biomolecules. The technique is capable of covering motions ranging from picosecond (ps) to millisecond (ms). Other advantages of EPR are that small amount of material (typically 0.05–1.0 nmol) is required, samples are not limited by molecular size, and the measurements can be carried out under physiological conditions. However, RNA does not contain stable unpaired electrons and, therefore, spin-centers must be introduced into the RNA in order to conduct EPR studies. This chapter first summarizes methods for attaching spin labels to nucleic acids, with a special focus on RNA (Section 1). Detailed protocols for these labeling methods have been described in a series of recent publications (Edwards and Sigurdsson, 2007; Qin et al., 2007; Schiemann et al., 2007), and will not be repeated here. Acquisition and processing of cw-EPR spectrum will then be described in detail (Section 2), followed by a brief discussion of spectral analysis (Section 3). Specific examples of using cwEPR to study dynamics and interactions in RNA will subsequently be given (Section 4).
1. Site-Directed Spin-Labeling Since RNA is diamagnetic, EPR studies of RNA require incorporation of unpaired electrons into the biopolymer. Nitroxides in five- or sixmembered rings that are flanked by methyl groups are stable organic free radicals that are commonly used for spin-labeling (Fig. 15.1). For a free
305
cw-EPR Studies of RNA
O
O
O
O
O
N
N
N
N
N
Figure 15.1 Structures of nitroxides that have been used for spin-labeling.
radical, the nitroxide is chemically unusually stable because the unpaired electron is shared by the nitrogen and the oxygen atoms. The radical is also sterically protected from reaction with solvent and other molecules by the methyl groups. The labeling must be specific, that is, directed to a specific site in the RNA to yield meaningful results pertaining to specific nucleotides. This is commonly referred to as site-directed spin-labeling (SDSL) (Altenbach et al., 1989; Barhate et al., 2007; Edwards et al., 2001; Kim et al., 2004; Qin et al., 2001, 2003; Schiemann et al., 2004). Therefore, incorporation of multiple labels through enzymatic RNA synthesis (e.g., triphosphate polymerization with polymerases; Keyes et al., 1997) is of limited value. Instead, labels are generally introduced chemically, either during chemical synthesis of the nucleic acid or by postsynthetic modification of the polymer. After selecting a specific nucleotide (or nucleotides for multiple labels) for spin-labeling in the RNA of interest, the point of attachment is chosen. The label can be incorporated into the base (Barhate et al., 2007; Fischhaber et al., 1997; Keyes et al., 1997; Piton et al., 2007; Prisner et al., 2001; Qin et al., 2003; Schiemann et al., 2004; Spaltenstein et al., 1988), sugar (Edwards et al., 2001; Kim et al., 2004), or the phosphodiester (Fidanza et al., 1992; Nagahara et al., 1992; Qin et al., 2001) (Figs. 15.2 and 15.3). In addition, the structure of the label needs to be taken into consideration. For example, if spin labels are linked to the RNA with a flexible linker, the nitroxide ring may have considerable motions independent of the parent nucleic acid, which must be accounted for during data analysis. Excess linker flexibility may give rise to difficulties in detecting nucleic acid dynamics or measuring interspin distances. As with all chemical modifications of RNA, spin-labeling may interfere with the structure and function of the nucleic acid. Therefore, one must carefully evaluate the effect of the label itself by running appropriate control experiments.
1.1. Spin-labeling during chemical synthesis of the nucleic acid Nucleic acids are synthesized chemically in laboratories around the world using automated synthesizers. Therefore, if one can prepare a phosphoramidite of a spin-labeled nucleoside, which is the building block for
306
Xiaojun Zhang et al.
N
O N
HN
HN N
N HO
N
O
O
HO
NH2 N
N
N
O
1
OH
HO
N
O
2
OH
N NH2
O OH X
N
HO
4
O OH X
N
O
N
NH O
N O
3
O
O
N N
N
N
OH OH
O N
O
HO
O
NH
N H HO
O 5
O OH
N
O 6
O N
O N
NH O
N HO
OR
O OH
7
HO
O OH
N N
O
8
Figure 15.2 Spin-labeled nucleosides that have been incorporated into nucleic acids during chemical synthesis of the oligomer. In structures 4 and 5, X is either H or OH.
oligonucleotide synthesis, the spin label can be incorporated into any sequence and at any site in the sequence. Many different spin-labeled nucleosides have been prepared, some of which are shown in Fig. 15.2, and incorporated into nucleic acids. For example, spin labels have been attached to the exocyclic amino groups of C (1) and A (2), the 2-position of A (3), and the 5-position of pyrimidines (4–6). The most important advantage of this approach over the postsynthetic method described below is that it enables incorporation of spin label modifications that require an elaborate synthetic effort, such as C-nucleosides (7) (Miller et al., 1995) and the rigid spin label C ¸ (8) (Barhate et al., 2007). The drawbacks of the synthetic approach include the effort in preparation of the spin-labeled phosphoramidites and exposure of the spin label to the conditions of the chemical synthesis. The latter also applies for methods that use on-column modification during the synthesis (Piton et al., 2007; Schiemann et al., 2004). For example, the iodine solution that is commonly used for the phosphorous oxidation during the synthesis reduces nitroxides
307
cw-EPR Studies of RNA
A S
S N
N RO
N
O
H N
N O
S
RO
O
OR OH 9
N
O
N
O
O
O
N
O
HO
OR OH 10
O
N
O O P
RO
B
O O
O
N
O
RO
S P O O
O RO 15
N
O HN P
O
S
O
O
RO
O S P O S O
B
O RO
OH N O
16
B
O OR 14
N O
13
O
N O
B
O O
12
D
RO
P
H N
O HN O
RO
N H
N
C
RO
O
11
OH
B
N
NH
B
O
B
N
OH 17
N O
Figure 15.3 Nucleotides that have been spin-labeled by postsynthetic modification of the oligomer at either the base (A), the sugar (B), the internal phosphodiester (C), or the terminus (D). ‘‘B’’ indicates the nucleoside base.
(Cekan et al., 2008; Gannett et al., 2002; Piton et al., 2007), although this problem can be circumvented by using tert-butyl hydroperoxide (Cekan et al., 2008). Partial decomposition of nitroxides is also observed due to the acid treatment that is used to remove the dimethoxytrityl protecting groups during the synthesis cycle (Abakumov and Tikhonov, 1969; Cekan et al., 2008; Piton et al., 2007).
1.2. Postsynthetic spin-labeling Chemical synthesis of RNA is more challenging than for DNA, due to the presence of the 20 -hydroxyl group, which needs to be protected during oligomer synthesis and subsequently deprotected. Thus, postsynthetic labeling has been more extensively used for the preparation of spin-labeled RNA. In this approach, the nucleic acid is first synthesized and subsequently incubated with a spin-labeling reagent. To ensure site-specific labeling, the RNA must contain a reactive group that can be specifically modified with the labeling reagent. Such modified RNAs and spin-labeling reagents are in many cases commercially available. Therefore, scientists who do not have
308
Xiaojun Zhang et al.
extensive training in organic synthesis can readily prepare spin-labeled oligomers with the postsynthetic method. Another advantage of postsynthetic labeling is that the spin label is not subjected to the reagents used in oligomer synthesis, thus avoiding partial reduction of the nitroxide. Incorporation of spin labels through chemical synthesis of the nucleic acid has focused on spin-labeled nucleobases. On the other hand, postsynthetic labeling has been used for attaching nitroxides to different parts of nucleotides: at nucleoside bases (Fig. 15.3A) (Hara et al., 1970; Okamoto et al., 2004; Qin et al., 2003; Varani et al., 1999), sugars (Fig. 15.3B) (Edwards et al., 2001; Kim et al., 2004), and the internal phosphate backbone (Fig. 15.3C) (Fidanza et al., 1992; Nagahara et al., 1992; Qin et al., 2001). Spin labels have been attached to the 50 -end of RNA by modification of a 50 -phosphorothioate (15, 16, Fig. 15.3D) (Grant et al., 2008; Macosko et al., 1999) and the 30 -end by periodite oxidation and reductive amination (17, Fig. 15.3D) (Hermann, 1977). However, the most general spin-labeling approach is incorporation of nitroxides at internal sites of the oligomer (Fig. 15.3A–C), which has been used extensively to prepare spinlabeled RNA. For example, incorporation of 20 -amino groups at selected nucleotides enables selective reaction with aliphatic isocyanates, resulting in 20 -urea-labeled oligomers (Fig. 15.4A) (Edwards et al., 2001). The reactivity of sulfur atoms has also been utilized for postsynthetic labeling of pyrimidine bases (Fig. 15.4B) (Qin et al., 2003) and at the phosphodiester (Fig. 15.4C) (Qin et al., 2001).
2. Acquisition and Processing of cw-EPR Spectrum A typical layout of an EPR spectrometer is shown in Fig. 15.5. While the general principles of operation and detection are the same for different spectrometers, specific steps may vary. The readers are encouraged to consult the manual and vendor regarding each individual spectrometer. Examples used in discussions below are based on a Bruker EMX X-band spectrometer that is equipped with an ER-041X microwave bridge and a high-sensitivity cavity (ER-4119HS, Bruker Biospin, Inc.). There are four general steps in acquiring an undistorted cw-EPR spectrum with the desired signal-to-noise ratio (S/N): (1) sample preparation and insertion; (2) cavity coupling; (3) spectral acquisition; and (4) spectral processing. Depending on the spectral line-shape, one may be able to obtain a cw-EPR spectrum with a sufficient S/N in less than 15 minutes using as low as 50 pmol of nitroxide-labeled samples. Data averaging, typically in the order of tens of minutes to hours, can be used to measure weaker signals.
309
cw-EPR Studies of RNA
O
A
O NH OCN
O
N
O O
O
O
O
P
S S
O
N
O O
O
O
S
N
O
O
O NH
O H N
O HN
O
O
B
N
O O
NH2
P O
NH
N O
S
S N O
N
N O O
OH
N
O O
O
OH
O
C
O NH
O O
O P
S
N
O
O
I
O
H
NH N O
O
O N
O
N
O
O S
O P
O
H
O
Figure 15.4 Postsynthetic labeling of internal nucleotides in RNA. (A) Reaction of a nitroxide containing an aliphatic isocyanate with 20 -amino groups in RNA. (B) Reaction of methanethiosulfonate nitroxide with a 4-thiouridine nucleotide in RNA. (C) Reaction of iodomethyl nitroxide with deoxyribo-phosphorothiolate linkage in RNA.
Bridge Console Magnet Cavity
Computer
Figure 15.5 The general layout of an EPR spectrometer. The spectrometer consists of a microwave bridge, a cavity (or resonator), magnets, and a console. The microwave bridge houses the microwave source and the detector, together with other control and support components. The cavity is where resonance between the electron spins and the electromagnetic radiation takes place, and is stationed between the magnets. The console contains signal processing units and electronics that control the operation of the bridge. The console is generally connected to a computer for spectral acquisition and data analysis.
310
Xiaojun Zhang et al.
2.1. Sample preparation and insertion The required total number of spins in a measurement is typically 1013–1014, which translates into 100 pmol of nitroxide. Biological samples are usually prepared in buffer solutions, and it is important to ensure that no undesired EPR signal comes from the buffer. For nitroxide-labeled samples, diamagnetic (EPR silent) metal ions (e.g., Kþ, Naþ, Mg2þ, Ca2þ) are acceptable. Paramagnetic metal ions, such as Cu2þ, Fe3þ, Mn2þ, have EPR signals and thus may interfere with SDSL studies. In addition, compounds such as sucrose, glycerol, or ficoll are often added to solution to increase solvent viscosity and reduce the overall rotational tumbling motion of the macromolecular (vide infra). They may also serve as cryoprotecting agents to ensure a glass-like homogeneous mixture when measuring frozen samples. EPR samples are held in glass or quartz capillaries/cells. Glass is much cheaper than quartz, and often sufficient for cw-EPR measurements. However, glass typically gives a much more prominent background signal at low temperature due to the presence of impurities (e.g., defects or paramagnetic metal ions). EPR capillaries/cells vary in size and shape depending on the resonator used with the spectrometer. They are chosen to place the sample at the ‘‘active’’ volume of the resonator (usually at the center), where the magnetic field is at its maximum and the electric field is at the minimum. To obtain an optimal signal, the sample should fill the resonator active volume as much as possible. This dictates the sample volume and the positioning of the capillary/cell in the cavity. For example, with the Bruker high-sensitivity cavity, a tilted capillary covers a smaller fraction of the active volume and therefore reduces signal intensity. It should be noted that water absorbs microwave radiations via interacting with the electric field, thus lowering the cavity Q value and making it difficult or impossible to tune. Therefore, aqueous samples, which include almost all RNA samples, tend to be ‘‘lossy’’ with low Q values. This may restrict the size of the capillaries/cells and the sample volume.
2.2. Critical coupling of cavity After sample insertion, the cavity is ‘‘tuned’’ to achieve critical coupling, so that the microwave power stored in the cavity is maximized while its dissipation is minimized. Tuning generally involves two steps: ‘‘roughtuning’’ to approximately match the microwave frequency to the cavity resonance frequency; followed by ‘‘fine-tuning’’ to establish critical coupling. In rough-tuning, the spectrometer is put at the ‘‘tune’’ mode. A low microwave power (e.g., 25 dB, 0.63 mW) is applied to generate the ‘‘model pattern’’, which is the microwave power reflected from the resonator as a function of the microwave frequency. Microwave frequency is adjusted to
311
cw-EPR Studies of RNA
position the ‘‘dip’’, which reports the frequency at which the microwave power is absorbed, at the center of the model pattern. This indicates an approximate match between the microwave frequency and the cavity resonance frequency. The resonator is then further adjusted (e.g., move the iris and/or change the ‘‘signal phase’’ if using a Bruker EMX system) to make the dip symmetrical and as deep as possible (maximal power absorbed). One should also allow the cavity and sample to come to thermal equilibrium during rough-tuning in order to minimize drifts in later finetuning. To fine-tune the cavity, the spectrometer is put in the ‘‘operate’’ mode. Adjust the microwave frequency, the iris position (resonator parameter), and the reference arm current (‘‘bias’’) so that the analog indicators for the automatic frequency control (‘‘AFC’’) and the ‘‘diode’’ always stay at the center as the microwave power is increased from minimum (e.g., 50 dB, 2 mW) to maximum (e.g., 0 dB, 200 mW). This indicates that at all power levels, the majority of microwave power is stored in the resonator and very little is reflected. Adjust the ‘‘signal phase’’ to let the ‘‘diode’’ indicator reach the maximum, and then decrease the ‘‘bias’’ if necessary to put ‘‘diode’’ back to center again.
2.3. Spectral acquisition After cavity tuning, acquisition parameters need to be set up before acquiring a spectrum. Table 15.1 lists these parameters together with example values used for acquiring a nitroxide spectrum on a Bruker EMX system. These values vary significantly depending on the spectrometer, the resonator, and the sample, and thus should be optimized in each individual case. Improper Table 15.1 Examples of acquisition parameters used with a Bruker EMX spectrometer Parameter
Sample value
Trouble-shooting notes
Center field Scan width Microwave power Modulation amplitude Modulation frequency Modulation phase Number of points Conversion time Time constant Number of scans Receiver gain
3324 G 100 G 2 mW (20 dB) 1G 100 kHz 42 2048 40.96 ms 20.48 ms 16 6.3 103
Section 2.3.1.2 Section 2.3.1.2 Section 2.3.2.2 Section 2.3.2.1 Section 2.3.2.1 Section 2.3.1.3 Section 2.3.2.3 Section 2.3.2.3 Section 2.3.2.3 Section 2.3.1.5 Section 2.3.1.4
312
Xiaojun Zhang et al.
acquisition parameters lead to low quality spectra, with the two major symptoms being: (1) no signal or very weak signal; and (2) spectral distortion. These pitfalls and their remedies are discussed in the following sections. 2.3.1. No signal or very small signal If sufficient amount of spins (>1013 for nitroxide-labeled sample) is present but no signal or very weak signal is detected, one should check the following parameters. 2.3.1.1. Sample positioning The capillary/cell should be straight and not tilted, with the sample occupying the resonator active volume (e.g., cavity center). To ensure proper sample positioning, one can insert the sample with the spectrometer set at the ‘‘tune’’ mode, and adjust the capillary so that the dip displacement is maximized. 2.3.1.2. Center field and scan width These two parameters dictate the center and the width of the measured spectrum, and should be set to cover the entire region of all desired signals. For nitroxides, the splitting of peaks can be up to 70 Gauss (G), and most studies use a convenient scan width of 100 G. The center field is related to the resonating microwave frequency according to hn H¼ ð15:1Þ gbe
where h is the Planck constant (h ¼ 6.626 10 34 J s), n is the microwave frequency, be is the Bohr magneton (the intrinsic unit of an electron magnetic moment, be ¼ 9.274 10 24 J/T), and g is the electron spin g-factor. The nitroxide g-factor is very close to that of a free electron (ge 2.0023), and Eq. (15.1) can be rewritten as: H ¼ 356:8n
ð15:2Þ
where H is expressed in Gauss, and n in 109 Hz (GHz). For a microwave frequency of 9.5 GHz (X-band), the corresponding magnetic field is 3390 G. 2.3.1.3. Modulation phase EPR signal is detected using phase-sensitive lock-in amplification (Blair and Sydenham, 1975; Cova et al., 1979; Poole, 1996), and therefore depends on the cosine of the phase angle (phase difference) between a reference signal and the modulated EPR signal. The parameter ‘‘modulation phase’’ is a phase shifter that changes the phase of reference signal in order to optimize detection of the desired EPR signal. For cw-EPR, the in-phase signal is detected, with the phase shifter adjusted
cw-EPR Studies of RNA
313
to give a phase angle of 0 (maximal EPR signal). To determine the proper modulation phase, one can measure a sample with a varying modulation phase. The minimal signal corresponds to the out-of-phase position (phase angle ¼ 90 or 270 ) from which the in-phase value can be deduced by subtracting (or adding) 90 . 2.3.1.4. Receiver gain Sufficient receiver gain is required for observing the signal. However, a high receiver gain does not improve S/N, as noises are amplified to the same degree as the signal. One should also avoid excessive receiver gain to prevent ‘‘clipping’’ of the spectrum (signal outside of maximal detection range). 2.3.1.5. Number of scans (signal averaging) EPR S/N is proportional to the square root of number of scans. In theory, a desirable S/N can always be achieved by increasing the number of scans. In practice, however, over a prolonged period, system instability (temperature variations, air drift, etc.) and background signals (e.g., impurities at capillary and/or resonator) may give a baseline with variable features. Therefore, very weak signals may not be detectable even with a very large number of scans.
2.3.2. Spectral distortion The line-shape of a true (undistorted) EPR spectrum should be independent of the acquisition parameters, and therefore to assess spectral distortion one can compare spectra acquired with different parameters. Figure 15.6 illustrates the effect of modulation amplitude on EPR line-shape. The central line-width (peak-to-peak width DHpp ¼ 1.6 G) remains unchanged when the modulation amplitude is increased from 0.5 to 1 G; while at a modulation amplitude of 10 G, distortion and line-broadening (DHpp ¼ 6.4 G) can be clearly observed. The main sources of spectral distortions are modulation amplitude, microwave power, and scanning rate (speed). These are discussed in the following sections. 2.3.2.1. Modulation amplitude and modulation frequency High modulation amplitude increases EPR signals intensity, but may also distort the line-shape (Fig. 15.6). It is advised to use a modulation amplitude that is approximately equal to the line-width (DHpp) of the narrowest peak one tries to resolve. This amounts to 1–2 G for most nitroxide systems. Moreover, a modulation frequency of 100 kHz is typically used. 2.3.2.2. Microwave power A microwave power of 2 mW (20 dB) is typically used to acquire a nitroxide spectrum. Without saturation, EPR signal is proportional to the square root of the microwave power. As microwave power increases, the rate of excitation may become greater than the rate of relaxation. This decreases the difference in spin populations
314
Xiaojun Zhang et al.
M.A. = 0.5 G
ΔHpp
M.A. = 1.0 G
M.A. = 10 G
20 G ΔHpp
Figure 15.6 Effects of modulation amplitude on cw-EPR line-shape. Spectra shown were obtained on a Bruker EMX spectrometer equipped with a high-sensitive cavity. The sample was an aqueous solution of tempol (4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl) (12.5 mM, 5 ml) placed in a round glass capillary (0.6 mm ID, 0.8 mm OD; Vitrocom, Inc., Mountain Lakes, NJ) sealed at one end. Acquisition parameters listed in Table 15.1 were used, except that the modulation amplitude (M.A.) was varied as respectively indicated for each spectrum. Note that these spectra have not been normalized, and differences in the amplitude of the spectrum are due to the different modulation amplitudes used.
between the ground state and the excited state, thus reducing the EPR signal. High power may also cause line-broadening (Fajer, 2000). 2.3.2.3. Scanning speed: Number of points, conversion time, and time constant A greater number of points in a spectrum gives better resolution, but requires longer sweep time. A general rule is to have at least 10 points for the narrowest line in a spectrum. ‘‘Conversion time’’ is the amount of time the detector spends integrating at one field position before moving to next one, and ‘‘time constant’’ is a measure of the cutoff frequency of the lock-in detector, which will filter out signals (noises) with frequencies greater than 1/(time constant). The time constant is generally kept below the conversion time to prevent spectral distortion, although for weak signal a long time constant may be used. Conversion time should be sufficiently
cw-EPR Studies of RNA
315
long to ensure that a weak signal is captured during the digitization steps. However, lengthy conversion time may lead to a prolonged sweep time (sweep time ¼ conversation time number of points) and may cause baseline drift. One may avoid this by acquiring more scans with short conversion time. In doing so, the baseline drift may only cause a constant offset, which will not distort the spectrum.
2.4. Spectral processing After acquisition of the spectral data, a number of processing steps, including spectrum averaging, baseline correction, and spectrum integration, are executed to prepare the spectrum for further analysis. These steps can be carried out using software provided by the vendor (e.g., WinEPR from Bruker). The spectrum can also be exported using the ‘‘ASCII’’ format, then processed and analyzed using customized programs relying on commercially available software, such as Microsoft Excel, Matlab, or Labview. The general ideas behind each processing step, illustrated using examples generated with Microsoft Excel, are discussed in the following sections. 2.4.1. Spectrum averaging Spectrum averaging is probably the most effective way to increase S/N (Fig. 15.7), and is particularly important when measuring spectrum with broad lines. 2.4.2. Baseline correction An ideal baseline should be completely straight without any feature. This may not be the case, especially for samples with weak signals, due to a number of reasons. The cavity, the cryostat, or the capillary/cell may have a small amount of contamination, the signal of which may be reduced or even eliminated by cleaning. In addition, system instability (temperature variations, air drift, etc.) may also yield broad features in the baseline. One may acquire a ‘‘background spectrum’’ and then subtract it from the averaged experimental signal (Fig. 15.7). In doing so, the control should resemble the real sample as much as possible: use the same buffer and the same exact type of capillary; insert to the same position of the cavity; and performing the scan immediately after sample measurement with the same parameters. A reasonably reproducible background spectrum can be obtained in many cases. In extreme situations (e.g., large variations in background signal due to capillary changes), one might have to quench the EPR signal in situ (e.g., injecting ascorbic acid to reduce the nitroxide) to obtain the background spectrum.
316
Xiaojun Zhang et al.
Single scan
16 scan average
Background
Corrected 20 G
Figure 15.7 Examples of spectral processing. Acquisition parameters listed in Table 15.1 were used to measure a solution of tempol. The background spectrum was collected using water, which shows a feature due to a cavity defect.
2.4.3. Spectral integration With the use of lock-in detection, the measured EPR signal corresponds to the first derivative of the radical absorption spectrum. The 1st integral of a measured EPR spectrum as a function of the magnetic field is computed according to ni X Y ðHi Þ ¼ yj DH ð15:3Þ j¼1
with ni ¼
Hi Hb DH
ð15:3aÞ
in the calculations, yj is the amplitude of the detected EPR signal at magnetic field Hj, DH is a fixed interval (e.g., DH ¼ 1 G; smaller DH gives more precise integral), and ni is the number of intervals between the beginning of the spectrum (Hb) and the current field position (Hi). The 1st integral (Yi vs. Hi) gives the absorption spectrum (Fig. 15.8A). The 1st integral can be integrated again to give the area under the absorption spectrum between Hb and Hi: AreaðHi Þ ¼
ni X j¼1
Yj DH
ð15:4Þ
317
cw-EPR Studies of RNA
A Derivative Amplitude(y)
B
Before normalization 20 G
Intensity(Y)
1st integral
Area
2nd integral A
Magnetic field
After normalization
C 7.5
A = 0.066 × [tempol] + 0.028 Asample
A
5.0
20 G
2.5 [spinsample] 0 0
50 [Tempol] (mM)
100
Figure 15.8 Examples of spectral integration and normalization. Spectra shown were obtained with nitroxide label 14 (Fig. 15.3C). Acquisition parameters are listed in Table 15.1, except that number of scans ¼ 4 and number of points ¼ 1024. (A) Spectrum of an aqueous sample of a 23-nt RNA, together with its 1st and 2nd integrals. (B) Spectral comparison between a 23-nt RNA (40 mM, dotted line) and a 49-nt RNA (30 mM, solid line). Comparison of the normalized spectra is not skewed by the different amount of labeled RNAs used in the measurement, and reports different nitroxide behavior due primarily to the difference in RNA size. (C) An example of spin counting. The calibration curve was generated by linear fitting (solid line) of data points (solid square) obtained using tempol solutions of various concentrations. Using this calibration curve, the sample measured in (A) was found to contain 37.5 mM of spins (Asample ¼ 2.5). Based on an RNA concentration of 40 mM, the nitroxide labeling efficiency was determined to be 93.6%.
The 2nd integral (designated as A) generally refers to the area computed over the entire spectrum (n ¼ (Hf Hb)/DH, with Hf being the magnetic field at the end point of the spectrum), and is directly proportional to the total number of spins (Fig. 15.8A). It should be noted that the actual value of
318
Xiaojun Zhang et al.
the 2nd integral is affected significantly by the baseline, and therefore is meaningful only if the baseline is acceptable or has been corrected. The 1st and 2nd integrals are used in a number of processing steps described in the following sections. 2.4.3.1. Baseline corrections With an ideal baseline, the 1st integral should be completely flat in the outer low-field and high-field regions (Fig. 15.8A). Therefore, the 1st integral can be used in baseline correction, with the rules of thumb in selecting the correct baseline being that: (i) no negative regions in the 1st integral (absorption spectrum); and (ii) low-field and high-field regions beyond the absorption peaks are flat and equals to 0. 2.4.3.2. Spectral normalization Many spectral analysis procedures compare spectrum per spin, therefore the measured spectrum has to be normalized (Fig. 15.8B). Normalization is carried out by dividing the measured spectrum by its 2nd integral, which is proportional to the total number of spins. 2.4.3.3. Spin counting While the 2nd integral reports the total number of spins, the proportional constant depends on acquisition parameters and the specific spectrometer. To obtain the proportional constant, one can generate a calibration curve, where samples with known amount of spins are measured (Fig. 15.8C). Spin counting generally yields a ‘‘spin concentration’’, which is proportional to the total number of spins as the spectrometer activity volume is considered constant. An important reason for spin counting is to assess labeling efficiency, which is the ratio of the spin concentration to that of the RNA (e.g., determined by UV–Vis absorption measurement, Fig. 15.8C).
3. Spectral Analysis The line-shape of a cw-EPR spectrum is dictated by the reorientation dynamics (motions) of the nitroxide, which average the anisotropic nitroxide magnetic tensors. Figure 15.9A shows simulated X-band EPR spectra of a nitroxide undergoing isotropic tumbling, with the nitroxide motion characterized by a rotational correlation time t. As nitroxide motions reduce (longer t), the EPR spectrum shows broader lines, and extra features appear at the low-field and high-field regions. The goal of assessing nitroxide dynamics is to obtain information at the labeling sites of the macromolecule. In SDSL studies, the macromolecule may influence nitroxide dynamics via a combination of three modes (Fig. 15.9B): (i) the overall rotational motion of the macromolecule
319
cw-EPR Studies of RNA
A 0.1
B Mobile
1.0
tR N O
Molecular tumbling
2.5 ti Internal motion
5.0 10
N O
30 Immobile
50 t(ns)
tB
Local dynamics N
DHpp
2Aeff
O
Figure 15.9 Nitroxide dynamics and cw-EPR spectral line-shape. (A) Simulated X-band EPR spectra of nitroxides undergoing isotropic rotation at different rotational correlation time t. (B) The three modes of motion that contribute to nitroxide dynamics. Adopted from Sowa and Qin (2008) with permission.
(characterized by a rotational correlation time tR); (ii) torsional rotations about bonds that connect the nitroxide ring to the macromolecule (ti); and (iii) segmental motions of the macromolecule at or near the labeling point (tB). The ti and tB motions may be influenced by site-specific macromolecular structural and dynamic features, while the tR motion is uniform across the macromolecule, and depends on the molecular size (weight) as well as solvent viscosity. As demonstrated by the examples of application, either ti/tB (example 1) or tR (example 2) effects can be used to study RNA. A number of methods have been developed for assessing nitroxide dynamics based on the cw-EPR spectrum (see review by Sowa and Qin, 2008). In the semiquantitative approach, parameters measured directly from the EPR spectrum, such as the central line-width (DHpp, Fig. 15.9A), the splitting of the resolved hyperfine extrema (2Aeff, Fig. 15.9A), and the second moment (H2, characterizing how broad the spectrum is), are used to characterize nitroxide dynamics (Columbus and Hubbell, 2002, 2004; Mchaourab et al., 1996). These parameters report on the nitroxide mobility, which describes a combined effect of the rate and the amplitude of motion. For example, a broad center line gives a small (DHpp) 1 value and indicates low mobility, which can result from low frequency but large amplitude motions, or small amplitude motions with fast rates. The line-shape parameters can be easily measured on a properly processed EPR spectrum, and
320
Xiaojun Zhang et al.
provide a means to quickly access relative nitroxide mobility at different sites. This approach, specifically by measuring 2Aeff, was used in example 1 to probe site-specific dynamic features of the HIV TAR RNA. In addition to the semiquantitative approach, more quantitative analytical approaches have been reported. For example, in the fast motion regime (t 10 11–10 9 s at X-band), one can compute the nitroxide rotational correlation time based on the measured line-widths and amplitudes (Marsh, 1981; Qin et al., 2001; Xi et al., 2008). Furthermore, it is possible to simulate a nitroxide spectrum based on quantum mechanics and specific motional models (Columbus et al., 2001; Grant et al., 2009; Hustedt et al., 1993; Liang et al., 2000; Qin et al., 2006; Schneider and Freed, 1989). The details of these advanced analysis techniques are not discussed here, interested readers are instead referred to a recent review (Sowa and Qin, 2008) and the relevant literatures.
4. Examples of Application 4.1. Example 1: cw-EPR studies of the HIV TAR RNA 4.1.1. Variable dynamic signatures in ligand-bound TAR RNA The trans-activation response (TAR) RNA is a structural motif at the 50 -end of the HIV RNA, which interacts with the Tat protein during HIV transcription. The binding site of the Tat protein is an internal loop in the TAR RNA (Fig. 15.10A). The TAR–Tat interaction assures an efficient transcription and thereby facilitates replication of the HIV virus (Frankel, 1992). Therefore, interference with the TAR–Tat interaction has been pursued as a strategy to combat HIV. Below is a description of how cw-EPR spectroscopy has yielded information about RNA dynamics that are correlated with the structure of the TAR RNA receptor bound to various ligands. Four spin-labeled TAR RNAs were prepared using the 20 -labeling strategy described in Fig. 15.4A, each RNA having one spin label in the 20 -position position of either U23, U25, U38, or U40 (Fig. 15.10A) (Edwards et al., 2001). Using the width of the X-band EPR spectra (2Aeff) as a measure of mobility (Fig. 15.10B), it became clear that U23 and U25 had similar mobility, which was faster than that observed for U38 and U40. This was consistent with the fact that U23 and U25 were located in a flexible loop and showed that the spin label was a good reporter of the mobility of nucleotide to which it was attached (Edwards et al., 2001). Thus, these results encouraged investigation of the changes in dynamics of individual nucleotides in the TAR RNA upon ligand binding. The effects of several ligands, such as metal ions (Edwards and Sigurdsson, 2003; Edwards et al., 2002), small organic molecules (Edwards
321
cw-EPR Studies of RNA
A 25
U C 23 U
5'
Change in spectral width (G)
C
Li+
G C G A G
G C U38 C
B U23
U25 A G A C C G G
Na+
U40 C U G G C C C
U38 U40
Mg2+
Hoechst
Neomycin CGP 40336A Cl
5
N
OH N N
H N
H N
O NH N
N H2N
N
NH2
0 U23 U25 U38 U40
U23 U25 U38 U40
U23 U25 U38 U40
U23 U25 U38 U40
U23 U25 U38 U40
U23 U25 U38 U40
NH2 O
HO HO
H2N
–5
O
H2N NH2
O
HO
OH
O O
OH O
H2N
NH2 OH
HO
–10
Figure 15.10 (A) The TAR RNA construct used in this study. Spin-labeled nucleotides are shown in bold. (B) EPR spectra of spin-labeled TAR RNA samples (Edwards et al., 2001). The position of the spin label is indicated on each spectrum. The samples were prepared in 50% sucrose/100 mM NaCl, 10 mM sodium phosphate, 0.1 mM Na2EDTA, pH 7.0. A line indicating the spectral width of U23 TAR RNA is extended through the other spectra, to allow for visual qualitative comparison of the EPR spectral widths. (C) Changes in EPR spectral width plotted as a function of spin-labeled position (U23, U25, U38, and U40) to give a dynamic signature for each metal ion and small molecule that binds to the TAR RNA (Edwards and Sigurdsson, 2002, 2003). Chemical structures of Hoechst, CGP 40336A, and neomycin are shown.
and Sigurdsson, 2002), and peptides (Edwards et al., 2002, 2005) on TAR RNA dynamics were studied by EPR. The EPR data were analyzed by plotting the change in the spectral width (2Aeff) upon ligand binding as a function of spin label position, which gives a dynamic signature for each ligand (Fig. 15.10C). The dynamic signatures are very sensitive to the structure of ligand–receptor complex; of the six examples shown, no two are alike. The first three dynamic signatures are representatives of three classes of metal ions observed when studying the binding of 10 different metal ions to the TAR RNA (Edwards and Sigurdsson, 2003). Not much change is observed in the spectral width for Liþ, whereas a large narrowing
322
Xiaojun Zhang et al.
of the spectral width for U25 and U40 was observed for Mg2þ binding to the TAR RNA. The fact that Naþ and Mg2þ show different dynamic signatures (a large difference in the spectral width of U25) indicates that the two structures are different, which has been verified by NMR spectroscopy (Al-Hashimi et al., 2003). The latter three dynamic signatures in Fig. 15.10C are those of small molecules that have been found to inhibit the TAR–Tat interaction (Edwards and Sigurdsson, 2002). Each one of these molecules has been shown to bind to the TAR RNA in a different manner and all give different dynamic signatures. Therefore, determination of dynamic signatures by cw-EPR spectroscopy can readily identify binding sites in RNA and should be useful for screening libraries of ligands for binding to specific sites in drug-discovery programs. 4.1.2. Dynamic signatures and structural requirements in the TAR–Tat interaction Given the sensitivity of dynamic signatures for the structure of ligands binding to the TAR RNA, this approach was applied to study the interaction of the TAR RNA with its biological partner, the Tat protein. The Tat protein is ca. 86 residues long, but a short peptide spanning this region, YGRKKR52RQRRR (Tat 47–58), binds to the TAR RNA with high affinity (dissociation constant Kd 200 nM ) (Hamy et al., 1993; Roy et al., 1990). Arginine 52 (R52) has been identified as an essential amino acid for the TAR–Tat interaction. Although the NMR structure of the TAR RNA bound to a peptide has not been solved, a high-resolution NMR structure is available for argininamide, which is the simplest analog of the Tat protein, complexed with the TAR RNA (Brodsky and Williamson, 1997). Figure 15.11A shows dynamic signatures for argininamide, the Tat wildtype peptide (YGRKKR52RQRRR) and a Tat mutant peptide (YKKKKR52KKKKA). The mutant peptide has been shown to bind to the TAR RNA with similar affinity as the wild-type sequence (Edwards et al., 2002). The three ligands have similar dynamic signatures: U23, U38, and U40 have lower mobility, while U25 is faster than the unbound TAR RNA. This indicates a similar structural motif for all three complexes. There is, however, a striking difference between the mutant peptide and the wild type: an enormous reduction of mobility of U23 and U38 for the wild type. Interestingly, U23 and U38 are involved in formation of a triple base pair upon binding argininamide (Brodsky and Williamson, 1997). This indicates that amino acids flanking R52 in the wild-type sequence make specific contacts to the RNA, which greatly affects its internal motion. To determine which amino acids promoted specific TAR–Tat complex formation, a series of peptide mutants were prepared and their dynamic signatures collected (Edwards et al., 2005). Figure 15.11B shows the sequence of the peptides and the changes in spectral width for U38 upon
323
cw-EPR Studies of RNA
A
Change in spectral width (G)
20
10
0
U23 U25 U38 U40
U23 U25 U38 U40
YKKKKRKKKKA (Mutant)
Argininamide
U23 U25 U38 U40
YGRKKRRQRRR (Wild-type)
B Argininamide YKKKKRKKKKA YGRKKRRQRRR
YKKKKRRQRRR YGRKKRRQ YGRKKRKKKKK YGRKKRKQRRR YGRKKRRQKRR YGRKKRRQRKR YGRKKRRQRRK 0 R52
10
20
Change in U38 EPR spectral width (G)
Figure 15.11 (A) Dynamic signatures of argininamide and two Tat-derived peptides, the wild-type sequence, and a mutant. (B) Sequences of the Tat-derived peptides and changes in EPR spectral width of U38 in the presence of each peptide. Amino acid mutations are shown in bold. Arrow shows the location of the essential R52 in the peptide sequence.
324
Xiaojun Zhang et al.
binding the peptides. The data clearly show that amino acids flanking R52 at the carboxyl terminus participate in this specific interaction. In particular, the R56K mutant (Fig. 15.11B, second peptide from the bottom) shows a similar spectral width as argininamide, pointing to the importance of this amino acid in the formation of a specific complex. Inspection of the crystal structure of the Tat protein shows that R52 and R56 are the extremes of an apparent palm on the protein surface that is a likely binding site (Edwards et al., 2005). Thus, information about the dynamics of nucleotides in the TAR RNA obtained by cw-EPR has enabled identification of specific structural requirements in the Tat protein and demonstrates the utility of EPR spectroscopy for investigating RNA-protein interactions.
4.2. Example 2: Studying RNA/RNA interactions via cw-EPR reported tR effects The global tumbling of a 25 nucleotide (nt) oligonucleotide (7500 Da) is estimated to have a tR of approximately 4 ns in aqueous solution at room temperature (Cantor and Schimmel, 1980; Qin et al., 2003), which affects the X-band cw-EPR spectrum. In such cases, changes in the size of the system due to interactions with other macromolecules, such as a complimentary strand (Keyes and Bobst, 1998), a partner nucleic acid (Qin et al., 2001), or a protein (Keyes and Bobst, 1998; Xi et al., 2008; Zhang et al., 2008), will change the EPR spectrum. Such spectroscopic changes provide a reporter to monitor the interactions. In the following example, the tR effect was used to study a frequently used RNA tertiary interaction motif— binding between a GAAA tetraloop and its RNA receptor (Qin et al., 2001) (Fig. 15.12). In this study, nitroxide label 14 (Fig. 15.3C) is attached near the 50 terminus of a 12-nt RNA hairpin (TL1, Fig. 15.12A) containing the GAAA tetraloop (boxed region within TL1, Fig. 15.12A). TL1 spectra were obtained in a Mg2þ containing buffer with various concentrations of a 23-nt RNA (TLR, Fig. 15.12A) that includes the GAAA tetraloop receptor (boxed region within TLR, Fig. 15.12A). Line-broadening was observed when comparing spectra obtained in the absence and presence of TLR (Fig. 15.12B). This indicates reduced nitroxide motions. As the nitroxide was attached far away from the sites of interactions, no changes in ti/tB motions are expected, and reduced nitroxide motion report increased tR due to the formation of a higher molecular weight TL1/TLR complex (Fig. 15.12A). This provided the first direct evidence that the GAAA tetraloop can dock into an isolated receptor in solution. Furthermore, spectra obtained in the presence of various TLR concentrations were analyzed using a spectral decomposition procedure (see an example in Fig. 15.12C). The normalized TL1-alone spectrum was scaled and subtracted from normalized ‘‘TL1 þ TLR’’ spectra. At each TLR
Figure 15.12 SDSL studies of GAAA tetraloop binding to its RNA receptor. (A) Schematics of the experimental design, with R indicating the nitroxide label. (B) Normalized TL1 EPR spectra obtained in the absence (i) and presence (ii) of TLR (1.75 mM). Due to weak binding, the ‘‘TL1 þ TLR’’ spectrum is a composite of the unbound TL1 and the complex. (C) Spectral decomposition by subtracting scaled TL1 spectrum from that of ‘‘TL1 þ TLR.’’ The value used for properly subtraction (middle, 21%) represents the fraction of unbound TL1 in the presence of 1.75 mM TLR. If the scaling factor is too small (top, under subtraction), the resulting spectrum shows remaining features of the free TL1, which is most clearly observed by comparing the low-field regions between the under subtracted spectrum (dotted line) and the properly subtracted spectrum (solid line) (see inset). In the opposite case (bottom, over subtraction), the resulting spectrum (dash line) is distorted (indicated by the triangle). (D) Binding curve between TL1 and TLR. Data reproduced from Qin et al. (2001) with permission.
326
Xiaojun Zhang et al.
concentration, the fraction of unbound TL1 was represented by the largest possible scaling factor that gives a difference spectrum resembling an EPR spectrum with the lowest possible mobility. TLR concentration dependence of these fractions thus generates a binding curve, and gave a measured Kd of 0.4 mM between the GAAA tetraloop and its receptor (Fig. 15.12D). This is a weak interaction and is difficult to measure using other methods (Qin et al., 2001). This study highlights two important facets of using the tR effect. First, the nitroxide may be attached at a remote site and therefore will not perturb the native interaction. In addition, as an EPR experiment generally requires a nitroxide concentration greater than 10 mM, the method is suitable for measuring interactions where the Kd is approximately 10–1000 mM.
ACKNOWLEDGMENTS This research was supported by the National Institutes of Health (R01 GM069557, PZQ), Nation Science Foundation (MCB 054652, PZQ), and the Icelandic Research Fund (60028021, STS).
REFERENCES Abakumov, G. A., and Tikhonov, V. D. (1969). Reaction of stable radical 2,2,6,6-tetramethylpiperidone-4-oxyl-1 with acids. Biol. Bull. Acad. Sci. USSR 796–801. Al-Hashimi, H. M., et al. (2003). Mg2þ-induced variations in the conformation and dynamics of HIV-1 TAR RNA probed using NMR residual dipolar couplings. J. Mol. Biol. 329, 867–873. Altenbach, C., et al. (1989). Structural studies on transmembrane proteins. 2. Spin labeling of bacteriorhodopsin mutants at unique cysteines. Biochemistry 28, 7806–7812. Barhate, N., et al. (2007). A nucleoside that contains a rigid nitroxide spin label: A fluorophore in disguise. Angew. Chem. Int. Ed. Engl. 46, 2655–2658. Blair, D. P., and Sydenham, P. H. (1975). Phase sensitive detection as a means to recover signals buried in noise. J. Phys. E 8, 621–627. Brodsky, A. S., and Williamson, J. R. (1997). Solution structure of the HIV-2 TARargininamide complex. J. Mol. Biol. 267, 624–639. Cantor, C. R., and Schimmel, P. R. (1980). Biophysical Chemistry. Vol. 2, pp. 460–564. W.H. Freeman, San Francisco. Cekan, P., et al. (2008). Rigid spin-labeled nucleoside C ¸ : A nonperturbing EPR probe of nucleic acid conformation. Nucleic Acids Res. 36, 5946–5954. Columbus, L., and Hubbell, W. L. (2002). A new spin on protein dynamics. Trends Biochem. Sci. 27, 288–295. Columbus, L., and Hubbell, W. L. (2004). Mapping backbone dynamics in solution with site-directed spin labeling: GCN4–58 bZip free and bound to DNA. Biochemistry 43, 7273–7287. Columbus, L., et al. (2001). Molecular motion of spin-labeled side chains in a-helices: Analysis by variation of side chain structure. Biochemistry 40, 3828–3846. Cova, S., et al. (1979). Versatile digital lock-in detection technique: Application to spectrofluorometry and other fields. Rev. Sci. Instrum. 50, 296.
cw-EPR Studies of RNA
327
Edwards, T. E., and Sigurdsson, S. T. (2002). Electron paramagnetic resonance dynamic signatures of TAR RNA—Small molecule complexes provide insight into RNA structure and recognition. Biochemistry 41, 14843–14847. Edwards, T. E., and Sigurdsson, S. T. (2003). EPR spectroscopic analysis of TAR RNAmetal ion interactions. Biochem. Biophys. Res. Commun. 303, 721–725. Edwards, T. E., and Sigurdsson, S. T. (2007). Site-specific incorporation of nitroxide spinlabels into 20 -positions of nucleic acids. Nat. Protoc. 2, 1954–1962. Edwards, T. E., et al. (2001). Site-specific incorporation of nitroxide spin labels into internal sites of the TAR RNA: Structure-dependent dynamics of RNA by EPR spectroscopy. J. Am. Chem. Soc. 123, 1527–1528. Edwards, T. E., et al. (2002). Investigation of RNA-protein and RNA-metal ion interactions by electron paramagnetic resonance spectroscopy. The HIV TAR–Tat motif. Chem. Biol. 9, 699–706. Edwards, T. E., et al. (2005). Identification of amino acids that promote specific and rigid TAR RNA-tat protein complex formation. Chem. Biol. 12, 329–337. Fajer, P. G. (2000). Electron spin resonance spectroscopy labeling in proteins and peptides analysis. In ‘‘Encyclopedia of Analytical Chemistry,’’ (R. Meyers, ed.), pp. 5725–5761. Wiley, Chichester. Fidanza, J. A., et al. (1992). Site-specific labeling of DNA-sequences containing phosphorothioate diesters. J. Am. Chem. Soc. 114, 5509–5517. Fischhaber, P. L., et al. (1997). Synthesis of duplex DNA containing a spin labeled analog of 2’-deoxycytidine. Nucleosides Nucleotides 16, 365–377. Frankel, A. D. (1992). Activation of HIV transcription by Tat. Curr. Opin. Genet. Dev. 2, 293–298. Gannett, P. M., et al. (2002). Probing triplex formation by EPR spectroscopy using a newly synthesized spin label for oligonucleotides. Nucleic Acids Res. 30, 5328–5337. Grant, G. P., et al. (2008). Diastereomer characterizations of nitroxide-labeled nucleic acids. Biochem. Biophys. Res. Commun. 371, 451–455. Grant, G. P. G., et al. (2009). Motions of the substrate recognition duplex in a group I intron assessed by site-directed spin labeling. J. Am. Chem. Soc. 131, 3136–3137. Hamy, F., et al. (1993). Hydrogen-bonding contacts in the major groove are required for human immunodeficiency virus type-1 tat protein recognition of TAR RNA. J. Mol. Biol. 230, 111–123. Hara, H., et al. (1970). 4-Thiouridine-specific spin-labeling of E. coli transfer RNA. Biochem. Biophys. Res. Commun. 38, 305–311. Hermann, D. (1977). Spin-labeled nucleic acids. Acc. Chem. Res. 10, 47–54. Hustedt, E. J., et al. (1993). Motions of short DNA duplexes: An analysis of DNA dynamics using an EPR-active probe. Biochemistry 32, 1774–1787. Keyes, R. S., and Bobst, A. M. (1998). Spin-labeled nucleic acids. In ‘‘Biological Magnetic Resonance,’’ (L. J. Berliner, ed.), Vol. 14, pp. 283–338. Plenum Press, New York. Keyes, R. S., et al. (1997). Overall and internal dynamics of DNA as monitored by fiveatom-tethered spin labels. Biophys. J. 72, 282–290. Kim, N. K., et al. (2004). A distance ruler for RNA using EPR and site-directed spinlabeling. Chem. Biol. 11, 939–948. Klug, C. S., and Feix, J. B. (2008). Methods and applicants of site-directed spin labeling EPR spectroscopy. Methods Cell Biol. 84, 617–658. Liang, Z., et al. (2000). An electron spin resonance study of DNA dynamics using the slowly relaxing local structure model. J. Phys. Chem. 104, 5372–5381. Macosko, J. C., et al. (1999). A novel 5 displacement spin-labeling technique for electron paramagnetic resonance spectroscopy of RNA. RNA 5, 1158–1166. Marsh, D. (1981). Electron spin resonance: Spin labels. Mol. Biol. Biochem. Biophys. 31, 51–142.
328
Xiaojun Zhang et al.
Mchaourab, H. S., et al. (1996). Motion of spin-labeled side chains in T4 lysozyme. Correlation with protein structure and dynamics. Biochemistry 35, 7692–7704. Miller, T. R., et al. (1995). A probe for sequence-dependent nucleic acid dynamics. J. Am. Chem. Soc. 117, 9377–9378. Nagahara, S., et al. (1992). Spin-labeled oligonucleotides site specifically labeled at the internucleotide linkage—Separation of stereoisomeric probes and EPR spectroscopical detection of hybrid formation in solution. Nucleosides Nucleotides 11, 889–901. Okamoto, A., et al. (2004). Nitroxide-labeled guanine as an ESR spin probe for structural study of DNA. Bioorg. Med. Chem. 14, 3415–3418. Piton, N., et al. (2007). Base-specific spin-labeling of RNA for structure determination. Nucleic Acids Res. 35, 3128–3143. Poole, C. P. (1996). Electron spin resonance: A comprehensive treatise on experimental techniques. Dover Publications, Inc., Mineola, NY. Prisner, T., et al. (2001). Pulsed EPR spectroscopy: Biological applications. Annu. Rev. Phys. Chem. 52, 279–313. Qin, P. Z., et al. (2001). Quantitative analysis of the isolated GAAA tetraloop/receptor interaction in solution: A site-directed spin-labeling study. Biochemistry 40, 6929–6936. Qin, P. Z., et al. (2003). Monitoring RNA base structure and dynamics using site-directed spin-labeling. Biochemistry 42, 6772–6783. Qin, P. Z., et al. (2006). A model system for investigating lineshape/structure correlations in RNA site-directed spin labeling. Biochem. Biophys. Res. Commun. 343, 117–124. Qin, P. Z., et al. (2007). Measuring nanometer distances in nucleic acids using a sequenceindependent nitroxide probe. Nat. Protoc. 2, 2354–2365. Roy, S., et al. (1990). A bulge structure in HIV-1 TAR RNA is required for Tat binding and Tat-mediated trans-activation. Genes Dev. 4, 1365–1373. Schiemann, O., et al. (2004). A PELDOR-based nanometer distance ruler for oligonucleotides. J. Am. Chem. Soc. 126, 5722–5729. Schiemann, O., et al. (2007). Spin-labeling of oligonucleotides with the nitroxide TPA and use of PELDOR, a pulse EPR method, to measure intramolecular distances. Nat. Protoc. 2, 904–923. Schneider, D. J., and Freed, J. H. (1989). Calculating slow motional magnetic resonance spectra: A user’s guide. In ‘‘Spin Labeling: Theory and Applications,’’ (L. J. Berliner, ed.), Vol. 3, pp. 1–76. Plenum Press, New York. Sowa, G. Z., and Qin, P. Z. (2008). Site-directed spin labeling studies on nucleic acid structure and dynamics. Prog. Nucleic Acid Res. Mol. Biol. 82, 147–197. Spaltenstein, A., et al. (1988). A rigid and nonperturbing probe for duplex DNA motion. J. Am. Chem. Soc. 110, 1299–1301. Varani, L., et al. (1999). Structure of tau exon 10 splicing regulatory element RNA and destabilization by mutations of frontotemporal dementia and parkinsonism linked to chromosome 17. Proc. Natl. Acad. Sci. USA 96, 8229–8234. Xi, X., et al. (2008). HIV-1 nucleocapsid protein NCp7 and its RNA stem loop 3 partner: Rotational dynamics of spin-labeled RNA stem loop. Biochemistry 47, 10099–10110. Zhang, Z., et al. (2008). Rotational dynamics of HIV-1 nucleocapsid protein NCp7 as probed by a spin label attached by peptide synthesis. Biopolymers 89, 1125–1135.
C H A P T E R
S I X T E E N
Mapping Global Folds of Oligonucleotides by Pulsed Electron–Electron Double Resonance Olav Schiemann Contents 330 332 332 333 334 335 335 335 338 339 340 340 341 345 346 347 347
1. Introduction 2. Overview 2.1. The PELDOR pulse sequence 2.2. The PELDOR time trace 2.3. Extracting the distance 3. The Experiment 3.1. EPR-sample preparation 3.2. The PELDOR experiment 3.3. Data analysis 4. Structure Generation 5. Beyond Distances 5.1. Spin counting 5.2. Spin label orientations 5.3. Through-bond and through-space coupling 6. Comparison with Other Methods Acknowledgments References
Abstract The understanding of structure–dynamics–function relationships in oligonucleotides or oligonucleotide/protein complexes calls for biophysical methods that can resolve the structure and dynamics of such systems on the critical nanometer length scale. A modern electron paramagnetic resonance (EPR) method called pulsed electron–electron double resonance (PELDOR or DEER) has been shown to reliably and precisely provide distances and distance distributions in the range of 1.5–8 nm. In addition, recent experiments proved that a PELDOR experiment also contains information on the orientation of labels, Centre for Biomolecular Sciences, Centre of Magnetic Resonance, University of St Andrews, St Andrews, United Kingdom Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69016-9
#
2009 Elsevier Inc. All rights reserved.
329
330
Olav Schiemann
enables easy separation of coupling mechanisms and allows for counting the number of monomers in complexes. This chapter briefly summarizes the theory, describes how to perform and analyze such experiments and discusses the limitations.
1. Introduction During the last 10 years, RNA and its structure–dynamics–function relationship has become a major research area, in particular, triggered by the finding of RNAi (Denli and Hannon, 2003; Voinnet, 2001) and of riboswitches (Winkler and Breaker, 2003), the X-ray structures of the ribosome (Ban et al., 2000; Nissen et al., 2000), and new insights into the spliceosome (Luehrmann and Stark, 2009). It appears that the folding of RNA and its conformational changes, induced by interactions with proteins, metal ions, or small molecules, are essential for its biological function. Thus, current RNA research engages ever larger and more complex systems, driven by the realization that a full understanding of RNA function needs to target the multicomponent assembly in which it is embedded. It is thus a fundamental challenge for modern biophysics to provide methods by which these nanometer-size architectures can be quantitatively analyzed. Electron paramagnetic resonance (EPR) methods are ideally suited to address these challenges, because they can be applied to biomolecules of any size, provide distances in the critical range of 110 nm, and resolve conformational distributions and dynamics. For proteins, modern EPR spectroscopy has proven its efficiency in characterizing the local structures of binding sites of, for example, metal ions, metal clusters, or organic cofactors ( Jeschke, 2005; Prisner et al., 2001; Silakov et al., 2009, Ubbink et al., 2002). For RNA, examples include the work on the local structure of metal(II) ion binding sites in the minimal (Morrissey et al., 1999; Schiemann et al., 2003; Vogt and DeRose, 2006) and tertiary stabilized hammerhead ribozymes (Kim et al., 2005; Kisseleva et al., 2005; Wolfson et al., 2008) and the Diels–Alder ribozyme (Kisseleva et al., 2007). For both, proteins and nucleic acids, EPR has also been extensively used to probe local and overall dynamics (Berliner, 1998; Grant et al., 2009). However, whereas EPR studies on the arrangements of domains in proteins and protein complexes are numerous (Borbat et al., 2002; Grote et al., 2008; Hagelu¨ken et al., 2009; Hubbel et al., 2000; Hustedt and Beth 2000; Klare et al., 2004; Liu et al., 2001; Park et al., 2006, Ward et al., 2009), they are apart from oligonucleotide duplexes (Cai et al., 2006; Kuznetsov et al., 2009; Schiemann et al., 2004; Sicoli et al., 2008; Sowa and Qin 2008), rare for complex folds of RNA (Kim et al., 2004; Macosko et al., 1999; Qin and Dieckmann, 2004). One reason for the
331
Mapping Global Folds of Oligonucleotides by PELDOR
lack of EPR studies on more complex nucleic acid structures is that they require efficient and site-directed labeling with nitroxides and it is only recently that such labeling strategies have been developed (see Chapter 15 in this volume). Furthermore, EPR methods capable of reliably measuring the distance between two of these nitroxides were needed. Continuous wave (CW) EPR methods work up to 2 nm but their reliability imposes problems (Berliner et al., 2000). A pulsed EPR technique nowadays routinely used for precise nanometer distance measurements in the range of 1.5–8 nm is pulsed electron–electron double resonance (PELDOR or DEER); a more specialized one is double quantum coherence EPR (DQC-EPR) (Borbat et al., 2004). Milov et al. (1981) were the first to introduce the three-pulse version of PELDOR (Fig. 16.1A) and the groups of Spiess (Martin et al., 1998) and Jeschke (Pannier et al., 2000) extended it to the dead-time-free four-pulse version (Fig. 16.1B). In addition to the mean distance, four-pulse PELDOR provides also distance distributions which have been correlated with dynamics (Godt et al., 2006; Margraf et al., 2007) and enables counting the number of monomers in a complex as demonstrated on model systems (Bode et al., 2007; Jeschke et al., 2009) and for a 320-kDa channel protein (Hagelu¨ken et al., 2009). It also yields angular information as shown first at high-frequency/high-field on the ribonucleotide reductase (Denysenkov A n1
p
p /2 t
t
n2
p T
B n1
p
p
p/2 t1
n2
t
t1
t2
t
t2
RE
p
Figure 16.1 Pulse sequences for (A) three-pulse PELDOR and (B) four-pulse PELDOR (RE denotes the wanted refocused echo).
332
Olav Schiemann
et al., 2006) and subsequently for DNA using a commercial X-band spectrometer (Schiemann et al., 2009). Interestingly, the spin labels can also be paramagnetic metal centers, allowing, for example, the triangulation of metal ions. Examples were published for model systems (Bode et al., 2008; Narr et al., 2002) and for proteins (Amsterdam et al., 2003; Elsa¨sser et al., 2002; Kay et al., 2007). This chapter will first give an overview of the approach and briefly summarize the PELDOR theory and then describe how to set up and analyze a PELDOR experiment. Two review articles covering the field of EPR-based nanometer distance measurements have been published recently; one summarizes the whole range of EPR methods (Schiemann and Prisner, 2007), the other is focused on PELDOR ( Jeschke and Polyhach, 2007).
2. Overview Section 2 is intended to provide a flow chart of the whole procedure without discussion or experimental details, these follow then in Sections 3–5. This might make it easier for the reader to follow why certain steps are made in the way they are made. The general approach is based on the idea that the arrangement of two domains in a biomolecule can be resolved by measuring a set of distances across these domains and then mapping the collected distances onto a structural ensemble. The structure that agrees the best with the long-range constrains is the right one. Thus, one has to define for each distance measurement two points, one on each domain, between which the distance is measured. For oligonucleotides these points will most commonly be two site-directedly attached nitroxide spin labels (see Chapter 15 in this volume). EPR methods as for example PELDOR are then used to measure the dipolar spin–spin coupling ndd between the two unpaired electrons localized in the N–O bond of the nitroxides. Using the point-dipole approximation (Riplinger et al., 2009), this coupling is then parameter free converted into the distance between the two unpaired electrons.
2.1. The PELDOR pulse sequence The four-Pulse PELDOR pulse sequence is in principle a pump-probe experiment that selects from all contributions to the EPR spectrum dominantly the dipolar coupling between the two unpaired electrons. The detection sequence p/2–t1–p–t1–echo1–t2–p (Fig. 16.1B) applied at a microwave frequency nA creates at time t2 after the last p-pulse a refocused echo (RE) for the spins being in resonance with nA. The position in time of
333
Mapping Global Folds of Oligonucleotides by PELDOR
the RE does not change during the experiment, as the time intervals of the detection sequence are not changed. Introduction of the inversion pulse at pump frequency nB during the time interval T flips spins which are resonant with this second frequency. Incrementing the pump pulse from t ¼ 0 to t ¼ T leads to an oscillation of the amplitude V of the RE in dependence of t, if the two spins are coupled. The frequency of this oscillation is the dipolar coupling ndd between both spins.
2.2. The PELDOR time trace Since PELDOR is not a single molecule experiment, the PELDOR signal V(t) contains not only the contribution from the interaction between the two spins in one molecule but also from the interaction between spins on different molecules and is thus considered a product of two contributions (Eq. (16.1), Fig. 16.2A) (Milov et al., 1984). V ðtÞ ¼ V ðtÞintra V ðtÞinter
A
B
1.0 0.9
V intra(t)
V(t)
1.0 0.9
0.8 0.7 0.6 0.5
0.8 0.7 0.6 0.5
0.4 0
1
2 t (ms) C 1.0 Normalized intensity
0.3
ð16:1Þ
3
4
q ⊥B0
0.4
0
1
2 t (ms)
3
4
q ⊥B0
0.8 0.6 0.4 0.2
q ||B0
q ||B0
0.0 –2
–1 0 1 Frequency (ndd)
2
Figure 16.2 (A) The green curve shows the original PELDOR time trace V(t). The red curve is the fitted monoexponential decay function, representing V(t)inter. (B) V(t)intra after division of V(t) by V(t)inter. (C) Sketch of a dipolar Pake pattern obtained from the time trace in (B) after Fourier transformation.
334
Olav Schiemann
V(t)intra describes all spins coupled intramolecularly, whereas V(t)inter takes into account the signal decay caused by intermolecular interactions between homogeneous distributed spins in the sample. V(t)intra can be described by Eq. (16.2): + * n Y n 1 X V ðtÞintra ¼ ½1 lB ð1 cosðndd tÞÞ ð16:2Þ n A¼1 B ¼1 B 6¼A
with ndd ¼
m0 hgA gB ð1 3 cos2 yÞ 3 4p rAB
ð16:3Þ
where n is the number of radicals per biomolecule, lB is the fraction of B spins inverted by the pump pulse, t is the time delay of the pump pulse, ndd is the full dipolar splitting, m0 is the vacuum permeability, gA and gB are the magnetogyric ratios of the two spins, h is the Planck constant, rAB is the distance between the spins, y is the angle between the distance vector rAB and the external magnetic field B0, and h. . .i is the averaging over values of rAB and y. Equation (16.3) is based on the point dipole approximation, which is full-filled for large distances and nondelocalized systems (Riplinger et al., 2009). V(t)inter can be written in the form of a monoexponential decay function (Eq. (16.4)), where c is the radical concentration in spin/m3 and h the Planck constant divided by 2p. 2pgA gB m0 h pffiffiffi clB t : V ðtÞinter ¼ exp ð16:4Þ 9 3 Dividing V(t) by this decay function leaves the frequency modulated signal V(t)intra (Fig. 16.2B) of the wanted intramolecular interaction (Eq. (16.2)).
2.3. Extracting the distance The frequency of this oscillation is ndd and can be obtained by Fourier transformation of V(t)intra. This yields not a single frequency but a frequency/intensity pattern that is called the Pake pattern (Fig. 16.2C), because the molecules are randomly oriented with respect to the outer magnetic field B0 and ndd depends on the angle y between the distance vector rAB and B0. The peaks and edges of the Pake pattern correspond to y ¼ 90 ¼ y? and y ¼ 0 ¼ yk, respectively. Reading off the frequency at y? and substituting it into Eq. (16.5), the rearranged version of Eq. (16.3) yields the distance rAB between the two nitroxide spins: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 52:16 3 ð16:5Þ rAB ðnmÞ ¼ nðy? Þ ðMHzÞ
Mapping Global Folds of Oligonucleotides by PELDOR
335
Instead of Fourier transforming the time traces, they are commonly analyzed with the program DeerAnalysis2008 from Jeschke et al. (2006), which is based on Tikhonov regularization and the assumption that the full Pake pattern is excited.
3. The Experiment 3.1. EPR-sample preparation Commonly the EPR sample is prepared as a 50 mM RNA solution (100 mM in spins) in the appropriate sterile buffer. To ensure a homogeneous sample and thus a good signal-to-noise ratio and a long time window for T, a cryoprotectant is added to the buffer. The amount and type of cryoprotectant may vary from RNA to RNA and depends on a compromise between long PELDOR time window (especially needed for long distances) and structural stability. In several cases, 20% (v/v) ethylene glycol have been used although percentages of up to 50% have been reported (Ward et al., 2007). Time windows can additionally be increased by using deuterated water and deuterated cryoprotectant and by deoxygenating the sample. Then the sample is transferred into a sterilized quartz EPR tube using a sterile Hamilton syringe with Teflon tubing to reach the bottom of the tube. Freeze the sample by immersing the tube quickly into a precooled mixture of methyl-cyclohexane:isopentane 1:4 (cooled close to the freezing point); this freezes the sample quicker compared to immersing it directly into liquid nitrogen and, thus, reduces the formation of ice crystals and makes the sample more homogeneous. The sample can be stored in liquid nitrogen for months.
3.2. The PELDOR experiment The following protocol is worked out for the commercial pulsed X-band EPR spectrometer ELEXSYS E580 from Bruker (also see the Bruker manual) but can also be applied in modified versions to home-built spectrometers. It is based on a protocol published earlier (Schiemann et al., 2007). 3.2.1. Switching on the spectrometer Switch on the pulsed EPR spectrometer with the two microwave (mw) sources for the pump and probe pulses. Cool the sample cryostat down to 50 K and insert the frozen sample tube. Overcouple the resonator to a Q-value of approximately 50 and adjust the mw frequency to the center of the resonator frequency.
336
Olav Schiemann
3.2.2. Optimize the echo signal Program a two-pulse Hahn-echo sequence using 16 and 32 ns pulses, a pulse separation time of 200 ns and a shot repetition time of 6 ms. Set the detection video bandwidth to 25 MHz. Switch the TWT on and check that there is no cavity ringing observable after the protection gate with full mw power. If you see ringing, problems might be: (a) The Q of the cavity is too high. In this case reduce it by overcoupling the resonator more. (b) If (a) does not apply, the receiver protection pulse might be too short. In this case adjust the length of the receiver protection gate, if you are in doubt contact Bruker. Sweep the magnetic field until an echo appears in the Specjet. Optimize the echo intensity by varying the mw power and by adjusting the phase of the detection channels. Decrease the shot repetition time until the echo intensity decreases. Sweep the magnetic field until the echo is maximal. This magnetic field is kept throughout the PELDOR experiment. Readjust the sample position in the cavity to maximize the echo intensity. 3.2.3. Optimize the PELDOR inversion pulse Place a 20-ns pulse in the ELDOR channel 500 ns before the probe Hahn-echo sequence in Section 3.2.2 and use for the ELDOR frequency the same frequency as the probe frequency. Give the inversion pulse a length of 20 ns and decrease its attenuation to 0 dB. Optimize the inversion of the Hahn-echo by decreasing the length of the inversion pulse. Usual values are 0 dB attenuation and a pulse length of 12–18 ns, depending on the performance of the spectrometer/cavity. The length of the inversion pulse is critical for the deepness of the modulation. A longer inversion pulse leads to less modulation amplitude and a shorter inversion pulse will lead to a stronger overlap of detected and pumped spins and thus to a decreased signal-to-noise ratio and artifacts. 3.2.4. Set the probe pulse frequency Switch off the high power pulses by increasing the global attenuation and the attenuation of the ELDOR channel to 30 dB and switch the TWT to standby. Increase the frequency of the probe pulses by 80 MHz. Switch the TWT back on. The value by which the frequency is increased depends on the sample. With a frequency offset of 80 MHz, the detection pulses are applied at the low-field edge of the nitroxide field sweep spectrum. This decreases slightly the echo intensity, but gives rise to a more pronounced modulation since the pulses are more selective at this field position. However, frequency offsets of 70 or 60 MHz maybe better if the DeerAnalysis2008 program is used since the program works the best if the whole Pake pattern is excited.
Mapping Global Folds of Oligonucleotides by PELDOR
337
3.2.5. Optimize the detection pulses Create a Hahn-echo sequence with 16/32 ns in one channel and optimize the echo intensity. Do the same in a second channel and adjust its phase 180 to the first cannel. Both channels are used for the two-step phase cycle. Running the experiment with the two-step phase cycle eliminates receiver offsets and allows reading off the real echo signal intensity, which is important for accurately determining Vl in spin counting experiments (Section 5.1). On the ELEXSYS E580, this step is done via PulseSpel. One loads the program and the variable file . Pulse lengths and timings are typed into the variable file. Then the subroutine <<þx><x>observer> of the program are run together with the phase programs <þx> and <x>. The amplitudes and phases of both channels are adjusted to optimize the echoes in both channels with a 180 shift between them. If one wants to use the same pulse length for the p/2 and p pulses, one needs three MPFU channels and the protocol is different (Schiemann et al., 2007).
3.2.6. Set up the four-pulse PELDOR sequence On the ELEXSYS E580, the complete four-pulse PELDOR sequence is set up using the subroutine <standing DEER> together with the phase program . Type the length of the inversion pulse (determined in Section 3.2.4) and the pulse separation times T and t1 into the variable file. T and t are called d1 and d2 in PulseSpel. The time interval T has to be chosen long enough to detect at least one oscillation frequency of the PELDOR signal (depends on the expected distance rAB). Set the attenuation of the ELDOR source to 0 dB. Run the pulse sequence, adjust the global phase until the echo points upwards and optimize the video amplifier gain. In principle, the video amplifier gain can be increased to its maximum; however, care has to be taken that neither the noise nor the echo are clipped in the single shot mode of the Specjet. The pulse separation time t1 can be either set to a fixed value (a typical value would be 136 ns, which suppresses hydrogen modulation, or to 380 ns for the suppression of deuterium modulations) or varied over a range of values to strongly suppress hyperfine modulations (eight steps of 8 ns to suppress the free Larmor frequency of hydrogen). The latter option is included in the program of PulseSpel and just requires to enter the values for the number of steps and the step width into the variable file. The pump inversion pulse starts with a delay of 80–100 ns after the first p-pulse of the detection sequence to avoid interferences and is swept within the experiment up to 80–100 ns before the second p-pulse. The step width depends on the period of the modulation, a good starting point might be 12 ns. The acquisition trigger is placed on the
338
Olav Schiemann
RE (trigger delay time d0 in PulseSpel). Be careful to choose the right echo, there are four unwanted echoes. The right one is the one that points into the opposite direction of the unwanted echoes. 3.2.7. Run the experiment Switch to the subroutine and the phase program and start the experiment. The number of scans and shots per point depend on the signal-to-noise ratio. After the experiment is finished store the time trace.
3.3. Data analysis 3.3.1. Getting started Adjust the t-axis of the experiment until the actual time point zero is at the maximum amplitude of the echo decay function. The now negative time part is usually disregarded. With mathematical procedures it is also possible to correct the phase of the time trace using the real and imaginary part of the data. In DeerAnalysis2008 both steps are done automatically but can also be adjusted. 3.3.2. The background function The next step, the fitting of the intermolecular background, is the most critical step of the data analysis. This is rather easy if modulations can be seen and if they are damped to the end of the time trace. For a homogeneously three-dimensional distributed sample, the background is fitted as a monoexponential decay to the part of the time trace where all modulation is damped. The original time trace is then divided by this background function. A way to find out if the fit was good is to perform a Fourier transformation and to check the resulting Pake pattern. If a peak or a hole appears at the zero frequency then the background was too shallow or too steep, respectively. This is done automatically in DeerAnalysis2008. If the time trace shows no modulations, the signal-to-noise is bad, the PELDOR effect is weak and the time trace is short, it will by tremendously difficult to fit the background. In such cases the background fit will easily be a source for a wrong analysis; wrong in terms of both mean distance and distance distributions. In addition, for biomolecules that tend to aggregate, the background will not be a simple monoexponential decay function and it is thus no problem to introduce or erase modulations by fitting the wrong background. Between this extreme and the high-quality PELDOR data with modulations, there are various shades of gray and every scientist has to decide on a case-to-case basis whether and how to treat such data. However, the analysis of such data requires great care, comparison
Mapping Global Folds of Oligonucleotides by PELDOR
339
measurements, and experience. Sometimes it can help to run a sample in which the biomolecule is only labeled with a single spin label, but be aware of the fact that the distribution function of a singly labeled sample is different compared to a doubly labeled sample. This author advocates that instead of analyzing such critical data, it would be better to try to improve the sample quality and/or to check whether all settings of the spectrometer were correct. It is clear that a flexible region of the biomolecule will not give a modulation but in this case one may want to stick to the information that the region is flexile instead of drawing tables of mean values with errors of 0.1 nm and precise values for distributions. PELDOR is a very reliable, precise, and parameter free method as long as modulations are observed; if they are not observed this is not the case and one is prone to completely wrong structures. 3.3.3. Determining the distance and the distance distribution A first estimate of the distance can be obtained via Fourier transformation of the time trace into the frequency domain. Reading off the frequency value for y? and substituting it into Eq. (16.5) will yield a distance. More information, especially about the distribution in r can be obtained by using a mathematical method called Tikhonov regularization as included in DeerAnalysis2008. Note: the distribution in r is encoded in the damping of the modulation. The optimization of the distribution function P(r) is done by minimizing the deviation of the simulated time trace from the experimental one and by searching for a smooth P(r) distribution. The regularization parameter weighting these two contributions influences the solution strongly and can be optimized with the so-called L-curve.
4. Structure Generation Having collected a set of long-range distances one has to convert these distances into a structure. This can be done by constraining, for example, MD simulations, molecular modeling approaches, or docking models. Important for this approach is to include the spin label, as the label has a certain length and may adopt different conformations and orientations. This can either be done by including the labels into the MD simulation directly (Schiemann et al., 2004), molecular modeling (Bo¨hme et al., 2009), or by using rotamer libraries (Grote et al., 2008; Hilger et al., 2009). For RNA labels, rotamer libraries and a structure generation program have been developed by Qin and coworkers (Cai et al., 2007) and are both available via the internet.
340
Olav Schiemann
5. Beyond Distances Below it is shown that PELDOR spectra contain also information about numbers of spins coupled in a complex, spin label orientation and coupling mechanisms and how to extract this information.
5.1. Spin counting In order to count the monomers in RNA/protein complexes, one can use the following protocol. First, the constituting monomers are spin labeled with one label per monomer, then they are mixed under conditions yielding the complex, a PELDOR time trace is acquired, the intermolecular decay divided off and the intensity normalized to unity. This follows the routine as described above. However, now it is not the frequency of the modulation that is of interest but the value Vl one reads off on the y-axis for long times t where all modulation is damped (Fig. 16.3) (Milov et al., 1984). From the value of Vl, the number of spins coupled in one complex and thus the number of monomers can be calculated according to Eq. (16.6): n¼
ln Vl þ1 lnð1 lB Þ
ð16:6Þ
The only free parameter in the calculation of n is lB, the fraction of spins inverted by the pump pulse. The value of lB is determined from an independent PELDOR experiment of a standard biradical measured 1.0
Mono
Normalized echo amplitude
0.9 0.8 0.7 0.6
Bi
0.5 0.4 0.3
Tri
0.2
Tetra
0.1 0.0 0
1000
2000 3000 t (ns)
4000
5000
Figure 16.3 PELDOR time traces for a mono- (Vl ¼ 1), bi- (Vl ¼ 0.54), tri- (Vl ¼ 0.31), and tetraradical (Vl ¼ 0.20) (taken from Bode et al., 2007).
341
Mapping Global Folds of Oligonucleotides by PELDOR
under the same buffer and spectrometer conditions (pulse lengths, cavity coupling, etc.). From its Vl value, lB is calculated as 1 Vl. Care has to be taken that: (1) All modulation is damped. This requires long time traces which can be a problem for large distances. (2) The background has to be fitted carefully. Bad fits can give wrong Vl values. (3) In the case of mixtures of oligomers one has to work with the weighted ratios and account for dipolar relaxation (Bode et al., 2007). Attempts to solve this problem have been published recently (Hagelu¨ken et al., 2009; Jeschke et al., 2009). (4) Higher order spin correlations are not included in DeerAnalysis2008. Using this program for multispin systems may give wrong distance distributions ( Jeschke et al., 2009).
5.2. Spin label orientations Usually, the attached spin labels have enough motional freedom that their orientations are not correlated with each other. However, if the labels are buried in an RNA fold or hydrogen bonded to the RNA, then their mobility is strongly reduced and their orientation will be fixed with respect to each other. This is not a bad case but enables not only to determine the distance distribution but also the orientation of the labels. Indeed, recently a spin label has been developed for oligonucleotides that is rigid to allow for such orientation-selective PELDOR experiments (Fig. 16.4, Barhate et al., 2007). The general idea is that the spectrum is broader than the pulse excitation band width. The pulses will then select certain orientations depending on which position on the field sweep spectrum they are applied. At high field/high frequencies the full orientation can be resolved (Denysenkov et al., 2006) but also at X-band orientation information can be obtained (Schiemann et al., 2009).
O N N N
O N
H H
O N Ç
N
N
H O
N N H G
Figure 16.4 Spin label C ¸ base-paired to G.
342
Olav Schiemann
At X-band and for nitroxides, the experiment is performed by applying the pump pulse on the maximum of the nitroxides spectrum (Fig. 16.5A) at the center of the cavity. At this position the pump pulse selects all orientations (Fig. 16.5B). The frequency of the detection sequence is applied at 90 MHz higher frequency than the pump pulse frequency. At this frequency the pulses excite the low-field edge of the nitroxides field sweep spectrum, which is the Azz component of the hyperfine tensor (Fig. 16.5C). Stepping the frequency offset in 10 MHz steps down to 40 MHz (each frequency offset is an independent PELDOR experiment that is set up as described
A
I I I Axx III Ayy I
I
1.0
I Azz
V/Vmax
0.5
0.0 336
338
340
B
342 344 B0 /mT C
346
348
D
350 z
x
–y 0
0.2
0.4
0.6
0.8
1
Figure 16.5 (A) Field swept EPR spectrum of DNA I at 40 K with the excitation profiles of the observer pulses (purple) and pump pulse (olive) and the corresponding 14 N stick spectrum. Arrows indicate observer pulse positions varying from Dn ¼ 40 MHz (orange) to 90 MHz (black). (B) Excited orientations for the pump pulse and (C) the detection sequence for Dn ¼ 90 MHz and (D) 40 MHz (taken from Schiemann et al., 2009).
343
Mapping Global Folds of Oligonucleotides by PELDOR
above) deselects Azz and excites more of the Axx, Ayy, and off-diagonal elements of the hyperfine tensor (Fig. 16.5D). If the oligonucleotide is rigid, the hyperfine tensor has a fixed orientation with respect to the dipolar distance vector r and selection of a certain hyperfine tensor component selects a specific component of the dipolar Pake pattern, which depends on the orientations of the spin labels. Thus, if excitation of Azz yields the perpendicular component of the Pake pattern and excitation of Axx and Ayy more of the parallel component then Azz is perpendicular to the distance vector and the two spin labels are arranged in the same plane and point in the same or opposite direction (model system 1 in Fig. 16.6C). If selection of Azz yields the parallel component and the perpendicular appears at smaller frequency offsets then the two labels are standing on top of each other pointing in the same direction (DNA I in Fig. 16.6A). If the angle between the distance vector r and Azz is 45 , the frequency of the modulation does not change, only the modulation depth is a function of the frequency offset (DNA II in Fig. 16.6B). Other orientations manifest themselves in intensity variations of these three cases (Fig. 16.7). Beyond this rough by-eye analysis, a more sophisticated analysis by simulating the
B
A Azz
r
Azz
r
j = 5⬚
j Azz
Azz j
t-Bu
C
t-Bu
N
N
N
N
N
N
O–N
N–O
t-Bu
t-Bu
Azz
j
j = 45⬚
Azz
r j = 90⬚
Figure 16.6 (A, B) DNA I and DNA II, respectively, with the spin label C ¸ highlighted in red and simplified schematic structures showing the orientation of Azz with respect to r. (C) Structure of 1 and a simplified scheme of the orientation of Azz with respect to r (taken from Schiemann et al., 2009).
A 1.0
1
40 50 60 70 80 90
MHz MHz MHz MHz MHz MHz
0.6
C
Simulation
FT
q = 90°
1.0
1.0
0.8 0.8 V/V0
0.8 V/V0
B
Experiment
0.6
q = 0°
0.4
0.6
0.2 0.4
t(ns)
1000
1500
E
1.0
DNA I
0.8
0.6
0.6
0.4 0
500
t(ns)
1000
H
DNA II
1.0
500
t(ns)
1000
−10 −8 −6 −4 −2 0 2 4 6 8 10 Frequency(MHz)
1500
F 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2
0.4 0.2
1500
0
1.0
0.8
0.2
G
500
V/V0
V/V0
D
0
0.4
0
500
t(ns)
1000
q = 90° q = 0°
−6
1500
−4 −2 0 2 Frequency(MHz)
4
6
I
1.0
2.0 0.8 V/V0
V/V0
0.8 0.6
1.5
0.6
0.5 0.4
0
500
t(ns)
1000
1500
0.4
q = 90°
1.0
0
500
t(ns)
1000
1500
Figure 16.7 (Continued )
q = 0°
−20 −15 −10 −5 0 5 10 15 20 Frequency(MHz)
Mapping Global Folds of Oligonucleotides by PELDOR
345
time traces allows a precise determination of the angle between r and Azz. Knowing the orientation of the labels makes a translation of the spin label to spin label distance much easier and less distance measurements are required. Unfortunately, at the moment only simulation programs that are not freely available are capable to analyze such data. Applying DeerAnalysis2008 to orientation, selected time traces will yield wrong distance distributions and mean distances.
5.3. Through-bond and through-space coupling In cases in which the distance between both nitroxides is below 1.5 nm or if the nitroxide group of the spin label is part of a conjugated system (e.g., spin label C ¸ ), one might not only encounter the dipolar coupling through space but an additional through bond coupling called exchange coupling J (analogue to Fo¨rster and Dexter mechanism in FRET). In this case the observed coupling nAB is the sum of both contributions. nAB ¼ ndd þ J
ð16:7Þ
In this case, the frequencies of the Pake pattern are shifted by J and the frequency of the edge (yk) and peak (y6¼) of the Pake pattern do not behave like 2:1 (Fig. 16.8) what they do if J is zero (Fig. 16.2C). However, if the full Pake pattern is observed both contributions can be disentangled and quantified. From the frequencies at which the peaks and edges of the Pake pattern appear, the dipolar coupling constant ndd and J can be calculated using Eqs. (16.8) and (16.9) derived from Eqs. (16.7) and (16.3) (Milov et al., 1998): n? njj ndd ¼ ð16:8Þ 3 2n? þ njj ð16:9Þ 3 Knowing ndd allows then calculating rAB by applying Eq. (16.3). Care has to be taken for: (1) The program DeerAnalysis2008 does not include J at the moment, thus if applied it would yield wrong results. (2) The above analysis only works if the whole width of the coupling tensor is observed, if this is not the case neither J nor r can be determined. (3) J can shift the two halves J¼
Figure 16.7 PELDOR data. (A), (D), (G) PELDOR time traces recorded with Dn-values ranging from 40 to 80 MHz for 1, and 40–90 MHz for DNA I and DNA II, respectively. (B), (E), (H) Simulated PELDOR time traces for 1, DNA I and DNA II, respectively. (C), (F), (I) Fourier Transformations of the experimental PELDOR time traces. The peak at about 1 MHz in the Fourier Transformation of DNA II is attributed to end-to-end stacking of two DNAs (taken from Schiemann et al., 2009).
346
Olav Schiemann
1.0 t-Bu
O–N
Simulation Experiment
0.8
N
N
N
N t-Bu
D+J
PELDOR
0.6 0.4
N
O
2D + J
0.2 0.0 10 20 −30 −20 −10 0 Frequency (MHz)
30
Figure 16.8 Pake pattern of a biradical with dipolar and exchange coupling. The edges and peaks do not behave like 1:2 (taken from Margraf et al., 2009).
of the Pake pattern in two different directions, depending on its sign. Thus, one obtains a priori two pairs of J and r. These can be distinguished on the basis of whether one observes an intensity around the zero frequency or not. For details see Margraf et al. (2009).
6. Comparison with Other Methods High-resolution liquid-state NMR has been extensively used to determine biomolecular structures in solution, down to the atomic level and with high precision and reliability. However, high-resolution NMR is currently limited to 80 kDa and measures mainly short-range distances. In contrast, PELDOR has no such size restriction, because it is only sensitive to the unpaired electron spin. Thus, PELDOR is most suited for large systems or in combination with NMR providing critical long-range distance constrains. Solid state NMR might be a method to use for large RNA folds, but this method is still under development and would have to prove its suitability first. Cryo electron microscopy has also no size restriction, does not require labeling, and has yielded already exciting insight into huge protein complexes. Yet, it has to deal with depositing the complexes onto surfaces and freeze-drying the sample; both can lead to structural perturbations. In PELDOR, the RNAs are kept in the buffer solution, leading to a frozen ensemble which has been shown to nicely resemble the distribution and dynamics found in the liquid state at room temperature (Godt et al., 2006; Margraf et al., 2007; Schiemann et al., 2004). Like PELDOR FRET requires labeling with two labels. Commonly, these labels are smaller for PELDOR and PELDOR does not need two different labels for high-quality data. Yet, FRET can be performed on the single molecule level, in liquid solution, and provides real-time dynamics
Mapping Global Folds of Oligonucleotides by PELDOR
347
with a resolution down to picoseconds. In contrast, PELDOR is done in the micromolar region and the frozen state. But it is the frozen state that enables easy separation of coupling mechanisms and circumvents the k2 problem, making PELDOR-derived distances generally more precise. In addition, PELDOR does not need a reference measurement. Thus, both methods are nicely complementary to each other. PELDOR is also a useful tool even if an X-ray structure of an RNA complex is already available. On the one hand, it can be used to ascertain whether the overall structure in the crystal is an image of the structure in solution. On the other hand, PELDOR can give added value to existing structures by providing the overall structure of the initial or final state of a folding or catalytic event, which may not be possible to crystallize. Beyond this PELDOR might even be used to follow structural changes through the folding or catalytic event by using freeze quench snapshots. Overall, each method has its strength and weaknesses making them complementary to each other. Which method one uses or which methods one combines should depend on the problem one wants to solve.
ACKNOWLEDGMENTS B. E. Bode is acknowledged for providing Figs. 16.1 and 16.2. T. F. Prisner and G. Jeschke are thanked for helpful discussions.
REFERENCES Amsterdam, I. M. C.v., Ubbink, M., Canters, G. W., and Huber, M. (2003). Measurement of a Cu–Cu distance of 26 A˚ by a pulsed EPR method. Angew. Chem. Int. Ed. Engl. 42, 62–64. Ban, N., Nissen, P., Hansen, J., Moore, P. B., and Steitz, T. A. (2000). The complete atomic structure of the large ribosomal subunit at 2.4 A˚ resolution. Science 289, 905–920. Barhate, N., Cekan, P., Massey, A. P., and Sigurdsson, S. T. (2007). A Nucleoside that contains a rigid nitroxide spin label: A Fluorophore in disguise. Angew. Chem. Int. Ed. Engl. 46, 2655–2658. Berliner, L. J. (ed.) (1998). In ‘‘Spin labeling. The next millennium, Biological Magnetic Resonance.’’ Vol. 14. Plenum Press, New York. Berliner, L. J., Eaton, S. S., and Eaton, G. R. (eds.) (2000). In ‘‘Distance Measurements in Biological Systems by EPR, Biological Magnetic Resonance.’’ Vol. 19. Kluwer Academic/Plenum Publishers, New York. Bode, B. E., Margraf, D., Plackmeyer, J., Du¨rner, G., Prisner, T. F., and Schiemann, O. (2007). Counting the monomers in nanometer sized oligomers. J. Am. Chem. Soc. 129, 6736–6745. Bode, B. E., Plackmeyer, J., Prisner, T. F., and Schiemann, O. (2008). PELDOR measurements on a nitroxide labeled Cu(II) porphyrin: Orientation selection. spin-density distribution and conformational flexibility. J. Phys. Chem. A 112, 5064–5073. Bo¨hme, S., Padmavathi, P. V. L., Holterhues, J., Ouchni, F., Klare, J. P., and Steinhoff, H.-J. (2009). Phys. Chem. Chem. Phys. 11, 6770–6777.
348
Olav Schiemann
Borbat, P. P., Mchaourab, H. S., and Freed, J. H. (2002). Protein structure determination using long-distance constraints from double-quantum coherence ESR: Study of T4 lysozyme. J. Am. Chem. Soc. 124, 5304–5314. Borbat, P. P., Davis, J. H., Butcher, S. E., and Freed, J. H. (2004). Measurement of large distances in biomolecules using double-quantum filtered refocused electron spin-echoes. J. Am. Chem. Soc. 126, 7746–7747. Cai, Q., Kusnetzow, A. K., Hubbell, W. L., Haworth, I. S., Gacho, G., Van Eps, N., Hideg, K., Chambers, E. J., and Qin, P. Z. (2006). Site-directed spin labeling measurements of nanometer distances in nucleic acids using a sequence-independent nitroxide probe. Nucleic Acid Res. 34, 4722–4730. Cai, Q., Kusnetzow, A. K., Hubbell, W. L., Haworth, I. S., Hideg, K., and Qin, P. Z. (2007). Nanometer distance measurements in RNA using site-directed spin labeling. Biophys. J. 93, 2110–2117. Denli, A. M., and Hannon, G. J. (2003). RNAi: An ever-growing puzzle. Trends Biochem. Sci. 28, 196–201. Denysenkov, V. P., Prisner, T. F., Stubbe, J., and Bennati, M. (2006). High-field pulsed electron–electron double resonance spectroscopy to determine the orientation of the tyrosyl radicals in ribonucleotide reductase. Proc. Natl. Acad. Sci. USA 103, 13386–13390. Elsa¨sser, C., Brecht, M., and Bittl, R. (2002). Pulsed electron–electron double resonance on multinuclear metal clusters: Assignment of spin projection factors based on the dipolar interaction. J. Am. Chem. Soc. 124, 12606–12611. Godt, A., Schulte, M., Zimmermann, H., and Jeschke, G. (2006). How flexible are poly (paraphenyleneethynylene)s? Angew. Chem. Int. Ed. Engl. 45, 7722–7726. Grant, G. G., Boyd, N., Herschlag, D., and Qin, P. Z. (2009). Motions of the substrate recognition duplex in a group I intron assessed by site-directed spin-labeling. J. Am. Chem. Soc. 131, 3136–3137. Grote, M., Bordignon, E., Polyhach, Y., Jeschke, G., Steinhoff, H., and Schneider, E. (2008). A comparative electron paramagnetic resonance study of the nucleotide-binding domains’ catalytic cycle in the assembled maltose ATP-binding cassette importer. Biophys. J. 95, 2924–2938. Hagelu¨ken, G., Ingledew, W. J., Huang, H., Petrovic-Stojanovska, B., Whitfield, C., ElMkami, H., Schiemann, O., and Naismith, J. H. (2009). PELDOR distance fingerprinting of the octameric outer-membrane protein Wza from Escherichia coli. Angew. Chem. Int. Ed. Engl. 121, 2948–2950. Hilger, D., Polyhach, Y., Jung, H., and Jeschke, G. (2009). Backbone structure of transmembrane domain IX of the Naþ/proline transporter PutP of Escherichia coli. Biophys. J. 96, 217–225. Hubbel, W. L., Cafiso, D. S., and Altenbach, C. (2000). Identifying conformational changes with site-directed spin labeling. Nat. Struct. Biol. 7, 735–741. Hustedt, E. J., and Beth, A. H. (2000). Structural Information from CW-EPR spectra of dipolar coupled nitroxide spin labels. In ‘‘Biological Magnetic Resonance: Distance Measurements in Biological Systems by EPR,’’ (L. J. Berliner, S. S. Eaton, and G. R. Eaton, eds.), Vol. 19. Kluwer Academic/Plenum Publisher, New York. Jeschke, G. (2005). EPR techniques for studying radical enzymes. Biochim. Biophys. Acta 1707, 91–102. Jeschke, G., and Polyhach, Y. (2007). Distance measurements on spin-labelled biomacromolecules by pulsed electron paramagnetic resonance. Phys. Chem. Chem. Phys. 9, 1895–1910. Jeschke, G., Chechik, V., Ionita, P., Godt, A., Zimmermann, H., Banham, J., Timmel, C. R., Hilger, D., and Jung, H. (2006). DeerAnalysis2006—A comprehensive software package for analyzing pulsed ELDOR data. Appl. Magn. Reson. 30, 473–498, http://www.epr.ethz.ch/software/index.
Mapping Global Folds of Oligonucleotides by PELDOR
349
Jeschke, G., Sajid, M., Schulte, M., and Godt, A. (2009). Three-spin correlations in double electron–electron resonance. Phys. Chem. Chem. Phys. 11, 6580–6591. Kay, C. W. M., Mkami, H. E., Cammack, R., and Evans, R. W. (2007). Pulsed ELDOR determination of the intramolecular distance between the metal binding sites in dicupric human serum transferrin and lactoferrin. J. Am. Chem. Soc. 129, 4868. Kim, N.-K., Murali, A., and DeRose, V. J. (2004). A distance ruler for RNA using EPR and site-directed spin labeling. Chem. Biol. 11, 939–948. Kim, N.-K., Murali, A., and DeRose, V. J. (2005). Separate metal requirements for loop interactions and catalysis in the extended hammerhead ribozyme. J. Am. Chem. Soc. 127, 14134–14135. Kisseleva, N., Khvorova, A., Westhof, E., and Schiemann, O. (2005). Binding of manganese (II) to a tertiary stabilized hammerhead ribozyme as studied by electron paramagnetic resonance spectroscopy. RNA 11, 1–6. Kisseleva, N., Kraut, S., Ja¨schke, A., and Schiemann, O. (2007). Characterizing multiple metal ion binding sites within a ribozyme by cadmium-induced EPR silencing. HFSP J. 1, 127–136. Klare, J. P., Gordeliy, V. I., Labahn, J., Bu¨ldt, G., Steinhoff, H.-J., and Engelhard, M. (2004). The archaeal sensory rhodopsin II/transducer complex: A model for transmembrane signal transfer. FEBS Lett. 564, 219–224. Kuznetsov, N. A., Milov, A. D., Koval, V. V., Samoilova, R. I., Grishin, Y. A., Knorre, D. G., Tsvetkov, Y. D., Fedorova, O. S., and Dzuba, S. A. (2009). PELDOR study of conformations of double-spin-labeled single- and double-stranded DNA with non-nucleotide inserts. Phys. Chem. Chem. Phys. 11, 6826–6832. Liu, Y.-S., Sompornpisut, P., and Perozo, E. (2001). Structure of the KcsA channel intracellular gate in the open state. Nat. Struct. Biol. 8, 883–887. Luehrmann, R., and Stark, H. (2009). Structural mapping of spliceosomes by electron microscopy. Curr. Opin. Struct. Biol. 19, 96–102. Macosko, J. C., Pio, M. S., Tinoco, I., and Shin, Y. K. (1999). A novel 5 displacement spinlabeling technique for electron paramagnetic resonance spectroscopy of RNA. RNA 5, 1158–1166. Margraf, D., Bode, B. E., Marko, A., Schiemann, O., and Prisner, T. F. (2007). Conformational flexibility of nitroxide biradicals determined by X-band PELDOR experiments. Mol. Phys. 105, 2153–2160. Margraf, D., Cekan, P., Prisner, T. F., Sigurdsson, S. T., and Schiemann, O. (2009). Ferroand antiferromagnetic exchange coupling constants in PELDOR spectra. Phys. Chem. Chem. Phys. 11, 6708–6714. Martin, R. E., Pannier, M., Diederich, F., Gramlich, V., Hubrich, M., and Spiess, H. W. (1998). Determination of end-to-end distances in a series of TEMPO diradicals of up to 2.8 nm length with a new four-pulse double electron electron resonance experiment. Angew. Chem. Int. Ed. Engl. 37, 2834–2837. Milov, A. D., Salikhov, K. M., and Shchirov, M. D. (1981). Application of the double resonance method to electron spin echo in a study of the spatial distribution of paramagnetic centers in solids. Soviet. Phys. Solid State 23, 565–569. Milov, A. D., Ponomarev, A. B., and Tsvetkov, Y. D. (1984). Electron–electron double resonance in electron spin echo: Model biradical systems and the sensitized photolysis of decalin. Chem. Phys. Lett. 110, 67–72. Milov, A. D., Maryasov, A. G., and Tsvetkov, Y. D. (1998). Pulsed electron double resonance (PELDOR) and its applications in free-radicals research. Appl. Magn. Reson. 15, 107–143. Morrissey, S. R., Horton, T. E., Grant, C. V., Hoogstraten, C. G., Britt, R. D., and DeRose, V. J. (1999). Mn2þ-nitrogen interactions in RNA probed by electron spinecho envelope modulation spectroscopy: Application to the hammerhead ribozyme. J. Am. Chem. Soc. 121, 9215–9218.
350
Olav Schiemann
Narr, E., Godt, A., and Jeschke, G. (2002). Selective measurements of a nitroxide-nitroxide separation of 5 nm and a nitroxide–copper separation of 2.5 nm in a terpyridine-based copper(II) complex by pulse EPR spectroscopy. Angew. Chem. Int. Ed. Engl. 41, 3907–3910. Nissen, P., Hansen, J., Ban, N., Moore, P. B., and Steitz, T. A. (2000). The structural basis of ribosome activity in peptide bond synthesis. Science 289, 920–930. Pannier, M., Veit, S., Godt, A., Jeschke, G., and Spiess, H. W. (2000). Dead-time free measurement of dipole–dipole interactions between electron spins. J. Magn. Reson. 142, 331–340. Park, S. Y., Borbat, P. P., Gonzalez-Bonet, G., Bhatnagar, J., Pollard, A. M., Freed, J. H., Bilwes, A. M., and Crane, B. R. (2006). Reconstruction of the chemotaxis receptorkinase assembly. Nat. Struct. Mol. Biol. 13, 400–407. Prisner, T. F., Rohrer, M., and MacMillan, F. (2001). Pulsed EPR spectroscopy: Biological applications. Annu. Rev. Phys. Chem. 52, 279–313. Qin, P. Z., and Dieckmann, T. (2004). Application of NMR and EPR methods to the study of RNA. Curr. Opin. Struct. Biol. 14, 350–359. Riplinger, C., Kao, J. P. Y., Rosen, G. M., Kathirvelu, V., Eaton, G. R., Eaton, S. S., Kutateladze, A., and Neese, F. (2009). Interaction of radical pairs through-bond and through-space: Scope and limitations of the point-dipole approximation in electron paramagnetic resonance spectroscopy. J. Am. Chem. Soc. 131, 10092–10106. Schiemann, O., and Prisner, T. F. (2007). Applications of electron paramagnetic resonance to distance measurements in biomolecules. Q. Rev. Biophys. 40, 1–53. Schiemann, O., Fritscher, J., Kisseleva, N., Sigurdsson, S. T., and Prisner, T. F. (2003). Structural investigation of a high-affinity MnII binding site in the hammerhead ribozyme by EPR spectroscopy and DFT calculations. Effects of neomycin B on metal-ion binding. ChemBioChem 4, 1057–1065. Schiemann, O., Piton, N., Mu, Y., Stock, G., Engels, J. W., and Prisner, T. F. (2004). A PELDOR-based nanometer distance ruler for oligonucleotides. J. Am. Chem. Soc. 126, 5722–5729. Schiemann, O., Piton, N., Plackmeyer, J., Bode, B. E., Prisner, T. F., and Engels, J. W. (2007). Spin labeling of oligonucleotides with the nitroxide TPA and use of PELDOR, a pulse EPR method, to measure intramolecular distances. Nat. Protoc. 2, 904–923. Schiemann, O., Cekan, P., Margraf, D., Prisner, T. F., and Sigurdsson, S. T. (2009). Relative orientation of rigid nitroxides by PELDOR: Beyond distance measurements in nucleic acids. Angew. Chem. Int. Ed. Engl. 121, 3342–3345. Sicoli, G., Mathis, G., Delalande, O., Boulard, Y., Gasparutto, D., and Gambarelli, S. (2008). Double electron–electron resonance (DEER): A convenient method to probe DNA conformational changes. Angew. Chem. Int. Ed. Engl. 120, 747–749. Silakov, A., Wenk, B., Reijerse, E., and Lubitz, W. (2009). 14N HYSCORE investigation of the H-cluster of [FeFe] hydrogenase: Evidence for a nitrogen in the dithiol bridge. Phys. Chem. Chem. Phys. 11, 6592–6599. Sowa, G. Z., and Qin, P. Z. (2008). Site-directed spin labeling studies of nucleic acid structure and dynamics. Prog. Nucleic Acid Res. Mol. Biol. 82, 147–197. Ubbink, M., Worrall, J. A. R., Canters, G. W., Groenen, E. J. J., and Huber, M. (2002). Paramagnetic resonance of biological metal centers. Annu. Rev. Biophys. Biomol. Struct. 31, 393–422. Vogt, M., and DeRose, V. J. (2006). Coordination environment of a site-bound metal ion in the hammerhead ribozyme determined by 15N and 2H ESEEM spectroscopy. J. Am. Chem. Soc. 128, 16764–16770(and work cite there in). Voinnet, O. (2001). RNA silencing as a plant immune system against viruses. Trends Genet. 17, 449–459.
Mapping Global Folds of Oligonucleotides by PELDOR
351
Ward, R., Keeble, D. J., El-Mkami, H., and Norman, D. G. (2007). Distance determination in heterogeneous DNA model systems by pulsed EPR. ChemBioChem 8, 1957–1964. Ward, R., Bowman, A., El-Mkami, H., Owen-Hughes, T., and Norman, D. G. (2009). Long distance PELDOR measurements on the histone core particle. J. Am. Chem. Soc. 131, 1348–1349. Winkler, W. C., and Breaker, R. R. (2003). Genetic control by metabolite-binding riboswitches. ChemBioChem 4, 1024–1032. Wolfson, A., Kisseleva, N., Khvorova, A., Westhof, E., and Schiemann, O. (2008). Correlation of the fast cleavage behaviour of the minimal and tertiary stabilized hammerhead ribozymes derived from Tobacco Ringspot Virus RNA with EPR data. Oligonucleotides 18, 101–110.
C H A P T E R
S E V E N T E E N
Laser-Induced Temperature Jump Infrared Measurements of RNA Folding R. Brian Dyer* and Eric B. Brauns† Contents 354 356 356 357 358 361 361 361 362 362 363 366 366 369 371 371
1. Introduction 2. Infrared Spectral Properties of RNA 2.1. Sensitivity to structure 2.2. Spectral assignments 2.3. Thermodynamics 2.4. Practical considerations 3. Experimental Methods 3.1. Samples and sample preparation 3.2. Sample cells 3.3. FTIR spectroscopy 3.4. Time-resolved spectroscopy 4. Examples 4.1. T-jump IR measurements of tRNA 4.2. T-jump IR measurements of tetraloop formation 5. Conclusions References
Abstract Probing a sample using infrared spectroscopy following a laser-induced temperature jump is a powerful method to monitor fast relaxation kinetics. Here, we describe how this approach is used to study the kinetics of RNA folding. We begin with a concise summary of the infrared spectral properties of RNA in the 1500–1800 cm 1 region. The infrared transitions in this region are directly related to the double bond stretching vibrations and ring modes of the nucleotide bases. When RNA undergoes a conformational change, the local environments of the nucleotides are altered. Consequently, the changes in the corresponding infrared spectrum are associated with the structural changes. * Department of Chemistry, Emory University, Atlanta, Georgia, USA Corresponding author; Department of Chemistry, University of Idaho, Moscow, Idaho, USA
{
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69017-0
#
2009 Elsevier Inc. All rights reserved.
353
354
R. Brian Dyer and Eric B. Brauns
Experimentally, temperature is used to systematically vary the RNA structure. When a short laser pulse is used to produce a rapid temperature increase in the sample, the structural changes that ensue can be followed in real time. In this contribution, we discuss experimental methods including sample preparation, instrumentation, and data analysis. We conclude with several experimental examples that highlight usefulness of the technique.
1. Introduction Most types of optical spectroscopy exhibit sensitivity to molecular structure. For example, UV–visible absorption and fluorescence spectroscopies are commonly used to monitor RNA folding and unfolding. What distinguishes infrared (IR) spectroscopy however, is its structural specificity. UV–visible and fluorescence spectra each typically display a single broad band, the intensity of which reports on the conformational changes of the RNA. In contrast, an IR spectrum displays numerous absorption bands, spanning a broad wavenumber range from 800 to 1800 cm 1. Each transition is due to specific structural moieties in the RNA. Collectively, there is an IR spectral signature for nearly every molecular group within an RNA molecule. In particular, the molecular groups involved in base pairing and stacking (C¼O, C¼N, and C¼C of the individual bases) display strong IR absorptions in the 1500–1800 cm 1 range. When RNA structure changes, the local environments of the involved groups are altered (e.g., ionic strength, proximity to hydrogen bond donors or acceptors, etc.). Explicable changes in the corresponding IR spectrum result (intensity changes, band shifts, etc.) that reflect the local environmental change experienced by that group. The signal from each individual IR band provides comparable information to that found in a UV–visible or fluorescence spectrum. The difference is that an IR spectrum has multiple signals, each of which reports on a slightly different structural component of the RNA. The utility of IR spectroscopy is illustrated by the following analogy. In a FRET (fluorescence resonance energy transfer) experiment, a fluorescent donor is attached to one end of the RNA while an acceptor is attached to the other end. The strength of the fluorescence signal is related to the distance between the donor and acceptor. Folding is monitored by observing this signal as a function of time. Similarly, spectral characteristics of fluorescent base analogs substituted at various positions along the chain can be used to track folding. In both cases, the results provide a global indicator of RNA conformation. While inferences can be made, the behavior of unlabeled residues or the residues between the FRET pair cannot be determined directly. Now, imagine that we can place ‘‘FRET pairs’’ all along an RNA oligomer or simultaneously label all of the residues (ignoring the fact that this
355
T-Jump IR Measurements of RNA Folding
would profoundly affect the structure). In addition, assume that the overlap between the individual spectral signals is marginal so that each can be distinguished. Now, when the RNA folds, we not only know when the ends approach one another, but we also have indicators that tell us when the proximities of other residues change. Taking the combined data together paints a very detailed picture of structural progression. The need for multiple probes to fully map the energy landscape of small RNAs has been recognized in several recent papers (Ballin et al., 2007; Hyeon and Thirumalai, 2008; Ma et al., 2008). In another paper, a reference is made to a hypothetical ‘‘dual fluorophore’’ quenching experiment where one fluorophore is placed in the loop and the other near the stem (Nivon and Shakhnovich, 2004). Since the intrinsic IR absorption bands of nucleic acids are structurally specific (described in Section 2), we can achieve the desired effect without the addition of nonnative probes that may ultimately alter the structure. Experimentally, temperature is commonly used to bring about systematic structural changes in the RNA. To illustrate, consider the thermal equilibrium between a population of folded and unfolded RNA molecules, F Ð U (for now ignore the existence of intermediates), governed by the equilibrium constant Keq ¼ cU =cF . It can be shown that the change in concentration of the folded population is related to a temperature change by (Bernasconi, 1976) DcF ¼
Keq c0 DH RT 2 ð1
þ DKeq Þ2
DT
ð17:1Þ
Here, c0 is the total concentration cF þ cU, DH is the enthalpy of unfolding, and R is the gas constant. Accordingly, an increase in temperature will result in a decrease in the folded population, provided that the unfolding enthalpy is both nonzero and positive. Implicit in this equation is that the concentration changes are proportional to key IR absorptions bands. Thus, Eq. (17.1) provides the link between RNA structure and the observable IR spectrum. This equation also shows that the greatest changes will be observed when temperatures near the melting temperature (Tm) are studied (i.e., K 1). If the temperature is changed suddenly (i.e., during a ‘‘temperature jump’’), then it becomes possible to monitor DcF (or DcU) as a function of time. For this simple two-state example, the concentration as a function of time would show exponential behavior Dc ¼ et=t
ð17:2Þ
(we have used Dc rather than DcF or DcU to be more general). The experimentally obtained quantity is the relaxation time, t ¼ ðkF þ kU Þ1 . Neither the folding rate, kF, nor the unfolding rate, kU, is determined directly. Strictly speaking, the experiment measures RNA ‘‘relaxation’’ rather than folding or unfolding.
356
R. Brian Dyer and Eric B. Brauns
2. Infrared Spectral Properties of RNA 2.1. Sensitivity to structure IR spectroscopy is extremely sensitive to RNA conformational changes and base sequence. This is illustrated in Fig. 17.1 using the IR spectra of two different RNA tetraloops. The lower panel shows IR spectra of 50 -gcUGCGgc-30 (‘‘HPG’’) and the upper panel corresponds to 50 gcUCCGgc-30 (‘‘HPC’’). (The bases written in uppercase indicate the unpaired bases of the loop while those in lowercase are the bases that comprise the stem.) In each, the dashed line is the IR spectrum for the unfolded oligonucleotide (acquired at T Tm ) while the solid line is for
60 5¢-gcUCCGgc-3¢
Abs (mOD)
50 40 30 20 10 0 50 5¢-gcUGCGgc-3¢
Abs (mOD)
40 30 20 10 0 1550
1575
1600 1625 1650 1675 Wavenumber (cm−1)
1700
1725
Figure 17.1 Equilibrium FTIR spectra of two RNA tetraloops. The sequences are indicated in the figures. Bases written in uppercase are the unpaired bases found in the loop while those written in lowercase are the bases found in the stem. The dashed lines are spectra of the unfolded RNA and the solid lines are the folded spectra. Note the similarities between the two unfolded spectra and the differences between the folded spectra.
357
T-Jump IR Measurements of RNA Folding
the corresponding folded conformation (acquired at T Tm ). The spectral changes that develop as the hairpins fold are substantial. Furthermore, despite differing by only a single base in the loop, the folded spectra for each hairpin are very different (this will be explained later). By comparison, the UV–visible or fluorescence spectra of the same two hairpins would be practically indistinguishable whether folded or unfolded. This is an example of the structural sensitivity afforded by IR spectroscopy. Since the IR spectrum of an RNA molecule has numerous features, IR spectroscopy is not limited to the determination of the fraction folded only. Depending upon the wavenumber of the transition, IR spectroscopy can distinguish AU base pairs from GC base pairs as well as the more complex base pairing schemes found in triple helices (Banyay et al., 2003; Brauns and Dyer, 2005). Similarly, base stacking interactions can be distinguished from base pairing interactions. It has even been shown that transfer RNAs of different species can be distinguished from their IR spectra (Thomas, 1969).
2.2. Spectral assignments The IR spectra of RNA have a useful property that can be used to facilitate spectral interpretation as well as to more clearly illustrate spectral assignments. When fully unfolded, the IR spectrum of an RNA molecule is roughly equal to the sum of the contributions of its component nucleotides. For example, the IR spectrum of unfolded HPC would be approximately equal to AUMP þ 3AGMP þ 4ACMP. Figure 17.2 shows the IR spectra of the four mononucleotides and is provided as a visual aid to the reader for the description of the spectral assignments that follow. 1.2 AMP CMP GMP UMP
e ⫻ 10−3 (M−1cm−1)
1.0 0.8 0.6 0.4 0.2 0.0 1500
1550
1650 1600 Wavenumber (cm−1)
1700
1750
Figure 17.2 Equilibrium FTIR spectra of the four nucleotide bases. Band assignments are given in the text.
358
R. Brian Dyer and Eric B. Brauns
The following IR spectral assignments are taken from the same references or our own work (Banyay et al., 2003; Brauns and Dyer, 2005). Between 1678 and 1689 cm 1 is a transition due to the C6¼O6 stretch of guanine that decreases in intensity and red shifts when duplexed with cytosine. There is a strong absorption band 1655 cm 1 that is due to the C4¼O4 stretch of uracil that decreases in intensity when it forms a base pair with adenine. Around 1650 cm 1 lies a transition attributed to the C2¼O2 stretch of cytosine. This band red shifts and decreases in intensity when involved in a base pair. The region between 1620 and 1632 cm 1 is dominated by a strong, narrow transition due to C¼N and C¼C ring vibrations of adenine. This band blue shifts and decreases in intensity when base paired. At 1525 cm 1 is a cytosine in-plane ring vibration that drastically decreases in intensity when duplexed with guanine. Around 1566 cm 1 is a transition that is due to a C¼N ring vibration of guanine with contributions from a combination of C6¼O6, C5–C6, and C4¼C5 stretches (all guanine). Near 1584 cm 1 is an additional very weak inplane ring vibration of cytosine. Finally, a second C¼N vibration of guanine lies between 1575 and 1590 cm 1 that decreases when base paired. To some degree, all of these transitions are also sensitive to base stacking interactions, in particular, the ring vibrations. The preceding assignments can be used to interpret the spectra shown in Fig.17.1. In the unfolded state, interbase interactions are negligible and the spectral differences between the two hairpins are solely due to base composition. Hence, the slight differences between the unfolded spectra shown in Fig.17.1 are simply because HPG has more guanine and less cytosine. Likewise, HPC has more cytosine and less guanine. However, in the folded state, interbase interactions are abundant and structural differences between the hairpins manifest in the spectra.
2.3. Thermodynamics Plotting the absorption at a specific wavenumber as a function of temperature produces a melt curve. By fitting the curves to a mathematical model, the melting temperature and other thermodynamic parameters can be determined. In the equilibrium limit, folding exhibits two-state behavior and the fraction unfolded fU at any temperature is related to the absorption by fU ¼
AðT Þ AU AF AU
ð17:3Þ
where AU is the absorbance of the unfolded RNA and AF is the absorbance of the folded RNA. Equation (17.3) provides a phenomenological model that describes the temperature dependence of the absorption. Equation (17.3) is related to the equilibrium constant by
359
T-Jump IR Measurements of RNA Folding
fU ¼
Keq 1 þ Keq
ð17:4Þ
The equilibrium constant can then be written in terms of the Gibbs free energy according to DG Keq ¼ exp ð17:5Þ RT Finally, the enthalpy of unfolding DH and melting temperature Tm are obtained from the temperature dependence of the Gibbs energy, given by the Gibbs–Helmholtz equation T T DGðTÞ ¼ DðT Þ 1 þ DCp T Tm T ln ð17:6Þ Tm Tm Equation (17.6) also provides a way to measure the change in heat capacity DCp between the folded and unfolded states. By combining the previous relationships, the raw melt curves are fit directly to eDGðT Þ=RT ð17:7Þ 1 þ eDGðTÞ=RT In this expression, AU, AF, DH(T ), DCp, and Tm are all fitting parameters. Once the raw data has been fit, the curves can be ‘‘normalized’’ for comparison by converting the absorbance data to fraction unfolded using Eq. (17.3). Examples are shown in Fig. 17.3. Determining DCp from optical melting experiments warrants further discussion. While calorimetric methods measure heat capacity directly, spectroscopic data is fit to a model function that includes DCp as a fitting parameter. Imposing a two-state model on a non-two-state system is a potential source of error. One way to ‘‘circumvent’’ this is to make the assumption that DCp is negligible and simply ignore it. Although this is frequently done, doing so is not completely justifiable (Mikulecky and Feig, 2004). Changes in solvent and electrostatic environments when RNA folds can result in heat capacity differences between the two different states. The alternative is to proceed with the fit and accept the possibility of error. In principle, this is the preferred choice since the error will likely be small. However, as we will explain in the next section, this must be done cautiously. The goal of curve fitting is to obtain a single set of fit parameters that uniquely describes the data. An overparameterized model can jeopardize the reliability of the fit and give ambiguous results. One solution is to reduce the number of parameters by holding DCp constant at zero. As we described previously, this may sacrifice a bit of accuracy. However, the reliability of the fit is increased substantially and general conclusions can still be drawn. AðT Þ ¼ AU þ ðAF AU Þ
360
R. Brian Dyer and Eric B. Brauns
1.0 0.8
fU
0.6
tRNAphe 1620 cm−1 Fit to Eq. (17.7) 1661 cm−1 Fit to Eq. (17.7)
0.4 0.2 0.0
1.0 0.8
fU
0.6
5¢-gcUCCGgc-3¢
1574 cm−1 1661 cm−1 Fit to Eq. (17.7)
0.4 0.2 0.0 280
300
320 340 Temperature (K)
360
Figure 17.3 Melt curves for the indicated RNAs. All data have been normalized by converting them to the fraction unfolded fU using Eq. (17.3).
This is acceptable provided that the assumption is stated up front. The melt curves in the lower part of Fig. 17.3 were fit this way. Alternatively, the full model can be used, with additional steps taken to ensure the validity of the results. An example is shown in the upper part of Fig. 17.3. Here, the experiments were repeated multiple times (minimizing the measurement error), the data were fit simultaneously (using a global fitting algorithm), and the results were corroborated using a separate singular value decomposition analysis. In the lower panel of Fig. 17.3 are melt curves for HPC at 1574 cm 1 (squares) and 1661 cm 1 (circles). The upper panel shows melt curves for tRNAphe at 1620 cm 1 (squares) and 1661 cm 1 (circles). For HPC (lower), these wavenumbers correspond to base stacking (1574 cm 1; ring vibrations) and base pairing (1661 cm 1; GC pairing). The melt curves (and subsequent thermodynamics) are virtually identical. It will be shown later, however, that the kinetics at the two different wavenumbers are very different. The wavenumbers for the tRNA (upper) correspond to AU base
T-Jump IR Measurements of RNA Folding
361
pairing (1620 cm 1) and GC base pairing (1661 cm 1). Unlike the oligonucleotide, the melt curves for tRNA are different at each wavenumber. The differences were verified using a much more rigorous analysis that included singular value decomposition (Brauns and Dyer, 2005). Different melt curves imply that the thermodynamics are also different. We were able to use this result in conjunction with the relaxation kinetics to suggest a parallel pathway folding mechanism for tRNA.
2.4. Practical considerations Because of the strong IR absorption of H2O in the 1500–1800 cm 1 region, all experiments must be carried out in D2O solutions. Still, even in D2O, there is a nonnegligible absorption in our spectral window of interest. The strong residual absorption necessitates the use of thin optical path lengths. A path length of 50 mm has been shown to be optimal (Venyaminov and Prendergast, 1997). However, thin path lengths further require that the sample be present at a relatively high concentration. The molar absorbtivity of RNA in the 1500–1800 cm 1 region is roughly 700 M 1 cm 1 (per nucleotide). This estimate is an average of the peak absorptivities from a number of samples over a range of temperatures. It is an approximation only, and should be treated as such. Using this value, an absorption between 0.01 and 0.1 requires mononucleotide concentrations between 3 and 30 mM. Anything lower, and the signal would simply be too weak to be practical. On the other hand, higher concentrations should be avoided since this would increase the likelihood of dimer formation (or aggregation in general). This is especially true for oligonucleotides.
3. Experimental Methods 3.1. Samples and sample preparation For practical reasons, samples are initially prepared in H2O solutions and then deuterated prior to performing experiments. High purity, deionized water is always used. A buffer is chosen with the only stipulation being that it does not interfere with the IR spectrum of the RNA. We typically use phosphate or Tris buffers at pH 7.2 (upon deuteration, the solution becomes slightly more basic with a pD ¼ 7.6 at 25 C). Depending upon the nature of the experiment, buffer concentrations in the 1–100 mM range are used. Additional salts (e.g., Naþ, Mg2þ) can also be added in accordance with a particular experiment.1 Usually two buffers are prepared: a primary 1
If the sample contains Mg2þ then Tris buffer must be used since MgPO4 will form a precipitate in aqueous solutions.
362
R. Brian Dyer and Eric B. Brauns
(containing EDTA at 0.1 mM) and a secondary (no EDTA). Samples are prepared by first dissolving them in the primary buffer. The EDTA is used to sequester tightly bound divalent cations such as Mg2þ. This is followed by dialysis against the secondary buffer that removes the EDTA from the sample (the carbonyls of EDTA could interfere with the sample spectra). Once the dialysis is complete, a known volume is lyophilized at least three times against D2O to remove the labile protons. To avoid contamination from atmospheric H2O, the samples are handled in a makeshift ‘‘glovebox’’ (a Plexiglas case that is purged with dry air, but not sealed from the surroundings). This precaution seems to be more than sufficient to guard against contamination. The final step is the degassing of the sample to guard against cavitation artifacts in the temperature jump (T-jump) experiments (explained later). We degas the sample by placing it under a mild vacuum for up to 15 min with occasional agitation.
3.2. Sample cells Reliable comparison between the two experimental methods (T-jump and equilibrium FTIR) is facilitated by using the same cells for both. The cells are a custom design comprised two CaF2 windows separated by a Teflon spacer. The spacer defines the optical path length and also divides the cell into two compartments; one for the sample (RNA and buffer) and one for the reference (buffer alone). In this way, sample and reference measurements are obtained by simply translating the cell from side to side. Path lengths are typically 50 mm. The CaF2 windows are placed in a copper housing. The windows are secured by fastening a copper faceplate to the housing. The relative path lengths of the sample and reference compartments are determined by measuring interference fringes in an empty cell. If the path length variation between each side of the cell exceeds 0.1 mm, the cell is dismantled and reassembled until the desired tolerance is met. The cell is then mounted on a larger copper block that is coupled to a circulating water bath for temperature control. The cell temperature is maintained to within 0.1 C.
3.3. FTIR spectroscopy Equilibrium FTIR spectra are recorded using a commercial rapid scanning interferometer coupled to a custom made sample chamber. The copper block with affixed cell is mounted to a computer-controlled translation stage and placed in the sample chamber. The chamber is purged with dry air for a minimum of 3 h prior to collecting spectra. The computer-controlled stage allows sample and reference spectra to be recorded without compromising the sample compartment purge. To obtain an acceptable signal-to-noise level, a single spectrum may require up to 512 coadded
363
T-Jump IR Measurements of RNA Folding
scans (256 scans for the reference and 256 scans for the sample). Rather than recording 256 scans for the reference followed by 256 sample scans, the cell is translated back and forth so that sample and reference spectra are recorded alternately. This is done to minimize any long-term baseline drift. Temperature-dependent spectra are usually recorded at 1 C increments from 20 to 90 C. A LabVIEW program has been written that automates the collection of temperature-dependent spectra. This is advantageous considering that a full melt profile can take up to 18 h to complete.
3.4. Time-resolved spectroscopy Time-resolved spectroscopy is performed using a pump-probe method in which a short-pulsed laser is used to initiate a T-jump and a mid-IR probe laser is used to monitor the transient IR absorbance in the sample. A schematic of the entire instrument is shown in Fig. 17.4. For clarity, only key components are shown. In the description that follows, only those components will be described. A continuous-wave (CW) lead-salt (PbSe) diode laser (output power <1 mW) tuned to a specific vibrational mode of the RNA molecule probes the transient absorbance of the sample. The linewidth of the probe laser is quite narrow (<0.5 cm 1) and sets the spectral resolution of the time-resolved experiments. The divergent output of the diode laser is collected and collimated by a gold coated off-axis l; 1.064 mm
l; 1.9 mm
Raman shifter
PB P
Nd:YAG 10 Hz
≤
20 m 10 J ns
L
OAP Diode laser
Pump/probe overlap: Probe Pump
S
MCT (20 MHz) 14-bit, high speed digitizer (100 MHz, 200 MS/s
Figure 17.4 Schematic of the T-jump spectrometer described in the text. OAP, off-axis parabolic mirror; PB, Pellin–Broca prism; P, polarizer; L, lens; S, sample; MCT, mercury cadmium telluride detector. The size of the pump relative to the probe at the point of overlap is shown in the lower left corner.
364
R. Brian Dyer and Eric B. Brauns
parabolic mirror (OAP) (off-axis angle, 90 ; parent focal length, 25.4 mm; effective focal length, 50.8 mm). The mirror is mounted to a kinematic mount, which is in turn mounted to a three-axis linear translation stage for precise alignment control. The collimated probe diameter is 1 cm and is focused tightly on the sample (diameter <100 mm) using a second OAP. The transmitted probe beam is collected by a third OAP and finally focused onto a 20 MHz MCT detector (15 ns rise time) using a fourth OAP. The T-jump pulse is generated by Raman shifting the fundamental output of a Q-switched Nd:YAG laser (l ¼ 1.064 mm) in a 1-m-long Raman cell filled with H2 at 500 psi. The laser operates at a repetition rate of 10 Hz and has a pulse temporal width of <10 ns (FWHM). The Stokes shift in H2 is 4155 cm 1. As a result, the first Stokes line is at a wavelength of 1.9 mm that is partially absorbed by the D2O solvent, and thus serves as the T-jump pulse. A Pellin–Broca prism is used to separate the 1.9 mm radiation from the remaining Stokes, anti-Stokes, and residual Rayleigh radiation. The T-jump pulse is carefully overlapped with the probe laser at the sample. A long focal length lens ( f ¼ 250 mm) is used to adjust the diameter of the T-jump pulse. At the point of overlap, the pump diameter should be roughly five times the diameter of the probe beam (this is shown in the lower left corner of Fig. 17.4). This assures that the entire probe volume experiences uniform heating and also minimizes the effects of any beam drift. D2O has an absorption band centered around 1.9 mm that corresponds to the first vibrational overtone. The D2O transmits 75–80% of the T-jump pulse in a cell with a 50 mm path length. The energy from the fraction that is absorbed produces the T-jump. The specific heat capacity of D2O is 4.22 J K 1 g 1. Assuming 20% absorption and a sample volume of 3.9 10 5 mL, one can calculate that 20 mJ of 1.9 mm radiation is required to generate a 20 C T-jump. Since the thermal equilibration of the aqueous solvent occurs on a picosecond time scale, the T-jump is virtually instantaneous when induced by a 10 ns laser pulse. Hence, the time resolution of the experiment is limited by the laser pulse width and/or the detector rise time (20 ns for the instrument described here). Prior to collecting time-resolved data, the temperature dependence of the probe beam transmission through the reference side of the cell is measured. Using these data, the transient absorption of the probe through the reference serves as an internal thermometer to measure the magnitude of the T-jump. The arrival of the T-jump pulse defines t ¼ 0 and triggers data collection. The transient absorption profile subsequent to the T-jump is detected by the MCT detector described above. The resulting waveform from 0 to 1.4 ms is digitized at 5 ns intervals by a 14-bit high-speed digitizer (100 MHz bandwidth, 200 MS s 1). Each waveform is usually averaged over 3000 laser shots depending upon the signal strength. The timedependent absorption is calculated for each sample/reference pair from
365
T-Jump IR Measurements of RNA Folding
T-jump signal (mV)
12 10 8 6 4
Raw signals
2 0 0
Abs (mOD)
−5 −10
Absorbance
−15 −20 6
ΔA (mOD)
5 4 3
Asample- Areference
2 1 0 0.01
0.1
1
10 Time (μs)
100
1000
Figure 17.5 Example of kinetic data. The digitized MCT signals (top; gray is the sample and black is the reference) are converted to absorbance (middle) then the reference absorbance is subtracted from the sample absorbance (bottom).
the digitized MCT signal. Finally, the time-resolved absorbance of the sample is obtained by subtracting the reference absorption from the sample absorption. Figure 17.5 shows this procedure graphically. Digitizing the waveforms at 5 ns increments over nearly 6 decades yields in excess of 260,000 data points. This has two drawbacks. One is that it is simply inconvenient; the excessive number of data points is cumbersome and impedes data analysis and display. The second drawback has a more important consequence. Digitizing a long time span in small intervals leads to unnecessary point density at longer times. For example, at t ¼ 25 ns, the
366
R. Brian Dyer and Eric B. Brauns
next point would be at 30 ns. However, at t ¼ 500 ms, the next point would be at 500.005 ms. Aside from being impractical, subsequent data analysis is heavily weighted toward longer times and could lead to misleading results. We circumvent this by interpolating the raw data (e.g., the data in the upper plot of Fig. 17.5) using a logarithmically spaced time axis.2 Using this method, we reduce the number of data points to 3004. Before proceeding, we verify that the interpolated data exactly overlays the original data. A possible complication that can arise in a T-jump experiment is ‘‘cavitation’’ (Wray et al., 2002). Cavitation is a photoacoustic phenomenon where bubbles in the fluid medium can be produced in response to the rapid rise in temperature. When this occurs, artifacts between 0.1 and 1 ms manifest, rendering the data in this region unusable. Our experience is that cavitation can be controlled by ridding the sample cells and sample solutions of nucleation sites for bubble growth. The use of high quality, scrupulously clean CaF2 windows, sample degassing, and filtration of particulates from sample solutions greatly diminishes (to the point of eliminating) the occurrence of cavitation events. Using this technique we can achieve T-jumps up to 20 C without complications. In addition, cavitation superimposes distinct features on the data. Thus, a simple filter algorithm in the data collection software can be used to remove any contaminated transients in the event that cavitation does occur.
4. Examples 4.1. T-jump IR measurements of tRNA Several groups have used the laser-induced T-jump technique to study nucleic acid folding (Ansari et al., 2001; Dewey and Turner, 1980; Kuznetsov et al., 2008; Ma et al., 2006). However, the first T-jump IR measurements on RNA were carried out by us (Brauns and Dyer, 2005). This section briefly summarizes that work. The melting curves for tRNAphe at two different IR frequencies were shown previously in Fig. 17.3. Although the thermodynamic parameters for the two were shown to be statistically different, their melting temperatures are the same, 58.5 C. T-jump experiments were performed by holding the initial temperature constant at 47 C where >98% of the total population is folded. Transient absorption profiles were recorded for both probe frequencies over a range of T-jump magnitudes up to a maximum 20 C jump. A plot showing a series of transient absorptions recorded at 1661 cm 1 is shown in Fig. 17.6 (qualitatively, the data at 1620 cm 1 show the same behavior). 2
An alternative approach that we have also employed is to use a digitizer that can change the digitization interval ‘‘on the fly.’’
367
T-Jump IR Measurements of RNA Folding
ΔAbs (mOD)
25 20 15 10
T-jump magnitude
30
5 0 0.1
1
10 100 Time (ms)
1000
10,000
Figure 17.6 A series of transient absorptions recorded at 1661 cm 1 for tRNAphe. Qualitatively, the data at 1620 cm 1 are the same. The initial temperature was held constant at 47 C and the magnitude of the T-jump was varied up to a maximum of 20 C. Each transient was fit to a three exponential model from t ¼ 100 ns to the maximum transient absorption (t ¼ 1 ms).
The transients were all fit to an exponential model from t ¼ 100 ns to t ¼ 1 ms, using DA ¼
N X
ai et=t
ð17:8Þ
i¼1
(The long time limit was chosen to be the time where the transient absorption reaches a maximum.) We observe three relaxation processes (i.e., N ¼ 3). The kinetic phases are well separated in time, spanning hundreds of nanoseconds to hunderds of microseconds. The amplitudes contribute more or less equally to the total decay, the slowest phase contributing the least. What is particularly interesting, however, is that the relaxation times for the 1620 cm 1 data are faster than those at 1661 cm 1. This is easily seen in Fig. 17.7 where the observed rate constants (1/t ¼ kobs ¼ kF þ kU) for each phase are plotted in Arrhenius coordinates (i.e., ln kobs vs. 1/T). The activation enthalpy DH{ can then be determined from the slope according to DH { þ const: ð17:9Þ RT The constant is the y-axis intercept and is related to the preexponential factor. The activation barrier for the slow phase and its average relaxation time suggest that we are observing base pair formation. The fact that the relaxation recorded at 1620 cm 1 is slightly faster than the relaxation at ln kobs ¼
368
R. Brian Dyer and Eric B. Brauns
15
ln kobs
14 13 12
1620 cm−1 1661 cm−1
11 10 10.0
ln kobs
9.5 9.0 8.5 8.0
1620 cm−1 1661 cm−1
7.5 2.92 2.94 2.96 2.98 3.00 3.02 3.04 3.06 3.08 3.10 1/T (K ⫻ 10−3)
Figure 17.7 Arrhenius plots of the tRNA T-jump data at 1620 cm 1 (squares) and 1661 cm 1 (circles). The upper plot shows the fast and intermediate phases. These correspond to nonactivated processes. The lower plot shows the slow phase. In addition to corresponding to an activated process, the kinetics at 1620 cm 1 is faster than those at 1661 cm 1.
1661 cm 1 is not surprising. The former corresponds to AU base pairs which are weaker than their GC counterparts. On the other hand, the activation barriers for the fast and intermediate relaxations are approximately zero. This implies that the barrier is mostly entropic. From these data we proposed a parallel pathway folding model to describe tRNA folding. According to our model, the unfolded tRNA is partitioned into two similar (but distinctly different) structural ensembles. Each ensemble folds via a separate pathway; both converging at the native state. The rate limiting step in each is the initial collapse of the unfolded states into compact (but misfolded) intermediates. Following this collapse, the misfolded intermediates undergo fast, localized reorganizations where nonnative interactions are broken and native contacts are formed.
369
T-Jump IR Measurements of RNA Folding
4.2. T-jump IR measurements of tetraloop formation More recently, we have begun studying the fast relaxation kinetics of small RNA tetraloops (Stancik and Brauns, 2008). The structures of small RNA tetraloops are deceptively simple. At first glance, one might be inclined to expect loop formation to follow a two-state model. However, recent results (including our own work) have shown that this is not the case (Ma et al., 2006; Moody et al., 2004; Proctor et al., 2004; Siegfried et al., 2007). We are particularly interested in the conserved tetraloop sequence, UNCG. The oligonucleotides introduced earlier, HPC and HPG, are members of this group (N¼C and N¼G). The short, two base pair stem is chosen so that the oligonucleotides will melt at lower temperatures allowing fully denaturing conditions to be achieved. Kinetic data for these tetraloops are shown in Fig. 17.8. The kinetics for each was probed at two different probe frequencies following a 10 C T-jump. Prior to the T-jump, both samples were equilibrated at their respective melting
1.0
Abs (a.u.)
0.8 0.6
1574 cm−1 1669 cm−1
0.4 0.2
5¢-gcUCCGgc-3¢ 0.0 1.0
Abs (a.u.)
0.8 0.6
1660 cm−1 1574 cm−1
0.4 0.2
59-gcUGCGgc-39 0.0
0
20
40 60 Time (ms)
80
100
Figure 17.8 Transient absorption profiles for the indicated oligonucleotides. The absorbance has been normalized to facilitate direct qualitative comparison.
370
R. Brian Dyer and Eric B. Brauns
temperatures (58 C for HPC and 60 C for HPG). The melting data for HPG are not shown. Qualitatively, two features of the data are particularly noteworthy: (1) Each sequence displays very different relaxation kinetics. For example, the overall relaxation for HPC is quite a bit faster than the relaxation for HPG. Additionally, the kinetic traces for HPC are both fit to biexponential functions while those for HPG are fit to a triexponential function. (2) The kinetics for each depend on the probe wavenumber. For example, in both cases, the overall relaxation rate is faster when the kinetics is probed at 1574 cm 1 than at the higher probe wavenumber (for HPC, the difference is substantial). The implication from (1) above is that a single base substitution is sufficient to significantly alter the folding landscape. Two possibilities (or a combination of both) can explain the second feature from above. One is that the wavenumber-dependent kinetics highlights sequential events. The lower probe wavenumber (1574 cm 1) is due almost entirely to ring vibrations of guanine while the higher probe wavenumber is due mainly to guanine carbonyl stretch. Chronologically, base stacking interactions would precede the formation of base pair hydrogen bonds. The lower wavenumber kinetics is faster because the absorption at 1574 cm 1 is more sensitive to stacking interactions while the higher wavenumber is a direct measure of hydrogen bonding. An alternative explanation is that there is more than one folding population and that the wavenumber dependence arises due to the differential absorption of each. The latter scenario is similar to what was observed for tRNAphe. The results from HPG are relatively new and still under investigation. However, the kinetic results for HPC have been analyzed in greater detail and have been published. For the sake of illustration, we will briefly summarize those findings. The consensus view is that the ruggedness of the energy landscape is due to misfolded structures (on or off pathway). However, almost invariably, misfolding is attributed to partial or incorrect stem contacts. In a small hairpin such as HPC, there are not enough degrees of freedom to access a wide variety of stem structures. Despite this, multiexponential kinetics are observed. Since misfolding in the stem is insufficient to explain the folding complexity, we postulate that loop interactions and/or single strand stacking fluctuations also contribute to the overall folding complexity. It is already known that the earliest stages of hairpin folding involve base stacking. However, our results provide more detail. Before the chain can collapse permitting loop closure, the stacks rearrange leading to partial conformational ordering. With the bases properly oriented relative to their neighbors, the chain can collapse and zipping of the stem base pairs can occur. We justify our conclusion based on the probe frequency dependence of the kinetics in conjunction with activation parameters (data not shown). Our conclusions are similar to the recent findings of Zewail and coworkers (Ma et al., 2008). They refer to these intermediates as ‘‘labile in destacking but compact in nature.’’
T-Jump IR Measurements of RNA Folding
371
5. Conclusions IR spectroscopy is extremely sensitive to RNA conformational changes. The IR spectrum of an RNA molecule exhibits many absorption transitions, each of which can be assigned to a specific molecular group of the RNA. If considered individually, each transition would provide a useful metric to monitor conformational changes. However, it is the ability to monitor the collective response of several transitions simultaneously that is unique to IR spectroscopy. Doing so allows for a more detailed analysis of structural changes. Moreover, folding kinetics can be studied by monitoring the IR spectrum as a function of time following a rapid laser-induced T-jump.
REFERENCES Ansari, A., Kuznetsov, S. V., and Shen, Y. (2001). Configurational diffusion down a folding funnel describes the dynamics of DNA hairpins. Proc. Natl. Acad. Sci. USA 98, 7771–7776. Ballin, J. D., Bharill, S., Fialcowitz-White, E. J., Gryczynski, I., Gryczynski, Z., and Wilson, G. M. (2007). Site-specific variations in RNA folding thermodynamics visualized by 2-aminopurine fluorescence. Biochemistry 46, 13948–13960. Banyay, M., Sarkar, M., and Gra¨slund, A. (2003). A library of IR bands of nucleic acids in solution. Biophys. Chem. 104, 477–488. Bernasconi, C. F. (1976). Relaxation Kinetics. Academic Press, New York. Brauns, E. B., and Dyer, R. B. (2005). Time-resolved infrared spectroscopy of RNA folding. Biophys. J. 89, 3523–3530. Dewey, T. G., and Turner, D. H. (1980). Laser temperature jump study of solvent effects on poly(adenylic acid) stacking. Biochemistry 19, 1681–1685. Hyeon, C., and Thirumalai, D. (2008). Multiple probes are required to explore and control the rugged energy landscape of RNA hairpins. JACS 130, 1538–1539. Kuznetsov, S. V., Ren, C.-C., Woodson, S. A., and Ansari, A. (2008). Loop dependence of the stability and dynamics of nucleic acid hairpins. Nucleic Acids Res. 36, 1098–1112. Ma, H., Proctor, D. J., Kierzek, E., Kierzek, R., Bevilacqua, P. C., and Gruebele, M. (2006). Exploring the energy landscape of a small RNA hairpin. JACS 128, 1523–1530. Ma, H., Wan, C., Wu, A., and Zewail, A. H. (2008). DNA folding and melting observed in real time redefine the energy landscape. Proc. Natl. Acad. Sci. USA 104, 712–716. Mikulecky, P. J., and Feig, A. L. (2004). Heat capacity changes in RNA folding: Application of perturbation theory to hammerhead ribozyme cold denaturation. Nucleic Acids Res. 32, 3967–3976. Moody, E. M., Feerrar, J. C., and Bevilacqua, P. C. (2004). Evidence that folding of an RNA tetraloop hairpin is less cooperative than its DNA counterpart. Biochemistry 43, 7992–7998. Nivon, L. G., and Shakhnovich, E. I. (2004). All-atom Monte Carlo simulation of GCAA RNA folding. J. Mol. Biol. 344, 29–45. Proctor, D. J., Ma, H., Kierzek, E., Kierzek, R., Gruebele, M., and Bevilacqua, P. C. (2004). Folding thermodynamics and kinetics of YNMG RNA hairpins: Specific
372
R. Brian Dyer and Eric B. Brauns
incorporation of 8-bromoguanosine leads to stabilization by enhancement of the folding rate. Biochemistry 43, 14004–14014. Siegfried, N. A., Metzger, S. L., and Bevilacqua, P. C. (2007). Folding cooperativity in RNA and DNA is dependent on position in the helix. Biochemistry 46, 172–181. Stancik, A. L., and Brauns, E. B. (2008). Rearrangement of partially ordered stacked conformations contributes to the rugged energy landscape of a small rna hairpin. Biochemistry 47, 10834–10840. Thomas, G. J. Jr. (1969). Determination of the base pairing content of ribonucleic acids by infrared spectroscopy. Biopolymers 7, 325–334. Venyaminov, S. Yu., and Prendergast, F. G. (1997). Water (H2O and D2O) molar absorptivity in the 1000–4000 cm 1 range and quantitative infrared spectroscopy of aqueous solutions. Anal. Biochem. 248, 234–245. Wray, W. O., Aida, T., and Dyer, R. B. (2002). Photoacoustic cavitation and heat transfer effects in the laser-induced temperature jump in water. Appl. Phys. B 74, 57–66.
C H A P T E R
E I G H T E E N
Probing Nucleic Acid–Ion Interactions with Buffer Exchange-Atomic Emission Spectroscopy Max Greenfeld* and Daniel Herschlag† Contents 375 377 379 382 383 384 388 388
1. Introduction 2. Description of BE-AES 3. Buffer Equilibration 4. ICP-AES 5. Measuring Anions 6. Example Protocol Acknowledgments References
Abstract The ion atmosphere of nucleic acids directly affects measured biochemical and biophysical properties. However, study of the ion atmosphere is difficult due to its diffuse and dynamic nature. Standard techniques available have significant limitations in sensitivity, specificity, and directness of the assays. Buffer exchange-atomic emission spectroscopy (BE-AES) was developed to overcome many of the limitations of previously available techniques. This technique can provide a complete accounting of all ions constituting the ionic atmosphere of a nucleic acid at thermodynamic equilibrium. Although initially developed for the study of the ion atmosphere of nucleic acids, BE-AES has also been applied to study site-bound ions in RNA and protein.
1. Introduction RNA folding and catalysis requires nonspecifically associated cations and often specifically bound ions for function. Specificity and affinity of DNA-binding proteins has been shown to be affected by ions in solution. * Department of Chemical Engineering and Biochemistry, Stanford University, Stanford, California, USA Departments of Biochemistry and Chemistry, Stanford University, Stanford, California, USA
{
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69018-2
#
2009 Elsevier Inc. All rights reserved.
375
376
Max Greenfeld and Daniel Herschlag
Higher order assemblies including the ribosome and chromatin also depend on counterions (Herrera et al., 1996; Tremethick, 2007). This strong functional dependence on counterions arises from the fact that RNA and DNA are polyelectrolytes with each residue having a formal negative charge. This results in a sheath of nonspecifically associated ions, called the ion atmosphere, forming around nucleic acids in solution (Draper, 2004; Manning, 1969a; Sharp and Hongin, 1995). For instance, in the absence of counterions the Tetrahymena Group I Intron would need to overcome a repulsion of 103 kcal/mol to fold into its active conformation (Bai et al., 2005). The ion atmosphere greatly mitigates this repulsive force, although exact quantification remains difficult. Extensive study has established that the ion atmosphere is a diffuse and dynamic structure (Misra and Draper, 2000) that cannot be described by simple ion-binding models (Manning, 1969a,b,c). This means that structural techniques such as crystallography or chemical footprinting cannot be used to study the ion atmosphere. Therefore, the techniques available for studying the ion atmosphere are severely restricted. Line broadening in NMR spectra has been used to study 23Naþ association with DNA. However, no rigorous relationship exists between line broadening and Naþ/DNA interactions, and the method is limited by available isotopes (Anderson and Record, 1990). Gel electrophoresis measures changes in the mobility of nucleic acids as affected by the presence of bulk ions, yet no formal relationship exists between mobility and the composition of the ion atmosphere (Li et al., 1998). Anomalous small-angle X-ray scattering provides unique insight to the spatial distribution of ions in the ion atmosphere, but it does not directly count the number of ions in the atmosphere and is applicable for only select ions (Das et al., 2003). Equilibrium dialysis read out by atomic absorption spectroscopy is an elegant approach for assaying the ion atmosphere. However, the original implementation for measuring Mg2þ association with tRNA was hindered in accuracy and throughput (Bina-Stein and Stein, 1976; Stein and Crothers, 1976a,b). Perhaps, the most extensively used technique involves fluorescent indicator dyes to assay the concentration of ions in solution. In practice, this technique only allows monitoring of a few select divalent ions, and as such has not been able to provide a complete accounting of the ion atmosphere (Grilley et al., 2009; Krakauer, 1971; Romer and Hach, 1975). Nevertheless, this technique has proven an informative approach for studying the effect of the ion atmosphere on the energetics of tertiary structure formation in RNA (Grilley et al., 2006, 2007). To overcome the limitations of other techniques, buffer exchangeatomic emission spectroscopy (BE-AES) was created (Bai et al., 2007). Measurements of the ion atmosphere done with BE-AES provide a rigorous thermodynamic measure of the number of ions associated with a nucleic acid. BE-AES is sensitive to a wide number of elements and has proven to
377
Probing Nucleic Acid–Ion Interactions with BE-AES
50 40
# ion/DNA
30 20 10 0 -10 0
10-1
100
101
[Mg2+] (mM)
Figure 18.1 Competitive association between Mg2þ (D) and 20 (mM ) Naþ (○) with a 24 bp DNA duplex. Depleted anions are shown by (r) and the net charge is given by (□). Solid lines are fitted to the Hill equation (Eq. (18.3)), while dotted lines are predictions from the nonlinear Poisson–Boltzman model. Reprinted from Bai et al. (2007).
be highly sensitive and accurate. BE-AES was used to provide the first rigorous experimental evidence that ion size, a parameter not included in the widely used nonlinear Poisson–Boltzmann electrostatic theory, affects the occupancy of the ion atmosphere (Bai et al., 2007). BE-AES has also been used to test for site-specific ion binding in RNA and protein model systems (Das et al., 2005; Zalatan et al., 2008). Figure 18.1 indicates the type of data that can be gathered from a typical BE-AES experiment (Bai et al., 2007). In this experiment the ions associated with a 24-bp DNA duplex (net charge þ46) were measured for a series of Mg2þ concentrations. As expected, Mg2þ displaces Naþ with increased Mg2þ concentration. At the same time, the net charge on the system remains neutral. This chapter details the general and technical principles necessary for performing experiments analogous to those shown in Fig. 18.1. In addition, a detailed protocol for the experiments shown in Fig. 18.1 is presented.
2. Description of BE-AES Key to studying the thermodynamics of nucleic acid ion association is establishing an equilibrium with a well-defined reference solution. BE-AES accomplishes this through the use of ultrafiltration spin columns that allow multiple buffer exchanges to be completed on one sample without
378
Max Greenfeld and Daniel Herschlag
significant loss of sample. This approach achieves equilibrium much faster than standard equilibrium dialysis. Once equilibrium is reached, the number of ions associated with a nucleic acid is simply the enrichment of ion concentration, over the reference buffer, normalized by the concentration of macromolecules. Equation (18.1) formally shows this relationship: N#ion=na ¼
na buffer cion cion cna
ð18:1Þ
na indicates the concentration of the ionic atmosphere species in the cion buffer nucleic acid containing sample, cion indicates the concentration of the ionic atmosphere species in the reference buffer, and cna indicates the concentration of nucleic acid (i.e., determined as the phosphorus concentration divided by the number of formal negative charges on the nucleic acid) in the nucleic acid containing sample. Experimental realization of the relationship outlined in Eq. (18.1) has been historically limited by the techniques available for determining ion concentrations. This limitation has been largely overcome through the application of modern inductively coupled plasma atomic emission spectroscopy (ICP-AES) for the determinations of the concentrations in Eq. (18.1). The advantages afforded by ICP-AES are twofold: (1) simultaneous quantification of multiple elements and (2) high accuracy over a broad range of ion concentrations. The simultaneous concentration determination of multiple ions in the ion atmosphere (e.g., sodium and magnesium) in addition to phosphorous (i.e., the nucleic acid) allows internally calibrated measurements that greatly increases the accuracy of BE-AES experiments. Figure 18.2 diagrams the workflow of a typical BE-AES experiment. There are two major experimental steps: (1) buffer equilibration and (2) ICP-AES concentration determination. Both sample preparation (i.e., buffer exchange) and sample concentration determination (i.e., AES) must be successfully completed in order to make meaningful measurements. To maximize the utility of BE-AES, experimental design must carefully balance practical issues such as the availability and behavior of the nucleic acid being studied with the desire to get high precision and accuracy in the final measurements.
Figure 18.2 Schematic depicting the three major steps of BE-AES. Reprinted from Bai et al. (2007).
379
Probing Nucleic Acid–Ion Interactions with BE-AES
3. Buffer Equilibration As depicted in Fig. 18.2, the first step in BE-AES is sequential buffer exchanges using ultrafiltration spin columns. Buffer exchanges are necessary to get the analyte into a known reference buffer. The last round of buffer exchange is used to concentrate the sample. The accuracy of BE-AES measurements is directly related to the accuracy with which a solution concentration can be determined. ICP-AES, requires milliliter quantities of sample to accurately determine a concentration. However, the technique is sensitive to low micromolar concentrations of most elements. The high analyte concentrations obtained during the final concentrating step are necessary for maximizing the difference in concentrations between the two buffer chambers and as such must be diluted prior to ICP-AES concentration determination. Table 18.1 indicates approximate concentrations and volumes present during buffer equilibration and ICP-AES concentration determination. Unless otherwise noted, concentrations are the actual concentration in the buffer equilibrium step prior to dilution for ICP-AES measurement. For any particular ion being studied, Eq. (18.1) indicates three concentrations that must be determined. However, the difference in concentration na buffer between cion and cion must be sufficiently large that it can be determined buffer with accuracy comparable to cna . The value of cion can be determined with the highest precision because there is a large quantity of this material which can be measured. In practice, the analyte must be concentrated to a concentration of hundreds of micromolar to afford a sufficiently large differential in DNA buffer cion and cion . At this concentration an amount of analyte of 20–40 mL is sufficient to accurately determine the concentration. Although buffer exchanging and concentration is technically simple, careful experimental design is necessary for ensuring that meaningful results are obtained and to prevent the use of unnecessarily large quantities of a nucleic acid. There are tradeoffs between working at high concentrations of a nucleic acid, which ensures good signal, with the complications that arise Table 18.1 Typical sample concentrations and volumes
Concentration analyte Volume
I. Buffer equilibration
II. AES determination
0.01–10 (mM) 100 (mL)
0.01–20 (mM) 4 (mL)
Actual concentration ranges of analyte in the buffer equilibration step and the AES concentration determination step. Unless otherwise noted concentrations listed in the text are the actual concentrations in the buffer equilibration step prior to dilution for AES measurements.
380
Max Greenfeld and Daniel Herschlag
with working at high concentrations. BE-AES experiments must be designed such that the following criteria are met: (1) results are independent of the final macromolecule concentration; (2) equilibrium has been reached; and (3) analyte concentrations are stable between sample preparation and ICP-AES measurements. The high final concentrations of an analyte can result in at least two problems. First, precipitation of nucleic acids can be a problem at high concentrations. The experimentalist should be vigilant during sample preparation to ensure that no signs of precipitation are observed. Second, interactions between nucleic acids can still occur without precipitation. This possibility necessitates controls where BE-AES results are shown to be the same over a range of nucleic acid concentrations. Establishing that equilibrium has been reached is central to measuring any thermodynamic quantity. In BE-AES, equilibrium can be verified by measuring the composition of the flow-through buffer of a spin column. The flow-through buffer of a spin column should remain constant over sequential rounds of buffer exchanges and should be of identical composition to the buffer being used for buffer exchanges. Figure 18.3 shows the change in ion concentration of the flow-through for sequential rounds of buffer exchanges. The number of rounds of buffer exchange needed to reach equilibrium depends on the composition of the buffer being used. The stability of the analyte concentrations can be disturbed by unintended sources. Two sources that have been identified are evaporation and
B Ion concentration (mM)
Ion concentration (mM)
A
20
15 0.1 0
1 2 3 4 5 6 7 8 Round
20
15 1 0
1 2 3 4 5 6 7 8 Round
Figure 18.3 Ensuring equilibrium has been reached requires monitoring the concentration of the flow-through in sequential buffer exchanges. Plots depict the ion concentration in the flow-through for sequential rounds of buffer exchanges. (A) 0.1 mM Mg2þ (D), 20 mM Naþ (○), and 16 mM cacodylate (□). (B) 1 mM Mg2þ (D), 20 mM Naþ (○), and 16 mM cacodylate (□). A final concentration of 0.5 mM DNA was targeted. Note that at lower Mg2þ more rounds of buffer exchange were needed to achieve equilibrium. Reprinted from Bai et al. (2007).
381
Probing Nucleic Acid–Ion Interactions with BE-AES
nonspecific loss during ICP-AES analysis. It is prudent to design experiments and controls with these potential sources of error in mind since data obtained with solutions of poorer stability will not necessarily look wrong. Rather, the experimentalist might have undesirable irreproducibility of measurements or systematic error not consistent with a well-known property of the system, such as charge neutrality. Evaporation can become a problem when small sample volumes are used for the final concentration of a sample. The large exposed surface area of spin columns and the time required for concentration are sufficient for evaporation to affect the final measured concentration. Figure 18.4 shows that some sample preparation protocols can result in up to a 15% error due to evaporation. In contrast, protocols designed to minimize evaporation can reduce error due to evaporation to negligible amounts. The easiest way to get around this problem is to use a large sample volume. Unfortunately, to maintain the same signal, large quantities of the nucleic acid being studied must be used for the buffer exchanges. This solution results in a large fraction of unanalyzed sample that can typically be recovered and used in subsequent experiments. Additionally, care can be taken to minimize evaporation, such as preparing samples in a cold room, or not letting them sit for a long time before the final dilution. 20% Deviation of flow-through relative to bulk
Mg2+ 15% 10% 5% 0
0
10 20 Time (min)
30
Figure 18.4 The time required for buffer exchanges can be sufficient for evaporation to affect the concentration of ions. Demonstration that the collected Mg2+ concentration can systematically increase with time. Flow-through was collected at 1, 10, and 20 or 30 min (corresponding to 450, 250, and 100 or 20 mL of solution remaining in the top chamber) during the final round of buffer equilibration. Deviations from the starting buffer concentration (1 mM Mg2þ, 20 mM Naþ, and 16 mM cacodylate) are plotted for equilibrium conditions that affect total evaporation. Buffer exchanges were carried out at 25 C and retained volume of 20 mL (○), 25 C and retained volume of 100 mL (D), and 4 C and retained volume of 100 mL (◊). Reprinted from Bai et al. (2007).
382
Max Greenfeld and Daniel Herschlag
The low concentrations of analyte in ICP-AES measurements can change significantly due to small nonspecific losses of sample through precipitation. This can potentially be dealt with by doing the final dilution of a sample immediately prior to ICP-AES analysis. This is a viable solution if the precipitation is not significant on the timescale of a few hours. Alternatively, it may be possible to stabilize the solution with different buffering conditions. Particularly relevant are low pH buffers, which can help prevent the formation of insoluble metal oxides. Verifying reproducibility with tubes made of different materials or made by different manufacturers is useful for verifying that erroneous results are not arising from nonspecific absorption to tubes or leaching of material from the tubes.
4. ICP-AES ICP-AES is a highly sensitive technique for determining the atomic composition of solutions. As shown in Table 18.2, ICP-AES is able to simultaneously detect many biologically relevant elements. As biophysical studies will inevitably rely on sample dilution to arrive at the concentration Table 18.2 (2007)
a
Selected elements detectable with ICP-AES. Reprinted from Bai et al.
Element
Detection limit (nM)
Linear rangea (mM )
Arsenic Barium Cadmium Calcium Chlorideb Cobalt Lithium Magnesium Manganese Phosphorous Potassium Rubidium Sodium Zinc
100 1 5 150 ND 50 300 80 300 65 250 ND 300 15
0.01–2 0 1–0.5 ND 0.01–0.5 ND ND 0.1–2.5 0.0 1–0.5 ND 0.025–2 0.1–2.5 0.1–2.5 0.2–2.5 ND
Linear regime was defined as the region for which the expected concentration deviated no more than 5% from the measured values. Detection of halides requires specialized detectors not available on many ICP-AES instruments. ND indicates not determined. b
383
Probing Nucleic Acid–Ion Interactions with BE-AES
of sample for ICP-AES analysis, the lower limit of detection is useful to keep in mind so as to avoid working at unnecessarily high concentrations that waste analyte. The 100–200-fold dilutions used for most measurements the authors have taken lie within the linear range reported in Table 18.2. The reported linear range was optimized for low concentrations of analyte. ICP-AES is capable of directly measuring higher concentrations, but this will sacrifice the accuracy of lower concentration measurements. If the details of a particular experiment allow for larger quantities of the analyte to be easily made, using a smaller dilution factor, which sacrifices the high sensitivity of ICP-AES, would likely yield more accurate results. Often ICP-AES instruments are run by operators who are highly knowledgeable about the intricacies of their particular instrument. Discuss with the operator what samples you are running. They will probably have suggestions of how to set up a sample run to make things run quickly and most accurately. However, ICP-AES has many analytical uses, probably least of which are biophysical measurements. This is not a problem; however, it should be remembered that some of the advice from standard ICP-AES references does not transfer directly to use with biological macromolecules.
5. Measuring Anions Most ICP-AES instruments are not equipped with a detector sensitive to Cl or other halides. To compensate for this limitation, cacodylate ((CH3)2AsO2), or potentially another detectable anion, can be used as a proxy for the other anions. As shown in Eq. (18.2), the total number of excluded anions can be inferred from measuring the number of excluded cacodylate anions and scaling that number by the ratio of total anions to cacodylate Mdepleted ¼ AsO2 ðCH3 Þ 2 depleted
½M ½AsO2 ðCH3 Þ 2
ð18:2Þ
This relationship assumes that all anions behave in the same way. To have confidence that the cacodylate concentrations are reflective of the actual anions of interest, the two controls outlined in Fig. 18.5 need to be completed. As cacodylate has a pKa of 6.3, it is necessary to make sure that the anion is the primary species present. As shown in Fig. 18.5A, this is done by verifying that the number of excluded anions does not vary as a function of the pH of the reference buffer. Additionally, as the size of the anion could affect how it is excluded, it is prudent to verify that the number of excluded anions does not change with the fractional abundance of the cacodylate. Figure 18.5B demonstrates that this assumption does hold for duplex DNA.
384
Max Greenfeld and Daniel Herschlag
B
A
[M -]/DNA
-2
-4
II
-6
pH = 7.4 pH = 8.0 pH = 8.5
I
AsO2(CH3)−2 depleted/DNA
AsO2(CH3)−2 depleted/DNA
0
-4 -6 -8 -10 0 -2 -4 -6 -8 0
0.8 0.2 0.4 0.6 [AsO2(CH3)−2 ]/[M -]
1
Figure 18.5 Most atomic anions are not detectable on common ICS-AES instruments. Addition of detectable anions can be used as proxies for more abundant anions. For a DNA duplex it was shown that cacodylate serves as a good substitute. (A) Excluded anions can be measured in a regime that is independent of pH for a 24-bp DNA duplex (I: 20 mM Naþ with 10 mM cacodylate and II: 1 mM Mg2þ 20 mM Naþ with 10 mM cacodylate). (B) The calculated number of excluded anions (see Eq. (18.2)) does not depend on the fractional abundance of cacodylate for three ionic conditions: (○) 20 mM Naþ, (□) 100 mM Naþ, and (◊) 1 mM Mg2þ and 20 mM Naþ. Lines are least-squares linear fits. Total excluded anions per DNA are calculated according to Eq. (18.2) and are plotted above the main plot with the same symbols. Reprinted from Bai et al. (2007).
6. Example Protocol This section outlines in detail the specific steps used to carry out the measurements shown in Fig. 18.1. Table 18.3 indicates the suppliers of the reagents and equipment used in gathering the example data. Alternative vendors have also been included for potential future reference. 1. Complementary DNA oligos (GGTGACGAGTGAGCTACTGGGCGG and CCGCCCAGTAGCTCACTCGTCACC) were purchased from commercial vendors (Integrated DNA Technologies or PAN Facility at Stanford) and purified using ion-exchange HPLC (Dionex, DNAPac PA-100). Collected fractions were concentrated and desalted using C-18 cartridges (Waters, Sep-Pak). Quantification of oligos was done using UV absorbance at 260 nm. 2. Equimolar amounts of complementary strands were annealed in 20 mM Na-EPPS (sodium 4-(2-hydroxyethyl)piperazine-10-propanesulfonic
385
Probing Nucleic Acid–Ion Interactions with BE-AES
Table 18.3 Vendors for experimental consumables Consumable
Vendor
Product description
ICP Standards
SPEXCertiPrep Sigma-Aldrich RICCA Chemical Company Millipore
AssuranceTM Standards TraceCERTTM ICP Standards MicroconTM Centrifugal Filter Units MF-MilliporeTM ‘‘V’’ Series Membranes Slide-A-LyzerTM MINI Dialysis Units Polystrene 13 100 mm Round-Bottom Polystrene 13 100 mm Round-Bottom Trace Metal Acids Trace Metal Acids Pipets and Flasks
Spin Columns
Millipore Pierce Tubes
Fisherbrand Falcon
Reagents Glassware
Fisher Sigma-Aldrich Kimax Pyrex
acid), pH 8.0. Purity was checked with Nondenaturing polyacrylamide gel electrophoresis of 1 mg of annealed DNA stained with StainsAll (Sigma-Aldrich) indicating that negligible quantities of free singlestranded DNA were present. 3. Reagents used for buffers and dilutions should be checked for purity using ICP-AES. It is also recommended to determine the concentration of buffers with ICP-AES, as this will lead to internally calibrated measurements. 4. Microcon YM-30 (Millipore) spin columns were used to carry out buffer exchanges with 1 mM Mg2þ (in a typical experiment, see Fig. 18.1, multiple Mg2þ concentrations would be chosen), 20 mM Naþ, 16 mM cacodylate, and 8 mM EPPS, pH 8.0 buffer. Buffer equilibration was conducted at 4 C, during which time the volume of DNA containing sample retained in the top chamber did not fall below 100 mL. Between six and eight rounds of buffer exchange were needed to reach equilibrium. Verification of equilibrium was determined as shown in Fig. 18.3. For the last round of buffer equilibration, the last 80–100 mL of flow-through was collected and is most representative the bulk solution of the DNA-containing sample. Buffer exchanges were done at a relative centrifugal force (RCF) of 10,000g. Used at three-quarters of the maximum RCF recommended by the manufacturer, good batches of these columns will quantitatively retain the DNA in the top chamber of the spin column.
386
Max Greenfeld and Daniel Herschlag
Deviation (%)
5. Aliquots of 20–40 mL from the top of the spin column (i.e., DNAcontaining samples) and the final flow-through were added to polystyrene tubes (Fisher Science) containing 4 mL of water. It has subsequently been determined that diluting the sample in ammonium acetate buffer, pH 5.2, is superior to water. Dilution factors were such that the final concentration of all ions in the analyte remained in the linear range of ICP-AES measurements. 6. ICP-AES was done using an IRIS Advantage 1000 radial ICAP Spectrometer (Thermo Jarrell Ash). A calibration standard solution was made using volumetric glassware typically containing 100 mM of each element being studied. The calibration standard was made from certified single atom reference standards (SPEXCertiPrep). Serial dilutions of this standard can be used to generate standard curves analogous to the one shown in Fig. 18.6. The standard curve shown in Fig. 18.6 is adjusted to the
6% 4% 0 -4%
[Mg2+] measured (mM )
100
10-1
10-2
10−3 10−3
10−2 10−1 [Mg2+] expected (mM )
100
Figure 18.6 Representative standard curve of ICP-AES. Expected concentrations are based on reference standards of certified accuracy. The residual error of measured concentrations is plotted above the standard curve. Error bars represent standard errors of three or more replicate measurements. The linear concentration regime is bracketed by arrows and represents deviations of expected and measured concentrations of less than 5%. Reprinted from Bai et al. (2007).
387
Probing Nucleic Acid–Ion Interactions with BE-AES
concentrations present during buffer exchange before the 100–200 dilution necessary for ICP-AES. Elements were monitored using the following atomic emission lines: As 189.042 nm, Mg 280.270 nm, Na 589.592 nm, P 213.618 nm. Prior to determining the concentration of each sample, the instrument was flushed with dilute nitric acid and then 1 mL of the analyte. The concentration of the remaining 3 mL of sample was determined three times sequentially. The average of the three measurements was used in the final analysis. However, spurious measurements among the three measurements can be an indication of a problem. The instrument was minimally programmed to measure a reference standard every 15 samples. There is significant flexibility to the order in which samples are run on an ICP-AES. A good ICP-ASE can be stable enough that samples can be run in arbitrary order. However, it would be exceedingly difficult to diagnose problems if for any particular sample there was no expectation of what the concentration was. Table 18.4 indicates the typical order in which samples were run. 7. The data should be inspected for systematic error and spurious data points. If a large number of samples are being run, drift in the instrument calibration can cause deviations in the absolute concentration of elements measured. If this drift is uniform for all elements, the results will be minimally affected. However, differential drift among the elements being monitored is possible and will not have a benign effect. Spin column leakage or evaporation can often be detected by looking for Table 18.4 Typical sample order for ICP-AES measurements Sample number
1 2 3 4 5 6 7 8 9 10 11 12 13
Buffer
Description
1 mM Mg
0.1 calibration standard 0.2 calibration standard 0.5 calibration standard 0.8 calibration standard 1 calibration standard Blank Buffer no spin columns Flow-through Top Buffer no spin columns Flow-through Top etc.
2 mM Mg
Celebration standard: typically 100 mM of P, Mg, Na, and As.
388
Max Greenfeld and Daniel Herschlag
spurious data points. Examples include phosphorus showing up in the flow-through, too little phosphorus being retained, and significantly higher concentrations of ions in the flow-through when compared to a reference sample. Any of these observations is an indication that one or more of the basic assumptions of BE-AES measurements is not met and are causes to reevaluate the procedure being carried out. It should be verified that the serial dilutions of the calibration have the expected values. If the deviations from a linear value are significant, a linear regime should be chosen and the concentrations recalibrated. Concentrations outside that range should not be used. Once concentrations are determined, Eqs. (18.1) and (18.2) can be used to determine the individual components of the ion atmosphere. The effectiveness of one cation species out competing another cation was summarized in Fig. 18.1 by monitoring the number of competing cations (CC) and background cations (BC) associated with the DNA in accordance with Eq. (18.3). In Eq. (18.3) the competition constant was determined as the midpoint ð½M1=2 Þ of the BC association assuming a two state model: N ¼ N1 þ
N0 N1 1 þ ð½M=½M1=2 Þn
ð18:3Þ
where [M] is the titrated CC concentration and N0 and N1 are the number of BC at the start and end states. The Hill coefficient n is fitted simultaneously with [M]1/2. The Hill analysis was carried out as an empirical description of the competition behavior and does not represent a physical model for ion binding (Das et al., 2005; Draper, 2004).
ACKNOWLEDGMENTS We are appreciative of technical assistance from Guangchao Li of the Stanford Geological and Environmental Sciences Department on ICP-AES measurements. We would like to thank Yu Bai for help in preparing the figures. Funding was provided by NIH program project grant P01-GM-66275 and an NIH grant GM49243 to D. H.
REFERENCES Anderson, C. F., and Record, M. T. J. (1990). Ion distributions around DNA and other cylindrical polyions: Theoretical descriptions and physical implications. Annu. Rev. Biophys. Biophys. Chem. 19, 423–465. 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(4), 1035–1040.
Probing Nucleic Acid–Ion Interactions with BE-AES
389
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(48), 14981–14988. Bina-Stein, M., and Stein, A. (1976). Allosteric interpretation of Mg2þ binding to the denaturable Escherichia coli tRNAGlu2þ. Biochemistry 15(18), 3912–3917. Das, R., Mills, T. T., Kwok, L. W., Maskel, G. S., Millett, I. S., Doniach, S., Finkelstein, K. D., Herschlag, D., and Pollack, L. (2003). Counterion distribution around DNA probed by solution X-ray scattering. Phys. Rev. Lett. 90(18), 188103. 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(23), 8272–8273. Draper, D. (2004). A guide to ions and RNA structure. RNA 10(3), 335–343. 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(38), 14003–14008. Grilley, D., Misra, V., Caliskan, G., and Draper, D. E. (2007). Importance of partially unfolded conformations for Mg2þ-lnduced folding of RNA tertiary structure: Structural models and free energies of Mg2þ interactions. Biochemistry 46(36), 10266–10278. Grilley, D., Soto, A. M., and Draper, D. E. (2009). Direct quantitation of Mg2þ-RNA interactions by use of a fluorescent dye. Methods Enzymol. 455, 71–94. Herrera, J. E., Correia, J. J., Jones, A. E., and Olson, M. O. (1996). Sedimentation analyses of the salt- and divalent metal ion-induced oligomerization of nucleolar protein B23. Biochemistry 35(8), 2668–2673. Krakauer, H. (1971). The binding of Mgþþ ions to polyadenylate, polyuridylate, and their complexes. Biopolymers 10(12), 2459–2490. Li, A. Z., Huang, H., Re, X., Qi, L. J., and Marx, K. A. (1998). A gel electrophoresis study of the competitive effects of monovalent counterion on the extent of divalent counterions binding to DNA. Biophys. J. 74(2 Pt 1), 964–973. Manning, G. (1969a). Limiting laws and counterion condensation in polyelectrolyte solutions I. Colligative properties. J. Chem. Phys. 51(3), 924–933. Manning, G. (1969b). Limiting laws and counterion condensation in polyelectrolyte solutions II. Self-diffusion of small ions. J. Chem. Phys. 51(3), 934–938. Manning, G. (1969c). Limiting laws and counterion condensation in polyelectrolyte solutions III. An analysis based on Mayer ionic solution theory. J. Chem. Phys. 51(8), 3249–3252. Misra, V., and Draper, D. (2000). Mg2þ binding to tRNA revisited: The nonlinear Poisson– Boltzmann model. J. Mol. Biol. 299(3), 813–825. Romer, R., and Hach, R. (1975). tRNA conformation and magnesium binding. A study of a yeast phenylalanine-specific tRNA by a fluorescent indicator and differential melting curves. Eur. J. Biochem. 55(1), 271–284. Sharp, K., and Hongin, B. (1995). Salt effects on nucleic-acids. Curr. Opin. Struct. Biol. 5(3), 323–328. Stein, A., and Crothers, D. M. (1976a). Equilibrium binding of magnesium(II) by Escherichia coli tRNAfMet. Biochemistry 15(1), 157–160. Stein, A., and Crothers, D. M. (1976b). Conformational changes of transfer RNA. The role of magnesium(II). Biochemistry 15(1), 160–168. Tremethick, D. J. (2007). Higher-order structures of chromatin: The elusive 30 nm fiber. Cell 128(4), 651–654. Zalatan, J. G., Fenn, T. D., and Herschlag, D. (2008). Comparative enzymology in the alkaline phosphatase superfamily to determine the catalytic role of an active-site metal ion. J. Mol. Biol. 384(5), 1174–1189.
C H A P T E R
N I N E T E E N
Using Anomalous Small Angle X-Ray Scattering to Probe the Ion Atmosphere Around Nucleic Acids Suzette A. Pabit,* Kenneth D. Finkelstein,† and Lois Pollack* Contents 1. Introduction 2. Background: SAXS and ASAXS 2.1. Small angle X-ray scattering 2.2. Anomalous SAXS provides measurements of ion distributions 2.3. Computation and interpretation of ASAXS signal 3. Experimental Setup: Samples 3.1. Sample preparation 4. Experimental Setup: Beamline 4.1. From source to sample 4.2. Sample cells 4.3. From sample to detector 5. Data Acquisition 6. Results and Data Analysis 6.1. Comparison of spatial distributions for ions of different valence 6.2. ASAXS probes competition of different ionic species 6.3. Comparing experimental ASAXS profiles with NLPB simulations 7. Conclusion Acknowledgments References
392 393 393 393 395 398 398 398 398 399 400 400 403 403 403 405 407 409 409
Abstract Anomalous small angle X-ray scattering (ASAXS) exploits contrast variation methods to highlight the scattering from one elemental component in a multielement sample, such as one ion species in an ion–DNA system. The ASAXS
* {
School of Applied and Engineering Physics, Cornell University, Ithaca, New York, USA Cornell High Energy Synchrotron Source, Cornell University, Ithaca, New York, USA
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69019-4
#
2009 Elsevier Inc. All rights reserved.
391
392
Suzette A. Pabit et al.
method has been applied to measure ions condensed around short nucleic acid duplexes. This chapter, which briefly describes the origin of the ASAXS signal, focuses on the experimental methods required to carry out these measurements and the interpretation of the anomalous signals.
1. Introduction The high negative charge of DNA and RNA backbones is intimately linked to the structure and function of these macromolecules. To achieve electrical neutrality, this substantial charge is compensated by positively charged atoms/molecules, such as proteins, polyamines, or ions (Bloomfield et al., 2000). For many RNAs, counterions provide sufficient screening to overcome repulsive forces; large RNAs fold to biologically functional structures following the addition of millimolar quantities of divalent ions to low ionic strength solution. Clearly, ion association to nucleic acids is important in facilitating biologically relevant interactions. The importance of quantifying ion–nucleic interactions has long been recognized; numerous techniques have been applied to detect ions associated to (condensed around) nucleic acid strands. Past studies have monitored changing NMR relaxation rates for cations condensed around DNA (Anderson and Record, 1990). A second method detects associated counterions by monitoring energy transfer between luminescent lanthanide ions, which depends on the collision frequencies of condensed counterions (Wensel et al., 1986). Ion-sensitive dyes provide precise calibration of free ion concentrations; ion binding can alter interaction with these indicators (Grilley et al., 2009). Finally, inductively coupled plasma-atomic emission spectroscopy (ICP-AES or AES) (Bai et al., 2007; Plum and Bloomfield, 1988) has been applied to determine the elemental composition of carefully buffer-exchanged samples containing DNA relative to a control sample which does not contain DNA. Differences in ion population in these two samples report the number of excess counterions or excluded coions due to the presence of the nucleic acid. The subject of this chapter is a technique that provides information about the spatial distribution of ions associated to DNA or RNA: anomalous SAXS (or ASAXS). These measurements enable direct comparison with theoretical predictions of ion spatial distribution. To date, much experimental effort has focused on measuring ions associated to short strands of DNA, due to the enhanced robustness (and reduced expense) of DNA relative to RNA. ASAXS results confirm the predicted dependence of ion spatial distribution on ion valence and competition. Most recently, ASAXS studies of ion–RNA systems illustrate key differences between ion association to RNA as opposed to DNA, highlighting the importance of helix topology.
ASAXS Probes Ion Atmospheres
393
2. Background: SAXS and ASAXS 2.1. Small angle X-ray scattering We begin with a brief discussion of small angle X-ray scattering (SAXS). This valuable experimental technique provides information about the size, shape, compactness, and molecular weight of molecules in solution when their characteristic size is larger than the wavelength (l) of the X-rays used, ˚ (Doniach, 2001; Svergun and Koch, 2003). Thus, it is an ideal typically 1 A probe of nucleic acid conformation. SAXS experiments on macromolecules are typically conducted in solution, where molecules can freely change conformation; however, the orientation of molecules relative to the incident beam is unconstrained, so the SAXS signal reports a spherical average of molecular orientations. Some of the most recent analysis methods are discussed in Chapter 11. Small angle scattering from RNA or DNA arises from phase differences between X-rays scattered from electrons at different locations within the macromolecule. The intensity of the scattered radiation is largest along the direction of the incident beam, and decreases as the scattering angle, 2y, increases. The angular dependence of the scattered X-rays reflects the size and shape of the sample; at a given scattering angle the measured intensity relative to the maximum value at 2y ¼ 0 is smaller for larger objects, due to the greater path length (hence phase) difference of X-rays scattered from widely separated sources (Glatter, 1982). Thus, the angular dependence of the scattered radiation reflects the spatial distribution of electrons inside the sample, in this case the macromolecule. Figure 19.1 illustrates the geometry of a typical scattering experiment. The strength of the scattering signal arises from the difference in electron density (or contrast) of the macromolecule relative to the background solvent. The measured intensity I(2y) is directly proportional to the product of the molecular concentration and the square of the electron density difference between macromolecule and solvent (Svergun and Koch, 2003) at the corresponding length scale. When the scattering is from DNA or RNA, an additional component must be considered: the high-density cloud of counterions closely associated to the nucleic acid strand (Manning, 1969). For these systems, scattering arises from the excess electron density (contrast) of both the nucleic acid and the counterions relative to the uniform solvent background.
2.2. Anomalous SAXS provides measurements of ion distributions Anomalous SAXS (ASAXS) takes advantage of contrast variation methods to highlight the small angle scattering signal from a single elemental component correlated with a larger system (Stuhrmann, 1981), for example, one
394
Suzette A. Pabit et al.
Scattering image CCD Incident X-rays
Converted to I(q) via radial integration
Sample 2q
I
Beam stop
2q
2q 2q
Figure 19.1 Schematic of a small angle X-ray scattering experiment. The top panel illustrates how a SAXS experiment is carried out. An X-ray beam, incident from the left, is scattered by the sample. The angular dependence of the scattered radiation is recorded on a two-dimensional CCD detector placed about 1 m downstream of the sample. A typical image from such a detector is shown at top, right. To avoid detector damage, the direct beam is blocked by a beam stop, which can be seen at the lower left of the image. The lower panel illustrates how the size of objects in the sample affects the scattering profiles. Scattering from smaller objects persists to higher angle than scattering from larger objects because the phase differences between scattered X-rays are smaller. For smaller objects destructive interference occurs at larger angles. In addition, the scattering in the forward direction (at zero angle) reflects the molecular weight of the sample. Thus, scattering from the larger (pink) object has a higher intensity at zero angle, but decreases more rapidly with angle than scattering from a smaller, less massive (green) object.
species of ion in an ion–DNA system. All elements possess a unique set of characteristic energies, corresponding to electron binding energies and can be individually targeted by tuning the energy of an X-ray beam until it exactly equals, or is resonant with a specific electronic transition. Near its resonant energy the scattering power of this element is reduced (e.g., Creagh, 1999). Thus, if scattering is measured far from a resonant edge and a second profile is measured close to a resonant edge, scattering from the resonant element will be altered. For these experiments, the contrast, or scattering strength of the ions alone has been modified. By precise subtraction of normalized signals, it is possible to obtain information about the spatial distribution of the resonant elements surrounding nonresonant structures. If ions with accessible resonant edges are employed, counterions around DNA can be targeted. This latter condition places severe restrictions on the elements that can be probed. Since most biologically interesting counterions (e.g., Naþ, Kþ, Mg2þ, Ca2þ) are derived from low Z
395
ASAXS Probes Ion Atmospheres
(atomic number) elements, this lower bound is of interest. Due to experimental challenges associated with measurements at X-ray energies below about 6 keV, elements with Z greater than 24 (e.g., Mn) are most easily accessible.
2.3. Computation and interpretation of ASAXS signal The amplitude of X-rays scattered into a particular direction (characterized by a momentum transfer q ¼ (4p/l)siny ) is expressed as the product of a form factor, F(q) that describes the shape of the ‘‘scatterer’’ and a scattering factor, f which reflects the electron density of the ‘‘scatterer’’ relative to the background. The latter factor is of critical importance to ASAXS. Far from the absorption edge, f has a constant value f0, that of the X-ray atomic form factor that (in vacuum) is proportional to atomic number (Z); near the edge f changes dramatically. Due to ‘‘resonant scattering’’, the real part of f decreases by an amount denoted f 0 , while an imaginary component, f 00 appears at (and above) the edge, and accounts for absorption by the atom. Near an ion’s absorption edge, the total scattering factor is energy dependent and expressed as fion ¼ f0 þ f 0 ðEÞ þ if 00 ðEÞ
ð19:1Þ 0
The energy dependent contributions to the scattering factor, f and f 00 are illustrated in Fig. 19.2 for cobalt. Cobalt is of interest to our work because of its presence within cobalt hexammine, a trivalent ion commonly used in nucleic acid studies. For a multicomponent system consisting of ions plus DNA, the scattering amplitude is expressed as: As ¼ fDNAFDNA(q) þ fIONSFIONS(q). Here, the f ’s represent the effective number of electrons that contribute to the scattering while the F’s reflect the spatial arrangement of the scattering particles, thus fIONS ¼ Nions*fion where Nions equals the number of ions associated to (condensed on) the nucleic acid. (Note that the effective number of electrons contributed to the scattering is computed relative to the solvent background. See Das et al. (2003b) for more information about this contrast.) Since detectors measure intensity, not amplitude, and the intensity I (q) is the product of the scattering amplitude As and its complex conjugate A*s : IðqÞ ¼ ð fDNA FDNA ðqÞ þ fIONS FIONS ðqÞÞð fDNA FDNA ðqÞ þ fIONS FIONS ðqÞÞ* ð19:2Þ To compute I(q, E ) near the ion absorption edge (and far from any resonant edge for the DNA), the energy dependent expression for fion is substituted into the above equation. With the restriction that measurements are made below the edge, so f 00 (E) 0, the intensity can be expressed as a quadratic function of f 0 (E) (Patel et al., 2004):
396
Suzette A. Pabit et al.
5 Scattering factors (electrons)
f⬘ f ⬘⬘ 0
Eoff
-5
Eon
-10 7000
7200
7400
7600
7800
8000
8200
X-ray energy (eV)
Figure 19.2 Changes in the scattering factor of cobalt near the cobalt absorption edge. At energies approaching the cobalt K-edge (7.71 keV), the effective scattering factor of cobalt changes due to anomalous or resonant effects. The lower curve reflects changes in the real part of the scattering factor, while the upper curve reflects changes in the imaginary part of the scattering factor (absorption). The circles represent typical working energies for an ASAXS experiment at this edge. Energies are selected to maximize changes in f 0 while keeping changes in f 00 minimal.
Iðq; EÞ ¼ c þ b f 0 ðEÞ þ að f 0 ðEÞ2 Þ
ð19:3Þ
where 2 2 2 FIONS ; b ¼ Nions *ð2fDNA FDNA FIONS þ 2f0 Nions FIONS Þ; a ¼ Nions 2 FIONS Þ2 c ¼ ð fDNA FDNA Þ2 þ 2fDNA f0 Nions FDNA FIONS þ ð f02 Nions
Thus, at each value of q, I(q, E) varies with f 0 (E) to the 0th, 1st, and 2nd order. The total number of electrons associated with the DNA is significantly larger than the total number of electrons associated with the counterion system, although the effects of hydration and electrostriction make it challenging to compute exact values for scattering factors a priori (Das et al., 2003b). The third term in Eq. (19.3) is significantly smaller than the second, because the change in the contrast is typically a small fraction of the total contrast. Thus, if measurements are carried out at only two energies below the edge, the intensity difference is well approximated by linear variation: DIðqÞ ¼ bð f 0 ðE2 Þ f 0 ðE1 ÞÞ FDNA FIONS
ð19:4Þ 2 FIONS
term are This approximation assumes that contributions from the smaller than those resulting from the FDNAFion terms because in general fDNAFDNA is larger than f0NionsFIONS.
397
ASAXS Probes Ion Atmospheres
Three products of form factors occur in Eq. (19.3): FDNAFDNA, FDNAFIONS, and FIONSFIONS. Figure 19.3 provides guidance for interpreting each of these terms. The interference term of Eq. (19.4) depends on both the structure of the DNA and the spatial distribution of the counterion cloud surrounding it. In simple terms, it reflects the set of all vectors that have one end inside the DNA and the other end in the ion cloud (such as the red arrow in Fig. 19.3). In this chapter, we report on anomalous signals derived from the difference between scattering intensities taken at two energies (Eq. (19.4)). Extraction of all three terms in Eq. (19.3) requires 2 precise measurement of the (smaller) FIONS term. As discussed in the literature ( Jusufi and Ballauff, 2006), we note that the ‘‘a’’ term can be extracted from a quadratic fit, if measurements are carried out at several energies such that f 0 (E)2 varies by an order of magnitude. The coefficient of 2 this term, FIONS directly reports the spatial distributions of the ions without reference to the structure of the DNA. Efforts are currently underway to measure this term for ions around DNA. In summary, ASAXS has the potential to provide information both about the association of counterions to a macroion, as well as about the
B-DNA Associated counterion
Figure 19.3 Interpretation of form factor products in Eq. (19.3). This figure illustrates the meaning of the cross-terms of Eq. (19.3), using an atomic representation of B-DNA, along with condensed counterions (blue circles). The FDNAFDNA term, discussed in the text, is related to the set of all vectors that begin and end in the DNA, for example, the yellow arrow. The FDNAFIONS term is related to the set of all vectors with one end in the DNA and the other end in the counterion cloud, for example, the red arrow. Finally, the FIONSFIONS term is related to the set of all vectors connecting ions, represented by the cyan arrow in the figure.
398
Suzette A. Pabit et al.
structure of the ion cloud alone. Analysis and interpretation of the data is discussed in a later section.
3. Experimental Setup: Samples 3.1. Sample preparation Samples discussed here are annealed duplexes of single-stranded 25 nucleotide desalted and purified DNA (IDT, Coralville, IA) and RNA molecules (Lafayette, CO) (Andresen et al., 2004, 2008; Das et al., 2003b; Pabit et al., 2009). The duplex strands were dialyzed extensively to a specific bulk salt concentration (e.g., 100 mM Rb(CH3COO) and 0.5 mM Co(NH3)6Cl3) and buffer solutions using microcon centrifugal filter units (Millipore Corporation, Billerica, MA). This dialysis step is important because it ensures that the condensed ions around the nucleic acids and ions in the bulk solution were able to attain equilibrium. Note that we favor the lower molecular weight anion acetate, as opposed to Cl to minimize background scattering. After dialysis, the nucleic acid solutions are brought to a final volume of 40 ml and final duplex concentrations of 0.2–0.6 mM.
4. Experimental Setup: Beamline 4.1. From source to sample The following information is based on experience using a single beamline (CHESS C-line), over approximately 5 years. A bend magnet source of synchrotron radiation X-rays has proved quite adequate for a wide range of ASAXS measurements. The beamline delivers flux at the sample of up to 5 1011 photons/s/mm2, in bandwidth DE/E 0.00025 at 10 keV, using a silicon (220) double crystal monochromator (mono). The mono is set to collect up to two horizontal milliradians that is focused (approximate ratio 3:1) at a convenient point between the sample and detector. This point is determined by the maximum horizontal sample size. One can reduce horizontal divergence at the cost of intensity. Before a recent upgrade, a flat rhodium coated, postmono mirror was used in conjunction with monocrystal detuning to suppress high-energy harmonics. The upgrade provides vertical focusing capability upstream of the mono, increasing flux by two to three times, without appreciable change in divergence. The effect of the horizontal divergence is to ‘‘blur’’ the scattering pattern in this direction. Although the blurring is pronounced only along the horizontal direction, we accept scattering data that fall within 45 around the vertical. The width of a peak in the scattering profile from our polycrystalline calibrant
399
ASAXS Probes Ion Atmospheres
Upstream flight tube
Slit
Slit
Detector Downstream flight tube
Figure 19.4 A schematic view of the ASAXS setup; downstream is to the right. Ion chambers (blue) monitor beam intensity before and after the main slit, and can be inserted for measurement of sample transmittance between downstream flightpath (green) and CCD X-ray detector. A guard slit with built-in shutter and sample is located between the flight tubes. All slits are moved under user control. The flight tube upstream of the sample is helium filled and contains a built-in ion chamber; the downstream path is vacuum terminated by silicon nitride at the front end and Kapton at the detector end. A beamstop with transverse position control is located in the vacuum; it is made from amorphous metal that greatly reduces energy dependent powder diffraction structures that can compromise the beamstop scattering used to monitor beam transmission through the sample.
silver stearate displays no broadening on this scale. A schematic drawing of the ASAXS beamline is shown in Fig. 19.4.
4.2. Sample cells Small volume (30 ml) sample cells are manufactured by machining a rectangular hole through a flat piece of acrylic. The X-ray energy (determined by the resonant energy for ASAXS experiments) dictates the acrylic (and sample) thickness. Thin cells are better suited for lower energies due to increased absorption. For measurements near 8 keV (Co K-edge) 1 mm is optimal while 3 mm sample thickness is used at 15–16 keV (Rb and Sr K-edges). The cells are fitted with silicon nitride windows, fabricated at the Cornell Nanoscale Science and Technology Facility using a wet-etch process. Ultrathin windows provide good signal-to-noise by minimizing background absorption and scattering and allowing accurate background subtraction (Andresen et al., 2004). Window fabrication is discussed in more detail in Chapter 12. Small holes are drilled through the acrylic into the sample volume to enable filling by syringe. Figure 19.5 shows a photograph of the sample cells, with details provided in the caption.
400
Suzette A. Pabit et al.
5 mm
Figure 19.5 The ASAXS sample cells. This photograph shows three adjacent sample cells used for ASAXS measurements. As described in the text, the cells are machined from acrylic. The long fill holes that run up and down are used to load and unload the cells. The leftmost cell in the photograph is unsealed; the two cells at the right have been sealed with silicon nitride windows. The silicon nitride free standing membrane is the light green rectangle at the center of the silicon supporting frame.
4.3. From sample to detector After passing though the sample cell, the scattered X-rays travel to the detector through an evacuated flight tube. The use of vacuum eliminates additional background arising from air scatter. Near the end of the flight tube, a motorized (XIA, Hayward, CA) beam stop prevents the direct beam from reaching the CCD detector (homebuilt 1 K, fiber-optically bonded CCD; Tate et al., 1995) used in collecting the scattering data. Compton scattering from the beam stop passes though a vacuum-sealed Kapton window on one side of the flight tube to an Xflash detector (Rontec, Carlisle, MA) which monitors the X-ray intensity transmitted through the sample cell and allows intensity normalization of the data. The distance between the sample cell and the CCD is about 1 m.
5. Data Acquisition For anomalous scattering experiments, data are acquired near the K absorption edge energies of chosen elements. The energy constraints, discussed above, require that we study monovalent Rbþ (edge at 15.2 keV), divalent Sr2þ (16.1 keV), and trivalent cobalt hexammine (Co edge at 7.71 keV) to monitor ion distributions around DNA. The energy resolution of the beamline monochromator varies from 2 to 4 eV between 7.5 and 16.1 keV. To obtain the anomalous signal, DI(q), Eq. (19.4), we acquire
ASAXS Probes Ion Atmospheres
401
scattering data at two X-ray energies. The first is well below (Eoff-edge or Eoff), and the second near (Eon-edge or Eon) the absorption edge. From Eqs. (19.3) and (19.4), we show that the difference of Eon and Eoff scattering profiles removes the energy independent terms and yields the anomalous scattering signal Eq. (19.4), written explicitly as DIðqÞ¼Iðq;Eoff ÞIðq;Eon Þ2fDNA FDNA ðqÞNions ðf 0 ðEoff Þf 0 ðEon ÞÞFIONS ð19:5Þ Energies are selected from transmission scans of a reference solution containing the ion of interest. Care must be taken to find the X-ray edge, which may be shifted from the edge of a pure metal reference. The on energy is set to maximize the magnitude of Df 0 ðEÞ while Df 00 ðEÞ and fluorescence contributions to the image are very small (see Fig. 19.2). To illustrate how X-ray energies, Eon and Eoff, are determined, consider measurements near the Rb edge. From measurements of the energy-dependent transmission through a Rb–Acetate solution the absorption peak was determined to be 15.230 keV. We set the ‘‘on’’ energy just below the onset of absorption, Eon ¼ 15.213 keV, while Eoff ¼ 15.113 keV is well below the absorption edge. To minimize time-dependent systematic errors, data are collected using the following acquisition sequence: two images at Eoff, four images at Eon, two images at Eoff. We repeat this sequence either eight (or 16) times. For effective background subtraction, anomalous scattering data acquisition of the sample is bracketed by four (or eight) sequences of buffer images, taken immediately before and after the DNA measurements. Buffer background images are acquired by measuring SAXS profiles of solutions used for dialysis. This procedure yields a total of 32 (or 64) Eoff images and 32 (or 64) Eon images per sample. At higher (i.e., Rb and Sr) energies where there is reduction in background scattering, 32 images per energy per sample suffice to achieve good signal to noise ratio. For the lower energy Co experiments, 64 images per energy per sample are required. The X-ray exposure time per image is 10 s, and the total exposure is below the threshold for radiation damage at CHESS C-line. Images are radially integrated to produce a one-dimensional I versus q curve. Decay of X-ray beam intensity over time, as well as any energy dependent differences in transmission are accounted for by transmitted beam intensity normalization. Figure 19.6 illustrates how we process the two-energy data and generate the ASAXS profiles. First, we average the set of all DNA and buffer intensity profiles at each of the two energies. The averaged data are displayed in Fig. 19.6A. Note that both DNA and buffer profiles, taken at Eon, have slightly elevated background relative to Eoff data; this is due to X-ray fluorescence. Most of this fluorescence background is removed by buffer subtraction (see Fig. 19.6B). The remaining on-edge fluorescence results
402
A
Suzette A. Pabit et al.
B
0.025 Eoff Eon
0.02
Eoff Eon
0.01
I(q)
0.015 0.01
0.005
0.005 0
0.1
0.3
0.2
0
0.4
C
0.1
0.2
0.3
0.4
D ⫻10–4 Eoff Eon
8 6 DI
I(q)
0.01
0.005
4 2 0
0
0.1
0.2
0.3
q (Å−1)
0.4
-2
0.1
0.2 q (Å−1)
0.3
Figure 19.6 Demonstration of data processing to generate ASAXS profiles. Panel (A) shows the raw data, acquired at the two ASAXS energies, along with the corresponding buffers. Panel (B) shows the buffer subtracted curves at each energy. Panel (C) shows the data after corrections have been applied to account for energy dependent effects and for fluorescence, detailed in the text. Panel (D) illustrates the measured difference signal: I(Eoff) I(Eon).
from the slightly elevated concentration of ions in the DNA-containing solution compared to the buffer solution. We compute the additive correction resulting from this offset, fluoroffset, by matching Eon and Eoff profiles at high q. It is also critically important to account for the changing transmission of all beamline components as the energy is varied, as well as the energy response of all detectors. A multiplicative correction factor lcorrection is computed, based on differences between on- and off-edge profiles of DNA in nonresonant buffers of identical ionic strength. The correction factor lcorrection ¼ Inonrsnt ðq; Eoff Þ=Inonrsnt ðq; Eon Þ is typically computed from a sample dialyzed against for example, 100 mM NaCl instead of 100 mM Rb–Acetate. Here, Inonrsnt is the scatting from the nonresonant sample. Typically, this multiplicative factor is equal or very close to unity. The corrected SAXS signals are represented by Eq. (19.6) and shown in Fig. 19.6C.
ASAXS Probes Ion Atmospheres
Icorr ðq; Eon Þ ¼ Iðq; Eon Þ lcorrection flouroffset Ianom ðqÞ ¼ Iðq; Eoff Þ Icorr ðq; Eon Þ
403
ð19:6Þ
The corresponding ASAXS profiles, Ianom(q), are shown in Fig. 19.6D. Note that the difference signal is small. This signal arises from approximately 40 Rbþ ions per 25 bp DNA and typically is one order of magnitude smaller than the on- or off-energy SAXS profiles. However, these ASAXS data collection methods produce results that are robust and reproducible between runs.
6. Results and Data Analysis ASAXS provides valuable spatial information about ions associated to nucleic acids. Because ion resonance is X-ray energy specific, a particular ionic species can be targeted in ASAXS experiments (even in conditions where there are multiple ionic species in the bulk solution). In this section, we provide several examples of how ASAXS experiments can be used to study counterion atmospheres around DNA and RNA.
6.1. Comparison of spatial distributions for ions of different valence The shape and extent of the anomalous signal can used to probe the spatial distribution of selected valence ions around DNA. In Fig. 19.7, we show the increase in high-angle (high q) anomalous scattering as the ion valence increases from 1 to 3. Note that X-ray scattering from a larger object falls off more rapidly than scattering from a smaller object (illustrated in Fig. 19.1). In ASAXS, the anomalous signals generally represent the set of all vectors with one end inside DNA and the other in the ion cloud. Thus, ASAXS signals persisting to higher q represent shorter vectors linking DNA to ions. Through this qualitative comparison, we find that higher valence ions are more tightly bound (more localized) to the DNA surface than lower valence ions (Andresen et al., 2004, 2008).
6.2. ASAXS probes competition of different ionic species When a solution contains multiple ionic species, ASAXS can be used to determine the fractional contribution of each cationic species to the ion atmosphere around nucleic acids. As an illustration, we show monovalent and trivalent ion competition around DNA. Figure 19.8A displays anomalous difference signal from Rbþ ions bound to DNA in two solutions. In a
404
Suzette A. Pabit et al.
ΔI
100
10-1 Rb Co
0.05
0.15
0.1 q (Å-1)
Figure 19.7 Monovalent and trivalent ASAXS signal comparison. This figure illustrates differences in shape between the ASAXS signals for monovalent Rb around DNA and trivalent cobalt hexammine around DNA. The curves have been scaled to match at the lowest q to enable shape comparison. The increase in high-angle scatter associated with the trivalent relative to monovalent ion signal, suggests that the trivalent ions are more tightly localized to the DNA. A
B 0.14
Rb energy: 100 mM Rb+ with
0.12
ΔI
0.08
0 mM CoHex 0.5 mM CoHex
0.08
0 mM CoHex 0.5 mM CoHex
0.1
Co energy: 100 mM Rb+ with
0.1
0.06
0.06
0.04
0.04 0.02
0.02
0
0 −0.02
0.05
0.1
0.15 q (Å−1)
0.2
0.25
−0.02
0.05
0.1 0.15 q (Å−1)
0.2
Figure 19.8 Anomalous signals of DNA in 100 mM Rbþ (blue curves) and in a solution containing both 0.5 mM [Co(NH3)6]3þ and 100 mM Rbþ (green curves). Panel (A) shows the ASAXS signal acquired at the Rb energy for both of these samples. In the pure Rb sample (blue), all of the condensed counterions are Rbþ. The Rb anomalous signal drops after even small amounts of cobalt hexammine are added, reflecting the very efficient replacement of monovalent by trivalent ions. The bulk Rb ion concentration is identical for both samples: the anomalous signal reports on the condensed counterions. Panel (B) shows ASAXS curves acquired on the same samples at the Co energy. In the pure Rb sample (blue curve), no cobalt hexammine is present and the signal intensity is zero. The green curve shows the increasing cobalt anomalous signal, reflecting the presence of cobalt in the counterion cloud.
ASAXS Probes Ion Atmospheres
405
solution containing DNA in 100 mM Rbþ ions, there is a large anomalous signal; yet when even small amounts of trivalent ions are added to the solution (0.5 mM [Co(NH3)6]3þ ion), the monovalent ions are rapidly displaced from around DNA and the Rb anomalous signal decreases significantly. Although the bulk concentration of Rbþ is equal for both experiments, the decrease in bound Rbþ occurs because of strong competition from trivalent [Co(NH3)6]3þ. The corresponding increase in the Co anomalous signal is shown in Fig. 19.8B. The change in amplitude of the ASAXS signal is quantified by integrating the total anomalous scattering profile and scaling this signal to reflect the DNA charge. Within the approximations discussed herein, this yields the relative number of ions around the DNA. This value was further verified by comparing the integrated magnitude of the ASAXS signal with results from ICP experiments on identically prepared samples. The latter measurements report the number of excess cations, and are in good agreement with the ASAXS results. Figure 19.9 shows the integral of the ASAXS signals at the Rb and Co energies and the comparison to the predictions to the nonlinear Poisson–Boltzmann (NLPB) equation, using the APBS package for comparison (Baker et al., 2001). Note the marked decrease in the number of bound Rbþ ions as the concentration of the competing trivalent ion is increased while the bulk monovalent ion concentration is kept at 100 mM. Details of this experiment can be found in Andresen et al. (2008).
6.3. Comparing experimental ASAXS profiles with NLPB simulations While most previous ASAXS studies have focused on DNA, we have recently used ASAXS to probe the counterion distribution around double-stranded RNA molecules. In previous work, we showed that numerical calculations using an ion-size corrected NLPB model adequately describes ASAXS data (Andresen et al., 2004; Das et al., 2003a,b). By comparing experimental ASAXS profiles with simulations based on the NLPB equation (Pabit et al., 2009), we found that the RNA anomalous signal is more sensitive to the choice of probe ion radius used in the calculations ˚ is necessary to (see Fig. 19.10). In fact, an ion radius upper bound of 4 A adequately describe RNA ASAXS data. In the case of DNA, a wider range ˚ ion radii sufficiently describes the data (see Fig. 19.9). We attribute of 2–6 A the difference between DNA and RNA to A-form helix topology. The RNA A-form helix has a deeper and narrower major groove compared to the DNA B-form helix. We find that counterions penetrate the A-form major groove as suggested by past crystallographic (Robinson et al., 2000) and theoretical (Chin et al., 1999; Mills et al., 1992) studies. In general, the counterions are more closely localized to the RNA central axis than DNA as shown in the ASAXS profiles displayed in Fig. 19.11A. RNA anomalous
406
Suzette A. Pabit et al.
40 APBS r=2Å r=3Å r=6Å
35
Excess cation charge
30 25 20 15 10
Rb ASAXS for Rb-CoHex Rb ASAXS for Rb-Spd
5 ,
0 0
0.1
0.2
0.3
0.4
0.5
Co ASAXS for CoHex-Na Co ASAXS for CoHex-Rb 0.6
0.7
0.8
0.9
1
Trivalent ion concentration (mM)
Figure 19.9 Comparison of counterion competition data with the numerical solution of the PB equation from APBS. Upper half: DNA charge compensated by Rbþ ions in competition with (Co(NH3)6)3þ (circles) and spermidine3þ (squares) in 100 mM RbCl, 0.2 mM [DNA]. Lower half: DNA charge compensated by (Co(NH3)6)3þ ions in competition with 100 mM NaCl at 0.2 mM [DNA] (triangles) and in competition with 100 mM NaCl (diamonds) or RbCl (stars) at 0.6 mM [DNA]. Data without error bars have errors smaller than symbol size. The dashed lines represent APBS computations ˚ ion radius underestimates the for DNA surrounded by ions with different radii. A 2 A fraction of monovalent ions in the atmosphere. The data are consistent with ion radii of ˚ or greater with an upper bound of 6 A ˚ . (Figure and caption reproduced with 3 A permission from Andresen et al. (2008).)
signals persist at higher q suggesting closer association of counterions to the RNA molecules. Another comparison of experimental and theoretical data is shown in Fig. 19.11B and C. Here, we compute the radial Patterson inversions, U(R), of the experimental data (Fig. 19.11B) for comparison with the simulations (Fig. 19.11C). The radial Patterson inversion of the anomalous difference signal is calculated using ð R UðRÞ ¼ 2 qIanom ðqÞ sinðqRÞ dq ð19:7Þ 4p
407
ASAXS Probes Ion Atmospheres
1
RNA, data RNA, calculated
ΔI(q)
0.8
r = 4 Å probe ion radius
0.6 0.4 0.2 0 0.05
0.1
0.15 0.2 q (Å-1)
0.25
0.3
Figure 19.10 RNA ASAXS data compared with APBS predictions. This figure shows the measured anomalous difference signal for Rb-RNA in comparison with predictions generated using APBS. To obtain good agreement, the finite size of the ion must be ˚ (pictured) optimizes the included. In this case, an ion probe radius of 3 or 4 A comparison.
as described by Engelman et al. (1975). In the ideal case, U(R) reports on the distribution of vector lengths correlating nucleic acids with the condensed counterions, supporting our qualitative result that ions are more closely localized to the RNA than the DNA. However, some caution must be employed when using Patterson inversions in describing ASAXS data because the anomalous signal contains small contributions from ion–ion distances, as shown in Eqs. (19.3) and (19.4). We note that ion–ion distances should mimic the DNA– or RNA–ion distances since ions that are tightly bound to the nucleic acid are also closely associated with each other. Due to the complexities in separating out the different terms illustrated in Fig. 19.3, it is most straightforward to analyze and interpret the data by direct comparison to simulations.
7. Conclusion Overall, this chapter shows how ASAXS data can be acquired and carefully interpreted to yield information related to the spatial distribution and number of counterions associated to DNA. Due to the small size of the signal, ASAXS requires careful attention to experimental detail. Background scattering must be minimized to the greatest possible extent.
408
Suzette A. Pabit et al.
A
ΔI(q)
101
100
10-1
Rb energy RNA DNA 0.05
0.15 q (Å-1)
0.2
0.25
C
⫻10-2
U(R)
B
0.1
5 4 3 2 1 0
100 mM Rb+
r=3 Å
RNA calc DNA calc
RNA data DNA data 0
20 30 40 10 Correlation vector length (Å)
0
10 20 30 40 Correlation vector length (Å)
50
Figure 19.11 Comparison of anomalous signals obtained for Rb-RNA and Rb-DNA. Panel (A) shows the ASAXS signals for both DNA and RNA displayed on a log scale. Panels (B) and (C) illustrate an alternative method for comparing measured with predicted ion distributions. A radial Patterson inversion can be applied both to the data and to the curve generated from simulation. Although it is challenging to interpret this curve directly (see text), it is instructive to compare experiment with prediction in ˚ is used, the differences in ion distribution this form. When an ion probe radius of 3 A between RNA and DNA are clearly mirrored by simulation.
The most robust analysis methods involve direct comparison of the DI (q) term with theoretical computations. Agreement between data and prediction then validate the models used. The application of ‘‘inversion’’ approaches involves taking Fourier transforms of the data to yield the set of vectors connecting scattering particles. However, one must be cautious when interpreting the results of these inversions. Experiments are cur2 rently underway to measure the FIONS term independently, which will allow us to extract the pure ion–DNA cross-term which is more straightforward to interpret. In spite of challenges, ASAXS continues to provide unique information about ions associated to DNA or RNA.
ASAXS Probes Ion Atmospheres
409
ACKNOWLEDGMENTS We acknowledge those who have contributed to the ASAXS projects: K. Andresen, R. Das, S. Doniach, D. Herschlag, L. Kwok, J. Lamb, L. Li, G. Maskel, S. Meisburger, T. Mills, H. Y. Park, and X. Qiu. Funding for this work was provided by NSF through grant MCB0347220, and by the NIH through P01-GM066275. CHESS is supported by the NSF and the NIH/NIGMS under Grant No. DMR-9713424. The CNF is supported by the NSF, Cornell University and industrial affiliates.
REFERENCES Anderson, C. F., and Record, M. T. (1990). Ion distributions around DNA and other cylindrical polyions—Theoretical descriptions and physical implications. Annu. Rev. Biophys. Biophys. Chem. 19, 423–465. Andresen, K., Das, R., Park, H. Y., Smith, H., Kwok, L. W., Lamb, J. S., Kirkland, E. J., Herschlag, D., Finkelstein, K. D., and Pollack, L. (2004). Spatial distribution of competing ions around DNA in solution. Phys. Rev. Lett. 93, 248103. Andresen, K., Qiu, X., Pabit, S. S., Lamb, J. S., Park, H. Y., Kwok, L. W., and Pollack, L. (2008). Mono and tri-valent ions around DNA: A small angle scattering study of competition and interactions. Biophys. J. 95, 287–295. 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. Baker, N. A., Sept, D., Joseph, S., Holst, M. J., and McCammon, J. A. (2001). Electrostatics of nanosystems: Application to microtubules and the ribosome. Proc. Natl. Acad. Sci. USA 98, 10037–10041. Bloomfield, V. A., Crothers, D. M., and Tinoco, I. J. (2000). Nucleic acids: Structures, properties and functions. University Science Books, Sausalito, CA. Chin, K., Sharp, K. A., Honig, B., and Pyle, A. M. (1999). Calculating the electrostatic properties of RNA provides new insights into molecular interactions and function. Nat. Struct. Biol. 6, 1055–1061. Creagh, D. C. (1999). X-ray dispersion corrections. The International Union of Crystallography, Dordrecht/Boston/London. Das, R., Kwok, L. W., Millett, I. S., Bai, Y., Mills, T. T., Jacob, J., Maskel, G. S., Seifert, S., Mochrie, S. G. J., Thiyagarajan, P., Doniach, S., Pollack, L., et al. (2003a). The fastest global events in RNA folding: Electrostatic relaxation and tertiary collapse of the tetrahymena ribozyme. J. Mol. Biol. 332, 311–319. Das, R., Mills, T. T., Kwok, L. W., Maskel, G. S., Millett, I. S., Doniach, S., Finkelstein, K. D., Herschlag, D., and Pollack, L. (2003b). The counterion distribution around DNA probed by solution X-ray scattering. Phys. Rev. Lett. 90, 188103. Doniach, S. (2001). Changes in biomolecular conformation seen by small angle X-ray scattering. Chem. Rev. 101, 1763–1778. Engelman, D. M., Moore, P. B., and Schoenborn, B. P. (1975). Neutron-scattering measurements of separation and shape of proteins in 30s ribosomal-subunit of Escherichia coli—S2–S5, S5–S8, S3–S7. Proc. Natl Acad. Sci. USA 72, 3888–3892. Glatter, O. (1982). Small Angle X-Ray Scattering. Academic Press, New York. Grilley, D., Soto, A. M., and Draper, D. E. (2009). Direct quantitation of Mg(2þ)-RNA interactions by use of a fluorescent dye. Methods Enzymol. 455, 71–94. Jusufi, A., and Ballauff, M. (2006). Correlations and fluctuations of charged colloids as determined by anomalous small-angle X-ray scattering. Macromol. Theory Simul. 15, 193–197.
410
Suzette A. Pabit et al.
Manning, G. S. (1969). Limiting laws and counterion condensation in polyelectrolyte solutions. 2. Self-diffusion of small ions. J. Chem. Phys. 51, 934–938. Mills, P. A., Rashid, A., and James, T. L. (1992). Monte-Carlo calculations of ion distributions surrounding the oligonucleotide D(Atatatatat)2 in the B-conformations, a-conformations, and wrinkled D-conformations. Biopolymers 32, 1491–1501. Pabit, S. A., Qiu, X. Y., Lamb, J. S., Li, L., Meisburger, S. P., and Pollack, L. (2009). Both helix topology and counterion distribution contribute to the more effective charge screening in dsRNA compared with dsDNA. Nucleic Acids Res. 37, 3887–3896. Patel, M., Rosenfeldt, S., Ballauff, M., Dingenouts, N., Pontoni, D., and Narayanan, T. (2004). Analysis of the correlation of counterions to rod-like macroions by anomalous small-angle X-ray scattering. Phys. Chem. Chem. Phys. 6, 2962–2967. Plum, G. E., and Bloomfield, V. A. (1988). Equilibrium dialysis study of binding of hexammine cobalt(III) to DNA. Biopolymers 27, 1045–1051. Robinson, H., Gao, Y. G., Sanishvili, R., Joachimiak, A., and Wang, A. H. J. (2000). Hexahydrated magnesium ions bind in the deep major groove and at the outer mouth of A-form nucleic acid duplexes. Nucleic Acids Res. 28, 1760–1766. Stuhrmann, H. B. (1981). Anomalous small-angle scattering. Q. Rev. Biophys. 14, 433–462. Svergun, D. I., and Koch, M. H. J. (2003). Small-angle scattering studies of biological macromolecules in solution. Rep. Prog. Phys. 66, 1735–1782. Tate, M. W., Eikenberry, E. F., Barna, S. L., Wall, M. E., Lowrance, J. L., and Gruner, S. M. (1995). A large-format high-resolution area X-ray-detector based on a fiberoptically bonded charge-coupled-device (CCD). J. Appl. Crystallogr. 28, 196–205. Wensel, T. G., Meares, C. F., Vlachy, V., and Matthew, J. B. (1986). Distribution of ions around DNA, probed by energy-transfer. Proc. Natl Acad. Sci. USA 83, 3267–3271.
C H A P T E R
T W E N T Y
Simulations of RNA Interactions with Monovalent Ions Alan A. Chen,*,† Marcelo Marucho,‡,† Nathan A. Baker,*,‡,† and Rohit V. Pappu*,†,§,} Contents 1. Introduction 2. Finite Size Artifacts in All-Atom Simulations of Ion–Nucleic Acid Interactions 3. On the Use of ‘‘Neutralizing Counterions Only’’ in All-Atom Simulations 4. Results from Simulations of Canonical A-Form RNA and B-Form DNA Helices 5. Comparison with Predictions of the Nonlinear Poisson–Boltzmann Equation Acknowledgment References
412 414 421 421 426 430 430
Abstract RNA folding and binding reactions are mediated by interactions with ions that make up the surrounding aqueous electrolytic milieu. Although Mg2þ ions are often implicated as being crucial for RNA folding, it is known that folding is feasible in high concentrations of monovalent alkali-halide salts. Experiments have yielded important information regarding the salt dependence of RNA stability. Recent work has shown that molecular simulations based on explicit representations of solvent molecules and monovalent ions can also provide useful insights regarding the ionic atmospheres around model RNA systems. These insights can help rationalize intriguing observations regarding the dependence of RNA stability on cation type providing one pays attention to important considerations that go into the proper design of molecular * Computational and Molecular Biophysics Program, Washington University in St. Louis, St. Louis, Missouri, USA Center for Computational Biology, Washington University in St. Louis, St. Louis, Missouri, USA { Department of Biochemistry and Molecular Biophysics, Washington University in St. Louis, St. Louis, Missouri, USA } Department of Biomedical Engineering, Washington University in St. Louis, St. Louis, Missouri, USA } Washington University in St. Louis, St. Louis, Missouri, USA {
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69020-0
#
2009 Elsevier Inc. All rights reserved.
411
412
Alan A. Chen et al.
simulations. These issues are discussed in detail and the methods are applied to an A-form RNA and B-form DNA sequence. The results of these simulations are compared to previous work on a kissing-loop system with analogous sequence. In particular, ionic atmospheres obtained from molecular simulations are compared to those obtained using the nonlinear Poisson Boltzmann model for continuum electrostatics for these three different nucleic acid systems. The comparisons indicate reasonable agreement in terms of coarse-grained observables such as the numbers of counterions accumulated around the solutes. However, details of the ionic atmospheres, captured in terms of spatial free energy density profiles, are quite different between the two approaches. These comparisons suggest the need for improvements in continuum models to capture sequence-specific effects, ion–ion correlation, and the effects of partial dehydration of ions.
1. Introduction The folding and binding reactions of RNA molecules are mediated by mobile counterions in the surrounding milieu (Draper, 2004; Woodson, 2008). These reactions often require millimolar amounts of divalent Mg2þ ions (Laing et al., 1994; Pan et al., 1999; Rangan and Woodson, 2003; Takamoto et al., 2004). Monovalent ions such as Naþ/Kþ also play an important role in RNA folding although the required concentrations of monovalent ions are considerably larger than Mg2þ ions (Das et al., 2003; Draper, 2004; Perez-Salas et al., 2004; Takamoto et al., 2004). Several RNA systems are capable of folding/binding in the absence of Mg2þ and such systems have been the focus of recent investigations (Lambert et al., 2009). These studies have identified intriguing dependencies of RNA stability on monovalent counterion type that have merited closer scrutiny through computational methods. An elegant example is the Tar– Tar* kissing-loop system studied by Draper and coworkers (Lambert et al., 2009). Tar–Tar* is composed of a pair of 16 nucleotide hairpins, each containing a 5 bp stem and a six nucleotide loop (Chang and Tinoco, 1997). The two loops have complementary sequences that stabilize the complex via intermolecular Watson–Crick base pairs. This results in substantial deformation of the RNA backbone, with phosphorus atoms from ˚ . This distance of opposing strands (C6 and U22) approaching within 5.25 A approach is small in comparison to the smallest possible interstrand separa˚ between phosphorus atoms in canonical A-form duplexes. tion of 10.4 A Lambert et al. measured the stability of the Tar–Tar* complex in a series of alkali-chloride salts using isothermal titration and thermal melt experiments. The stability of the Tar–Tar* complex is inversely proportional to the crystallographic radius of the monovalent counterion. Conversely, folding of the isolated Tar* hairpin shows only marginal dependence on
413
Simulations of RNA Interactions with Monovalent Ions
counterion type. Chen et al. (2009) used all-atom molecular dynamics simulations with explicit representations of ions and water molecules to quantify differences in the ionic atmospheres of Naþ, Kþ, and Csþ, respectively around the Tar–Tar* complex. These simulations yielded insights that appear to rationalize the experimentally observed trends regarding the dependence of Tar–Tar* stability on the identity of monovalent counterions. The work of Chen et al. also yielded lessons for the optimal way to design molecular dynamics simulations that are focused on quantitative studies of ionic atmospheres around nucleic acids. These lessons encompass a range of issues including the selection of forcefield parameters for nucleic acids and ions, the choice of system size (numbers of water molecules and ions) for a target salt concentration, the specification of simulation length, and the choice of suitable analysis methods required to obtain detailed ionic atmosphere information. This chapter provides a discussion of these issues and the considerations that must be taken into account when setting up all atom molecular dynamics simulations to study ion interactions with nucleic acids. Following our detailed discussion of these considerations, we present representative results for A-form RNA and B-form DNA sequences and contrast these with results obtained by Chen et al. for the Tar–Tar* system. The sequences for the canonical RNA/DNA structures are analogous to that of Tar–Tar* (Fig. 20.1). This allows us to query the effects of threedimensional structural changes on ionic atmospheres. The results demonstrate the capabilities of modern simulation methods and the complementary role such simulations play to experimental approaches that interrogate the interactions of ions with nucleic acids.
3⬘ 5⬘ G C Tar* G C A U G C A U C G U A G C G C G C A U C G C G G C A U Tar G C 5⬘ 3⬘ Tar–Tar
3⬘ C G A C A G A C C C U C C G A G 5⬘
5⬘ G C U G U C U G G G A G G C U C 3⬘
A-form RNA
3⬘ C G A C A G A C C C T C C G A G 5⬘
5⬘ G C T G T C T G G G A G G C T C 3⬘
B-form DNA
Figure 20.1 Details of the sequences for the Tar–Tar*, A-form RNA, and B-form DNA systems discussed in this work.
414
Alan A. Chen et al.
Our analyses of what is needed to obtain meaningful results from all atom simulations with explicit representations of ions and solvent molecules also reveal the limitations of atomistic approaches—specifically, the difficulty of scaling up these simulations to study conformational transitions, binding– unbinding reactions, and the details of salt dependencies of such reactions. Therefore, in Section 5, we provide a comparative characterization of coarser-grained continuum solvation models based mainly on the nonlinear Poisson–Boltzmann equation. This comparison is done with a view toward assessing the suitability and accuracy of fast, novel, albeit empirical implicit solvent models (Vitalis and Pappu, 2009) for future use in modeling ionmediated processes for RNA molecules. Based on the encouraging results from our comparative studies, we propose a general approach that incorporates implicit solvent models in calculations with RNA molecules.
2. Finite Size Artifacts in All-Atom Simulations of Ion–Nucleic Acid Interactions Molecular dynamics simulations attempt to represent a macroscopic system using a combination of periodic boundary conditions and small, fixed numbers of solvent molecules in the central simulation cell. In these simulations of solvent plus a single macromolecular solute, the sizes of simulation cells determine the effective solute concentration. Hence, the size of the simulation cell needs to be chosen to ensure that interactions between images of macromolecules are minimal and the simulation mimics a dilute macromolecular solution. This is usually the first consideration in setting up the simulation. For systems with dipolar macromolecular solutes and solvents (such as water), the size of the central simulation cell determines whether the bath contains sufficient numbers of solvent molecules to represent the fraction of solvent molecules that are perturbed by the presence of the macromolecule and the fraction of molecules that display bulk-like behavior. Typical molecular dynamics simulations are run for hundreds of nanoseconds. The larger the simulation cell, the larger the number of force evaluations per time step, and hence the longer the ‘‘wall clock’’ time needed to complete the simulations. Given the computational cost of energy and force evaluations in molecular simulations, there is considerable motivation to use the smallest possible simulation cells to make simulations practically feasible. When the solvent is dipolar, for example water, the perturbation of solvent structure by a low-charge density macromolecule is short-range in nature and does not exceed two to three layers of solvent molecules (Smolin and Winter, 2004). Therefore, increasing box sizes beyond minimal requirements is not necessary, and in fact additional solvent molecules will not have a dominant effect on the behavior of the macromolecule.
Simulations of RNA Interactions with Monovalent Ions
415
The preceding discussion focused on considerations that go into choosing parameters for simulation cell sizes and numbers of solvent molecules when the systems comprise of low-charge density solutes and dipolar solvents. This discussion also applies, with appropriate modifications, to systems with solutes that have small numbers of charged moieties when compared to the number of dipolar entities. However, these considerations change when considering the setup for simulating systems that include highly charged solutes such as RNA or DNA. Such systems are characterized by significant intramolecular and solvent-mediated long-range electrostatic interactions and the solvent must include mono- and multivalent counter- and coions. Most methods for handling long-range electrostatic interactions require electroneutrality within the central simulation cell; this requirement sets the minimal number of counterions to be included in the simulation. For macromolecules such as RNA and DNA, there is a significant accumulation of counterions and a concomitant depletion of coions within a local volume around the macroion (Anderson and Record, 1995). Proper recovery of the partitioning between local and bulk ions can only occur if sufficiently large numbers of ions are present in the simulation. Recent work focused on the simulation of ionic atmospheres of monovalent alkali halide salts around a model 32-nucleotide kissing-loop system known as Tar–Tar* (Chen et al., 2009). The target salt concentrations for these simulations were 800 mm and the goal was to provide an explanation for the experimentally observed cation-specific stability of the Tar–Tar* system. Systematic investigations were performed to identify the minimal setup required to achieve satisfactory, if less than perfect partitioning between local and bulk ionic environments. Such a setup consisted of ˚ cubic boxes encompassing 55,000 water molecules, 800 ion 120 A pairs, and 30 neutralizing counterions. Similar parameters (103 excess ion pairs and 6 104 water molecules) were found to be necessary in comparative simulations of 16 bp RNA or DNA duplexes. The results of these simulations will be described in the following sections. The sizes of these simulation cells are substantially larger than the molecular dimensions of the RNA/DNA solutes, which are approximate cylinders with radii of ˚ and long axes of 50 A ˚ . We found that smaller simulation cells with 15 A 2 O(10 ) excess ion pairs result in incorrect partitioning between local and bulk ionic environments. Such inaccuracies can be detected using methods that are described below. The most easily diagnosed finite size artifact is the presence of a concentration mismatch between the counter- and coion to water ratio (molality) and the actual electrolyte concentration at the ‘‘edge’’ of the simulation box. This is detected by calculating the macromolecule-counterion and macromolecule–coion radial distribution functions (RDFs) from an equilibrated trajectory. Examples of RDFs based on previous work on the
416
Alan A. Chen et al.
Tar–Tar* system (Chen et al., 2009) are shown in Fig. 20.2. Panels A and B of Fig. 20.2 show the RNA phosphate-counter-/coion RDFs using simulation data based on 80 and 120 A˚ simulation cells, each containing 236 and ˚ box, both RDFs decay to 792 excess ion pairs, respectively. In the 120 A unity (within 1%) as the edge of the central simulation cell is approached. This indicates the presence of a clear local-bulk partition and the presence of a bulk-like reservoir with which ions can be exchanged. The phosphatecounterion and phosphate-coion RDFs differ by a few percent in this regime, but these differences are due to the fact that there are always 30 additional neutralizing counterions present to ensure net electroneutrality of the Tar–Tar* system. Conversely, at the boundary of the 80 A˚ box, the coion density exceeds unity by 15%. This artifact is the result of the
A 5 4
P-Na P-Cl
g(r)
3 2 1 0
0
1
2 r (nm)
3
4
B 5
g(r)
4
P-Na P-Cl
3 2 1 0
0
1
2
3 r (nm)
4
5
6
Figure 20.2 Illustration of how finite size artifacts lead to concentration mismatches in the bulk: (A) RNA phosphate-counterion radial distribution functions for the Tar–Tar* ˚ box of 800 mm NaCl. The anion concentration is 15% different complex in an 80 A than the cation concentration at large separations. (B) RNA phosphate-counterion ˚ box of 800 mm radial distribution functions for the Tar–Tar* complex in a 120-A NaCl. The anion and cation concentrations agree with each other to within 1% at large radii. Taken from Chen et al. (2009).
417
Simulations of RNA Interactions with Monovalent Ions
exclusion of coions in the immediate vicinity of the macroion. The even split between counterion accumulation and coion exclusion observed in these simulations is consistent with equilibrium dialysis experiments of polyelectrolytes at high concentrations ( 1 M ) of excess monovalent salt (Strauss et al., 1967). In the smaller simulation cell, there is a spurious boundary accumulation of the coions that have been depleted from the vicinity of the RNA. This spurious boundary accumulation arises because the small simulation cell cannot support a proper bulk environment and is unable to absorb the accumulated coions into a bulk-like reservoir. The anomalous repartitioning described above can also be quantified in terms of preferential interaction coefficients, Gþ and G, respectively. Details of how these coefficients are calculated have been published (Chen et al., 2009). At high concentrations (such as 800 mm), the split between Gþ and G should be roughly even (Strauss et al., 1967). Conversely, for lower concentrations macroscopic thermodynamics requires that Gþ > G (with the discrepancy increasing with decreasing concentration) and for higher concentrations Gþ < G; that is, excluded volume considerations lead to the dominance of coion depletion over counterion accumulation for salt concentrations that are greater than 1 m. Table 20.1 shows how Gþ and G vary with the size of the simulation cell. Contrary to the expectations from macroscopic thermodynamics, Table 20.1 demonstrates that that the smaller box exhibits a smaller degree of coion exclusion and a concomitant higher degree of counterion inclusion, and this is entirely a consequence of the artificially truncated ionic atmosphere of the smaller simulation box. In the quantification of ion accumulation around a macroion, it is assumed that the macroion is in contact with a thermodynamic bath of bulk electrolyte from which individual ions can be added or subtracted without significantly altering the properties of the bath itself. One way to assay the existence of two independent ion populations within the simulation is to quantify net-charge fluctuations as a function of time within spherical shells around the macroion. Ideally, the shells farthest from the macroion should correspond to the bulk population and should not be strongly coupled to ion fluctuations within the ionic atmosphere of the Table 20.1 Preferential interaction coefficients calculated as a function of box size for the Tar–Tar* systems and a target concentration of 800 mm of NaCl Box size (length of each side ˚) of the cubic box) (A
No. of excess ion pairs
Calculated Gþ
Calculated G
80 120
236 792
19.6 1.0 15.0 2.3
9.3 1.5 14.8 2.7
418
Alan A. Chen et al.
macroion. Panels A and B of Fig. 20.3 show the cumulative net charge contained within spherical shells of varying radii from the macromolecule, averaged over 100 ps intervals. This measure is calculated as the average RNA-counterion cumulative distribution function (CDF) minus the RNA-coion CDF. Since the macromolecule bears a net charge of 30e,
A 40
No. of cations−no. anions
400 mm NaCl, 80Å box 30
20
10.0–10.1 ns 10.1–10.2 ns 10.2–10.3 ns 10.3–10.4 ns 10.4–10.5 ns 10.5–10.6 ns 10.6–10.7 ns 10.7–10.8 ns 10.8–10.9 ns 10.9–11.0 ns
10
0
0
1
2 r (nm)
B
3
4
40
No. of cations−no. anions
800 mm NaCl, 120Å box 30
20
10.0–10.1 ns 10.1–10.2 ns 10.2–10.3 ns 10.3–10.4 ns 10.4–10.5 ns 10.5–10.6 ns 10.6–10.7 ns 10.7–10.8 ns 10.8–10.9 ps 10.9–11.0 ns
10
0
0
1
2
3 r (nm)
4
5
6
Figure 20.3 Box size dependencies of the accumulated net charge around the Tar– Tar* system. The quantity on the ordinate is plotted against shells of increasing radii and different families of curves are shown for different time intervals within the simulations.
Simulations of RNA Interactions with Monovalent Ions
419
the net-charge CDF for the ions around the RNA should reach þ30e at the length scale at which the charge is completely compensated by the ionic atmosphere. It can be seen that in the 80 A˚ box (panel A of Fig. 20.3), the net-charge CDF only reaches þ30e at the very edge of the box, with no visible plateau region. The lack of a plateau in the net-charge CDF is problematic because it indicates that even the solvent fraction farthest from the macroion is still technically part of the ‘‘local atmosphere’’ and not a truly independent population. Furthermore, the charge fluctuations appear to be coupled on all length scales, which is evident by the similarity of the net-charge CDFs both close to and far away from the macroion. In contrast, panel B of Fig. 20.3 shows that the situation is quite different for ˚ box. First, a clear plateau region emerges from 40 to 60 A ˚ in the the 120 A ˚ 120 A box. In addition, this region experiences large net-charge fluctuations that are completely uncoupled from what is occurring very close to the macroion, indicating that there exists a bulk-like population of electrolyte that lies beyond the local ionic atmosphere around the macroion. Finite size artifacts also impact the kinetics of ion relaxation around the macroion. Converged simulations of ion accumulation around a macroion can only be determined if the simulation is much longer than the intrinsic timescale of ion exchange between the local and bulk partitions. In prior work (Chen and Pappu, 2007b), it was shown that the reorganization time of bulk electrolytes in the absence of a macromolecule was approximately 100 ps. To assess how this timescale was altered by the presence of the RNA, we employed a measure that is similar to that used by Ponomarev et al. (2004) and earlier by Cheatham and Young (2000). They quantified convergence rates by comparing the similarity between the atom-resolved distributions of nucleic acid-counterion contacts between the two identical strands of the Drew–Dickerson dodecamer. In the simulations that employ only neutralizing Naþ counterions (i.e., no excess salt), Ponomarev et al. found that the ions exhibited very slow relaxation, characterized by a Pearson correlation coefficient (PCC) of 0.69 at 10 ns and 0.92 at 60 ns, leading to the extrapolation that 100 ns would be required for full convergence (PCC of 0.97) of the ionic atmosphere (Ponomarev et al., 2004). The Tar–Tar* sequence as well as the A-form RNA and B-form DNA equivalents of Tar–Tar* (see Fig. 20.3) do not contain palindromes. However, we have data from multiple independent simulations and the analysis of Ponomarev et al. can be repeated using data from pairs of simulations. The equivalent test is to compare the residue-resolved ion-contact distributions between replicate trajectories. Such comparisons are a stringent test if each simulation was separately initialized with randomized ion starting positions and velocities. The PCCs for the A-form RNA simulations are shown in Fig. 20.4. Error bars denote the standard errors across the six possible pairwise comparisons across four independent trajectories, each of
420
Alan A. Chen et al.
20 ns length. Within a 1 ns interval, the two trajectories have a PCC of 0.67, and this reaches 0.94 by 15 ns, nearly an order of magnitude faster than observed in the Drew–Dickerson dodecamer simulations of Ponomarev et al. Therefore, the use of only neutralizing counterions results in significantly retarded ion reorganization kinetics compared to simulations employing large numbers of excess ions. This observation of slower relaxation kinetics suggests that spurious ion–ion correlations are introduced into systems with small numbers of ions. This result is explained as follows: Counterions are predisposed to be in the vicinity of the macroion. If only neutralizing counterions are present, then there is no reservoir with which ions can be exchanged. In the absence of such a reservoir, ions within the local environment around the macroion have to exchange with each other rather than the bulk. This local exchange is inevitably slow because these ions are also restricted to be in the vicinity of the macroion. The preceding analysis shows that there is a distinct lower bound for the size of the simulation cell. Below this bound, finite size artifacts heavily bias the thermodynamics and kinetics of ion accumulation. Therefore, it seems unadvisable to design simulations using neutralizing counterions alone, especially if the objective is to understand ion-mediated conformational transitions or to understand how ions interact with a macroion. The minimal number of excess ions needed to simulate 800 mm excess monovalent salt was nearly identical ( 800 pairs) for all of the systems studied (32 nt A-RNA, B-DNA, and the Tar–Tar* RNA kissing-loop complex), indicating a simple dependence on net macroion charge. For other macromolecular solutes or other concentrations of excess salt, a box-size
Pearson correlation coefficient
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0
2
4
6
8 Time (ns)
10
12
14
Figure 20.4 Person correlation coefficients quantifying the similarities in ion occupancies around equivalent atoms of the A-form RNA system between different pairs of simulations.
Simulations of RNA Interactions with Monovalent Ions
421
‘‘titration’’ needs to be performed, with the macroion–ion RDFs providing a quick test to assess whether a reasonably sized ion bath has been used. Finally, a variant of the test suggested by Ponomarev et al. (2004), performed using data from independent trajectories with randomized ion starting positions, provides a useful measure of how long a trajectory must be run in order to obtain converged statistics.
3. On the Use of ‘‘Neutralizing Counterions Only’’ in All-Atom Simulations There are several recurring justifications regarding the utility of including only neutralizing counterions in all-atom simulations. The primary justification is one of practicality; that is, to maximize trajectory length by minimizing the volume of bulk solvent included in the central simulation cell. However, the above analysis shows that, in addition to suffering from finite size artifacts, counterion relaxation times are at least an order of magnitude slower when compared to simulations with sufficiently large boxes and clear local-bulk ion partitioning. The second, less-publicized justification is in response to the spurious salt cluster and lattice formation which has been observed to occur in simulations utilizing the AMBER force-field at concentrations well below the solubility limit (Auffinger et al., 2007; Chen and Pappu, 2007a). In these cases, the omission of the coion is used to mask cluster and lattice formation. However, these artifacts have since been shown to be the result of incorrect ion parameter adaptations, with two independently proposed fixes that eliminate this problem (Chen and Pappu, 2007a; Joung and Cheatham, 2008). Lastly, there is an analytical solution to the Poisson–Boltzmann equation for the polyelectrolyte cell model, which is only possible for the case of no excess salt. Simulations including only neutralizing counterions are therefore only appropriate in cases where a direct comparison to such theories is desirable.
4. Results from Simulations of Canonical A-Form RNA and B-Form DNA Helices Recent simulation work (Chen et al., 2009) focused on comparative quantification of the accumulation of different types of group I monovalent cations around the Tar–Tar* kissing loop. The analysis identified three salient features regarding the distributions of accumulated ions around Tar–Tar*: (1) The ionic atmosphere can be quantified in terms of free energy layers around the RNA using the methods described in detail below. The most favorable free energy levels (DG 1.5 kcal/mol) are occupied
422
Alan A. Chen et al.
by partially dehydrated ions. The first hydration shell statistics of these ions are markedly different from those in the bulk and for smaller cations, the extent of dehydration is greater because favorable water–ion interactions are replaced by favorable interactions with the Tar–Tar* complex. The guanine tract in the Tar hairpin facilitates this partial dehydration. (2) There is asymmetry in the extent of cation accumulation around each hairpin. However, the degree of the observed asymmetry is smallest for Naþ and largest for Csþ because the smaller cations can form a belt of positive-charge density across the loop–loop interface. (3) Finally, in the bulk-like free energy intervals (DG > 0.5 kcal/mol), there remain differences in cation accumulation that are explained in terms of the generic differences between the corresponding 1:1 electrolytes (NaCl, KCl, and CsCl). In particular, at high concentrations (ca. 800 mm), the degree of nonideality varies with cation size and is more prominent for the smaller cations. These collective observations were used to rationalize experimental measurements that demonstrate cation specificity for the free energy of complex formation (Lambert et al., 2009). Recent work (Semichaevsky et al., 2009) showed that asymmetry in counterion accumulation across the loop–loop interface is a characteristic that is shared among a wide variety of kissing-loop complexes. They hypothesized that this asymmetry in counterion accumulation facilitates rapid dissociation kinetics required in the function of naturally occurring antisense RNA feedback loops. This hypothesis needs further testing because of the cation specificity observed in ion asymmetric ion accumulation. The results summarized above were obtained through the analysis of spatial free energy density profiles using the following methods. Spatial distribution functions for ion occupancies around the macroion are computed on a cubic grid. These distribution functions can be converted to spatial free energy density profiles using the relation: rðx; y; zÞ DGðx; y; zÞ ¼ RT ln ð20:1Þ ro ˚ 3 cell, r0 is the bulk ion density here, r(x, y, z) is the density within each 1 A as measured in the periphery of the simulation cell, R ¼ 1.987 10 3 kcal/mol-K, and T is the simulation temperature. The resulting free energy contours provide an assessment of the spatial variations of the strengths of ion interactions with the macroion. The cumulative ion numbers within each contour cutoff are calculated by summation of r(x, y, z) over all the cells that result in a DG value above a specified value. The relationship shown in Eq. (20.1) is based on the fact that, in an equilibrium simulation, the distributions of ions in local and bulk partitions obey Boltzmann statistics. The ratio within the square brackets of Eq. (20.1) quantifies the relative probability—vis-a`-vis bulk—of realizing a specific ion density at a
Simulations of RNA Interactions with Monovalent Ions
423
particular location around the RNA. If the ratio is unity, then the logarithm of the ratio is zero, thus setting the bulk milieu as the reference state. Deviations from unity are either due to ion depletion or accumulation and the quantity shown in Eq. (20.1) quantifies the free energy change associated with depleting or accumulating ions vis-a`-vis the bulk. We wished to know if the features of ionic atmosphere around the Tar– Tar* were generalizable to typical nucleic acid architectures such as A- or B-form helices. We performed simulations of both an A-form RNA helix and a B-form DNA helix, each 32 nt/16 bp long, and containing nearly identical sequences as found in the Tar–Tar* kissing-loop complex (Fig. 20.1). The only modification to the sequences is that the loop region has been altered to form canonical duplexes instead of two separate hairpins; for the B-DNA duplex, the uracil residues have been substituted with thymine residues and ribose rings of RNA were replaced with deoxyribose rings. These systems were simulated under identical conditions as used in the Tar–Tar* kissing-loop study, that is, 800 mm excess NaCl and KCl, with 792 excess ion pairs in 120 A˚ boxes. The nucleic acids were modeled using the AMBER-99 forcefield (Cornell et al., 1995) employing the GROMACS implementation of Sorin and Pande (2005). The rigid threesite TIP3P model ( Jorgensen et al., 1983) was used to simulate water ˚ qvist (1990) molecules. Ions were modeled using the parameters of A according to the approach proposed by Chen and Pappu (2007a). Figure 20.5 shows the spatial free energy density profiles for the A- and B-form systems, respectively. Counterion densities were visualized by creating free-energy isocontours at 0.5, 1.0, 1.5, and 2.0 kcal/mol intervals, identical to the procedure described in previous work (Chen et al., 2009). The integrated numbers for the cumulative number of counterions accumulated within a free energy interval is calculated as a function of free energy cutoff and plotted in Fig. 20.6. It is known that the change in the melting temperature resulting from a change in the monovalent salt concentration is higher for A-RNA duplexes than B-DNA duplexes (Nakano et al., 1999), whereas the salt dependence of the Tar–Tar* complex is even higher still (Lambert et al., 2009). Therefore, we expected a systematic trend whereby Tar–Tar* accumulates more counterions than A-form RNA, which in turn accumulates more counterions than B-DNA—across all free energy cutoffs. Additionally, oligomers of B-DNA are known to exhibit identical stabilities in equimolar buffers of NaCl and KCl from 0.5 to 600 mM excess salt (Nakano et al., 1999). Since counterion specificity in Tar—Tar* is found to result from differences in the number of highly favorable (DG < 2.0 kcal/mol) accumulated counterions, it is expected that the B-form DNA will exhibit minimal differences between Naþ and Kþ in this free energy regime. We find that both of the canonical helices exhibit less accumulation within all free energy contours when compared to the RNA kissing-loop
424
Alan A. Chen et al.
Na+ around A-form RNA
ΔG
−2.0
−1.5
Na+ around B-form DNA
−1.0
−0.5
kcal/mol
Figure 20.5 Spatial free energy density profiles. Spatial free energy density profiles for counterion accumulation around the A-form RNA (left) and B-form DNA (right). Freeenergy isocontours (surfaces of zero thickness at specific free-energy values) are colored according to the key shown at the bottom of the figure. The volume encompassed by each isocontour contains regions of more negative free energy density than prescribed by the isocontour itself.
system, in the order Tar–Tar* > A-RNA > B-DNA (Fig. 20.6). Although all three structures systematically accumulate more Naþ counterions than Kþ counterions at any free energy cutoff, for B-DNA there is negligible counterion accumulation in the most favorable range (DG < 2.0 kcal/mol), which is the free energy interval that we expect to be most important for specificity in counterion accumulation. Differences in counterion accumulation that go beyond the lowest and intermediate free energy contours (DG > 1.0 kcal/mol) reflect the nonidealities in the bulk electrolytes at high concentrations. The region of highest accumulated counterion density in the A-form RNA lies within the major groove, along the cylindrical axis of the helix. In contrast, the B-form DNA exhibits more diffuse layers of counterion accumulation, which are mostly situated along the major and minor grooves. These conclusions are drawn based on the comparative spatial free energy density profiles shown in Fig. 20.5 where panel A shows the profile for Naþ
Simulations of RNA Interactions with Monovalent Ions
Cumulative counterion numbers
12
425
A-RNA - Na+ A-RNA - K+
10
B-DNA - K+ B-DNA - Na+ Kissing-loops Na+
8
Kissing-loops K+
6 4 2
−2
−1.5 −1 ΔG-free energy interval (kcal/mol)
−0.5
Figure 20.6 Comparative cumulative distribution functions to assess the amount of counterion accumulation around the Tar–Tar* complex, A-form RNA, and B-form DNA, respectively. Data were obtained from molecular dynamics simulations with explicit representations of water molecules and ions. The ordinate quantifies the number of counterions accumulated around the macroions within contours that have DG less than or equal to the value on the abscissa.
accumulation around the A-form RNA and panel B the corresponding profile for accumulation around B-form DNA. This reflects the inherent difference in helix architecture, as the A-form helix possesses an accessible ‘‘open core’’ compared to the completely occluded core of the canonical Bform helix. Both A-RNA and B-DNA exhibit prominent ‘‘spines’’ of moderate to weakly favorable ( 1.0 DG 0.5 kcal/mol) counterion accumulation threaded along the helical grooves. This is split roughly equally across major and minor grooves in B-DNA but is predominantly in the major groove in A-RNA. The tracking of counterion density with nucleotide sequence is much more pronounced in A-form RNA than in the B-form DNA. The region of most favorable counterion accumulation lies directly adjacent to the tract of tandem purines, including the three tandem guanines that were found to be key in the accumulation of ions at the loop–loop interface of the Tar–Tar* complex. This favorable counterion accumulation lies along the cylindrical axis of the A-form RNA and ends abruptly at the center of the structure, which is where the purine tract ends. It is at this point that the major groove develops a ‘‘spine’’ of moderate to weakly favorable ( 1.0 DG 0.5 kcal/mol) accumulated counterions, reminiscent of the diffuse ‘‘belt’’ found straddling the loop–loop interface in our previous characterization of the Tar–Tar* kissing-loop system. The similarity
426
Alan A. Chen et al.
between the kissing-loop complex and the A-RNA duplex is also evident in the quantification of number of accumulated counterions, in that the kissing-loop complex and A-RNA all converge to roughly the same number at the weakest free energy cutoff examined (DG 0.5 kcal/mol) whereas the B-DNA does not. Therefore, it would appear that the architecture of the A-form RNA is inherently predisposed for particular patterns of counterion accumulation, which are amplified in the context of a kissingloop complex. In recent work, Savelyev and Papoian examined the relative affinities of 15 Naþ and 15 Kþ ions to a 16 bp DNA helix, in which a systematic preference for accumulation of Naþ over Kþ is also observed (Savelyev and Papoian, 2008). The authors concluded that the accumulation of Naþ is favored due to the trade-off between the more favorable electrostatic potential accessible to smaller ions, which outweighs the greater desolvation cost for partial dehydration. Our results for the Tar–Tar* system and the two systems studied here agree with the trends identified by Savelyev and Papoian. However, the proposed mechanisms for cation specificity are different: whereas the ionic potential used in our studies result in K–Cl pairing constants that are systematically less favorable than Na–Cl pairing constants (Chen and Pappu, 2007b), the interionic potential employed by Savelyev and Papoian predict the reverse situation, resulting in a coion competition effect which sequesters Kþ ions from the vicinity of the macroion. These discrepancies in interpretation call for systematic comparative studies, and these are ongoing.
5. Comparison with Predictions of the Nonlinear Poisson–Boltzmann Equation Atomistic simulations based on explicit solvent models are informative and yield important insights regarding the origins for specificity in counterion accumulation. Spatial free energy density profiles are useful devices for quantifying the intricate details of the liquid-like organization of ions around the macroion. Integrals of these profiles yield numeric values for the numbers of ions in different layers, and these quantities are useful for extracting a range of colligative properties. However, as discussed in the section on simulation setups, it is clear that large simulation cells are needed in order to obtain thermodynamically accurate quantification of the ionic atmospheres around RNA and DNA molecules. To be truly effective, we need to be able to extend such calculations to a range of RNA systems, many of which are considerably larger than 32 nucleotides. Furthermore, to generate insights regarding driving forces for conformational changes and ligand binding, it is important to be able to simulate the salt dependence of
Simulations of RNA Interactions with Monovalent Ions
427
ion binding and the coupling between ion-interactions and conformational fluctuations. The ability to simulate such systems at a variety of salt concentrations would help elucidate the nature of the strong coupling between ion-macromolecule interactions and the macromolecular conformational equilibria. It is difficult to address these issues through simplistic scaling up of simulations based on explicit representations of solvent molecules. Continuum solvent models possess an intrinsic advantage because the effects of solvation can be captured implicitly, thereby allowing for the savings in computational expense. However, the issue of accuracy vis-a`-vis explicit, all-atom simulations remains unresolved. Accordingly, we assess the utility of the most commonly used continuum solvation model by comparing spatial free energy density profiles and their integrals to those obtained from simulations with explicit solvent for the Tar–Tar*, A-form RNA, and B-form DNA systems discussed above. Analysis of ion atmospheres around highly charged macromolecules has traditionally been performed using numerical solutions to the nonlinear Poisson–Boltzmann (P–B) equation (Anderson and Record, 1980; Bai et al., 2007; Baker, 2004), in which the macromolecule is approximated as a collection of point charges embedded in a low dielectric cavity surrounded by a high-dielectric solvent. This approach utilizes the precise three-dimensional structure of the macromolecule (albeit in a static sense). We would not expect such a framework to capture subtleties, which are dependent on the partial dehydration of ions. The spatial free energy density plots and the numbers of accumulated counterions as a function of free energy cutoff obtained using the APBS implementation (Baker et al., 2001; Dolinsky et al., 2004) of the nonlinear Poisson–Boltzmann equation are presented in Figs. 20.7 and 20.8, respectively. Calculations utilized a bulk salt concentration of 800 mM salt, with ˚, the hydrated radii of Naþ, Kþ, and Cl set to 3.33, 4.93, and 4.42 A respectively. The solute and solvent dielectric constants were set to 20 and 78.54, respectively. The solute–solvent dielectric interface was calculated ˚. using a molecular surface definition using a solvent probe radius of 1.4 A The nonlinear Poisson–Boltzmann equation was solved on a grid of 257 257 257 grid points, corresponding to a grid spacing of 0.469 A˚. The most representative single structure for each system was extracted from the molecular-dynamics trajectories through cluster analysis using the GROMACS g_cluster utility, which identified the structural snapshot most equidistant from all other snapshots in terms of root mean square deviation (RMSD). At first glance, the numbers of accumulated counterions in Fig. 20.8 are rather similar to the values from the all-atom simulations (Fig. 20.6). Both the all-atom molecular dynamics and the P–B calculations agree that B-DNA accumulates substantially fewer highly favorable (DG 1.5 kcal/mol) counterions than A-RNA. However, the P–B calculations are unable to capture the greater propensity for
428
Alan A. Chen et al.
Na+ around A-form RNA
Na+ around B-form DNA
Figure 20.7 Spatial free energy density profiles obtained using data from P–B calcula˚ 3 were used in these calculations. The color-coding for the free tions. Grid sizes of 1 A energy levels is based on the key shown at the bottom of Fig. 20.5.
Cumulative number of counterions
12 10 8
A-RNA-Na+ A-RNA-K+ B-DNA-K+ B-DNA-Na+ Kissing-loops- Na+ Kissing-loops K+
6 4 2
−2
−1.5 −1 ΔG-free energy interval (kcal/mol)
−0.5
Figure 20.8 Comparative cumulative distribution functions to assess the amount of counterion accumulation around the Tar–Tar* complex, A-form RNA, and B-form DNA, respectively. Data were obtained from P–B calculations and are to be compared with those shown in Fig. 20.6.
Simulations of RNA Interactions with Monovalent Ions
429
counterion accumulation around the Tar–Tar* kissing loop as compared to the A-form RNA. While the trends for the integrals of spatial density profiles are roughly similar to each other (cf. Figs. 20.6 and 20.8), close examination of the spatial free energy density profiles reveals that the trends in counterion accumulation as predicted by the P–B calculations do not arise from the same regions as in the molecular dynamics simulations. Specifically, the interior of both the Tar–Tar* kissing loops as well as that of the A-form RNA are completely occluded by the Stern layer, which was set to the hydrated radii of the ions. For all three structures, only a narrow band within the helical grooves is accessible to the ions in the P–B calculations. For B-form DNA, both the major and minor grooves are accessible to weak to moderately accumulated counterions ( 1.0 DG 0.5 kcal/mol) while only the deeper major groove of the A-form RNA is accessible to the favorable counterions at 1.5 kcal/mol. The greater accumulation around the A-form RNA in the P–B calculation is therefore a simple consequence of the major groove being deeper than that of B-DNA, which agrees with X-ray scattering measurements and P–B calculations (Pabit et al., 2009). Similarly, the accumulation differences between Kþ versus Naþ are captured solely on account of the larger Stern layer for Kþ, resulting in shallower grooves. It is notable that this aspect of counterion accumulation, although incomplete, is still in agreement with the ‘‘spines’’ and ‘‘belts’’ observed in the molecular dynamics simulations, which we have previously determined to consist of fully hydrated ions. For the Tar–Tar* kissing loops, the P–B calculations are unable to discern their propensity to accumulate counterions accumulation at the loop–loop interface (data not shown). This is because the fully hydrated ions as defined by the Stern layer cannot penetrate into the central cation ‘‘binding pocket’’ (data not shown). Similarly, the axial spine of counterion density observed in the A-RNA helix (Fig. 20.5) is not captured by the P–B calculation (Fig. 20.7). No noticeable sequence specificity is observed in the counterion accumulation patterns in the P–B calculations, even though the sequence effects are explicitly represented in the P–B calculation through the appropriate geometry and assignment of point-charges. This is because the sequence specificity observed in the molecular dynamics simulations usually involves first shell interactions of base moieties with partially dehydrated ions, which cannot be accurately represented in the P–B framework. Defining a Stern layer using hydrated radii of the ions is justified in the P–B framework, since ions must be fully hydrated in order for the uniform dielectric approximation to be valid. If ions were allowed to penetrate this layer, then additional physics would have to be introduced to account for ion–ion correlations, partial dehydration, and other considerations that fall well outside the realm of applicability for standard continuum solvent models. To generalize the P–B framework, new approaches are needed
430
Alan A. Chen et al.
and promising results have been reported along these lines (Xu et al., 2005; Yi et al., 2008). These include the size-modified Poisson–Boltzmann theory (Borukhov et al., 1997), in which a lattice-gas partition function is introduced to limit the maximum possible ionic density (Chu et al., 2007), as well as the tightly bound ion theory (Tan and Chen, 2005), in which additional ion–ion correlations are introduced in areas of high electrostatic potential. Spatial free energy density profiles extracted from analysis of data obtained using simulations based on explicit representations of solvent molecules suggest that hydrated ions make up a majority of the atmosphere around RNA and DNA structures (Fig. 20.5). As a result, integrals of spatial free energy density profiles obtained from P–B calculations yield numerical values that are in the right range and show many—although not all—of the trends obtained from integrals of profiles from more expensive simulations. This bodes well for the use of modified P–B models that have the ability of incorporating ion–ion correlations or the effects of partially hydrated ions. These results also suggest that the use of grand canonical ensemble Monte Carlo (GCMC) simulations (Vitalis et al., 2004) combined with empirical implicit solvation models such as the recently developed ABSINTH model (Vitalis and Pappu, 2009) might provide a way to combine the virtues of solvent coarse graining and explicit inclusion of ion–ion correlations and many, if not all the necessary effects of partially desolvated ions. In fact, initial results applying a slightly modified ABSINTH model have yielded promising results for the Tar–Tar* system (Chen and Pappu, unpublished data) and more work is necessary in this direction.
ACKNOWLEDGMENT This work was supported by grant MCB-0718924 from the National Science Foundation (RVP).
REFERENCES Anderson, C. F., and Record, M. T. Jr. (1980). The relationship between the PoissonBoltzmann model and the condensation hypothesis: An analysis based on the low salt form of the Donnan coefficient. Biophys. Chem. 11, 353–360. Anderson, C. F., and Record, M. T. Jr. (1995). Salt-nucleic acid interactions. Annu. Rev. Phys. Chem. 46, 657–700. ˚ qvist, J. (1990). Ion water interaction potentials derived from free-energy perturbation A simulations. J. Phys. Chem. 94, 8021–8024. Auffinger, P., Cheatham, T. E. III, and Vaiana, A. C. (2007). Spontaneous formation of KCl aggregates in biomolecular simulations: A force field issue? J. Chem. Theory Comput. 3, 1851–1859.
Simulations of RNA Interactions with Monovalent Ions
431
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. Baker, N. A. (2004). Poisson-Boltzmann methods for biomolecular electrostatics. Numer. Comput. Methods 383(Pt D), 94–118. Baker, N. A., Sept, D., Joseph, S., Holst, M. J., and McCammon, J. A. (2001). Electrostatics of nanosystems: Application to microtubules and the ribosome. Proc. Natl. Acad. Sci. USA 98, 10037–10041. Borukhov, I., Andelman, D., and Orland, H. (1997). Steric effects in electrolytes: A modified Poisson-Boltzmann equation. Phys. Rev. Lett. 79, 435–438. Chang, K.-Y., and Tinoco, I. Jr. (1997). The structure of an RNA ‘‘kissing’’ hairpin complex of the HIV TAR hairpin loop and its complement. J. Mol. Biol. 269, 52–66. Cheatham, T. E. III, and Young, M. A. (2000). Molecular dynamics simulations of nucleic acids: Successes, limitations and promise. Biopolymers 56, 232–256. Chen, A. A., and Pappu, R. V. (2007a). Parameters of monovalent ions in the AMBER-99 forcefield: Assessment of inaccuracies and proposed improvements. J. Phys. Chem. B 111, 11884–11887. Chen, A. A., and Pappu, R. V. (2007b). Quantitative characterization of ion pairing and cluster formation in strong 1:1 electrolytes. J. Phys. Chem. B 111, 6469–6478. Chen, A. A., Draper, D. E., and Pappu, R. V. (2009). Molecular simulation studies of monovalent counterion-mediated interactions in a model RNA kissing loop. J. Mol. Biol. 390, 805–819. Chu, V. B., Bai, Y., Lipfert, J., Herschlag, D., and Doniach, S. (2007). Evaluation of ion binding to DNA duplexes using a size-modified Poisson-Boltzmann theory. Biophys. J. 93, 3202–3209. Cornell, W. D., Cieplak, P., Bayly, C. I., Gould, I. R., Merz, K. M. Jr., Ferguson, D. M., Spellmeyer, D. C., Fox, T., Caldwell, J. W., and Kollman, P. A. (1995). A second generation force field for the simulation of proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc. 117, 5179–5197. Das, R., Kwok, L. W., Millett, I. S., Bai, Y., Mills, T. T., Jacob, J., Maskel, G. S., Seifert, S., Mochrie, S. G. J., Thiyagarajan, P., Doniach, S., Pollack, L., and Herschlag, D. (2003). The fastest global events in RNA folding: Electrostatic relaxation and tertiary collapse of the tetrahymena ribozyme. J. Mol. Biol. 332, 311–319. Dolinsky, T. J., Nielsen, J. E., McCammon, J. A., and Baker, N. A. (2004). PDB2PQR: an automated pipeline for the setup of Poisson-Boltzmann electrostatics calculations. Nucleic Acids Res. 32, W665–W667. Draper, D. E. (2004). A guide to ions and RNA structure. RNA 10, 335–343. Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W., and Klein, M. L. (1983). Comparison of simple potential functions for simulating liquid water. J. Phys. Chem. 79, 926–935. Joung, I. S., and Cheatham, T. E. III (2008). Determination of alkali and halide monovalent ion parameters for use in explicitly solvated biomolecular simulations. J. Phys. Chem. B 112, 9020–9041. Laing, L. G., Gluick, T. C., and Draper, D. E. (1994). Stabilization of RNA structure by Mg ions specific and non-specific effects. J. Mol. Biol. 237, 577–587. Lambert, D., Leipply, D., Shiman, R., and Draper, D. E. (2009). The influence of monovalent cation size on the stability of RNA tertiary structures. J. Mol. Biol. 390, 791–804. Nakano, S.-I., Fujimoto, M., Hari, H., and Sugimoto, N. (1999). Nucleic acid duplex stability: Influence of base composition on cation effects. Nucleic Acids Res. 27, 2957–2965.
432
Alan A. Chen et al.
Pabit, S. A., Qiu, X., Lamb, J. S., Li, L., Meisburger, S. P., and Pollack, L. (2009). Both helix topology and counterion distribution contribute to the more effective charge screening in dsRNA compared with dsDNA. Nucleic Acids Res. 37, 3887–3896. Pan, J., Thirumalai, D., and Woodson, S. A. (1999). Magnesium-dependent folding of selfsplicing RNA: Exploring the link between cooperativity, thermodynamics, and kinetics. Proc. Natl. Acad. Sci. USA 96, 6149–6154. Perez-Salas, U. A., Rangan, P., Krueger, S., Briber, R. M., Thirumalai, D., and Woodson, S. A. (2004). Compaction of a bacterial group I ribozyme coincides with the assembly of core helices. Biochemistry 43, 1746–1753. Ponomarev, S. Y., Thayer, K. M., and Beveridge, D. L. (2004). Ion motions in molecular dynamics simulations on DNA. Proc. Natl. Acad. Sci. USA 101, 14771–14775. Rangan, P., and Woodson, S. A. (2003). Structural requirement for Mg2þ binding in the group I intron core. J. Mo. Biol. 329, 229–238. Savelyev, A., and Papoian, G. A. (2008). Polyionic charge density plays a key role in differential recognition of mobile ions by biopolymeres. J. Phys. Chem. B 112, 9135–9145. Semichaevsky, A. V., Marlowe, A. E., and Yingling, Y. G. (2009). Comparing ion distributions around RNA and DNA helical and loop-loop motifs. Mater. Res. Soc. Symp. Proc. 1130, W05. Smolin, N., and Winter, R. (2004). Molecular dynamics simulations of staphylococcal nuclease: Properties of water at the protein surface. J. Phys. Chem. B 108, 15928–19537. Sorin, E. J., and Pande, V. S. (2005). Exploring the helix-coil transition via all-atom equilibrium ensemble simulations. Biophys. J. 88, 2472–2493. Strauss, U. P., Helfgott, C., and Pink, H. (1967). Interactions of polyelectrolytes with simple electrolytes. II. Donnan equilibria obtained with DNA in solutions of 1–1 electrolytes. J. Phys. Chem. 71, 2550–2556. 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, Z.-J., and Chen, S.-J. (2005). Electrostatic correlations and fluctuations for ion binding to a finite length polyelectrolyte. J. Chem. Phys. 122, 044903. Vitalis, A., and Pappu, R. V. (2009). ABSINTH: A new continuum solvation model for simulations of polypeptides in aquous solutions. J. Comput. Chem. 30, 673–699. Vitalis, A., Baker, N. A., and McCammon, J. A. (2004). ISIM: A program for grand canonical Monte Carlo simulations of the ionic environment of biomolecules. Mol. Simul. 30, 45–61. Woodson, S. A. (2008). RNA folding and ribosome assembly. Curr. Opin. Chem. Biol. 12, 667–673. Xu, D. R., Greenbaum, N. L., and Fenley, M. O. (2005). Recognition of the spliceosomal branch site RNA helix on the basis of surface and electrostatic features. Nucleic Acids Res. 33, 1154–1161. Yi, M., Nymeyer, H., and Zhou, H. X. (2008). Test of the Gouy-Chapman theory for a charged lipid membrane against explicit-solvent molecular dynamics simulations. Phys. Rev. Lett. 101, 038103-1-4.
C H A P T E R
T W E N T Y- O N E
Ion–RNA Interactions: Thermodynamic Analysis of the Effects of Mono- and Divalent Ions on RNA Conformational Equilibria Desirae Leipply,* Dominic Lambert,† and David E. Draper*,† Contents 434 435 435 439
1. Introduction 2. Thermodynamic Descriptions of Ion–RNA Interactions 2.1. Excess ions in an equilibrium dialysis experiment 2.2. Thermodynamic definition of interaction coefficients 2.3. Implications of G and Gþ for ion activities in the presence of RNA 2.4. Application of interaction coefficients to RNA conformational changes 3. Analysis of Experimental Salt Dependence Data 3.1. Salt dependence of RNA folding reactions 3.2. MgCl2 dependence of folding reactions Acknowledgment References
440 441 444 444 448 461 461
Abstract RNA secondary and tertiary structures are strongly stabilized by added salts, and a quantitative thermodynamic analysis of the relevant ion–RNA interactions is an important aspect of the RNA folding problem. Because of long-range electrostatic forces, an RNA perturbs the distribution of both cations and anions throughout a large volume. Binding formalisms that require a distinction between ‘‘bound’’ and ‘‘free’’ ions become problematic in such situations. A more fundamental thermodynamic framework is developed here, based on preferential interaction coefficients; linkage equations derived from this framework provide a model-free description of the ‘‘uptake’’ or ‘‘release’’ of cations and anions that accompany an RNA conformational transition. Formulas * Department of Biophysics, Johns Hopkins University, Baltimore, Maryland, USA Department of Chemistry, Johns Hopkins University, Baltimore, Maryland, USA
{
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69021-2
#
2009 Elsevier Inc. All rights reserved.
433
434
Desirae Leipply et al.
appropriate for analyzing the dependence of RNA stability on either mono- or divalent salt concentration are presented and their application to experimental data is illustrated. Two example datasets are analyzed with respect to the monovalent salt dependence of tertiary structure formation in different RNAs, and three different experimental methods for quantitating the ‘‘uptake’’ of Mg2þ ions are applied to the folding of a riboswitch RNA. Advantages and limitations of each method are discussed.
1. Introduction It is well known that the conformations and stabilities of nucleic acids are sensitive to the concentrations and identities of salts that are present. With regard to RNAs, the strongly stabilizing effect of Mg2þ ion on tertiary structures has been of particular interest (Stein and Crothers, 1976), but virtually any RNA conformational equilibrium may shift in response to changes in either mono- or divalent ion concentrations. In describing the effects of ions on RNA folding equilibria, a widely used formalism assumes that ions are simple ligands that bind to RNA sites according to stoichiometric mass action equations. For instance, the effect of Mg2þ on the equilibrium between a folded (F) and unfolded (U) RNA has been written as U þ nMg2þ Ð FðMg2þ Þn
K¼
½F ðMg2þ Þn ½U½Mg2þ n
ð21:1Þ
(Fang et al., 1999; Latham and Cech, 1989; Schimmel and Redfield, 1980) This particular equilibrium and its corresponding equilibrium constant have been the starting point for deriving equations used to calculate the n ions ‘‘taken up’’ in a conformational transition or to extrapolate the free energy of RNA folding to different Mg2þ concentrations. Reaction equilibria such as Eq. (21.1) apply to neutral ligands that bind stoichiometrically to defined sites on a macromolecule, but their interpretation becomes problematic when used to describe ions and charged macromolecules. Mass action schemes overlook the essential electrostatic character of these interactions in two ways. First, the long-range character of electrostatic interactions is neglected. Ions a considerable distance from an RNA surface interact significantly with the RNA (Garcı´a-Garcı´a and Draper, 2003), and the total free energy of such long-range interactions may constitute the major source of ion-induced stabilization for many RNAs (Soto et al., 2007). Second, ions can be added to a solution only as electroneutral salts. When cations are titrated into an RNA solution, the unfavorable electrostatic interactions between the accompanying anions and the RNA are part of the overall energetic picture and generally must be taken
Thermodynamics of Ion–RNA Interactions
435
into account. As developed in the next sections, only under special circumstances can the effects of the added anion be ignored. To circumvent the above problems with mass action schemes, it is necessary to use a more general thermodynamic formalism based on parameters known as interaction coefficients, also called Donnan coefficients in some contexts (Record et al., 1998). This approach is completely general; it requires no assumptions about the types of interactions the ions may make with the RNA or the kinds of environments the ions may occupy. Although interaction parameters are a fundamental concept in thermodynamics and have been widely applied to biophysical problems, the literature on this topic can be difficult to access for anyone not already familiar with the formalism, and the application of interaction coefficients to the mixed monovalent–divalent cation solutions commonly used for RNA studies has received only limited attention (Grilley et al., 2006; Misra and Draper, 1999). For these reasons, the following theory section sets out the main concepts of the preferential interaction formalism in some detail, and outlines derivations of formulas relevant to monovalent ion–RNA interactions. Section 3 presents example analyses of experimental data, and extends the preferential interaction formalism to solutions of mixed salts (i.e., KCl and MgCl2). The section includes discussions of potential sources of error and practical considerations in data analysis for experiments with both mono- and divalent ions.
2. Thermodynamic Descriptions of Ion–RNA Interactions 2.1. Excess ions in an equilibrium dialysis experiment The key concept of the analysis developed here is the interaction coefficient, which we will use to assess the net interactions (favorable or unfavorable) taking place between ions and an RNA. We first introduce interaction coefficients by describing the way they might be measured in an equilibrium dialysis experiment, and give an overview of their significance. These parameters are defined in more formal thermodynamic terms in Section 2.2 and are subsequently used to derive formulas useful in the interpretation of experimental data. Consider two chambers separated by a dialysis membrane permeable to ions and water. After the chambers are filled with a KCl solution and allowed to come to equilibrium, the concentration of Kþ and Cl ions will be the same on each side of the membrane. An RNA with Z negative charges is then added to the left chamber (Fig. 21.1A). Because the RNA must be added as a neutral salt, Z cations are also added. In the example diagrammed in Fig. 21.1A, Z ¼ 16 and the Kþ salt of the RNA, KZRNA,
436
Desirae Leipply et al.
A
Equilibrium
Initial
ΓKCl = −2 L B
R +
−
+
L
Γ+ = 24 − 10 = 14
−
KZRNA
Γ+ +
− Γ−
R
Γ− = 8 − 10 = −2
Γ relationships
Z = Γ+ − Γ− ΓKCl = −Z + Γ+ = Γ−
KCl
Linit
Leq
Req
Figure 21.1 Ion–RNA interaction coefficients defined with respect to equilibrium dialysis. (A) Cartoon of a dialysis experiment. An RNA with 16 negative charges (red), together with 16 Kþ ions, is added (gray box) to a solution of KCl previously in equilibrium across a dialysis membrane (blue, Kþ; yellow, Cl; different shades of blue are used to distinguish Kþ contributed as a counterion to RNA (dark blue) or KCl (light blue)). In the approach to equilibrium (arrow), two Kþ and two Cl ions migrate from left to right across the dialysis membrane. (The right side dialysis chamber is presumed to be very large compared to the left side, such that the ion migration does not appreciably change the right side salt concentration.) Calculations of the single ion interaction coefficients (Gþ, G) are shown as the differences between the left and right side ion concentrations (Eq. (21.3)). (B) Histogram illustrating charge balance in the equilibrium dialysis experiment cartooned in panel A. The vertical heights of the colored bars represent the total concentration of the various ions. Color coding of mobile ions and RNA phosphate is as in panel A. The left-most pair of bars (Linit) represents the addition of the potassium salt of an RNA with Z negative charges (KZRNA) to the left dialysis chamber. The middle pair of bars (marked Leq) represents the ion concentrations in the left chamber after dialysis equilibrium has been achieved with the concentrations of salts indicated in the right dialysis chamber (marked Req). The relations between the single ion coefficients (Gþ, G) and histogram heights are marked, and the relationships between the interaction coefficients and the number of RNA charges (Z) are listed.
437
Thermodynamics of Ion–RNA Interactions
is illustrated. The two dialysis chambers are no longer in equilibrium after the addition, because of the excess of 16 Kþ on the left side of the membrane as compared to the right side. The RNA itself is confined to the left side, but ions may migrate across the membrane to achieve equilibrium. For simplicity, suppose that the volume of the right side is so large that the migration of ions in or out of it does not appreciably change its KCl concentration; that is, the illustrated volume is only a small fraction of the total. (As a further simplification, suppose that the RNA molecule occupies a negligible fraction of the left side volume.) The excess Kþ ions in the left chamber will tend to flow down the concentration gradient into the right chamber. To maintain a net zero charge on each side of the membrane, an equivalent number of negative charges must accompany the flow of cations. In the Fig. 21.1A example, equilibrium is achieved after two Kþ–Cl ion pairs migrate from the left to the right. There are now two ion concentration gradients: the initial excess of Kþ ions accompanying the RNA persists, though reduced in magnitude, and a deficiency of Cl anions has been created in the RNA solution. (Figure 21.1A shows a reduction in the number of the dark blue Kþ ions that accompanied the RNA, but obviously there is no way to distinguish dissolved Kþ ions on the basis of their original ionic partners.) The two ion gradients that develop in the dialysis experiment are conveniently quantitated in terms of interaction coefficients. A histogram (Fig. 21.1B) diagrams the ion concentrations in a similar experiment as cartooned in Fig. 21.1A. The total number of KCl ion pairs that migrates across the membrane as equilibrium is established, when normalized by the number of RNA molecules present, becomes the interaction coefficient GKCl. Another way to find the same number is to count the total number of KCl ion pairs on each side of the membrane at equilibrium GKCl ffi
½KClL ½KClR ½RNA
ð21:2Þ
where the subscripts on the salt concentrations indicate the side of the dialysis membrane in Fig. 21.1A, and refer to concentrations at dialysis equilibrium. (Note that on the left hand side, each Kþ ion can be paired with either a Cl or RNA phosphate anion for the purposes of determining the concentrations of the two electroneutral salts, KCl and KZRNA. On either side of the membrane in the Fig. 21.1A example, [KCl] ¼ [Cl].) Another way of representing the ion concentration differences at equilibrium is by way of single ion interaction coefficients, which count the cations and anions separately: Gþ ffi
½Kþ L ½Kþ R ½RNA
G ffi
½Cl L ½Cl R ½RNA
ð21:3Þ
438
Desirae Leipply et al.
These three different G parameters are related to each other and to the total number of RNA charges, Z (see ‘‘G relationships’’ in Fig. 21.1B). First, the number of KCl pairs lost from the left hand side during the approach to equilibrium is the same as the equilibrium difference in the number of Cl ions between the two chambers, GKCl ¼ G. Electroneutrality also requires that Gþ ¼ G þ Z, that is, the left side excess of cations (relative to the right side solution) must be balanced by an equivalent number of negative charges (note that G < 0). Figure 21.1 has been drawn to suggest that Gþ > |G|. This inequality is generally true for nucleic acids in low to moderate salt, a phenomenon sometimes called the ‘‘polyelectrolyte effect’’ (Draper, 2008; Record and Richey, 1988). Any RNA conformational change that increases the density of phosphate charges will also increase Gþ at the expense of |G| (Record et al., 1998). However, Gþ may be similar to |G| at high salt concentrations: for instance, Gþ ¼ 0.46 and G ¼ 0.54 ions/nucleotide for DNA in 0.98 M NaBr (Strauss et al., 1967). The dialysis experiment is a convenient way to conceptualize the meaning of the interaction coefficient, but the formal definition of a G does not depend on the presence of a membrane (see Section 2.2 below). G essentially measures the tendency of an RNA molecule to create ion concentration gradients in solution, manifested as the accumulation of cations and depletion of anions in a volume surrounding the RNA, relative to the ‘‘bulk’’ concentration of ions a large distance away from the RNA surface. These gradients are related to the sum of all the attractive and repulsive forces experienced by the ions, and therefore to the overall energetics of ion–RNA interactions. The advantages of using this interaction formalism when considering ions are evident in a comparison with the interpretation of an equilibrium dialysis experiment by a binding formalism. n, the ‘‘binding density’’ of a ligand with a macromolecule, would be calculated from a dialysis experiment in exactly the same way as Gþ (Eq. (21.3)), but interpreted as the number of ligands ‘‘bound’’ to the macromolecule. However, as pointed out in Section 1, n fails to capture two aspects of ion–RNA interactions that are represented by G. First, GKCl (or G) properly represents the depletion of anions near the RNA. The concept of a ‘‘binding density’’ obviously breaks down when applied to Cl; a negative-binding density cannot be meaningfully discussed in terms of binding sites. Second, manipulations of n (i.e., in the calculation of equilibrium constants and interaction free energies) presume that n ligands are bound per macromolecule and the remaining ‘‘free’’ ligands have no interactions at all. This distinction between free and bound ligands is valid only if all ligand– macromolecule interactions are short range. Because of long-range interactions, all the ions in a solution interact with an RNA, and it is not possible to possible to parse the ions into distinct bound and free fractions. The interaction coefficients, by contrast, are completely model-independent, in
Thermodynamics of Ion–RNA Interactions
439
that they reflect the influences of all the energetic costs that are present: long-range electrostatic attraction and repulsion as described by Coulomb’s law, as well as hydration changes and all short-range factors.
2.2. Thermodynamic definition of interaction coefficients The preceding sections have used standard molar concentration units for RNA and ions, indicated by brackets or the abbreviation M. Thermodynamic definitions of interaction coefficients are made in terms of molal units, abbreviated m, the moles of solute per kilogram of solvent water. Molal units have the convenient properties that the concentration of water is a constant 55.5 m regardless of the amount of solute(s) present, and the molality of one solute is unaffected by addition of a second solute. For dilute solutions, M and m units are interchangeable. We use molal units for the thermodynamic derivations in this section, and indicate later (Section 3.1) the salt concentrations where a correction for molar–molal conversion is required. Interaction coefficients are formally defined as partial derivatives, for example: @mþ @m Gþ ¼ ; G ¼ ð21:4aÞ @mRNA mþ @mRNA m @mKCl GKCl ¼ ð21:4bÞ @mKZ RNA mKCl where concentrations of ions (mþ, m), RNA (mRNA), and neutral salts (mKCl and mKZ RNA ) are in molal units. The derivatives are the limits of Eqs. (21.2) and (21.3) as the added concentration of RNA becomes very small (see equation 1 in Record et al. (1998)). In the dialysis examples (Fig. 21.1), the chemical potentials mKCl or mþ and m are set by the concentration of KCl in the right chamber, and therefore are held constant as RNA is added. Hence Gþ reports on the number of cations that should accompany the introduction of a single RNA molecule, in order to keep a constant cation chemical potential in the solution, and G reports the number of anions that must be simultaneously removed to maintain constant m. Note that the neutral salt coefficient is defined in terms of the addition of the neutral RNA salt, as illustrated in Fig. 21.1, while the single ion coefficients consider the hypothetical addition of a charged RNA. In the definitions of G, two variables in addition to the ion chemical potential must also be specified as constant. In an equilibrium dialysis experiment, these are temperature and the chemical potential of water. This partial derivative is known as the Donnan coefficient. (Note that the hydrostatic pressure is higher in the RNA-containing solution.) In making connections between G and the Gibbs free energy, it is more convenient if temperature
440
Desirae Leipply et al.
and pressure are held constant instead. At the concentrations of RNA and ions typically used in experiments, the quantatitive difference between these two kinds of interaction coefficients is inconsequential (Anderson, et al., 2002). For all partial derivatives in the following sections, the conditions of constant temperature and pressure apply but are not explicitly written.
2.3. Implications of G and Gþ for ion activities in the presence of RNA The effective concentration of an ion, better known as its activity, must be the same on each side of a dialysis membrane at equilibrium; in other words, there is no driving force for net ion migration from one side of the membrane to the other. The actual ion concentrations, used to calculate G and Gþ, are different on either side of the membrane because of RNA– ion interactions. The ratio of ion activity to concentration in the presence of an RNA will provide a starting point for derivations that link interaction coefficients to the effects of ions on RNA folding transitions. As background for these derivations, the relationships between interaction coefficients, ion concentrations, and ion activities are outlined here. In Fig. 21.1, the condition for thermodynamic equilibrium is that the chemical potential of each membrane-permeable ion; is identical between the left and right side solutions. The chemical potential m can be defined either for the 1:1 salt or for the individual ions mKCl ¼ mKCl þ RT ln aKCl
mKþ ¼ mK þ þ RT ln aKþ ;
mCl ¼ mCl þ RT ln aCl
ð21:5aÞ ð21:5bÞ
where m is a standard state defined for a particular temperature and pressure, and a is the activity of the salt or ion. Activities of cations and anions usually cannot be measured separately; instead, a mean ionic activity, a, is used: aKCl ¼ ðaKþ ÞðaCl Þ ¼ a2
ð21:6Þ
Activities are related to concentrations by the activity coefficient g, aKCl ¼ gKþ mKþ gCl mCl ¼ g2 mKþ mCl
ð21:7Þ
where g is the mean ionic activity coefficient. Because ion chemical potentials (and thus the ion activities) must be the same on each side of a dialysis membrane at equilibrium, the ion concentration differences in the Fig. 21.1A example imply that the ion activity coefficients are different in the presence and absence of RNA, and the following sets of inequalities apply: ðmKþ ÞL > ðmKþ ÞR ;
Gþ > 0;
gþL =gþR < 1
ð21:8aÞ
441
Thermodynamics of Ion–RNA Interactions
ðmCl ÞL < ðmCl ÞR ;
G < 0;
gL =gR > 1
ð21:8bÞ
These inequalities are a consequence of the net attractive (21.8a) or repulsive (21.8b) interactions between RNA and cations or anions, respectively. The reciprocal effect of the ions on the RNA activity coefficient is taken up in Section 2.4.
2.4. Application of interaction coefficients to RNA conformational changes The interconversion between any two conformations of an RNA, here named folded (F) and unfolded (U) for simplicity, is associated with a free energy change which is expressed in terms of the chemical potentials of the two conformations aF aF DG ¼ mF mU ¼ mF mU þ RT ln ¼ DG þ RT ln aU aU ð21:9Þ where m is the chemical potential, m is the standard state chemical potential, and a is the activity; subscripts specify the RNA conformation. Note the distinction between the actual free energy change, DG, and the free energy difference between U and F in their standard states, DG (more about the definition of standard states below). The condition for equilibrium between U and F is DG ¼ 0; in this circumstance, the thermodynamic equilibrium constant is defined as aF g mF Keq ¼ ¼ F ð21:10Þ aU gU mU and the standard relation DG ¼ RT ln Keq applies. The definition of the RNA standard state (m0) in Eq. (21.9) is critical. Thermodynamic standard states refer to hypothetical ideal states in which the molecular interactions of interest (those that cause nonideal behavior in the real world) are turned off. If it were possible to carry out a dialysis experiment with such an ‘‘ideal’’ RNA, one that had no net interactions with the ions of a salt, then the salt activity coefficients would be the same on each side of the dialysis membrane ðgKCl ÞL ¼ ðgKCl ÞR
ð21:11Þ
(cf. Eqs. (21.8a) and (21.8b)). (This relation does not imply that electrostatic interactions are absent, only that the net interactions of the RNA with all ions in solution sums to zero. Under this condition, added salt cannot affect any RNA conformational transition; Keq is therefore independent of salt
442
Desirae Leipply et al.
concentration.) In this hypothetical ideal state, the RNA activity is the same as its concentration (aRNA ¼ mRNA and gRNA ¼ 1). Deviation from ideal behavior in the form of net favorable interactions between the salt ions and RNA will cause gRNA to take on values less than 1. In experiments, only the concentrations of various RNA conformers can be measured directly, not their activities. An ‘‘observed’’ equilibrium constant is expressed in terms of the actual concentrations mF Kobs ¼ ð21:12Þ mU By omitting activity coefficients, this formula implies that the corresponding free energy change (DGobs ¼ RT ln Kobs ) does not refer to the ideal standard states described above, but to RNAs interacting with salt under the specific set of solution conditions used for evaluating Kobs. If the salt concentration changes, both the standard states and Kobs change. Substituting Eq. (21.12) into (21.10) gives the relation between thermodynamic and observed equilibrium constants as Keq ¼
gF Kobs gU
ð21:13Þ
This equation will give us a way to use changes in an experimental observable, Kobs, to access changes in RNA activity coefficients. At this point, we have defined an ideal reference state for the RNA in which there are no net interactions with ions, and introduced the RNA activity coefficient as a factor that assesses the deviation of the RNA from ideal behavior due to its interactions with all the ions in solution. No assumptions have been made about the nature of the ion interactions: anions and cations, long- and short-range interactions all contribute. The ion interaction coefficients (Eqs. (21.4a) and (21.4b)) also reflect the ion–RNA interactions that create concentration differences in a dialysis experiment, and there is an intimate relationship between activity coefficients (g) and interaction coefficients (G), as developed below. This relationship will be extremely useful: g comes from the chemical potential and gives access to free energies and other thermodynamic functions, while G is directly accessible by both experiment and computation (see Pappu et al., this volume, III.20). To show how g, G, and free energies are linked, we first find an expression for GKCl in terms of gRNA, and then connect GKCl to Eq. (21.13). We begin by expanding the definition of GKCl using a standard property of partial derivatives, the Euler chain rule @mKCl @ ln aKCl @mKCl GKCl ¼ ¼ ð21:14Þ @mRNA ln aKCl @mRNA mKCl @ lnaKCl mRNA
Thermodynamics of Ion–RNA Interactions
443
(It is equivalent whether mKCl or ln aKCl is held constant.) Because interactions are always mutual (if A attracts B, B must attract A), any ion–RNA interaction must cause both the ion and RNA activity coefficients to change. This principle is embodied in Euler reciprocity relations (discussed in standard physical chemistry textbooks, e.g., p. 116 of Levine, 2002); in this case the relation takes on the form @ ln aKCl @ ln aRNA ¼ ð21:15Þ @mRNA mKCl @mKCl mRNA Substituting Eq. (21.15) into (21.14) and simplifying by application of the standard chain rule, we obtain @ ln aRNA @ ln gRNA GKCl ¼ ¼ ð21:16Þ @ ln aKCl mRNA @ ln aKCl mRNA (The second equality follows because mRNA is being held constant.) If this equation is now applied to a difference between GKCl for the U and F states of an RNA @ lnðgF =gU Þ F U DGKCl ¼ GKCl GKCl ¼ ð21:17Þ @ ln aKCl mRNA Differentiation of Eq. (21.13) with respect to ln aKCl while holding mRNA constant gives @ lnðgF =gU Þ @ ln Kobs ¼ ð21:18Þ @ ln aKCl mRNA @ ln aKCl mRNA (Recall that Keq is independent of salt activity, and so does not appear in the derivative.) Combining Eqs. (21.17) and (21.18) and making use of the standard relation between DG and an equilibrium constant, the final equation relating a change in the RNA–salt interaction coefficient to a change in the observed RNA folding free energy is @ ln Kobs 1 @DGobs DGKCl ¼ ¼ ð21:19Þ @ ln aKCl @ ln aKCl RT (The specification of constant mRNA has been dropped because we are assuming that these equations are being applied to dilute enough ranges of RNA concentrations that GKCl can be considered a constant with respect to mRNA. Constant temperature and pressure still apply.)
444
Desirae Leipply et al.
A last point to consider is the relation of DGKCl to single-ion coefficients. If aKCl is put into terms of the mean ionic activity a (see Eq. (21.6), Section 2.3), Eq. (21.19) becomes @ ln Kobs ¼ 2DGKCl ¼ DGþ þ DG ¼ 2DG @ ln a
ð21:20Þ
The way the second equality is written is justified by two relationships: GKCl is numerically equivalent to G (cf. Fig. 21.1A and Eqs. (21.2) and (21.3)), and electroneutrality (Fig. 21.1B, Gþ ¼ G þ Z ) requires that DGþ ¼ DG. Suppose an RNA conformational change causes Gþ to decrease, in effect a ‘‘release’’ of cations that (in terms of the Fig. 21.1A diagram) flow out of the left dialysis chamber. The same number of anions must flow across the membrane in order to maintain charge neutrality, which causes G to become more negative. (Mass action equations which suppose that an RNA ‘‘takes up’’ or ‘‘releases’’ only cations, as Eq. (21.1), violate the principle of electroneutrality by ignoring the corresponding changes in anions.) We call the final result 2DG to emphasize that both cations and anions are affected to the same extent when an RNA changes conformation. Equations (21.19) and (21.20) are connections between the sensitivity of an RNA folding equilibrium to added salt, and the ‘‘uptake’’ or ‘‘release’’ of cations and anions as measured by the preferential interaction parameter. A similar kind of thermodynamic linkage between an added solute and a macromolecular conformational change was derived by Wyman for the case of O2 binding to specific sites on hemoglobin. A binding formalism (as in Eq. (21.1)) was originally used (Wyman, 1948), but later generalized in a way which bypasses specific binding models, as described above (Wyman, 1964). The modifications of linkage equations necessary when equilibria involve ions and charged polymers has been extensively considered by Record and colleagues (equation 47 in Record et al. (1998) is identical to Eq. (21.20) above).
3. Analysis of Experimental Salt Dependence Data 3.1. Salt dependence of RNA folding reactions RNA secondary structures and many tertiary structures are stable in the absence of any divalent cation, and may require only moderate concentrations of a monovalent salt such as KCl or NaCl. In this section, we give two examples of the calculation of DG for such RNAs.
445
Thermodynamics of Ion–RNA Interactions
3.1.1. Example 1: A-riboswitch RNA The adenine riboswitch RNA (A-riboswitch) adopts a folded tertiary structure upon binding a purine ligand; in the absence of the ligand, only secondary structure remains (Serganov et al., 2004). In melting experiments with this RNA, inclusion of a purine ligand results in the appearance of a new unfolding transition that is generally well resolved from the unfolding of secondary structure (Lambert and Draper, 2007, 2009). From the melting temperature (Tm) and DH of the folding transition, Kobs is readily calculated at a desired temperature (see legend to Fig. 21.2; analysis of melting curves has been reviewed by Draper et al. (2000)). Values of Kobs extrapolated to 20 C for the A-riboswitch in various KCl concentrations are shown in Fig. 21.2A. The solutions were originally made volumetrically using molar concentration units, as shown; two manipulations were needed to convert molarity to activity on the molal scale, as needed for application of the linkage Eq. (21.20):
B 4.0
0.8
3.5
0.6
3.0
0.4 log (Kobs)
log (Kobs)
A
2.5 2.0 1.5
0 −0.2 −0.4
1.0 0.5
0.2
−0.6 0.04
0.1
0.3
mK+ ( ) or a± ( )
0.5
−0.8
−0.6
−0.4
−0.2
Log (a±)
Figure 21.2 Monovalent salt dependences for the formation of RNA tertiary structures. For both RNAs shown, log(Kobs) was calculated at 20 C from the melting temperature (Tm) and enthalpy (DH ) of the tertiary folding transition observed in melting curves by the formula ln(Kobs) ¼ (H /R)(1/Tm 1/T0), where R is the gas constant and T0 is 293 K. (A) log(Kobs) for folding of the A-riboswitch (buffer: 4 mM 2,6-diaminopurine, 20 mM MOPS adjusted to pH 6.8 with KOH, 0.1 mM EDTA, and various KCl concentrations). Data are taken from Lambert et al. (2009). Each point is the average of three experiments; error bars are smaller than the data points. The salt molality and mean ionic activity are calculated from the total Kþ concentration contributed by both KCl and MOPS buffer. (B) log(Kobs) for formation of the complex between tar and tar* hairpins (buffer: 5 mM cacodylic acid adjusted to pH 6.4 with KOH, 0.1 mM EDTA, and various KCl concentrations). Data are taken from Lambert et al. (2009). Error bars are standard deviations from the average of three experiments.
446
Desirae Leipply et al.
1. Molar concentration units were converted into molal units using the partial molar volume of the salt. (We assume that RNA and other buffer components are present in such low concentrations that the density of the solution is not significantly different than a solution of water and salt alone.) The formula is mMCl ¼
½MCl ð1=r1 Þð1 þ ½MClV MCl =1000Þ
ð21:21Þ
where V MCl is the partial molar volume of the salt MCl, in units of ml/ mol; r1 is the density of water, 0.998 g/ml at 20 C; the brackets indicate molar concentrations, and m is the molal concentration. See Section 2.1 of Eisenberg (1976) for a full discussion of the relation between molal and molar units. Table 21.1 lists partial molar volumes of different salts. The difference between molar and molal units is about 1% for a 0.3 M KCl solution, and rises to only 3% for 1 M KCl. These percentages are smaller than the error in most measurements of Kobs. 2. Molal salt concentrations must be translated into mean ionic activity (Eq. (21.7)). The mean ionic activity coefficients of many salts are compiled in the literature as a function of salt concentration (see appendix 8.10 Table 21.1 Solution properties of some commonly used monovalent salts
a
b
Salt
V MCl (cm3/mol)b
LiCl
16.0
NaCl
12.9
KCl
23.6
RbCl
28.76
CsCl
36.13
NH4Cl
33.92
Dependence of ln(g) on ln(mMCl)a a1
a2
a3
a4
0.2609 0.2946 0.4215 0.4423 0.5081 0.5052 0.5429 0.5339 0.6140 0.1739 0.5099 0.5048
0.1098 0.0283 3.3369e2 8.3134e2 0.1139 0.1059 0.1306 0.1159 0.1739 0.1369 0.1147 0.1064
8.2973e2 0.02344 3.6126e2
1.3270e2 8.0234e3
2.8418e3 4.9786e3 1.2411e2 2.8509e3
The listed coefficients allow calculation of ln(g) from a polynomial, ln(g) ¼ a1 þ a2[ln(mMCl)] þ a3[ln(mMCl)]2 þ a4[ln(mMCl)]3 For each salt, the first line is appropriate for the range 0–0.8 m, and the second line for the range 0–0.5 m. The coefficients were obtained by least squares fitting of polynomials to the activity coefficient data compiled by Robinson and Stokes (2002). Partial molar volumes are taken from Millero (1971). See also Marcus (1994) for an extensive compilation and evaluation of V values for salts.
Thermodynamics of Ion–RNA Interactions
447
in Robinson and Stokes 2002). Table 21.1 provides polynomial coefficients needed to calculate ln(g) at a given ln(m) for common 1:1 chloride salts. A plot of ln(g) as a function of ln(m) generally turns concave upward as the salt concentration approaches 1 m; this curvature is particularly severe for LiCl, which has a minimum in g at about 0.5 m. Therefore two sets of coefficients are given in Table 21.1, one adequate for salt concentrations up to 0.5 m, and a higher order polynomial adequate up to 0.8 m. Figure 21.2A contains a second set of data points for which the molar units used in carrying out the experiments have been recast as activities on the molal scale. Both sets of data are linear within experimental error, but the slopes differ by about 10%: @ ln Kobs @ ln Kobs ¼ 3:33 0:09; ¼ 2DG ¼ 3:67 0:10 @ ln mKCl @ ln a A second way to find 2DG from salt dependence data takes advantage of the following application of the standard chain rule to Eq. (21.20): @ ln Kobs @ ln Kobs @ ln mKCl 2DG ¼ ¼ ð21:22Þ @ ln a @ ln mKCl @ ln a The first term on the right is the salt dependence of ln Kobs, expressed in molal concentration units. The second term can be calculated from the dependence of the salt activity coefficient on salt concentration @ ln a @ ln g ¼1þ ¼1þe ð21:23Þ @ ln mKCl @ ln mKCl For many salts, e is nearly constant over a wide concentration range; (1 þ e) is then a simple correction factor that can be applied to the slope of the salt dependence. This approximation is valid for KCl in the 0–0.5 m range; from the constant a2 in Table 21.1, e ¼ 0.1059. Applying this factor to the Fig. 21.2A data, we obtain @ ln Kobs 1 ¼ ð3:33Þð1 0:1059Þ1 ¼ 3:72 @ ln mKCl 1 þ e which is within experimental error of the result obtained above from the plot of ln(Kobs) as a function of ln(a). 3.1.2. Example 2: tar–tar* A so-called ‘‘kissing-loop’’ complex forms by Watson–Crick base pairing between the loops of two hairpins. Melting experiments with one such complex, between hairpins named tar and tar*, easily resolve disruption of
448
Desirae Leipply et al.
the bimolecular complex from denaturation of the individual hairpins at higher temperatures (Chang and Tinoco, 1994; Lambert and Draper, 2007). The dependence of log(Kobs) on log(a) for the tar–tar* complex is plotted in Fig. 21.2B. In contrast to the A-riboswitch data in Fig. 21.2A, these data are best fit by a second order polynomial. Although the dependences of RNA folding transitions on salt tend to be linear, there is no necessity that they be so. Whether a polynomial should be used to describe the salt dependence depends on the size of the errors in ln(Kobs) and the range of salt concentrations over which data have been collected. In this case, there are systematic deviations of the data from a linear fit that are larger than the error in the data, and the polynomial is a better description of the data. The fitted polynomial for the tar–tar* data is logðKobs Þ ¼ 0:9084 þ 0:6983 logða Þ 1:238ðlogða ÞÞ2 therefore 2DG ¼ 0:6983 2:476 logða Þ The slope evaluated at the middle of the data range (a ¼ 0.268) is 2DG ¼ 2.11. It would of course be unwise to extrapolate values of 2DG to activities outside of the salt concentration range over which data were collected. In both of the above examples, we used an anionic buffer (MOPS or cacodylate). The buffer anions have only repulsive interactions with RNA and can be grouped with chloride ions when calculating mean ion activities. Thus, we apply mean ionic activity coefficients measured with KCl solutions to solutions in which Kþ ions are contributed both by KCl and K-buffer salts. We strongly advise against the use of cationic buffers such as Tris, because of its idiosyncratic interactions with nucleic acids as compared to group I ions, and particularly against mixing KCl with Tris buffer, which creates a cationic mixture of unknown activity.
3.2. MgCl2 dependence of folding reactions 3.2.1. Interaction coefficients in mixed divalent–monovalent cation solutions Suppose the Kþ salt of an RNA is added to a dialysis chamber that has been equilibrated with a mixture of MgCl2 and KCl (Fig. 21.3A). As in the monovalent salt example (Fig. 21.1A), KCl ion pairs will tend to diffuse into the right chamber, leaving an excess of Kþ and creating a deficiency of Cl in the left chamber. However, there will also be a tendency for the RNA to accumulate Mg2þ in preference to Kþ; the resulting net diffusion of Mg2þ ions into the left chamber must be accompanied by enough Cl ions to
449
Thermodynamics of Ion–RNA Interactions
A
Initial
Equilibrium
ΓMgCl2 = 2
L
R
L Γ+ = 20 − 10 = 10
ΓKCl = −6
R Γ− = 12 − 14 = −2
Γ2+ = 4 − 2 = 2 B
+
−
Γ relationships
+
− 2Γ2+ + Γ+
KZRNA
+
Z = 2Γ2+ + Γ+ − Γ− ΓMgCl = Γ2+ 2 ΓKCl = −Z + Γ+
− Γ−
Γ−
Γ+
KCl 2Γ2+ MgCl2 Linit
Leq
Req
L R
L R
L R
Figure 21.3 Definition of ion–RNA interaction coefficients in mixed monovalent– divalent salt solutions. Kþ and Cl ions are identified by the same colors as in Fig.21.1; green circles represent Mg2þ. (A) Migration of ions taking place after addition of an RNA and its neutralizing Kþ ions (in gray rectangle) to a mixture of KCl and MgCl2 previously in equilibrium across a dialysis membrane. As equilibrium is reestablished (arrow), there is a net migration of Kþ into the right chamber and a net migration of Mg2þ into the left chamber. The movement of each cation type is accompanied by a neutralizing number of anions; in this example, there is a net flow of two Cl out of the left chamber. (As in Fig. 21.1, the right side dialysis chamber is presumed to be very large compared to the left side, such that the ion migration does not appreciably change the right side salt concentration.) Calculations of all three single ion interaction coefficients (Gþ, G, G2þ) are shown as the differences between the left and right side ion concentrations (Eqs. (21.3) and (21.24)). (B) Histogram illustrating charge balance in the equilibrium dialysis experiment cartooned in panel A. The vertical heights of the colored bars represent the total concentration of charge contributed by various ions, with the same color coding as in panel A. (The Mg2þ ion concentration is half the charge concentration.) The pair of bars labeled Linit represents the combination of salts added to the ‘‘left’’ chamber of a dialysis apparatus: the potassium salt of an RNA with Z
450
Desirae Leipply et al.
neutralize the Mg2þ charge. There are now two neutral salt interaction coefficients; GMgCl2 is positive and GKCl is negative. The ion gradients can also be represented by three single ion coefficients; Gþ and G are defined as in Eq. (21.4a), and G2þ is G2þ ffi
½Mg2þ L ½Mg2þ R ½RNA
ð21:24Þ
This coefficient is always positive. The relations between the neutral salt coefficients, the single ion coefficients, and the number of RNA negative charges are illustrated by the histograms and equations in Fig. 21.3B. The neutral salt coefficient GMgCl2 is identical to G2þ, which becomes important in the interpretation of linkage expressions (Section 3.2). Another important relation is the condition for electroneutrality, which is now 2G2þ þ Gþ G ¼ Z
ð21:25Þ
Any RNA conformational change that alters one of the three G values must also alter one or both of the other two. This is another aspect of RNA–ion equilibria that is generally not included in mass action equilibrium expressions: any change in the interactions between one kind of ion and an RNA is always accompanied by reciprocal changes in the interactions of the other ion species. 3.2.2. The linkage relation in titrations with MgCl2 The linkage equation that quantifies the dependence of an RNA folding equilibrium on added divalent cation is @ ln Kobs DGMgCl2 ¼ ð21:26Þ @ ln aMgCl2 mKCl where the concentration of monovalent salt must be held constant. The derivation of this relation proceeds as outlined in Eqs. (21.14)–(21.19), except that Eq. (21.13) is differentiated with respect to ln aMgCl2 with the additional stipulation that mKCl is held constant. Equation (21.26) generally cannot be applied directly to experimental data, because of the problem of determining the activity of the added MgCl2
negative charges (KZ RNA), KCl, and MgCl2. The next pair of bars (marked Leq) represents the ion charge concentrations in the left chamber after dialysis equilibrium has been achieved with the concentrations of salts indicated in the right dialysis chamber (marked Req). The Leq and Req concentrations of the three ions are individually compared on the right side of the diagram, and relationships between the single ion coefficients, neutral salt coefficients, and the number of RNA charges are listed.
Thermodynamics of Ion–RNA Interactions
451
in the presence of a monovalent salt. (Note that it is not feasible to carry out a titration with Mg2þ as the only cation, because buffer always contributes a monovalent cation, usually at comparable or larger concentrations than the added Mg2þ.) The Cl that is introduced with the Mg2þ affects the activities of both Kþ and Mg2þ ions. Although it is possible to estimate mean ion activities in solutions of mixed electrolytes (Harned and Robinson, 1968), it is easier in this case to design the experiment in a way that minimizes changes in the MgCl2 activity coefficient during the course of a titration. The strategy is to insure that the solution always has a large excess of KCl over the added MgCl2, such that the Kþ and Mg2þ ions see a relatively constant concentration of anions during the course of the titration. Because salt activity coefficients at low to moderate salt concentrations are primarily determined by cation–anion interactions, the activity coefficients change very little despite the increasing Mg2þ concentration. The approximation that the MgCl2 mean ionic activity coefficient remains constant is introduced into the linkage relation by first expanding Eq. (21.26) by the standard chain rule: @ ln mMgCl2 @ ln Kobs @ ln Kobs ¼ ð21:27Þ @ ln aMgCl2 mKCl @ ln mMgCl2 mKCl @ ln aMgCl2 mKCl We next examine the last term, the dependence of mMgCl2 on aMgCl2 . The activity of MgCl2 in a mixed solution with KCl is defined as aMgCl2 ¼ ðaMg2þ ÞðaCl Þ2 ¼ gMgCl2 mMgCl2 ð2mMgCl2 þ mKCl Þ2
ð21:28Þ
which, upon differentiation with respect to ln mMgCl2 while holding mKCl constant, yields @ ln gMgCl2 @ ln aMgCl2 4mMgCl2 ¼ þ1þ ð21:29Þ @ ln mMgCl2 mKCl @ ln mMgCl2 mKCl 2mMgCl2 þ mKCl The last term is simply computed from the concentrations of the two salts. The partial derivative of lnðgMgCl2 Þ can be estimated from a theory of mixed electrolytes (Harned and Robinson, 1968), and is small and of opposite sign compared to the last term. We have previously estimated from Eq. (21.29) that (@ ln mMgCl2 =@ ln aMgCl2 ), 1 as long as the ratio [KCl]/[MgCl2] remains larger than 30 (Grilley et al., 2006). In an experimental test, the titration of a fluorescent chelator dye with MgCl2 gave an anomalous binding curve unless excess KCl was included in the titration buffer. A 30:1 Kþ:Mg2þ ratio suppressed most of the anomaly (Leipply and Draper, 2009). Upon substitution of Eq. (21.27) into the linkage Eq. (21.26), and including the above approximation that (@ ln mMgCl2 =@ ln aMgCl2 ) 1, as well as the additional approximation that molarity and molality units are interchangeable, the linkage relation for MgCl2 becomes
452
Desirae Leipply et al.
@ ln Kobs DGMgCl2 DG2þ @ ln½Mg2þ ½MCl
ð21:30Þ
The last equivalence comes from the definitions of neutral salt and single ion interaction coefficients (Fig. 21.3). Because ions must always be added to a solution as neutral salts, it is usually impossible to distinguish the effects of the cation from those of the anion. Equation (21.30) is therefore an unusual result, in that a term containing the single ion interaction coefficient for Mg2þ may be extracted from the titration of an RNA with MgCl2. This outcome depends entirely on the fact that the effects of MgCl2 on RNA folding can usually be observed in the presence of a large excess of monovalent ion, which allows the approximation that Eq. (21.29) takes on the value of one. It has not generally been appreciated that the way to isolate the effect of Mg2þ on an RNA from other factors is not to minimize the monovalent cations in the experiment, but to make sure that they are in large excess. The first two of the following examples show how DG2þ may be determined by application of Eq. (21.30), using two different methods to find Kobs. In a third example, these DG2þ values are then compared with those obtained by a direct measurement which bypasses linkage relations and determination of Kobs. 3.2.3. Example 1: DG2 from isothermal titration with MgCl2 In a titration of RNA with MgCl2, the extent of folding can be monitored in several ways, for instance by chemical probing experiments (Koculi et al., 2007; Ralston et al., 2000) or by following the change in a spectroscopic signal such as UV absorbance or ellipticity (CD) (Pan and Sosnick, 1997). UV absorption data for a titration of the A-riboswitch RNA with MgCl2 are shown in Fig. 21.4A. Kobs can be calculated from the signal and used in Eq. (21.30) to extract DG2þ. With spectroscopic data, calculation of Kobs frequently requires correction for ‘‘baselines’’ at high and low Mg2þ concentrations. The most common origins of baselines are (i) Mg2þ-dependent extinction coefficients for either form of the RNA; (ii) instrument drift during the time of the titration. The two possibilities are easily distinguished by altering the concentration and timing of the titrant schedule, for instance by adding fewer aliquots of larger volume (or higher concentration) over a shorter time span. If instrument drift is a factor, an increase in the elapsed time of the titration should increase the slopes of the baselines. In Fig. 21.4A, the slight positive slope of the 260 nm dataset at high adenine concentrations is attributable to instrument drift, but the larger slope of the 295 nm data is mostly due to the effect of Mg2þ on the RNA extinction at this wavelength.
453
Thermodynamics of Ion–RNA Interactions
A
C 12.0 10.0
295 nm
Γ2+, ions per RNA
Relative absorbance
1.10 1.05 1.00 260 nm
0.95 0.90
10−5
10−4
8.0 6.0
−DAP 4.0 2.0 0.0 10−6
10−3
[Mg ] (M ) 2+
B
D 12.0 ΔΓ2+, ions per RNA
8.0 ln Kobs
10−5 10−4 2+ [Mg ] (M )
10−3
10−5
10−3
3.5 3.0
10.0
6.0 4.0 2.0 0.0
+DAP
2.5 2.0 1.5 1.0 0.5
10−5
10−4 10−3 2+ [Mg ] (M )
10−2
0.0 10−6
[Mg2+]
10−4 (M )
Figure 21.4 Three different measurements of DG2þ for the A-riboswitch RNA, all reported under the same temperature and ionic conditions (50 mM KCl, 20 mM K-MOPS, pH 6.8, 20 C). (A) Isothermal titration of A-riboswitch RNA (3 mM) with buffer containing MgCl2. 11 mM adenine was also present. Absorbances at 260 and 295 nm were collected after each addition. After correction of the absorbance for dilution, the data were normalized to the initial reading before MgCl2 addition. The curves are least squares fits of Eqs. (21.31) and (21.34), with the assumption that mU ¼ mF. (B) ln(Kobs) was calculated from the Tms and DH (average value, 56.3 2.4 kcal/mol) obtained from melting experiments with A-riboswitch RNA in buffer containing 11 mM 2,6-diaminopurine and various concentrations of MgCl2. The curve is a least squares fit of Eq. (21.37). Errors in ln(Kobs) are principally the uncertainty in DH . (C) Direct measurement of G2þ by titration with MgCl2 in the presence of a fluorescent dye that senses Mg2þ activity. The ‘‘þDAP’’ titration contained 250 mM 2,6-diaminopurine. The points and error bars are the averages and standard deviations of five different titrations for each data set. (D) Comparison of DG2þ obtained by three different methods. Data points with black or gray centers are from isothermal titrations observed at 260 and 295 nm, respectively, in which different ligands (purine, adenine, 2-aminopurine, or 2,6-diaminopurine) were used to vary the midpoint of the folding transition. The fitted midpoint of the folding transition, [Mg2þ]0 in Eq. (21.34), is plotted on the x-axis. Errors on both axes are standard
454
Desirae Leipply et al.
To correct for a time-dependent baseline, the data points are simply plotted as a function of the time of addition and a linear fit to the longest time points is subtracted from the dataset. After normalization, the baselinesubtracted dataset is then fit to Eq. (21.34), derived below. As a decision must be made as to which points to include in the linear fit, data must be collected over a long enough time span to identify unambigously a linear portion of the curve. For many datasets, including the 260 nm data in Fig. 21.4A, the correction for instrument drift is small enough that subtraction of either a time-dependent baseline (as just described) or a concentration-dependent baseline (described below, Eq. (21.31)) give the same set of RNA folding parameters, within the reproducibility of the experiment. When the effect of Mg2þ on the RNA extinction must be factored out of the dataset, a single equation that incorporates both baseline(s) and the folding transition can be used. The approach is similar to the methods used to fit baselines and thermodynamic parameters to melting data (Albergo et al., 1981). The equation assumes a linear dependence of the folded and unfolded RNA extinction coefficients on Mg2þ concentration: Abs ¼ ðaU þ mU ½Mg2þ Þð1 yÞ þ ðaF þ mF ½Mg2þ ÞðyÞ
ð21:31Þ
where aU, mU, aF, and mF are the intercepts and slopes of the unfolded and folded RNA baselines, respectively, and y is the fraction of the RNA in the folded form. To obtain an expression for the dependence of y on the Mg2þ concentration, we start with the integrated form of Eq. (21.30) ln Kobs ¼ DG2þ ln½Mg2þ þ C
ð21:32Þ
where C is an integration constant. Note that the integration assumes DG2þ is a constant which may be factored out of the integral; whether this assumption is warranted will be discussed at the end of this section and in Example 3. When Kobs ¼ 1, that is at the midpoint of the folding transition, C takes on the value (DG2þ)ln[Mg2þ]0. The desired expression for Kobs is then ½Mg2þ ln Kobs ¼ DG2þ ln ð21:33Þ ½Mg2þ 0 y is related to Kobs as Kobs ¼ y/(1 y). Making this substitution and solving for y gives deviations from three independent experiments. The solid curve is the calculated slope of the fitted line in panel B; the single error bar is the uncertainty in the slope at the inflection point. Solid gray data points are the difference between the two curves in panel C, with the corresponding cumulative error. Data in panels A, C, and D were taken from Leipply and Draper (2009).
Thermodynamics of Ion–RNA Interactions
y¼
ð½Mg2þ =½Mg2þ 0 ÞDG2þ
1 þ ð½Mg2þ =½Mg2þ 0 ÞDG2þ
455
ð21:34Þ
Both 295 and 260 nm datasets in Fig. 21.4A have been fit to Eqs. (21.31) and (21.34). A problem with these datasets is that the folding transition takes place at low Mg2þ concentrations, leaving insufficient data to determine a lower baseline. Fitting six variables (four variables in Eq. (21.31) and two variables in Eq. (21.34)) gave unrealistic values for mU that were an order of magnitude larger than mF. However, essentially identical values of [Mg2þ]0 and DG2þ are obtained whether mU is fixed as 0 or made to take on the same value as mF. Equation (21.34) can only be used to fit data for which Kobs is small, ≲0.02, under the initial conditions of the experiment ([Mg2þ] ¼ 0). We use a melting experiment under the buffer conditions of the isothermal titration to establish a value for Kobs at the titration temperature. If a significant fraction of the RNA is folded in the absence of Mg2þ, baseline (s) are subtracted by an appropriate method and the data are normalized to the appropriate range of y, y ¼ Kobs/(1 þ Kobs) for the initial zero Mg2þ concentration point and 1.0 for the maximum value of y. The following equation works well for fitting these curves: y¼
½ð½Mg2þ þ C0 Þ=Kn 1 þ ½ð½Mg2þ þ C0 Þ=Kn
ð21:35Þ
where the exponent n, the offset Co, and K are empirical parameters without specific physical significance. When Kobs < 1, the midpoint of the titration curve (y ¼ 0.5) is given by [Mg2þ]0 ¼ K C0. The value of DG2þ at a particular value of [Mg2þ] is extracted from the fitted parameters by taking the derivative specified by Eq. (21.30): y ¼ ð1=KÞn ð½Mg2þ þ C0 Þn 1y @ ln Kobs ½Mg2þ ¼ ¼n @ ln½Mg2þ ½Mg2þ þ C0
Kobs ¼ DG2þ
ð21:36Þ
In contrast to Eq. (21.34), which assumes that DG2þ is a constant over the entire range of fitted Mg2þ concentrations, Eq. (21.36) more realistically allows DG2þ to be a function of [Mg2þ]. The hyperbolic dependence of DG2þ on [Mg2þ] that is assumed by the equation is a good approximation of the experimentally observed dependence, as measured in Example 3 below. The application of Eq. (21.36) to simulated datasets is shown in Fig. 21.5 as part of a discussion of the errors associated with different methods for measuring DG2þ.
456
Desirae Leipply et al.
1.0
q, fraction folded
0.8 0.6 0.4 0.2 0.0 0.01 0.00
Resdiauls
−0.01 0.01 0.00 −0.01 0.01 0.00 −0.01 10−6
10−5
10−4
[Mg2+] (M )
Figure 21.5 Potential errors introduced by the assumption that DG2þ is constant in the derivation of the linkage Eq. (21.30). The dependence of DG2þ on [Mg2þ] was modeled by a polynomial fit to the solid gray data points in Fig. 21.4D. The polynomial was used in the integration of Eq. (21.30) to give expressions for ln Kobs and y with [Mg2þ]dependent DG2þ (in contrast to Eqs. (21.33) and (21.34), which assume DG2þ is constant). y is plotted for the calculated titration curve when the midpoint of the titration, [Mg2þ]0, is 10 mM (circles), 30 mM (squares), or 100 mM (diamonds). DG2þ (as used to calculate the displayed curves) at the titration midpoints (y ¼ 0.5) is 1.45, 2.30, and 2.73, respectively. The simulated data points have been fit to either a modified version of Eq. (21.34) that assumes the y-intercept of the curve has the value y ¼ 0, y ¼ y0 þ ð1 y0 Þð½Mg2þ =½Mg2þ 0 Þn =½1 þ ð½Mg2þ =½Mg2þ 0 Þn , or to an equation that allows a nonzero y-intercept, Eq. (21.35). The residuals of the fits are shown in the lower three panels; closed symbols correspond to Eq. (21.34) and open symbols to Eq. (21.35). For the curve with a midpoint of [Mg2þ]0 ¼ 100 mM, Eq. (21.35) could not be fit to the data because the value of C0 became vanishingly small. The values of DG2þ at the titration curve midpoints obtained from the modified Eq. (21.34) are 1.99, 2.39, and 2.73, in order of increasing [Mg2þ]0. DG2þ obtained by fitting of Eq. (21.35) and application of Eq. (21.36) are 1.43 and 2.29 (10 and 30 mM transition midpoints, respectively).
Thermodynamics of Ion–RNA Interactions
457
By using different purine derivatives as ligands or different concentrations of ligand, the midpoint of the riboswitch folding transition, [Mg2þ]0, was varied by more than an order of magnitude (Fig. 21.4D) (Leipply and Draper, 2009). DG2þ is clearly dependent on [Mg2þ]0, in contrast to the assumption made in the derivation of Eqs. (21.33) and (21.34) that DG2þ is constant. More complete information about the variability of DG2þ is presented in Example 3 of this section, and the question of the errors incurred when DG2þ is obtained by application of Eq. (21.34) is taken up in the subsequent comparison of the different methods for finding DG2þ. Equation (21.34) has the same mathematical form as the Hill equation, which is commonly used to analyze RNA folding data (Fang et al., 1999; Latham and Cech, 1989; Schimmel and Redfield, 1980). It is frequently supposed that the Hill exponent n corresponds to a stoichiometric uptake of ions to defined sites on the RNA, as written in Eq. (21.1). But the derivation of Eq. (21.34) shows that such an interpretation of the Hill exponent is unwarranted. First, the Hill exponent becomes equivalent to DG2þ and takes on a molecular interpretation only when there is an excess of monovalent ion over Mg2þ (Eq. (21.29)), and when DG2þ can be considered a constant (Eq. (21.33)). Second, DG2þ represents a change in the excess number of Mg2þ ions associated with the RNA. When long-range electrostatic interactions are present, it is not possible to identify ions in any particular environment as the ‘‘excess’’ ions; all the ions present in the solution interact with the RNA to some degree. Only at extremely high monovalent salt concentrations or at very low Mg2þ concentrations may long-range Mg2þ–RNA interactions be reduced to the degree that DG2þ approaches the number of site-bound ions (Bukhman and Draper, 1997; Das et al., 2005). 3.2.4. Example 2: DG2 from melting data Melting curves for the A-riboswitch RNA were obtained at Mg2þ concentrations up to 10 mM (Fig. 21.4B), under the same buffer conditions as used for the isothermal titration in the previous example. The tertiary unfolding transition was well resolved from the unfolding of secondary structure up to 1 mM Mg2þ; global fitting of transitions to melting curves observed at 280 and 295 nm allowed deconvolution of the unfolding transitions at the higher Mg2þ concentrations (Draper et al., 2000). Kobs was calculated from the Tms of the melting transitions (Fig. 21.4B) in the same way as done for the data in Fig. 21.2A. According to Eq. (21.30), the slope of the curve in Fig. 21.4B is DG2þ. The middle part of the plot, between Mg2þ concentrations of 50–500 mM, is approximately linear; the slope over this part of the graph gives DG2þ ¼ 2.8 0.1. However, the slope clearly approaches zero at both low and high Mg2þ concentrations. This behavior is expected: at extremely low Mg2þ neither the folded nor the unfolded form interacts significantly with Mg2þ, and at high enough
458
Desirae Leipply et al.
concentrations Mg2þ almost entirely displaces Kþ as the excess cation with both forms of the RNA. The melting experiments are able to access a much larger range of Kobs values than the isothermal titration (Fig. 21.4A), such that the dependence of DG2þ on Mg2þ concentration becomes evident. For this particular RNA, we are able to observe Tm over more than 40 , corresponding to a 5 order of magnitude range in the calculated Kobs. Isothermal titration experiments cannot accurately calculate Kobs over more than a 100-fold range (y varying between 0.1 and 0.9 in Eq. (21.34)), which is usually not enough to detect curvature in the dependence of ln Kobs on ln[Mg2þ]. To work out the dependence of DG2þ on [Mg2þ], we first fit the following curve to the Fig. 21.4B data: ð½Mg2þ =½Mg2þ 0 Þn K max lnðKobs Þ ¼ lnðK0 Þ þ ln ð21:37Þ 1 þ ð½Mg2þ =½Mg2þ 0 Þn K0 This is simply an empirical equation for fitting a smooth curve between the data points, where K0 and Kmax are the values of Kobs in the absence of Mg2þ and the presence of saturating Mg2þ, respectively, [Mg2þ]0 is the midpoint of the curve, and n controls the steepness of the curve at the midpoint. Differentiation of Eq. (21.35) with respect to ln[Mg2þ] yields an expression for DG2þ, 3 2 n 2þ 2þ @ ln Kobs K max 6 ð½Mg =½Mg 0 Þ 7 ¼ n ln DG2þ 4 2 5 2þ @ ln½Mg ½MCl K0 n 2þ 2þ 1 þ ð½Mg =½Mg 0 Þ ð21:38Þ This function is plotted in Fig. 21.4D, using parameters derived from the curve that was fit to the A-riboswitch data in Fig. 21.4B. Comments on the approximations and potential errors of this approach are deferred until after the next example. 3.2.5. Example 3: Direct measurement of G2 It is possible to measure the different single ion interaction coefficients directly by equilibrium dialysis (Bai et al., 2007; Strauss et al., 1967). An alternative method for measuring G2þ takes advantage of a fluorescent dye that chelates Mg2þ (Grilley et al., 2006). In essence, the method uses the dye to sense differences in the Mg2þ activity in the presence or absence of an RNA. Detailed theoretical justification, titration protocols, and data analysis for the method have been presented elsewhere (Grilley et al., 2009). For the A-riboswitch, titrations carried out in the presence or absence of ligand measure G2þ for the folded or unfolded state of the RNA, respectively
Thermodynamics of Ion–RNA Interactions
459
(Fig. 21.4C). As expected, G2þ is always larger for the folded RNA. The difference between these two curves is DG2þ (Fig. 21.4D), the identical thermodynamic quantity as obtained in the previous two examples but measured without recourse to a linkage relation (Eq. (21.30)). 3.2.6. Comparison of three different DG2 measurement methods Three methods for measuring DG2þ have been described in the preceding examples. Two of the methods depend on a linkage relation (Eq. (21.30)) but find Kobs in different ways; the third method is based on the direct measurement of G2þ for the two RNA conformations under consideration. The error bars in the Fig. 21.4D comparison of the sets of DG2þ values illustrate the difficulty in quantitating DG2þ very accurately by any method, but there are some systematic differences between the measurements (particularly at higher concentrations of Mg2þ) which most likely reflect the different assumptions made by each method. Here we first summarize the assumptions that go into the formulas and analyses, and then comment on the merits and drawbacks of each approach. Assumptions underlying measurements of DG2þ. Derivation of the linkage Eq. (21.30) and its integrated form Eq. (21.34) requires three assumptions:
that the folding transition being observed is two-state (Eq. (21.19)); that DG2þ is constant over the range of Mg2þ concentrations being analyzed; that sufficient excess monovalent salt is present to suppress changes in the Mg2þ ion activity coefficient as MgCl2 is titrated (Eq. (21.29)). This condition is an issue only at the highest concentration of MgCl2 reached in Fig. 21.4C. The isothermal UV titrations depend further on the assumption
that titration baseslines have a linear dependence on [Mg2þ].
The derivation of ln Kobs from UV melting profiles depends on the extrapolation of ln Kobs to a temperature different than the Tm using a value of DH
that is derived from the melting profile using the ‘‘two-state’’ assumption (i.e., only the folded (N) and unfolded (U) forms are ever present in significant concentrations); that is constant with temperature (i.e., DCp 0). In addition, comparison of data obtained by isothermal titration and melting profile methods must assume
that the structure of the unfolded state does not change with temperature.
Derivation of DG2þ from direct measurements of G2þ (Example 3) only requires the same assumption of excess monovalent salt as needed for the
460
Desirae Leipply et al.
other two methods. The calculation does not invoke the two-state assumption, as measurements on folded and unfolded RNAs are made separately. Advantages and disadvantages of the different measurement methods. In the application of linkage equations to folding equilibria, it can be difficult to know how much error is introduced by the two-state assumption. For the isothermal titrations, the difference in transition midpoints obtained when folding is monitored at 260 or 295 nm (Fig. 21.3A) suggests that the A-riboswitch may deviate from a strict two-state model. The presence of a significant concentration of intermediate RNA form(s) during the titration will generally broaden the curve and thereby reduce the apparent value of DG2þ. The tendency of the titration DG2þ values to land below those obtained by direct measurement (Fig. 21.4D) is possibly another indication that folding intermediates are present in the isothermal titrations. A second uncertainty associated with linkage equations is the degree to which the Mg2þ concentration dependence of DG2þ might affect the derived value of DG2þ. To see how large these potential errors might be, we simulated Mg2þ titration curves (y, the fraction of folded RNA, vs. [Mg2þ]) using an expression derived by integration of Eq. (21.30) with a polynomial function for DG2þ that approximated the [Mg2þ]-dependence of DG2þ displayed in Fig. 21.4D (solid gray data points). When the midpoint (where Kobs ¼ 1) of a simulated titration was set to 30 or 100 mM Mg2þ, a fit of Eq. (21.34) to the data points gave values of DG2þ that were very close to the actual values reached at the titration midpoints (see legend to Fig. 21.5). The Eq. (21.34) parameters are very sensitive to the data in the vicinity of y ¼ 0.5 and tend to report the slope at this point correctly, even when the fitted curve shows systematic deviations from the data (see plots of residuals in the lower panels of Fig. 21.5). This bias of the equation makes it surprisingly reliable for extracting DG2þ from titration data, though it should be emphasized that the fitted value of DG2þ applies only to the Mg2þ concentration at the midpoint of the titration curve. There is a caveat in using Eq. (21.34), illustrated by the simulated titration curve at 10 mM Mg2þ (Fig. 21.5). This titration curve shows a significant fraction of folded RNA in the absence of Mg2þ, where Kobs 0.15. Spectroscopic titrations and most other methods used to assess Kobs assume a baseline value of y ¼ 0 when [Mg2þ] ¼ 0 and normalize the titration curve accordingly. Two of the calculated curves in Fig. 21.5 have been fit to a modified Hill equation that incorporates this assumption. The apparent value of DG2þ at the midpoint of one curve (1.99) is larger than the value used to calculate the titration curve (1.45; see legend to Fig. 21.5). Because of the potential source of error just noted, it is advisable to have independent confirmation of Kobs in the absence of Mg2þ. For the A-riboswitch RNA, moderate concentrations of the most tightly binding ligand (2,6-diaminopurine) do in fact stabilize the native structure to some degree at 20 C; from the melting experiments used to generate Fig. 21.4B,
Thermodynamics of Ion–RNA Interactions
461
Kobs ¼ 1.3 when Mg2þ was omitted from the buffer. Equation (21.35) allows y > 0 when [Mg2þ] ¼ 0; when this equation was used to fit the simulated data in Fig. 21.5, much better fits to the data were obtained than with Eq. (21.34) (see residuals plotted in the lower panels of Fig. 21.5), and the DG2þ value calculated from the fitted parameters reproduced the values used to generate the curves. Direct measurements of G2þ are, in principle, the most reliable and informative way to obtain DG2þ; the two-state assumption is not required and the [Mg2þ]-dependence of DG2þ is observed. However, a drawback of the direct measurements is that the errors tend to be large: because DG2þ is a small difference between two large numbers, errors in the measurement of the individual G2þ are amplified. A second limitation is obtaining the RNA in both conformations of interest. Some RNAs do not adopt the folded state in the absence of Mg2þ, which limits the concentration range over which DG2þ can be obtained (Grilley et al., 2006). The A-riboswitch is particularly convenient in that the unfolded state can be obtained simply by omitting the purine ligand. For other RNAs it is necessary to use a mutant that is incapable of folding as a model for the unfolded state (Grilley et al., 2007; Soto et al., 2007); the mutations may perturb the ensemble of unfolded states and bias the measurements. Melting experiments have an advantage over isothermal titrations in that they are able to provide an absolute measurement of Kobs (there is no assumption that y ¼ 0 when [Mg2þ] ¼ 0) and extend the measurement of Kobs over several orders of magnitude, enough to detect the Mg2þ concentration dependence of DG2þ. However, uncertainties may arise in calculating Kobs from Tm and DH values. The partially unfolded state of the RNA is likely to be increasingly unstacked as the Tm of the RNA increases, and result in a temperature-dependent DH . This phenomenon has been observed as an apparent heat capacity change in reactions involving singlestranded polynucleotides (Ferrari and Lohman, 1994), but has not been thoroughly explored for RNA tertiary folding reactions. In the case of the melting data presented in Fig. 21.4B, there was no significant correlation of DH with Tm.
ACKNOWLEDGMENT This work was supported by NIH grant GM58545.
REFERENCES Albergo, D. D., Marky, L. A., Breslauer, K. J., and Turner, D. H. (1981). Thermodynamics of (dG-dC)3 double-helix formation in water and deuterium oxide. Biochemistry 20, 1409–1413.
462
Desirae Leipply et al.
Anderson, C. F., Felitsky, D. J., Hong, J., and Record, M. T. (2002). Generalized derivation of an exact relationship linking different coefficients that characterize thermodynamic effects of preferential interactions. Biophys. Chem. 101–102, 497–511. 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. Bukhman, Y. V., and Draper, D. E. (1997). Affinities and selectivities of divalent cation binding sites within an RNA tertiary structure. J. Mol. Biol. 273, 1020–1031. Chang, K. Y., and Tinoco, I., Jr. (1994). Characterization of a "kissing" hairpin complex derived from the human immunodeficiency virus genome. Proc. Natl. Acad. Sci. USA 91, 8705–8709. 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. Draper, D. E. (2008). RNA folding: Thermodynamic and molecular descriptions of the roles of ions. Biophys. J. 95, 5489–5495. Draper, D. E., Bukhman, Y. V., and Gluick, T. C. (2000). Thermal Methods for the Analysis of RNA Folding Pathways. In ‘‘Current Protocols in Nucleic Acid Chemistry,’’ (S. L. Beaucage, D. E. Bergstrom, G. D. Glick, and R. A. Jones, eds.), Wiley, New York, section 11.13. Eisenberg, H. (1976). Biological Macromolecules and Polyelectrolytes in Solution. Clarendon Press, Oxford. Fang, X., Pan, T., and Sosnick, T. R. (1999). A thermodynamic framework and cooperativity in the tertiary folding of a Mg2þ-dependent ribozyme. Biochemistry 38, 16840–16846. Ferrari, M. E., and Lohman, T. M. (1994). Apparent heat capacity change accompanying a nonspecific protein-DNA interaction. Eschericia coli SSB tetramer binding to oligodeoxyadenylates. Biochemistry 33, 12896–12910. Garcı´a-Garcı´a, C., and Draper, D. E. (2003). Electrostatic interactions in a peptide-RNA complex. J. Mol. Biol. 331, 75–88. 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. Grilley, D., Misra, V., Caliskan, G., and Draper, D. E. (2007). Importance of partially unfolded conformations for Mg(2þ)-induced folding of RNA tertiary structure: Structural models and free energies of Mg(2þ) interactions. Biochemistry 46, 10266–10278. Grilley, D., Soto, A. M., and Draper, D. E. (2009). Direct quantitation of Mg2þ–RNA interactions by use of a fluorescent dye. Methods Enzymol. 455, 71–94. Harned, H. S., and Robinson, R. A. (1968). Multicomponent Electrolyte Solutions Pergamon Press, Oxford. Koculi, E., Hyeon, C., Thirumalai, D., and Woodson, S. A. (2007). Charge density of divalent metal cations determines RNA stability. J. Am. Chem. Soc. 129, 2676–2682. Lambert, D., and Draper, D. E. (2007). Effects of osmolytes on RNA secondary and tertiary structure stabilities and RNA-Mg2þ interactions. J. Mol. Biol. 370, 993–1005. Lambert, D., Leipply, D., Shiman, R., and Draper, D. E. (2009). The influence of monovalent cation size on the stability of RNA tertiary structures. J. Mol. Biol. 390, 791–804. Latham, J. A., and Cech, T. R. (1989). Defining the inside and outside of a catalytic RNA molecule. Science 245, 245–276. Leipply, D., and Draper, D. E. (2009). The dependence of RNA tertiary structure stability on Mg(2þ) concentration: Interpretation of the Hill equation and coefficient (submitted). Levine, I. N. (2002). Physical Chemistry. McGraw Hill, New York.
Thermodynamics of Ion–RNA Interactions
463
Marcus, Y. (1994). A simple empirical model describing the thermodynamics of hydration of ions of widely varying charges, sizes, and shapes. Biophys. Chem. 51, 111–127. Millero, F. J. (1971). The molal volumes of electrolytes. Chem. Rev. 71, 147–176. Misra, V. K., and Draper, D. E. (1999). The interpretation of Mg2þ binding isotherms for nucleic acids using Poisson-Boltzmann theory. J. Mol. Biol. 294, 1135–1147. Pan, T., and Sosnick, T. R. (1997). Intermediates and kinetic traps in the folding of a large ribozyme revealed by circular dichroism and UV absorbance spectroscopies and catalytic activity. Nat. Struct. Biol. 4, 931–938. Ralston, C. Y., He, Q., Brenowitz, M., and Chance, M. R. (2000). Stability and cooperativity of individual tertiary contacts in RNA revealed through chemical denaturation. Nat. Struct. Biol. 7, 371–374. Record, M. T., Jr., and Richey, B. (1988). In ‘‘ACS sourcebook for physical chemistry instructors,’’ (E. T. Lippincott, ed.), pp. 145–159. American Chemical Society, Washington, DC. Record, M. T., Jr., Zhang, W., and Anderson, C. F. (1998). Analysis of effects of salts and uncharged solutes on protein and nucleic acid equilibria and processes: A practical guide to recognizing and interpreting polyelectrolyte effects, Hofmeister effects, and osmotic effects of salts. Adv. Protein Chem. 51, 281–353. Robinson, C. V., and Stokes, R. H. (2002). Electrolyte Solutions. Second Revised Edition. Dover Publications, Minoeola, NY. Schimmel, P. R., and Redfield, A. G. (1980). Transfer RNA in solution: Selected topics. Ann. Rev. Biophys. Bioeng. 9, 181–221. Serganov, A., Yuan, Y. R., Pikovskaya, O., Polonskaia, A., Malinina, L., Phan, A. T., Hobartner, C., Micura, R., Breaker, R. R., and Patel, D. J. (2004). Structural basis for discriminative regulation of gene expression by adenine- and guanine-sensing mRNAs. Chem. Biol. 11, 1729–1741. Soto, A. M., Misra, V., and Draper, D. E. (2007). Tertiary structure of an RNA pseudoknot is stabilized by "diffuse" Mg(2þ) ions. Biochemistry 46, 2973–2983. Stein, A., and Crothers, D. M. (1976). Conformational changes of transfer RNA. The role of magnesium(II). Biochemistry 15, 160–167. Strauss, U. P., Helfgott, C., and Pink, H. (1967). Interactions of polyelectrolytes with simple electrolytes. II. Donnan equilibria obtained with DNA in solutions of 1–1 electrolytes. J. Phys. Chem. 71, 2550–2556. Wyman, J. (1948). Heme proteins. Adv. Protein Chem. 4, 407–531. Wyman, J., Jr. (1964). Linked functions and reciprocal effects in hemoglobin: A second look. Adv. Protein Chem. 19, 223–286.
C H A P T E R
T W E N T Y- T W O
Predicting Electrostatic Forces in RNA Folding Zhi-Jie Tan* and Shi-Jie Chen†,‡ Contents 1. Introduction 2. Overview of Experimental Results for the Ion-Dependence of RNA Thermal Stability 2.1. General properties of ion binding 2.2. Helix–helix assembly 2.3. RNA tertiary structures and RNA–RNA complexes 3. Overview of Theoretical Methods for Predicting Ion Electrostatics for RNA Folding 3.1. Counterion condensation theory 3.2. Poisson–Boltzmann theory 3.3. Modified Poisson–Boltzmann theories 4. Tightly Bound Ion Model 4.1. General formalism of the theory 4.2. Free energy of the tightly bound ions 4.3. Free energy of the diffusive ions 4.4. Polarization energy of charges inside the tightly bound region 5. Enhancing the Computational Efficiency for the Numerical Calculations of the TBI Model 5.1. A hybrid treatment for the tightly bound modes 5.2. Computation of the free energy DGd for the diffusive ions 6. Applications of the TBI Model 6.1. General procedure for the numerical computations for the partition function 6.2. Ion-binding properties 6.3. Folding stability 7. Summary Acknowledgments References
466 467 470 470 471 471 471 472 472 474 474 475 476 476 477 477 477 478 478 480 481 482 484 484
* Department of Physics, Wuhan University, Wuhan, Hubei, China Department of Physics and Astronomy, University of Missouri, Columbia, Missouri, USA Department of Biochemistry, University of Missouri, Columbia, Missouri, USA
{ {
Methods in Enzymology, Volume 469 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)69022-4
#
2009 Elsevier Inc. All rights reserved.
465
466
Zhi-Jie Tan and Shi-Jie Chen
Abstract Metal ion-mediated electrostatic interactions are critical to RNA folding. Although considerable progress has been made in mechanistic studies, the problem of accurate predictions for the ion effects in RNA folding remains unsolved, mainly due to the complexity of several potentially important issues such as ion correlation and dehydration effects. In this chapter, after giving a brief overview of the experimental findings and theoretical approaches, we focus on a recently developed new model, the tightly bound ion (TBI) model, for ion electrostatics in RNA folding. The model is unique because it can treat ion correlation and fluctuation effects for realistic RNA 3D structures. For monovalent ion (such as Naþ) solutions, where ion correlation is weak, TBI and the Poisson–Boltzmann (PB) theory give the same results and the results agree with the experimental data. For multivalent ion (such as Mg2þ) solutions, where ion correlation can be strong, however, TBI gives much improved predictions than the PB. Moreover, the model suggests an ion correlation-induced mechanism for the unusual efficiency of Mg2þ ions in the stabilization of RNA tertiary folds. In this chapter, after introducing the theoretical framework of the TBI model, we will describe how to apply the model to predict ion-binding properties and ion-dependent folding stabilities.
1. Introduction Because RNA backbone is highly charged, the formation of a compact RNA structure is accompanied by massive charge build-up. How does an RNA molecule overcome the strong charge–charge repulsion to form a stable structure? How do metal ions facilitate the folding process via electrostatic interactions? How to accurately predict the electrostatic forces for different ions and ion concentrations? The answers to these questions are essential for understanding and predicting the stability, cooperativity, and kinetics for RNA folding (Chen, 2008; Chu and Herschlag, 2008; Draper, 2008; Walter et al., 2009). Mechanistic studies for RNA folding have revealed many important features of the ion-dependence of RNA folding. Adding metal ions into the solution would cause an increased ion binding and a reduction in backbone charge repulsion, which would lead to a higher flexibility of the RNA chain and a higher stability of RNA helices. An important consequence of the reduced charge repulsion is that it allows more frequent close approaches between the different RNA segments, causing the tertiary contacts to be kinetically accessible and thermodynamically stable. Further experimental and theoretical studies for specific systems have pointed to the importance of ion charge density (ion charge/ion size), in addition to the ion concentration, as a determinant for the ion effects in RNA stability
RNA-Ion Interactions
467
(Heilman-Miller et al., 2001; Koculi et al., 2007; Takamoto et al., 2002; Weixlbaumer et al., 2004). Mg2þ ions play an essential role in RNA tertiary structure folding (Chen, 2008; Chu et al., 2007; Draper, 2008; Walter et al., 2009). Extensive experimental studies have led to the conclusion that Mg2þ ions are unusually efficient than Naþ in stabilizing RNA tertiary folds. The effect of Mg2þ ions is unusual because it cannot be explained simply by the larger ionic strength of Mg2þ than Naþ. Although considerable progress has been made in mechanistic studies for the Mg2þ effect, owing to the complexity of the electrostatic interactions involving Mg2þ ions, many problems about quantitative predictions for the Mg2þ ion-mediated electrostatic interactions in RNA folding remain unsolved. For instance, how does [Mg2þ] modulate the stability of specific intermediate structures? How do the ion size, charge, and charge density determine the ion-specific folding energy landscape? How to quantify the ion dehydration and ion correlation effects in Mg2þ– RNA interactions? In this chapter, we describe a recently developed theory, namely, the tightly bound ion (TBI) theory (Tan and Chen, 2005), that aims to tackle these effects. In parallel to the investigations on the ion-dependence of RNA folding, systematic thermodynamic measurements for simple RNA folding systems at a fixed ionic condition (1 M NaCl) have led to quantitative insights into RNA folding energetics (Mathews and Turner, 2006; Serra and Turner, 1995). However, using the thermodynamic data at 1 M NaCl to predict RNA folding at a general ionic condition, such as the physiological solution condition, requires a theory for the ion effects, such as the TBI theory to be described in this chapter. In this chapter, we will give a brief overview for the experimental findings and the main theoretical approaches for the ion effects in RNA and DNA thermal stabilities. We will then describe the TBI theory. We will focus on both the theoretical formalism and the practical applications of the theory. Our aim here is to provide sufficient detail so that all the major issues in the theoretical derivations and numerical computations can be clearly understood and readily followed.
2. Overview of Experimental Results for the Ion-Dependence of RNA Thermal Stability Extensive experimental studies (see Table 22.1 for a brief summary) on the ion effects for a broad range of nucleic acids, from small hairpins and pseudoknots to large tetrahymena group I intron, have led to several important conclusions as well as quantitative estimations for the ion-mediated electrostatic forces in RNA folding, as described below.
468 Table 22.1 Measurements for the nucleic acids thermodynamics in ionic solutionsa Nucleic acids
References
Ionic conditions
Thermodynamic quantities
24-bp DNA duplex
Bai et al. (2007) Bai et al. (2007) Bai et al. (2007) Bai et al. (2007) Krakauer (1971) Clement et al. (1973) Tan and Chen (2007) Tan and Chen (2006a) Tan and Chen (2008b) Owczarzy et al. (2004) Vieregg et al. (2007) Soto et al. (2007) Bukhman and Draper (1997) Grilley et al. (2007) Romer and Hach (1975) Rialdi et al. (1972) Weixlbaumer et al. (2004) Lorenz et al. (2006) Nixon and Giedroc (1998) Theimer and Giedroc (2000)
Mixed Naþ/Mg2þ Mixed monovalent ions Mixed divalent ions Mixed Naþ/Mg2þ Mixed Naþ/Mg2þ Mixed Naþ/Mg2þ Naþ, Mg2þ Naþ, Mg2þ Naþ, Mg2þ Naþ Kþ, Naþ Naþ, mixed Naþ/Mg2þ Mixed NHþ/Mg2þ Mixed Kþ/Mg2þ Mixed Naþ/Mg2þ Mixed Naþ/Mg2þ Naþ, mixed Naþ/Mgþ Naþ, mixed Naþ/Mg2þ Kþ, mixed Kþ/Mg2þ Naþ, mixed Kþ/Mg2þ
fNaþ , fMg2þ f’s f’s fNaþ , fMg2þ fMg2þ fMg2þ DG, Tm DG, Tm DG, Tm Tm DG DH, DS, fMg2þ DG fMg2þ DG, fMg2þ fMg2þ Tm Tm DG Tm
24-bp DNA triplex Poly(AU) Calf thymus DNA RNA duplexesb DNA duplexesb DNA and RNA hairpinsb DNA duplexes (10–30 bp) RNA hairpins (49–124 nt) BWYV pseudoknotc 58-nt rRNA Yeast tRNAphe HIV-1type kiss complex 1 HIV-1type kiss complex 2 T2 pseudoknot MMTV pseudoknotd
T4-35 pseudoknot T4-32 pseudoknot T4-28 pseudoknot Tetrahymena ribozyme
Hairpin ribozyme
A RNA three-way junction a b c d
Qiu et al. (1996) Qiu et al. (1996) Qiu et al. (1996) Heilman-Miller et al. (2001) Takamoto et al. (2002) Koculi et al. (2007) Walter et al. (1999) Pljevaljcic et al. (2004) Bokinsky et al. (2003) Kim et al. (2002)
Mixed Naþ/Mg2þ Mixed Naþ/Mg2þ Mixed Naþ/Mg2þ Kþ, Naþ, Mg2þ, spermidine Naþ, Mg2þ Mg2þ, Ca2þ, Sr2þ, Ba2þ Mg2þ, Naþ Mg2þ Naþ, Mg2þ Naþ, Mg2þ
DG, Tm DG, Tm DG, Tm Fraction folded RH Fraction folded Fraction folded Fraction folded Fold and unfold rates Fold and unfold rates
The table summarizes the ion-dependent thermodynamics of nucleic acid secondary and tertiary structures, including ion-binding numbers f, folding free energy DG, melting temperature Tm, Stokes radius RH, fraction folded, and other folding properties. The ion-dependent thermodynamic data and corresponding experimental references for the various DNA duplexes, RNA duplexes, and DNA and RNA hairpins are collected/summarized in (Tan and Chen, 2006a, 2007, 2008b and references therein). BWYV: beet western yellow virus. MMTV: mouse mammary tumor virus.
469
470
Zhi-Jie Tan and Shi-Jie Chen
2.1. General properties of ion binding Multivalent ions such as Mg2þ ions are much more efficient than monovalent ions to neutralize RNA backbone charges and to screen the Coulombic repulsions (Draper, 2008). For example, 10 mM Mg2þ is approximately equivalent to 1 M Naþ in stabilizing short DNA and RNA oligomers (Tan and Chen, 2006a, 2007, 2008b; and references therein). Such unusually high efficiency of Mg2þ ions goes beyond the mean-field ionic strength effect. Furthermore, the high efficiency of Mg2þ is more pronounced for more compact/complex structures. For example, for a simple 24-bp DNA duplex, 0.4 mM Mg2þ is equivalent to 20 mM Naþ in ionic neutralization, that is, 0.4 mM Mg2þ and 20 mM Naþ can cause the same neutralization for the phosphate charges. In contrast, the 76-nt yeast tRNAphe, a more complex structure, 0.4 mM Mg2þ is equivalent to 32 mM, a much higher Naþ concentration (Bai et al., 2007; Romer and Hach, 1975).
2.2. Helix–helix assembly Ions can mediate nucleic acid helix–helix interactions and consequently affect the structure and stability of the helix assembly. Osmotic pressurebased experiments (Rau and Parsegian, 1992a,b) show that multivalent ions such as Co3þ can cause attraction between DNA helices (Rau and Parsegian, 1992a), while monovalent ions such as Naþ can only modulate the strength of helix–helix repulsion. For divalent ions, the experiments show that some ions such as Mn2þ can induce helix–helix attractive force, while others such as Ca2þ cannot. Mg2þ ions, with the addition of methanol in the solution, can cause attraction between DNA helices (Rau and Parsegian, 1992a). The different effects of the divalent ions may be attributed to the (solvent-mediated) ion-binding affinity to different groups of DNA. Small angle X-ray scattering (SAXS) experiments for a system with two short helices tethered by a neutral loop indicate that high-concentration ions (including 1þ, 2þ, and 3þ ions) cause a random conformational state (Bai et al., 2005), where Coulomb repulsion between the helices is relaxed. Furthermore, quantitative measurements suggest that the attraction is 0.21 kBT/bp between two 24-bp helices in a 0.6 M Mg2þ solution (Bai et al., 2005). Further measurements on the Mg2þ-induced folding of the system show that [Mg2þ] at the transition point is over 10 times lower than that predicted from the Poisson–Boltzmann theory (PB) (Bai et al., 2008). Because PB neglects ion correlation effect, the experimental finding suggests the possibility for ion correlation to be responsible for the unusually high efficiency of the Mg2þ ions.
RNA-Ion Interactions
471
2.3. RNA tertiary structures and RNA–RNA complexes One of the central issues for the ion effects in RNA tertiary structure folding is the unusually efficient role of Mg2þ ions. For example, for a short DNA/ RNA helix, 10 mM Mg2þ and 1 M Naþ can cause similar folding stabilities (Tan and Chen, 2007), while for a complex RNA tertiary structure such as tetrahymena ribozyme, a much lower concentration 0.5 mM of Mg2þ would be sufficient to induce folding transition as 0.5 M Naþ (HeilmanMiller et al., 2001). Even with high-concentration monovalent ion background, Mg2þ ions can make significant contribution to the folding stability of RNA tertiary structures. For example, for a 58-nt ribosomal RNA fragment, in the background of 1.6 M monovalent ions, 0.05 M Mg2þ can contribute about 5 kcal/mol to the global tertiary folding stability (Bukhman and Draper, 1997). Thermodynamic measurements for HIV-1 dimerization initiation signal (DIS) kissing complexes (Lorenz et al., 2006; Weixlbaumer et al., 2004) show that, compared to the corresponding duplex of the same sequence at the kissing interface, (a) the melting temperature Tm for kissing complexes increases with ion (Naþ and Mg2þ) concentrations much more rapidly and (b) the high efficiency of Mg2þ over Naþ is much more pronounced. Depending on the sequence, for a kissing complex, Mg2þ may bind to the specific site at the kissing interface. The specific binding can affect the [Mg2þ]-dependence of stability of the kissing complex (Lorenz et al., 2006).
3. Overview of Theoretical Methods for Predicting Ion Electrostatics for RNA Folding 3.1. Counterion condensation theory The counterion condensation (CC) theory was originally developed to treat an infinitely long line-charge system in an infinitely dilute ionic solution (Manning, 1978). The theory has led to several widely used conclusions such as the linear dependence of melting temperature Tm of nucleic helix on the logarithm of salt concentration (Manning, 1978). However, application of the CC theory to the interactions between two DNA helices predicts an attractive force between the helices even in a monovalent ion solution. This prediction as well as the predicted dependence of the attractive force on the ion charge (Ray and Manning, 2000) are not consistent with the experimental measurements (Qiu et al., 2007; Rau and Parsegian, 1992a,b) and computer simulations (Luan and Aksimentiev, 2008; Lyubartsev and Nordenskiold, 1995).
472
Zhi-Jie Tan and Shi-Jie Chen
3.2. Poisson–Boltzmann theory The history of PB theory can be traced back to the Gouy–Chapmann theory and Debye–Huchel theory in the early of 1900s (e.g., see Carnie and Torrie, 1984). These two theories represent special simplified forms of the PB theory: Gouy–Chapmann theory is a one-dimensional simplification for electric double-layer, while the Debye–Huchel theory is a special solution for spherical symmetric system. The PB equation can be derived based on the Poisson equation with a self-consistent mean electric potential c and a Boltzmann distribution for the ions ( ) X re0 ercðrÞ ¼ 4p rf þ ð22:1Þ za eca0 eza ecðrÞ=kB T ; a
where zaec is the electrostatic energy for an ion of species a with ionic charge zce in the mean electric potential c. e is the dielectric constant, rf is the charge density of fixed charges, and ca0 is the bulk concentration of ion species a. With proper boundary conditions, the solution of the PB equation gives the electric potential and the ion distribution for realistic nucleic acid structures (Baker et al., 2000; Gilson and Zhou, 2007; Gilson et al., 1987). PB equation is based on the mean-field approximation, where ions are treated as continuous fluid-like particles moving independently in a mean electric potential. The theory ignores the discrete ion properties such as ion size, ion–ion correlation and ion fluctuations. Fail to consider these properties can cause inaccurate predictions for RNA folding, especially in the presence of multivalent ions which are prone to ion correlation due to the strong, long-range Coulomb interactions. For example, PB cannot predict the experimentally observed attractive force between DNA helices in multivalent ion solutions.
3.3. Modified Poisson–Boltzmann theories Several attempts have been made to improve the PB by accounting for the effects of ion size, ion correlation, and fluctuations of ion distributions. The resultant modified models have led to improved predictions for the ion effects in RNA/DNA folding stability. 3.3.1. Size-modified Poisson–Boltzmann theory Motivated by the need to consider finite size of ions in the PB model, Chu et al. employed a lattice gas model for the ionic system, where ions of finite sizes are placed on the grid cells. In this way, the ion size can be conveniently represented by the cell size and the system can be treated with the lattice gas approximation (Chu et al., 2007). The modified PB theory
RNA-Ion Interactions
473
gives notable improvements in predictions for ion-binding properties of monovalent counterions, especially at high salt concentration, which involves saturation effect for ion binding. However, for nucleic acid solution containing multivalent ions, the model may not be accurate because the model, like PB, does not treat long-range interion correlations (Chu et al., 2007). 3.3.2. Modified PB based on Kirkwood/BBGY hierarchy Based on the Kirkwood/BBGY hierarchy, a fluctuation potential and an ion-size-exclusion term in the potential of mean force are used to represent the effect of ion correlations (Carnie and Torrie, 1984; Grochowski and Trylska, 2008). The fluctuation potential is associated with the energy for charging an ion, therefore, it implicitly takes into account the interion Coulomb correlations. The theory gives improved predictions for multivalent ion distributions near macroions of ideal shapes such as cylinder, sphere or plane (Carnie and Torrie, 1984). However, for realistic nucleic acids structures, the three-dimensional numerical solution requires exceedingly large computational cost. This is because the effective fluctuation potential itself is coupled to the electrostatic potential. As a result, the theory is computationally impractical for applications to realistic nucleic acid structures (Gavryushov, 2008). 3.3.3. Correlation-corrected Poisson–Boltzmann model Recently, Forsman developed a correlation-corrected PB model by introducing an effective potential between like-charge ions (Forsman, 2007). The effective potential at large ion–ion separation approaches the classical Coulomb potential and becomes a reduced effective repulsive Coulomb potential for small ion–ion separation. Such an effective potential represents liquid-like correlation behavior between the ions. For electric double layer with multivalent ions, the model makes improved predictions for the ion distribution and predicts an attractive force between two planes in the presence of multivalent ions (Forsman, 2007). However, for realistic nucleic acid structures, the model is computationally expensive. In addition, the ad hoc effective potential lacks validation for realistic nucleic acid structures. In addition to the above modified PB models, theories based on other approaches such as the density functional theory (e.g., Wu and Li, 2007) and the integral equation theory (e.g., Vlachy, 1999) have been developed to account for the interion correlation effects. However, the computational complexity for these approaches prohibits efficient applications to realistic RNA 3D structures. Recently, inspired by the experimental findings for the significance of ion size, ion correlation and fluctuations, the TBI model (Tan and Chen, 2005, 2008a) was developed. Comparisons between the TBI theory predictions and the experimental data on ion-binding properties (Stellwagen et al., 2007;
474
Zhi-Jie Tan and Shi-Jie Chen
Tan and Chen, 2005), thermal stabilities of DNA and RNA helices and hairpins in Naþ and Mg2þ solutions (Tan and Chen, 2006a, 2007, 2008b), and electrostatic energy landscapes for helix bending (Tan and Chen, 2008a) and helix assemblies (Tan and Chen, 2006b,c) indicate that the TBI model gives much improved predictions than previous models such as PB. In the following, we will focus on the TBI model and describe the theoretical framework as well as the details in numerical computations and applications of the model.
4. Tightly Bound Ion Model 4.1. General formalism of the theory Ion binding around a (polyanionic) nucleic acid structure causes counterion accumulation on the molecular surface. The strong, long-range Coulomb force between the ions cause the ions to become correlated (networked). Mathematically, the strong correlation can be characterized by the large correlation parameter G G¼
ðzeÞ2 Gc ; eakB T
ð22:2Þ
where z is the valency of the ion, e is the electronic charge, e is the dielectric constant of the solvent (’78 for aqueous solvent), a is the Wigner–Seitz distance between the ions, and kB ¼ 1.99 cal/mol K is the Boltzmann constant and T is the temperature. Gc ¼ 2.6 is the critical value for the gas–liquid transition for an ionic system (e.g., Ichimaru et al., 1987). Control tests for different Gc values suggest that the model is quite robust against small variations of Gc. The correlation strength distinguishes two types of ions: the tightly bound ions (TBIs) (G Gc, strong correlation) and the diffusive ions (G < Gc, weak correlation). The tightly bound and the diffusive ions are distributed in the vicinity of the nucleic acid (tightly bound region) and in the outer region (diffusive region), respectively. A central point in the TBI model is to treat the diffusive ions using the PB method while treating the TBIs with discrete ion distributions. A direct implication of the ion correlation is that the likelihood of finding an ion at a location is sensitive to the locations of other ions. Therefore, to treat the ion correlation effect, we must consider an ensemble of discrete ion distributions (i.e., fluctuations) (Ha and Thirumalai, 2003). In order to enumerate the different distributions for the TBIs, we divide the tightly bound region into discrete phosphate cells, each around a phosphate group. We generate the ensemble of distributions for the TBIs by
475
RNA-Ion Interactions
enumerating the different ways of partitioning the TBI among the different cells. We call each such distribution of the TBIs as a mode. The total partition function Z of the system is the sum of the partition function ZM for each mode M X Z¼ ZM : ð22:3Þ M
A mode is defined by a set of numbers (not the coordinates) of the TBIs in the different cells. The partition function ZM is given by the average over all the different spatial coordinates of the ions for a given mode M: ! ðY Nb pol ZM ¼ Z ðidÞ ðcz ÞNb dRi eðDGb þDGd þDGb Þ=kB T ; ð22:4Þ i¼1
Z(id )
is the partition function for the uniform ion solution (without where the nucleic acid), Nb is the total number of the TBIs, cz is the bulk concentration ofÐ the counterions, and Ri denotes the coordinate QNz-valent b of the ith TBI. dR is the volume integral over the tightly bound i i¼1 pol region for the TBIs. DGb and DGb are the Coulombic free energy and the (Born) self-polarization energy for the charges in the tightly bound region and DGd is the free energy for the interaction between the diffusive ions and between the diffusive and the TBIs (Tan and Chen, 2008a). From the partition function, we can predict the ion distribution from the probability distribution pM ¼
ZM Z
of the ion-binding modes and the electrostatic free energy of the system: X ZM Z E G ¼ kB T ln ðidÞ ¼ kB T ln : ð22:5Þ Z Z ðidÞ M
4.2. Free energy of the tightly bound ions Consider a TBI in a phosphate cell. Averaging over the possible positions in the respective phosphate cells for the TBIs (Tan and Chen, 2005, 2006c, 2008a) gives two useful potentials of mean forces: Fl(i ) for the interactions between charges within the cell i and F2(i, j ) for the interactions between different cells i and j: F1 ðiÞ ¼ kB T lnheuii ðRi Þ=kB T i;
F2 ði; jÞ ¼ kB T lnheuij ðRi ;Rj Þ=kB T i: ð22:6Þ
Here uii and uij are the Coulomb (generalized Born) potentials (Liu and Zou, 2006; Still et al., 1990) for the charges in cell i and between the charges
476
Zhi-Jie Tan and Shi-Jie Chen
in two different cells i and j, respectively. h. . .i denotes averaging over all the possible positions Ri (Rj) of the TBIs within the respective phosphate cells. For a give mode, we can readily compute DGb, the free energy for the charges inside the tightly bound region, as the sum of the potentials of mean force: X X DGb ’ F1 ðiÞ þ F2 ði; jÞ: ð22:7Þ i
ij
Because we can pretabulate the potentials of mean force before enumerating the modes, the computation of DGb for a give mode is efficient.
4.3. Free energy of the diffusive ions With the mean-field approximation for the diffusive ions, we compute DGd from the following equation (Tan and Chen, 2005): ð 0 1 X ca ðrÞza q½cðrÞ þ c ðrÞd3 r DGd ¼ 2 a ð X ð22:8Þ ca ðrÞ 0 3 þ kB T ca ðrÞln 0 ca ðrÞ þ ca d r; ca a where c(r) and c0 (r) are the electric potentials at position r for the system with and without the diffusive ions, respectively. ca ðrÞ and ca0 are the concentrations of ion species a at position r and in the bulk solvent, respectively. We use c0 (r) here because c(r) c0 (r) gives the contribution to the electrostatic potential from the diffusive ions. With the given charge distribution of the TBIs, we calculate c(r) and c0 (r) from the nonlinear PB (with salt) and the Poisson equation (without salt), respectively.
4.4. Polarization energy of charges inside the tightly bound region pol
The free energy DGb in Eq. (22.4) is the change in Born energies for the charges transferred from the bulk solvent to the tightly bound region. pol We compute DGb as a sum for all the charges, including the phosphates and the TBIs, inside the tightly bound region: X pol DGb ¼ F0 ðiÞ; ð22:9Þ i
where F0(i) is the Born energy for the charges inside the ith tightly bound cell. Assuming there is a TBI in the ith cell, we compute F0(i) by averaging over all the possible positions Ri of the ion (Tan and Chen, 2008a):
477
RNA-Ion Interactions
pol
F0 ðiÞ ¼ kB T lnheðDUP pol where DUP
pol pol DUP ði; Ri Þ and DUI
pol
þDUI Þ=kB T
i;
ð22:10Þ
pol ¼ DUI ði; Ri Þ are the (generalized i and of the ion (at position Ri),
¼ Born) self-energies of the phosphate respectively. h. . .i denotes averaging over all the possible ion positions Ri within the cell.
5. Enhancing the Computational Efficiency for the Numerical Calculations of the TBI Model 5.1. A hybrid treatment for the tightly bound modes The computational efficiency of the TBI model is limited by the enumeration of the large number of the binding modes, which scales with the number N of nucleotides as 2N, assuming each phosphate cell contains 0 or 1 tightly bound (multivalent) ions. Therefore, an exhaustive enumeration of all the modes for a large nucleic acid molecule is extremely computationally expensive. To improve the computational efficiency, we developed a hybrid method. The strategy of the hybrid method is to consider the complete ensemble of the low-energy modes and treat the vast number of the high-energy modes through a computationally efficient random sampling algorithm. The partition function is calculated from the following equation (Tan and Chen, 2006b): ! Ml Mr X X M0 Ml X Z¼ ð22:11Þ ZM þ ZM Mr Nb where Nb is the number of TBIs, M0 and Ml are the total number of the modes and the number of the low-energy modes for a given Nb, respectively, and Mr is the number of the high-energy modes that are selected for random sampling. The method is reliable because the most important modes, namely, the low-energy modes, are treated exactly. Because the low-energy modes form only a small fraction of the total ensemble of the modes, this approach leads to drastic improvements in the computational efficiency.
5.2. Computation of the free energy DGd for the diffusive ions Strictly speaking, the free energy of the diffusive ions is dependent on the TBI mode. For each mode, the computation of DGd requires the result for the electric potentials solved from the nonlinear PB and Poisson equations
478
Zhi-Jie Tan and Shi-Jie Chen
(see Eq. 22.8). Given the large number of the modes, the computation of DGd can be time-consuming. To improve the computational efficiency, we use a coarse grained method for the computation of DGd. The basic strategy of the method is to precalculate DGd for several typical number of the TBIs Nb (averaging over the different modes with the same Nb), then use the results to fit DGd as a polynomial function of Nb (Tan and Chen, 2005, 2007). For any given mode with Nb TBIs, we can efficiently compute DGd from the polynomial.
6. Applications of the TBI Model In this section, we show how to apply the TBI model to predict the ion-dependent thermodynamic properties, such as ion-binding numbers and electrostatic free energies for realistic RNAs and DNAs. For illustrative purpose, we choose a short DNA duplex as a paradigmatic system.
6.1. General procedure for the numerical computations for the partition function The computational procedures for the numerical calculations of the TBI model for specific systems were described in several previous publications (Tan and Chen, 2005, 2008a). Here, we summarize the general computational procedure using a 12-nt DNA duplex as an example. We assume temperature T ¼ 25 C, dielectric constant e ¼ 78 for solvent and 20 for the nucleic acid interior, and the radii of the (hydrated) ions 4.5 A˚ for Mg2þ ˚ for Naþ. The numerical calculations of the TBI model involve and 3.5 A the following five steps: 1. To construct the 3D structural model. We construct a 3D structure for the 12-nt DNA duplex. The current form of the TBI model is base on a coarse-grained structural model—grooved primitive model (Montoro and Abascal, 1995; Tan and Chen, 2005), where each nucleotide consists of two ‘‘united atoms’’ (spheres), representing the electrically neutral group and the charged phosphate group, respectively (see Fig. 22.1A). 2. To identify the tightly bound region. For the nucleic acid structure immersed in a given ionic solution, we solve the nonlinear PB equation for the spatial distribution of the ion concentration, from which we calculate the spatial distribution of the ion–ion distance (a in Eq. (22.2)). From the criteria for strong correlation (Eq. (22.2)), we determine the tightly bound region. For a 12-nt DNA helix immersed in a 0.01 M Mg2þ solution, we find that the tightly bound region is a thin layer of average ˚ surrounding the helix (Fig. 22.1A). thickness 1.7 A
479
RNA-Ion Interactions
Β
A i=6
i = 12
i=1
i=7
v(i) = 214, 240, 253, 195, 181, 185 3 Å 220, 279, 322, 366, 292, 160 Average thickness of the tightly bound region = 1.7 Å
Ml
D Z=
Z M+ Nb
E
E
M0–Ml Mr
G = 14.4 kBT
Mr
Φ0(i) = 0.36 kBT
0.00 –0.48 –0.39 –0.32 –0.28 –0.26 –0.28 –0.30 –0.34 –0.44 –0.54 –0.52 –0.40 0.00 –0.46 –0.38 –0.30 –0.26 –0.30 –0.26 –0.26 –0.30 –0.38 –0.49 –0.35 –0.38 0.00 –0.40 –0.35 –0.28 –0.42 –0.28 –0.24 –0.24 –0.27 –0.35 –0.26 –0.34 –0.39 0.00 –0.37 –0.34 –0.55 –0.33 –0.25 –0.23 –0.23 –0.27 –0.22 –0.26 –0.34 –0.39 0.00 –0.38 –0.57 –0.53 –0.32 –0.25 –0.23 –0.24 –0.23 –0.24 –0.28 –0.35–0.40 0.00 –0.39 –0.59 –0.64 –0.49 –0.35 –0.29 –0.29 –0.36 –0.52 –0.70 –0.64 –0.41 0.00 –0.36 –0.33 –0.27 –0.24 –0.23 –0.23 –0.23 –0.27 –0.38 –0.64 –0.66 –0.43 0.00 –0.43 –0.35 –0.27 –0.23 –0.28 –0.24 –0.24 –0.28 –0.39 –0.67 –0.36 –0.45 0.00 –0.46 –0.36 –0.28 –0.36 –0.29 –0.26 –0.27 –0.32 –0.50 –0.29 –0.38 –0.46 0.00 –0.46 –0.38 –0.51 –0.40 –0.31 –0.27 –0.27 –0.32 –0.27 –0.31 –0.39 –0.49 0.00 –0.46 –0.51 –0.55 –0.46 –0.36 –0.31 –0.29 –0.27 –0.28 –0.31 –0.38 –0.43 0.00
Φ2(i,j) for (1,1) = 0.00 0.18 0.33 0.27 0.22 0.18 0.00 0.25 0.33 0.26 0.33 0.25 0.00 0.34 0.33 0.27 0.33 0.34 0.00 0.36 0.22 0.26 0.33 0.36 0.00 0.23 0.23 0.26 0.32 0.29 0.26 0.27 0.25 0.23 0.30 0.22 0.22 0.25 0.28 0.26 0.24 0.22 0.22 0.25 0.29 0.26 0.24 0.22 0.22 0.25 0.29 0.27 0.25 0.22 0.23 0.37 0.32 0.30 0.27 0.24
C ZM
— Nb=2.28
Φ1(i) = −2.0 kBT
Φ2(i,j) for (1,0) =
0.23 0.23 0.26 0.32 0.29 0.00 0.35 0.28 0.21 0.25 0.27 0.26
0.26 0.27 0.25 0.23 0.30 0.35 0.00 0.30 0.30 0.25 0.23 0.24
0.22 0.22 0.25 0.28 0.26 0.28 0.30 0.00 0.28 0.33 0.26 0.23
0.24 0.22 0.22 0.25 0.29 0.21 0.30 0.28 0.00 0.22 0.31 0.27
0.26 0.24 0.22 0.22 0.25 0.25 0.25 0.33 0.22 0.00 0.18 0.30
0.29 0.27 0.25 0.22 0.23 0.27 0.23 0.26 0.31 0.18 0.00 0.16
0.37 0.32 0.30 0.27 0.24 0.26 0.24 0.23 0.27 0.30 0.16 0.00
ΔGd(Nb) = −18.3, −12.4, −7.7, −4.2, −1.8, −0.4, 0.0, −0.5, −1.7, −3.7, −6.6, −10.6, −15.2
for a 12-nt DNA duplex in 0.01 MMgCl2
Figure 22.1 A 12-nt DNA duplex in 0.01 M MgCl2 is used to illustrate the computational procedure of the TBI model. (A) The 3D structure and the tightly bound region (green region). The two strands are labeled from i ¼ 1 to 6 and from i ¼ 7 to 12, respectively. The molecule contains totally N ¼ 12 nucleotides (phosphate groups). The red and blue spheres surrounding the central cylinder represent the charged and neutral groups of the nucleotides, respectively. v(i) denotes the volume of each tightly bound cell i (1 i 12). (B) The results for the potentials of mean force (in kBT ). For F2(i, j) (i, j ¼ 1, 2, . . ., N ¼ 12), the row and column correspond to i and j, respectively. (C) The results for the free energy DGd.
3. To calculate and tabulate the pair-wise potentials of mean force F1(i), F2(i, j) (Eq.(22.6)) and the Born energy F0(i) (Eq.(22.10)). A key issue in the calculations for the average over ion positions inside the respective phosphate cells is to take into account the finite size (excluded volume) of the ions and the detailed shape of the molecular surface. For the 12-nt helix in 0.01 M Mg2þ solution, F1(i) is 2.0kBT, F0(i) is 0.36kBT, and F2(i, j) for the interaction between phosphate cells i and j varies with i and j; see Fig. 22.1B. F2(i, j) < 0 (attractive interaction) if one of the cells is vacant and the other is occupied by a tightly bound (Mg2þ) ion (denoted as (1, 0) in Fig. 22.1B). F2(i, j) < 0 (repulsive interaction) if both cells are occupied (denoted as (1, 1) in Fig. 22.1B). 4. To calculate the free energy DGd for the diffusive ions (Fig. 22.1C). In order to fit a polynomial for DGd as a function of the number of the TBIs Nb, for a general structure of N nucleotides (phosphates), we evaluate DGd for a
480
Zhi-Jie Tan and Shi-Jie Chen
set of Nb’s, such as Nb ¼ 0, N/8, N/4, 3N/8, N/2, 2N/3, 5N/6, and N. From such an eight-point dataset, we can readily calculate DGd for an arbitrary Nb through either polynomial-fitting or interpolation. 5. To compute the partition function. Using the above hybrid treatment for the ensemble of ion-binding modes, we calculate the partition function from the sum of the modes (Eq. (22.11)). For each mode, the free pol energies DGb and DGb for the TBIs are directly given by the pretabulated potentials of the mean force (Eqs. (22.7) and (22.9)). From the partition function, we can predict a variety of ion-dependent folding properties such as the ion-binding properties and the iondependence of the folding stability.
6.2. Ion-binding properties From the partition function Z, we can predict any experimentally measur able property A from the statistical average A: Z¼
X M
ZM ;
X ¼ 1 A AM ZM ; Z M
ð22:12Þ
where AM is the result of property A for mode M. An important application of the above general equation is the computation of the mean number of the TBIs Nb from the partition function Z. The experimentally measured number of bound ions includes both the TBIs and the diffusively bound ions: ð 0 NMg2þ ¼ Nb þ cMg2þ ðrÞ cMg2þ d3 r; ð22:13Þ ðh i 0 3 NNa ¼ cNaþ ðrÞ cNa þ d r: þ
ð22:14Þ
here c(r) and c0 are the local concentration at position r and the bulk concentration, respectively. c(r) is solved from the PB equation for the diffusive ions. The integrals account for the contributions from the diffusive ions. Because the correlation effect is weak and negligible for monovalent ions, Nb ’ 0 for Naþ. For an illustrative calculation, we apply the TBI model to predict the number of bound ions (Eq. (22.14)) for a 24-bp DNA duplex immersed in an ionic solution with fixed 2 mM [Naþ] and different [Mg2þ]. As shown in Fig. 22.2, comparison with the experimental data (Bai et al., 2007; Chu et al., 2007) shows that the predictions from the TBI model are more accurate than the results from PB. For example, for 2 mM [Mg2þ], the improvement for Mg2þ-binding number is 20%.
481
RNA-Ion Interactions
Fraction of bound ions
0.8
0.6 Na+ 0.4
0.2 Mg2+
0 1e–05
0.0001
0.001 [Mg2+]
0.01
0.1
(M)
Figure 22.2 The mean number of the bound Mg2þ and Naþ ions per nucleotide for a 24-bp DNA duplex, as functions of Mg2þ concentration in the presence of 20 mM Naþ. Symbols, the experimental data (Bai et al., 2007; Chu et al., 2007); solid lines, the TBI theory predictions; dotted lines, the PB predictions with exclusion layer of width ˚ (¼radii of hydrated Mg2þ); dashed lines, the PB predictions with exclusion layer 4.5 A ˚ . The mean ˚ . In the calculation, the radii of hydrated Naþ is taken as 3.5 A of width 3.5 A number of the bound ions per nucleotide is calculated as the total number.
6.3. Folding stability Folding stability, as quantified by the free energy difference between the folded state and the ensemble of the unfolded states, is determined by the free energy landscape—the distribution of the free energy for different structures. To predict the ion-dependent free energy landscape, we need to generate all the possible conformations, including all the folded and unfolded conformations, and compute the free energy, including the electrostatic free energy (Eq. (22.5)), for each conformation. For example, for a system of two helices tethered by a loop, we need to enumerate all the conformations with different helix orientations and interhelix distances as well as the different loop conformations (Bai et al., 2008; Tan and Chen, 2006c). From the ion dependence of the free energy landscape, we can predict how ions change the conformational distribution and folding stability to cause the structural transitions. To predict the ion-dependent folding stability for a short DNA helix, we use a two-state model. Specifically, we assume that the conformational ensemble of the system consists of two states: double-stranded (ds) helix as the folded state and single-stranded (ss) helices (stabilized by the singlestranded self-stacking) as the unfolded state. For a given ionic condition
482
Zhi-Jie Tan and Shi-Jie Chen
(Naþ, Mg2þ) and temperature T, using the TBI model (Fig. 22.1), we E compute the electrostatic free energies for the duplex Gds and the unfolded E state Gss , respectively. The electrostatic contribution to the folding stability is E GssE : DGE ¼ Gds
ð22:15Þ
The ion-dependence of DGE shows how the ionic condition affects the folding stability and how the changes in ionic conditions induce the folding/unfolding transition. To predict the total folding stability also requires the result for the nonelectric part of the free energy difference DGNE. Assuming the ionindependence of DGNE, we can estimate DGNE from the empirical parameters measured at 1 M Naþ (SantaLucia, 1998): ð22:16Þ DGT Naþ =Mg2þ ¼ DGTE Naþ =Mg2þ þ DGNE ¼ DGTE ðNaþ =Mg2þ Þ þ ½DGT ð1 M Naþ Þ DGE ð1 M Naþ Þ: Here DGT is the total folding free energy, including the electric and nonelectric free energies. From the folding free energy DGT(Naþ/Mg2þ), we can predict the iondependent melting temperature Tm from the following equation: DGT RT ln CS ¼ 0;
ð22:17Þ
where R (¼1.987 cal/K mol) is the gas constant, CS is the strand concentration. CS is replaced with CS/4 for noncomplementary sequences. Figure 22.2 shows the TBI-predicted folding free energy DGT and melting temperature Tm for short DNA duplexes as functions of [Mg2þ] (Tan and Chen, 2006a). Comparisons with the experimental data indicate that the TBI model gives much improved predictions than the PB theory (Fig. 22.3).
7. Summary Ion correlation can be potentially important for RNA folding. Because correlated states can reach energies much lower than the meanfield energy, correlation can significantly enhance stability. Neglecting correlation (such as in the PB theory) can cause notable under-estimation of the folding stability (Chen, 2008). Furthermore, the correlation effect may make significant contributions to the unusually high efficiency of the multivalent Mg2þ ions in RNA stability, especially for tertiary structures. In RNA tertiary structure folding process, the close approach of the helices
483
RNA-Ion Interactions
A
B
0
50 40
−4 Tm (°C)
ΔG37 (kcal/mol)
−2
−6
30
−8 −10
Symbol: expt. data TBI model PB theory
−12 0.0001
0.001
0.01 [Mg2+] (M)
Symbol: expt. data TBI model PB theory
20
0.1
1
10
0.001
0.01 [Mg2+] (M)
0.1
1
Figure 22.3 The folding free energy DG37 at 37 C (A) and the melting temperature Tm (B) as functions of [Mg2þ]. (A) The upper lines are for sequence GCATGC, and the bottom lines are for sequence GCCAGTTAA (Tan and Chen, 2006a; and references therein). (B) The upper lines are for sequence AGAAAGAGAAGA with total strand concentration CS ¼ 6 mm, and the bottom lines are for sequence GCCAGTTAA with strand concentration CS ¼ 8 mm (Tan and Chen, 2006a; and references therein). Here, the data for AGAAAGAGAAGA are taken from the duplex melting in the study of the triplex melting in Mg2þ and CS ¼ 6 mm is the total strand concentration for triplex formation. Tm is calculated from DG RT ln CS/6 ¼ 0 (Tan and Chen, 2006a).
and other structural units causes significant build-up of the phosphate charges, inducing a strong correlation for the multivalent counterions. The strong correlation results in a further correlation-enhanced stabilizing force for tertiary folds. Such a correlation-induced enhancement in RNA stability goes beyond the classical charge neutralization/screening effects predicted by the mean-field theories. Therefore, multivalent ions can have much higher efficiency in stabilizing RNA structures than that predicted from the PB theory, and the high efficiency is more pronounced for tertiary structures. This causes a correlation-induced force from the multivalent ions to promote the formation of the tertiary folds. The TBI model tackles ion correlation/fluctuation effects by evaluating the energetics based on the discrete all-ion distributions for the TBIs. This is contrastingly different from the mean-field (correlation-free) approaches which are based on an average distribution for a system of independent ions. The TBI model also accounts for the volume exclusion effects due to the finite sizes of the ions. Control tests indicate that the excluded volume correlation is weak as compared with the Coulomb correlation. For systems with weak correlation, such as RNAs in monovalent ion solutions, the TBI gives nearly identical results as the PB and both the TBI and PB results agree with the experiment. But for RNAs/DNAs in Mg2þ solutions, where correlation can be strong, the TBI theory gives much improved predictions than the PB as shown by the experimental comparisons.
484
Zhi-Jie Tan and Shi-Jie Chen
The TBI model is not based on any of such preassumptions, while other models such as the PB are based on a priori assumptions about the nonexistence of the ion correlation/fluctuation effect. If the ion correlation and fluctuation effects are strong, the TBI model will capture them, otherwise, the model gives identical predictions as other (PB-based) models. Therefore, the TBI model can effectively complement the PB-based methods. Ion dehydration effects can be responsible for a variety of ion-dependent RNA folding properties, especially for the tertiary structure folding (Draper, 2008). Refinement of the TBI model should include a more accurate treatment for the possible ion dehydration effect. Further development of the model should also consider the all-atom details of the RNA structure and the improvements of computational efficiency.
ACKNOWLEDGMENTS We are grateful to Drs. Donald Rau, Xiangyun Qiu, Richard Owczarzy, and Song Cao for helpful discussions. This work was supported by NIH grant R01GM063732 (to S.-J. C.) and by the University of Missouri Life Science Postdoctoral Fellowship (to Z.-J. T.) and the National Science Foundation of China through grant No. 10844007 (to Z. -J. T.).
REFERENCES Bai, Y., Das, R., Millett, I. S., 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. Bai, Y., Chu, V. B., Lipfert, J., Pande, V. S., Herschlag, D., and Doniach, S. (2008). Critical assessment of nucleic acid electrostatics via experimental and computational investigation of an unfolded state ensemble. J. Am. Chem. Soc. 130, 12334–12341. Baker, N. A., Sept, D., Joseph, S., Holst, M. J., and McCammon, J. A. (2000). Electrostatics of nanosystems: Application to microtubules and the ribosome. Proc. Natl. Acad. Sci. USA 98, 10037–10041. Bokinsky, G., Rueda, D., Misram, 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. Bukhman, Y. V., and Draper, D. E. (1997). Affinities and selectivities of divalent cation binding sites within an RNA tertiary structure. J. Mol. Biol. 273, 1020–1031. Carnie, S., and Torrie, G. M. (1984). The statistical mechanics of the electrical double layer. Adv. Chem. Phys. 56, 141–253. Chen, S. J. (2008). RNA Folding: Conformational statistics, folding kinetics, and ion electrostatics. Annu. Rev. Biophys. 37, 197–214. Chu, V. B., and Herschlag, D. (2008). Unwinding RNA’s secrets: Advances in the biology, physics, and modeling of complex RNAs. Curr. Opin. Struct. Biol. 18, 305–314.
RNA-Ion Interactions
485
Chu, V. B., Bai, Y., Lipfert, J., Herschlag, D., and Doniach, S. (2007). Evaluation of ion binding to DNA duplexes using a size-modified Poisson-Boltzmann theory. Biophys. J. 93, 3202–3209. Clement, R. M., Sturm, J., and Daune, M. P. (1973). Interaction of metallic cations with DNA VI. Specific binding of Mg2þ and Mn2þ. Biopolymers 12, 405–421. Draper, D. E. (2008). RNA folding: Thermodynamic and molecular descriptions of the roles of ions. Biophys. J. 95, 5489–5495. Forsman, J. (2007). Simple correlation-corrected theory of systems described by screened coulomb interactions. Langmuir 23, 5515–5521. Gavryushov, S. (2008). Electrostatics of B-DNA in NaCl and CaCl2 solutions: Ion size, interionic correlation, and solvent dielectric saturation effects. J. Phys. Chem. B 112, 8955–8965. Gilson, M. K., and Zhou, H. X. (2007). Calculation of protein-ligand binding affinities. Annu. Rev. Biophys. Biomol. Struct. 36, 21–42. Gilson, M. K., Sharp, K. A., and Honig, B. (1987). Calculating the electrostatic potential of molecules in solution: Method and error assessment. J. Comput. Chem. 9, 327–335. Grilley, D., Misra, V., Caliskan, G., and Draper, D. E. (2007). Importance of partially unfolded conformations for Mg(2þ)-induced folding of RNA tertiary structure: Structural models and free energies of Mg2þ interactions. Biochemistry 46, 10266–10278. Grochowski, P., and Trylska, J. (2008). Continuum molecular electrostatics, salt effects and counterion binding. A review of the Poisson-Boltzmann theory and its modifications. Biopolymers 89, 93–113. Ha, B. Y., and Thirumalai, D. (2003). Bending rigidity of stiff polyelectrolyte chains: A single chain and a bundle of multichains. Macromolecules 36, 9658–9666. Heilman-Miller, S. L., Pan, J., Thirumalai, D., and Woodson, S. A. (2001). Role of counterion condensation in folding of the Tetrahymena ribozyme. II. Counteriondependence of folding kinetics. J. Mol. Biol. 309, 57–68. Ichimaru, S., Iyetomi, H., and Tanaka, S. (1987). Statistical physics of dense plasmas: Thermodynamics, transport coefficients and dynamic correlations. Phys. Rep. 149, 91–205. Kim, H. D., Nienhaus, G. U., Ha, T., Orr, J. W., Williamson, J. R., and Chu, S. (2002). Mg2þ-dependent conformational change of RNA studied by fluorescence correlation and FRET on immobilized single molecules. Proc. Natl. Acad. Sci. USA 99, 4284–4289. Koculi, E., Hyeon, C., Thirumalai, D., and Woodson, S. A. (2007). Charge density of divalent metal cations determines RNA stability. J. Am. Chem. Soc. 129, 2676–2682. Krakauer, H. (1971). The binding of Mgþþ ions to polyadenylate, polyuridylate, and their complexes. Biopolymers 10, 2459–2490. Liu, H. Y., and Zou, X. (2006). Electrostatics of ligand binding: Parametrization of the generalized Born model and comparison with the Poisson-Boltzmann approach. J. Phys. Chem. B 110, 9304–9313. Lorenz, C., Piganeau, N., and Schroeder, R. (2006). Stabilities of HIV-1 DIS type RNA loop-loop interactions in vitro and in vivo. Nucleic Acids Res. 34, 334–342. Luan, B., and Aksimentiev, A. (2008). DNA attraction in monovalent and divalent electrolytes. J. Am. Chem. Soc. 130, 15754–15755. Lyubartsev, A. P., and Nordenskiold, L. (1995). Monte Carlo simulation study of ion distribution and osmotic pressure in hexagonally oriented DNA. J. Phys. Chem. 99, 10373–10382. Manning, G. S. (1978). The molecular theory of polyelectrolyte solutions with applications to the electrostatic properties of polynucleotides. Q. Rev. Biophys. 11, 179–246. Mathews, D. H., and Turner, D. H. (2006). Prediction of RNA secondary structure by free energy minimization. Curr. Opin. Struct. Biol. 16, 270–278.
486
Zhi-Jie Tan and Shi-Jie Chen
Montoro, J. C. G., and Abascal, J. L. F. (1995). Ionic distribution around simple DNA models. I. Cylindrically averaged properties. J. Chem. Phys. 103, 8273–8284. Nixon, P. L., and Giedroc, D. P. (1998). Equilibrium unfolding (folding) pathway of a model H-type pseudoknotted RNA: The role of magnesium ions in stability. Biochemistry 37, 16116–16129. Owczarzy, R., You, Y., Moreira, B. G., Manthey, J. A., Huang, L., Behlke, M. A., and Walder, J. A. (2004). Effects of sodium ions on DNA duplex oligomers: Improved predictions of melting temperatures. Biochemistry 43, 3537–3554. Pljevaljcic, G., Millar, D. P., and Deniz, A. A. (2004). Freely diffusing single hairpin ribozymes provide insights into the role of secondary structure and partially folded states in RNA folding. Biophys. J. 87, 457–467. Qiu, H., Kaluarachchi, K., Du, Z., Hoffman, D. W., and Giedroc, D. P. (1996). Thermodynamics of folding of the RNA pseudoknot of the T4 gene 32 autoregulatory messenger RNA. Biochemistry 35, 4176–4186. Qiu, X., Andresen, K., Kwok, L. W., Lamb, J. S., Park, H. Y., and Pollack, L. (2007). InterDNA attraction mediated by divalent counterions. Phys. Rev. Lett. 99, 038104. Rau, D. C., and Parsegian, V. A. (1992a). Direct measurement of the intermolecular forces between counterion-condensed DNA double helices. Evidence for long range attractive hydration forces. Biophys. J. 61, 246–259. Rau, D. C., and Parsegian, V. A. (1992b). Direct measurement of temperature-dependent solvation forces between DNA double helices. Biophys. J. 61, 260–271. Ray, J., and Manning, G. S. (2000). Formation of loose clusters in polyelectrolyte solutions. Macromolecules 33, 2901–2908. Rialdi, G., Levy, J., and Biltonen, R. (1972). Thermodynamic studies of transfer ribonucleic acids. I. Magnesium binding to yeast phenylalanine transfer ribonucleic acid. Biochemistry 11, 2472–2479. Romer, R., and Hach, R. (1975). tRNA conformation and magnesium binding. A study of a yeast phenylalanine-specific tRNA by a fluorescent indicator and differential melting curves. Eur. J. Biochem. 55, 271–284. SantaLucia, J. Jr. (1998). A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics. Proc. Natl. Acad. Sci. USA 95, 1460–1465. Serra, M. J., and Turner, D. H. (1995). Predicting thermodynamic properties of RNA. Methods Enzymol. 259, 242–261. Soto, A. M., Misra, V., and Draper, D. E. (2007). Tertiary structure of an RNA pseudoknot is stabilized by ‘‘ diffuse’’ Mg2þ ions. Biochemistry 46, 2973–2983. Stellwagen, E., Dong, Q., and Stellwagen, N. C. (2007). Quantitative analysis of monovalent counterion binding to random-sequence, double-stranded DNA using the replacement ion method. Biochemistry 46, 2050–2058. Still, W. C., Tempczyk, A., Hawley, R. C., and Hendrickson, T. (1990). Semianalytical treatment of solvation for molecular mechanics and dynamics. J. Am. Chem. Soc. 112, doi: tex 6127–6129. Takamoto, K., He, Q., Morris, S., Chance, M. R., and Brenowitz, M. (2002). Monovalent cations mediate formation of native tertiary structure of the Tetrahymena thermophila ribozyme. Nat. Struct. Biol. 9, 928–933. Tan, Z. J., and Chen, S. J. (2005). Electrostatic correlations and fluctuations for ion binding to a finite length polyelectrolyte. J. Chem. Phys. 122, 044903. Tan, Z. J., and Chen, S. J. (2006a). Nucleic acid helix stability: Effects of salt concentration, cation valency and size, and chain length. Biophys. J. 90, 1175–1190. Tan, Z. J., and Chen, S. J. (2006b). Ion-mediated nucleic acid helix-helix interactions. Biophys. J. 91, 518–536. Tan, Z. J., and Chen, S. J. (2006c). Electrostatic free energy landscape for nucleic acid helix assembly. Nucleic Acids Res. 34, 6629–6639.
RNA-Ion Interactions
487
Tan, Z. J., and Chen, S. J. (2007). RNA helix stability in mixed Naþ/Mg2þ solution. Biophys. J. 92, 3615–3632. Tan, Z. J., and Chen, S. J. (2008a). Electrostatic free energy landscapes for DNA helix bending. Biophys. J. 94, 3137–3149. Tan, Z. J., and Chen, S. J. (2008b). Salt dependence of nucleic acid hairpin stability. Biophys. J. 95, 738–752. Theimer, C. A., and Giedroc, D. P. (2000). Contribution of the intercalated adenosine at the helical junction to the stability of the gag-pro frameshifting pseudoknot from mouse mammary tumor virus. RNA 6, 409–421. Vieregg, J., Cheng, W., Bustamante, C., and Tinoco, I. Jr. (2007). Measurement of the effect of monovalent cations on RNA hairpin stability. J. Am. Chem. Soc. 129, 14966–14973. Vlachy, V. (1999). Ionic effect beyond Poisson-Boltzmann theory. Annu. Rev. Phys. Chem. 50, 145–165. Walter, N. G., Burke, J. M., and Millar, D. P. (1999). Stability of hairpin ribozyme tertiary structure is governed by the interdomain junction. Nat. Struct. Biol. 6, 544–549. Walter, N. G., Woodson, S. A., and Batey, R. T. (eds.) (2009). In ‘‘Non-Protein Coding RNAs’’, Springer, Berlin. Weixlbaumer, A., Werner, A., Flamm, C., Westhof, E., and Schroeder, R. (2004). Determination of thermodynamic parameters for HIV DIS type loop-loop kissing complexes. Nucleic Acids Res. 32, 5126–5133. Wu, J. Z., and Li, Z. D. (2007). Density-functional theory for complex fluids. Ann. Rev. Phys. Chem. 58, 85–112.
Author Index
A Abakumov, G. A., 307 Abascal, J. L. F., 478 Abelson, J., 72 Adams, P. L., 126, 131, 219 Addess, K. J., 27 Adilakshmi, T., 190 Aitken, C. E., 179 Akiyama, B. M., 27–45 Akiyama, S., 254 Aksimentiev, A., 471 Albergo, D. D., 454 Aleman, E. A., 179 Al-Hashimi, H. M., 322 Altenbach, C., 305 Amaral, P. P., 238 Amsterdam, I. M. C. v., 332 Anderson, C. F., 376, 392, 415, 427, 440 Andresen, K., 257, 398–399, 403, 405–406 Ansari, A., 366 A˚qvist, J., 423 Asseline, U., 291 Auffinger, P., 421 B Bain, J. D., 102 Baird, N. J., 211 Bai, Y., 218, 248, 376–377, 380–381, 384, 386, 392, 427, 458, 468, 470, 480–481 Baker, D., 248 Baker, N. A., 405, 411–430, 472 Ballauff, M., 397 Ballin, J. D., 355 Ban, N., 124, 330 Banyay, M., 357–358 Barhate, N., 305–306, 341 Barkema, G. T., 145 Bartley, L. E., 211, 300 Bassi, G. S., 172 Basu, S., 219 Batey, R. T., 4–5, 8, 11, 13, 15, 19, 22, 119–135, 155 Baugh, C., 131 Baum, D. A., 55, 95–114 Bellur, D. L., 198 Bennett, R., 160 Berliner, L. J., 330–331
Berman, H. M., 120–121 Bernasconi, C. F., 355 Berry, D., 83 Beth, A. H., 330 Betteridge, T., 69–89 Bevilacqua, J. M., 190 Bevilacqua, P. C., 190 Bhattacharyya, A., 146 Biertu¨mpfel, C., 155 Bina-Stein, M., 376 Biou, V., 82 Birikh, K. R., 56 Blackburn, E. H., 124 Blair, D. P., 312 Blanchard, S. C., 74, 182 Bloomfield, V. A., 191, 392 Bobst, A. M., 324 Bode, B. E., 331–332, 340–341 Bo¨hme, S., 339 Bokinsky, G., 211, 469 Borbat, P. P., 330–331 Bo¨rjesson, K., 167, 169 Borukhov, I., 430 Bouadloun, F., 70 Boukobza, E., 180 Brandt, G., 131 Brauns, E. B., 353–371 Brautigam, C. A., 232 Breaker, R. R., 35, 97, 132, 180–181, 330 Breslauer, K. J., 294 Brion, P., 49–50, 210, 244 Brodsky, A. S., 322 Brown, J. W., 124 Bruce, A. G., 102 Bru¨nger, A. T., 134 Bucci, E., 288, 292 Buchmueller, K. L., 193, 196, 200–201, 211 Bukhman, Y. V., 457, 468, 471 Burke, J. M., 160 Burley, S. K., 122 Bustamante, C., 50 C Cai, Q., 330, 339 Campbell, E. A., 130 Cann, J. R., 193 Cantor, C. R., 160, 211, 324 Carnie, S., 472–473
489
490
Author Index
Carrasco, N., 132 Cate, J. H., 28, 126, 131, 219 Cayley, S., 219 Cech, T. R., 28, 57, 102, 126, 193–194, 204, 211, 275, 434, 457 Cekan, P., 303–326 Cerutti, P., 75 Chacon, P., 241, 243, 264 Chang, K. Y., 412, 448 Chauhan, S., 275–276 Cheatham, T. E. I. I., 419, 421 Chen, A. A., 411–430 Chen, J. L., 49, 124 Chen, L. L., 262 Chen, S. J., 218, 430, 465–484 Cherepanov, A. V., 64 Chin, K., 405 Cho, H. D., 83 Chrambach, A., 192 Christian, E. L., 203 Chu, V. B., 246, 430, 466–467, 472–473, 480–481 Cilley, C. D., 194 Clegg, R. M., 160, 171–172 Clement, R. M., 468 Cochella, L., 73 Cochrane, J. C., 126, 131 Cohen, S. B., 28 Cole, J. L., 211 Coleman, T. M., 125 Cole, P., 210 Columbus, L., 319–320 Conn, G., 219 Cooper, J. P., 144, 146 Coppins, R. L., 109 Cornell, W. D., 423 Cornish, P. V., 45, 182 Costa, M., 126 Costantino, D., 211, 216, 219, 232 Cova, S., 312 Cowtan, K., 134 Creagh, D. C., 394 Crothers, D. M., 35, 102, 146, 190–191, 200, 210, 376, 434 Cruz, R. P. G., 98–100 Cudney, B., 127 D Dale, R. E., 160 Damberger, S. H., 124 Damha, M. J., 107 Dann, C. E. I. I., 211, 232 Das, R., 198, 210, 218, 240, 247–248, 254, 264–265, 376–377, 388, 395–396, 398, 405, 412, 457 Daugherty, M. A., 190–191 Davis, J. H., 275
Davis, R. H., 20 De´clais, A.-C., 155 de Gennes, P. G., 144 Delagoutte, B., 82 D. L.no, W. L., 243 Deniz, A. A., 179 Denli, A. M., 330 Dennis, P. P., 70 Denysenkov, V. P., 331–332, 341 D. R.se, V. J., 122, 330 Dervan, P. B., 28 Derwenskus, K. H., 82 Deutscher, M. P., 83 Deutsch, J. M., 145 de Vries, S., 64 Dewey, T. G., 366 Dieckmann, T., 330 Doherty, E. A., 126 Dolinnaya, N. G., 58 Dolinsky, T. J., 427 Domdey, H., 107 Doniach, S., 237–249, 253–267, 393 Dorywalska, M., 51, 174 Doty, P., 218 Doudna, J. A., 4, 34, 56, 102, 120, 122–123, 125 Downey, C. D., 48 Drak, J., 146 Draper, D. E., 146, 218–219, 246, 376, 388, 412, 433–461, 466–468, 470–471, 484 Drew, H. R., 125 Duckett, D. R., 144, 146, 149, 151 Duffy, D. C., 259 Duhamel, J., 288 Duke, T. A., 145 Dupasquier, M., 72, 83, 87 Dyer, R. B., 353–371 E Easton, L. E., 3–24 Edwards, A. L., 5, 9, 120, 131, 134 Edwards, T. E., 276, 288, 304–305, 308, 320–322, 324 Efron, B., 179 Efstratiadis, A., 107 Eichman, B. F., 144 Eisenberg, H., 446 Eisinger, J., 160 Eliezer, D., 262 Elsa¨sser, C., 332 Emerick, V. L., 190, 193, 197, 203–204 Emsley, P., 134 Engelman, D. M., 407 England, T. E., 29, 40 Ennifar, E., 131 Ermolenko, D. N., 174
491
Author Index F Fajer, P. G., 314 Fang, X. W., 55, 261, 434, 457 Fechter, P., 77 Fedor, M. J., 122 Fedorova, O. A., 58 Feig, A. L., 359 Fei, J., 74 Feix, J. B., 304 Ferat, J. L., 107 Ferguson, K. A., 192 Ferrari, M. E., 461 Ferre´-D’Amare´, A. R., 34, 56, 102, 122, 126, 151, 276 Fidanza, J. A., 305, 308 Finkelstein, K. D., 391–408 Fischer, H., 294 Fischhaber, P. L., 305 Flynn-Charlebois, A., 104 Forsman, J., 473 Fo¨rster, T., 159, 162 Frank, D. E., 298 Frankel, A. D., 320 Freed, J. H., 320 Freier, S., 218 Fresco, J., 218 Friederich, M. W., 190 Fried, M. G., 190–191, 200 Frisch, H. L., 145 Fujita, H., 211 Fukai, S., 82 G Gannett, P. M., 307 Garcı´a-Garcı´a, C., 434 Garner, M. M., 190 Garst, A. D., 199–135 Gavryushov, S., 473 Gherghe, C. M., 129 Giedroc, D. P., 468 Gilbert, S. D., 122, 125 Gilson, M. K., 472 Glatter, O., 393 Gluick, T. C., 203 Godt, A., 331, 346 Gohlke, C., 160 Golden, B. L., 120, 126, 130, 132, 219 Goldgur, Y., 82 Gold, L., 190 Goody, T. A., 145, 149, 152, 172, 174, 177–178 Gough, G. W., 144, 146 Grant, G. G., 330 Grant, G. P. G., 288, 308, 320 Greenfeld, M., 375–388 Green, R., 73 Griffiths-Jones, S., 124
Grilley, D., 218–219, 376, 392, 435, 451, 458, 461, 468 Grochowski, P., 473 Grosshans, C. A., 102, 204 Gross, H. J., 31 Grote, M., 330, 339 Gue´ron, M., 272, 280 Guest, C. R., 280 Guinier, A., 240 Guo, F., 131, 219 Gutell, R. R., 124 H Ha, B. Y., 474 Hach, R., 376, 468, 470 Hadden, J. M., 155 Hagelu¨ken, G., 330–331, 341 Hagerman, P. J., 144, 146, 190 Hallberg, B. M., 4 Hall, K. B., 269–284 Hamann, C. S., 72 Hamy, F., 322 Han, H., 28 Hannon, G. J., 330 Hansen, C. L., 134 Hara, H., 308 Harned, H. S., 451 Ha, T., 48, 53, 56, 179–180 Hauenstein, S., 72, 84 Haugland, R. P., 160 Hawkins, M. E., 291–293 Hedgcoth, C., 78–79 Heilman-Miller, S. L., 195, 197, 205, 467, 469, 471 Hengesbach, M., 28 Henley, D. D., 230 Henry, E. R., 263 Hermann, D., 308 Herrera, J. E., 376 Herschlag, D., 47–65, 261–262, 287–301, 375–388, 466 Heuer, D. M., 145 Hickerson, R., 174 Hilger, D., 339 Hirsh, D., 73 Hoadley, K. A., 104 Ho¨bartner, C., 132 Ho, C. K., 57 Hodak, J. H., 48, 56, 182 Hofrichter, J., 263 Hohng, S., 149, 151, 180, 182, 211 Holbrook, S. R., 120 Hongin, B., 376 Hou, Y. M., 69–89 Howlett, G. J., 211–212 Hubbell, W. L., 319–320, 330 Hustedt, E. J., 320, 330 Hyeon, C., 355
492
Author Index I
Ichimaru, S., 474 Ignacio Tinoco, J., 211 Ikawa, Y., 274 Inoue, H., 35 Iqbal, A., 165–169, 176–177 J Jacobson, M., 78–79 Jaeger, L., 49 Janiak, F., 82, 89 Jean, J. M., 270 Jeschke, G., 330–332, 335, 341 Johnson, M. L., 179 Jonikas, M. A., 211, 233 Joo, C., 179 Jordan, S. R., 127 Jorgensen, W. L., 423 Joung, I. S., 421 Joyce, G. F., 35, 56, 96–97 Jucker, F. M., 274 Jusufi, A., 397 K Kao, C., 77, 112, 125 Karbstein, K., 299 Kay, C. W. M., 332 Kazantsev, A. V., 124 Kazimirov, A., 260 Ke, A., 120, 122–123, 126 Keel, A. Y., 3–24, 131 Kern, J. A., 20 Keyes, R. S., 305, 324 Kieft, J. S., 3–24, 123, 131, 152, 211, 216, 219, 232 Kim, H. D., 211, 469 Kim, I., 4, 14 Kim, N. -K., 305, 308, 330 Kim, S. H., 72, 190 Kisseleva, N., 330 Klare, J. P., 330 Klein, D. J., 124, 130–133, 145, 149, 160, 172, 219 Klostermeier, D., 177 Klug, C. S., 304 Knight, J. B., 258 Koch, M. H. J., 393 Koculi, E., 197, 201–202, 211, 452, 467, 469 Komatsoulis, G. A., 72 Koo, H. S., 190 Kool, E. T., 291 Korostelev, A., 70 Kost, D. M., 103–104, 107 Kothe, U., 76 Kozin, M. B., 242 Krakauer, H., 376, 468
Kramer, G. F., 74 Kratky, O., 240 Kundrot, C. E., 120, 126 Kurschat, W. C., 57 Kuznetsov, N. A., 330 Kuznetsov, S. V., 366 Kwok, L. W., 210, 220, 254, 262 L Laederach, A., 211, 220 Lafontaine, D. A., 153, 172, 190 Laing, L. G., 279, 412 Lakowicz, J. R., 160, 282, 290 Lambert, D., 412, 422–423, 433–461 Lamb, J., 266 Lang, K., 276–277 Lang, M. J., 53 Lapham, J., 35, 102 Larsen, N., 124 Larsen, O. F. A., 280 Latham, J. A., 434, 457 Laue, T. M., 211–212 Lease, R. A., 51 Lebowitz, J., 211 Ledoux, S., 73 Lee, T.-H., 48, 50, 74, 182 Lehnert, V., 49 Leipply, D., 433–461 Lemay, J.-F., 180–181 Lerman, L. S., 145 Levene, S. D., 145, 191–192 Levine, I. N., 443 Lewis, F. D., 167 Liang, Z., 320 Li, A. Z., 376 L. C.ta, V. J., 288 Lilley, D. M. J., 143–155, 159–182, 191, 211, 244 Lim, F., 6 Ling, J., 75, 83, 88 Lipfert, J., 153, 237–249, 254, 257, 265 Liu, C., 69–89 Liu, G., 190 Liu, H. Y., 82, 475 Liu, J., 149–150, 174 Liu, Q., 124 Liu, Y.-S., 330 Li, Z. D., 473 Locke, B. R., 192 Loftfield, R. B., 70 Lohman, T. M., 461 Lorenz, C., 468, 471 Luan, B., 471 Luehrmann, R., 330 Lukavsky, P. J., 3–24, 123 Lumpkin, O. J., 145, 191 Lu, Y., 133 Lyubartsev, A. P., 471
493
Author Index M Macosko, J. C., 308, 330 Magnet, S., 120–121 Ma, H., 355, 366, 369–370 Major, F., 245, 248 Majumdar, Z. K., 174 Mandal, M., 180–181 Mandal, S. S., 88 Manning, G. S., 218, 376, 393, 471 Marcus, Y., 446 Margraf, D., 331, 346 Marino, J. P., 291 Marsh, D., 320 Martick, M., 125, 131 Martin, C. T., 83 Martin, R. E., 331 Marucho, M., 411–430 Mathews, D. H., 467 Matsumura, S., 149 Mattick, J. S., 238 M. C.llum, S. A., 279 M. C.rthy, T. J., 11 Mchaourab, H. S., 319 M. I.tosh, B., 74 M. K.nna, S. A., 4, 14 M. K.nney, S. A., 182 M. L.ughlin, L. W., 170, 192 M. P.erson, A., 127 Melcher, S. E., 152, 172–173, 177 Michel, F., 49, 107, 126 Mikulecky, P. J., 359 Millar, D. P., 177, 179 Miller, N., 75 Millero, F. J., 446 Miller, T. R., 306 Milligan, J. F., 4, 9, 55–56, 77, 102, 112, 125 Mills, P. A., 405 Milov, A. D., 331, 333, 340, 345 Mirzabekov, A., 55 Misra, V. K., 218–219, 376, 435 Mitra, S., 209–233 Moghaddam, S., 201 Mohanty, U., 192 Montange, R. K., 124, 129, 131–132 Montoro, J. C. G., 478 Moody, E. M., 369 Mooers, B. H. M., 125 Mookhtiar, K. A., 77 Moore, M. J., 29, 40, 56–57, 102, 276 Morrissey, S. R., 330 Mui, T. P., 107 Munro, J. B., 74 Murakami, H., 73 Murchie, A. I. H., 150, 160 Murphy, F. L., 126 Murthy, V. L., 218
N Nagahara, S., 305, 308 Nahas, M. K., 180 Nakano, S.–I., 423 Nandakumar, J., 40 Narr, E., 332 Neubauer, H., 169 Nissen, P., 330 Nivon, L. G., 355 Nixon, P. L., 468 Nordenskiold, L., 471 Nordland, T. M., 283 Norman, D. G., 164–165, 167 O Ogston, A. G., 192 Okamoto, A., 308 Okumus, B., 180, 182 Olson, W. K., 201 Ortiz-Lombardı´a, M., 144 Ortoleva-Donnelly, L., 203 Ott, G., 83 Oubridge, C., 126 Owczarzy, R., 468 Ozaki, H., 170 P Pabit, S. A., 391–408, 429 Padgett, R. A., 107 Pan, D., 81–82, 86 Pande, V. S., 423 Pan, J., 193, 195, 198, 201, 203–205, 412 Pannier, M., 331 Pan, T., 51, 102, 452 Papoian, G. A., 426 Pappu, R. V., 411–430 Pardi, A., 279 Parisien, M., 245, 248 Park, H. Y., 257–258 Park, S. Y., 330 Parsegian, V. A., 470–471 Patel, M., 395 Patte, J. C., 133 Paul, N., 106 Paulsen, H., 74–75 Peabody, D. S., 6 Pechukas, P., 160 Penedo, J. C., 172 Pereira, M. J., 182 Perez-Salas, U. A., 211, 412 Perrin, F., 159 Pestova, T. V., 24 Petoukhov, M. V., 238, 264 Philo, J. S., 215, 224 Pinard, R., 190 Pinol-Roma, S., 28
494
Author Index
Piton, N., 305–307 Pleiss, J. A., 55, 125 Pley, H. W., 126 Pljevaljcic, G., 179, 211, 469 Plum, G. E., 392 Po¨hler, J. R. G., 155 Pollack, L., 253–267, 391–408 Polyhach, Y., 332 Pomeranz-Krummel, D. A., 144, 149, 151 Ponomarev, S. Y., 419–421 Poole, C. P., 312 Porod, G., 240 Pratico, E. D., 110 Prendergast, F. G., 361 Price, S. R., 34, 56, 125 Prisner, T. F., 305, 330, 332 Proctor, D. J., 369 Proudnikov, D., 55 Puglisi, J. D., 4, 14, 16, 123 Purtha, W. E., 57, 102–103 Putnam, C. D., 238 Pyle, A. M., 97, 127, 190, 196 Q Qin, P. Z., 303–326, 330, 339 Qiu, H., 469 Qiu, X., 471 Query, C. C., 29, 40, 57, 102 Qu, X., 48 R Rachofsky, E. L., 270 Ralston, C. Y., 452 Ramakrishnan, B., 132 Rangan, P., 198, 205, 412 Rasnik, I., 52, 179 Rau, D. C., 470–471 Ray, J., 471 Record, M. T. Jr., 376, 392, 415, 427, 435, 438–439, 444 Redfield, A. G., 434, 457 Revzin, A., 190 Reyes, F. E., 119–135 Rialdi, G., 468 Richey, B., 438 Riordan, F. A., 146 Riplinger, C., 332, 334 Rist, M. J., 291 Robertson, J. M., 74–75 Robertson, M. P., 132 Robinson, C. V., 446–447, 451 Robinson, H., 405 Robinson, R. A., 446–447, 451 Rodbard, D., 192 Rodnina, M. V., 74–75 Rodriguez, J. R., 107 Romaniuk, P. J., 29, 40
Romero, D. P., 124 Romer, R., 376, 468, 470 Rose, G. D., 218 Roth, A., 11 Roy, R., 45, 48, 179 Roy, S., 322 Rueda, D., 177 Rupert, P. B., 151 Ruskin, B., 107 Russell, R., 48, 51, 182, 210–211, 254, 258, 261–264 Ryder, S. P., 190–191 S Saha, S., 145 Sampson, J. R., 77 Sanborn, M. E., 167, 169, 176 Sanderson, L. E., 89 Sandin, P., 291–29 Sandy, A. R., 260 SantaLucia, J. Jr., 482 Santoro, S. W., 56, 97 Sartori, A., 192, 205 Sase, I., 180 Sattin, B. D., 50, 211, 219 Savchenko, A., 125 Savelyev, A., 426 Sawaya, M. R., 130 Scaringe, S. A., 28 Schiemann, O., 304–306, 329–347 Schimmel, P. R., 211, 324, 434, 457 Schlatterer, J. C., 254, 261–262 Schlax, P. J., 203 Schleich, H. G., 74 Schlosser, K., 98–99 Schnare, M. N., 124 Schneider, D. J., 320 Schubert, S., 99 Schuck, P., 211, 224 Schurer, H., 56 Sclavi, B., 220, 239, 261 Scott, D. J., 211 Scott, K. S., 55 Scott, W. G., 125, 131–132 Segel, D. J., 262 Seif, E., 4 Seifert, S., 260 Selmer, M., 70, 124 Semichaevsky, A. V., 422 Semlow, D. R., 104 Serganov, A., 125, 130, 180, 276, 445 Serra, M. J., 467 Seth, M., 86 Severcan, I., 190 Shakhnovich, E. I., 355 Sharp, K., 376 Sharp, P. A., 29, 40, 56, 102, 120–121
495
Author Index
Shcherbakova, I., 211, 229 Sheldrick, G. M., 134 Shi, X., 287–301 Shuman, S., 57 Sicoli, G., 330 Siegfried, N. A., 369 Sigurdsson, S. T., 288, 303–326 Silakov, A., 330 Silverman, S. K., 35, 95–114, 211 Smith, G. J., 51, 174 Smolin, N., 414 Solomatin, S., 47–65 Sorin, E. J., 423 Sosnick, T. R., 452 Soto, A. M., 219, 434, 461, 468 Sowa, G. Z., 304, 319–320, 330 Spaltenstein, A., 305 Spitale, R. C., 124 Stafford, W. F. I. I., 211, 215, 224–226 Stahl, D. A., 190 Stahley, M. R., 219 Stancik, A. L., 369 Stark, H., 330 Stark, M. R., 102 Stein, A., 376, 434 Steiner, M., 182 Steiner, R. F., 288, 292 Steitz, T. A., 72 Stellwagen, E., 192, 473 Still, W. C., 475 Stokes, R. H., 446–447 Stone, M. D., 27–45, 56, 182 Strauss, U. P., 417, 438, 458 Streicher, B., 204 Strobel, S. A., 57, 125 Stryer, L., 160 Stuhrmann, H. B., 241, 264, 393 Sugimoto, N., 99 Sundaralingam, M., 132 Svergun, D. I., 238, 241–242, 264, 393 Switzer, C., 102 Sydenham, P. H., 312 Szewczak, A. A., 133, 193–194 Szymanski, M., 124 T Takamoto, K., 211, 216, 218, 220, 229, 231, 247, 412, 467, 469 Tan, E., 152, 180–181 Tanford, C., 211 Tang, R. S., 146 Tanner, M., 275 Tan, Z. J., 218, 430, 465–484 Tate, M. W., 400 Tereshko, V., 131 Theimer, C. A., 468 Thirumalai, D., 210, 355, 474 Thomas, G. J. Jr., 357
Thomas, J. C., 288 Thore, S., 276 Tibshirani, R. J., 179 Tikhonov, V. D., 307 Tinoco, I. Jr., 50, 412, 448 Tinsley, R. A., 83 Tomita, K., 72 Toor, N., 123, 131 Torarinsson, E., 124 Torrie, G. M., 472–473 Tremethick, D. J., 376 Trinh, S. H., 192 Trylska, J., 473 Tsuruta, H., 260 Tuerk, C., 190 Turner, B., 149, 174–175 Turner, D. H., 366, 467 Tuschl, T., 172 U Ubbink, M., 330 Uchida, T., 230 Uhlenbeck, O. C., 4, 29, 40, 55–56, 77, 89, 102, 122 V Va´mosi, G., 169 Vanderjagt, D., 70 Varani, L., 308 Venyaminov, S. Y., 361 Vicens, Q., 129 Vieregg, J., 468 Vitalis, A., 414, 430 Vlachy, V., 473 Vogt, M., 330 Voinnet, O., 330 Volkov, V. V., 242 W Wahl, M. C., 107 Walker, G. C., 29, 40 Walter, F., 149–150, 172 Walter, N. G., 37, 83, 160, 172, 177, 466–467, 469 Walther, D., 241–242, 264 Wan, C., 280 Wang, Y., 99, 102, 104–107, 109–110, 113, 182 Ward, R., 330, 335 Weeks, K. M., 196 Weiss, S., 48 Weixlbaumer, A., 467–468, 471 Wensel, T. G., 392 Westhof, E., 49–50, 122, 152, 210, 244 Widengren, J., 52 Wilhelmsson, L. M., 168–169, 291–292 Williams, D. J., 278–279, 282
496
Author Index
Williamson, J. R., 155, 194, 322 Wilson, J. N., 291 Wimberly, B. T., 73, 131 Winkler, W. C., 132, 276, 330 Wintermeyer, W., 74–75 Winter, R., 414 Wolfson, A., 330 Woodson, S. A., 190, 193, 197–198, 201–203, 205, 412 Woolhead, C. A., 74 Wowor, A. J., 288 Wozniak, A. K., 167 Wray, W. O., 366 Wriggers, W., 243 Wu, H. M., 190–191 Wu, J. Z., 473 Wu, Y. H., 241 Wyatt, J. R., 31 Wyman, J. Jr., 444 X Xie, Z., 48, 53, 56 Xing, F., 74, 78 Xiong, Y., 72 Xi, X., 320, 324 Xu, D. R., 430
Y Yarus, M., 203 Yi, M., 430 Young, M. A., 419 Yuan, C., 192 Yu, Y. T., 28 Z Zachau, H. G., 74–75 Zalatan, J. G., 377 Zaug, A. J., 197, 238 Zawadzki, V., 31 Zeitlin, S., 107 Zelin, E., 107, 109 Zhang, C. M., 72, 75, 83–84, 86–87 Zhang, Q., 288 Zhang, X., 303–326 Zhang, Z., 324 Zheng, W., 241 Zhou, H. X., 472 Zhuang, X. W., 45, 48, 51, 56, 179–182 Zimm, B. H., 145, 191–192 Zinkel, S. S., 191 Zou, X., 475
Subject Index
A Affinity chromatography method, RNA purification DNA templates preparation in vitro transcription, 9 PCR reaction, 10–11 T7 RNA polymerase, 9 universal DNA primers, 9 HMM protein preparation amylose resin, 9 buffers and media, 7 expression of, 8 59-kDa protein, 6 transcription and purification, RNA QIAGEN Ni-NTA spin column, 13 reagents and buffers, 12 Aminoacyl-tRNA (aa-tRNA), 70, 82 2-Aminopurine (2AP) advantages, 269–270 RNA oligonucleotides, 273 steady-state fluorescence group I ribozyme, 274 TPP riboswitch folding, 276–278 structure and photophysics fluorescence properties, 270 Jablonski diagrams, 271 lifetime measurement, 272 quenching, 270–272 use, 270 time-resolved fluorescence intensity decay, 278–284 Analytical ultracentrifugation (AUC), RNA molecules Beckman optima analytical ultracentrifuge, 215 DCDTþ analysis, 227 group I intron ribozyme, 225 IRES RNA, 225–226 materials and instrumentation cell assembly, 211 reagents, 210–211 M-box riboswitch, 226 RNA folding experiment SV and SE, 206 Tetrahymena thermophila, 229 tRNA, 224–225 Anion exchange chromatography, native purification RNAs
cell culture and plasmid purification, 16–19 in vitro transcription magnesium chloride optimization, 19 pyrophosphatase addition, 20 materials, 15 plasmid DNA template clone, 16 weak anion-exchange FPLC separation of, 23 T7 RNA polymerase bands, 22 Anomalous small angle X-ray scattering (ASAXS) analysis ionic species competition, 403–405 vs. NLPB simulations, 405–407 spatial distributions comparison, 403 beamline detector, 400 sample cells, 399–400 schematic view, 399 data acquisition absorption edge energy, 400–401 Eon and Eoff profiles, 401–403 time-dependent systematic error, 401 electrical neutrality, 393 ion distributions measurements counterions, 394–395 resonant edge, 394 sample preparation, 398 SAXS, 393 schematic view, 399 signal interpretation form factors, 397 intensity measurement, 395–396 scattering factor, 395 techniques, 393 ASAXS. See Anomalous small angle X-ray scattering B BE-AES. See Buffer exchange-atomic emission spectroscopy Buffer exchange-atomic emission spectroscopy (BE-AES) buffer equilibration analyte concentration, 379–380 concentration determination, 379 criteria, 380 exchanges, 379 limitation, 380–382
497
498
Subject Index
Buffer exchange-atomic emission spectroscopy (BE-AES) (cont.) cacodylate anion concentration, 383 equilibrium establishment, 377–378 ICP-AES, 382–3383 ion atmosphere, 376 Mg2þ and Naþ association, 377 protocol aliquots, 386 data analysis, 387–388 DNA oligos, 384 Microcon YM-30 spin column, 385 Na-EPPS solution, 384–385 reagents, 385 standard curve, 386–387 schematic workflow, 378 techniques, 376 thermodynamic measure, 376 ultrafiltration spin columns, 377 C Cavitation, 366 Complex RNAs assembly, splinted ligation dye-labeled telomerase RNA preparation, 30 RNA ligation method dye-labeling reaction, 39 FRET-labeled telomerase RNA, 42–43 HPLC purification, 39–40 protocol, 38 reagents, 39 reversed-phase HPLC, 37 strategies for, 41 T4 RNA ligase 2, 40 single-molecule FRET measurements applications, 44 biotin moiety, 45 Tetrahymena thermophila telomerase RNA, 29 unmodified RNA ligation precursor molecules preparation, 31–33 Continuous flow mixers, SAXS method fluid velocity, 258–259 hydrodynamic focusing mixer, 259 Mg2þ ions, 257 premixing time, 258 time resolution, 257–258 Continuous-wave electron paramagnetic resonance (cw-EPR) spectroscopy acquisition and processing Bruker EMX X-band, 308 cavity coupling, 310–311 modulation phase, 312–313 parameters, 311 receiver gain, 313 sample preparation, 310 signal averaging, 313 spectral distortion, 313–315
spectrometer layout, 309 steps, 308, 315–318 advantage, 304 analysis nitroxide dynamics, 318–319 semiquantitative approach, 319–320 spectral line-shape, 320 HIV TAR RNA 20 -labeling strategy, 320 ligand effects, 320–322 spin-labeled nucleotides, 321 Tat interaction, 322–324 RNA/RNA interactions, 324–326 GAAA tetraloop binding, 325 JR effects, 324–326 site-directed spin-labeling drawbacks, 306–307 nitroxides, 304–305 nucleic acid synthesis, 305–307 nucleotide selection, 305 phosphoramidite, 305 postsynthetic labeling, 307–308 Counterion condensation (CC) theory, 471 D Data analysis, SAXS 3-D shape reconstruction atomic resolution structure, 237 normalized spatial discrepancy (NSD), 236 SUPCOMB, 236 model free analysis Guinier analysis, 234 Kratky plot, 234 Porod’s law, 234 radius of gyration, 235 DEER. See Pulsed electron-electron double resonance (PELDOR) Deoxyribozymes use, RNA research branched RNA products formation individual deoxyribozymes, 109 lariat RNA, 107 synthesis of, 108 for cleavage procedure, 100–101 reagents, 100 deoxyribozyme-catalyzed labeling (DECAL), RNA labeled tagging RNA preparation, 111–113 overview of, 111 linear RNA products synthesis experimental preparation, 105–106 20 –50 RNA ligation, 104–105 30 –50 RNA ligation, 102–104 splint ligation, 102 1,3-Diaza-2-oxophenothiazine (tC), 291 1,3-Diaza-2-oxophenoxazine (tCo), 291
499
Subject Index E Electron-multiplying charge-coupled device (EMCCD) chips, 175 Electrostatic forces prediction, RNA folding correlation-corrected PB model, 473–474 counterion condensation (CC) theory, 471 features, 466–467 ion dehydration effects, 484 Kirkwood/BBGY hierarchy, 473 mean-field approximation, 472 Mg2þ ions, 467 Poisson–Boltzmann (PB) theory, 472 RNA thermal stability, ion-dependence helix–helix assembly, 470 ion binding properties, 470 tertiary structures and RNA–RNA complexes, 471 size-modified PB theory, 472 stabilizing force, correlation-enhanced, 483 thermodynamic measurement, 467–469 tightly bound ion model application, 478–482 computational efficiency, 477–478 correlation parameter, 474 diffusive ions, free energy, 476 free energy (DGd), 475–476 ion correlation effect, 474–475 partition function (Z), 475 polarization energy, 476–477 types, 474 volume exclusion effects, 483 Equilibrium tertiary folding, AUC group I intron ribozyme, 225 IRES RNA, 225–226 materials and instrumentation cell assembly, 211 reagents, 210–211 M-box riboswitch, 226 RNA folding experiment SV and SE, 206 tRNA, 224–225 F Fluorescence polarization anisotropy (FPA) helical dynamics study junction model construct, 294–299 Tetrahymena group I intron ribozyme, 299–301 measurement anisotropy and background signal, 289 factors, 290 L-format and T-format, 289 polarizer alignment, 290 wavelength selection, 289–290 11mer control RNA, simple duplex Fluorolog-3 spectrometer, 293 sample preparation, 292–293
sequence design, 292 normalization ratio, 301 probes advantages, 291 dyes, 291 fluorescent base analog, 291–292 properties, 292 Fluorescence resonance energy transfer (FRET), 354–355 chemical structure, fluorophore, 164 emission spectrum, Cy3, 165 fluorophore orientation effect Cy3 structure, 161 Energy transfer efficiency, 162 measurement, 44–45 RNA species construction, fluorophorelabeled, 166 single-molecule, 175–178 steady-state measurements, 167–171 theory of, 158–159 time-resolved measurements, 172–174 Fluorescent tRNA CCA sequence 2AP base, 88 2AP to position 76 incorporation, 85 position 75, 87 cysteine-specific tRNA synthetase (CysRS), 72 Dus1p, 71 labeling, D residues HPLC chromatography, 81 scheme of, 76 target tRNA preparation, 77 U to D conversion, 78–79 tRNACys recognition, 72 FPA. See Fluorescence polarization anisotropy Free energy (DGd), TBI model, 475–476 FRET. See Fluorescence resonance energy transfer FRET-labeled telomerase RNA, 42–43 Full-length telomerase RNA, in vivo transcription dye-labeling reaction, 39 HPLC purification, 39–40 protocol, 38 reagents, 39 reversed-phase HPLC, 37 G Gel electrophoresis analysis branched RNA, 153 discrete bend, 143–144 experimental strategies, 145–146 four-way RNA junctions, 147–150 kink turns, 147 long-short arm method, 144–145 principle and theory, 142–143 three-way RNA junctions, 150–152 Gibbs–Helmholtz equation, 359 Guinier analysis, 234
500
Subject Index H
Helical dynamics study, FPA junction model construct with LacI, 298–299 without LacI, 294–296 lac repressor, 296–298 Tetrahymena group I intron ribozyme, 299–301 His-tagged MBP-MS2 coat fusion protein (HMM), 6 I Inductively coupled plasma atomic emission spectroscopy (ICP-AES), 382–383 Interaction coefficient, 435 Ion-RNA interactions, thermodynamic analysis equilibrium dialysis experiment binding density, 438–439 interaction coefficient, 437 ion concentration, 437 parameters, 438 process, 435–437 interaction coefficients, 435, 439–440 ion activities, 440–441 Mg2þ effect, 434–435 RNA conformational changes, 441–444 activity coefficient, 442 equilibrium constant, 442 free energy change, 441 g, G, and free energies link, 442–444 neutral activity, 441 salt dependence data analysis MgCl2 dependence, 448–461 RNA folding reactions, 444–448 IRE hairpin RNA loop 2APA tetraloop, 283 decay spectrum, 281–282 FluoFit, 280–281 fluorescence emission decay component, 280 loop motions, 282–283 NMR structural data, 279–280 riboswitch, 283–284 sequence, 278–279 K 59-kDa protein, 6 Kink nucleic acids, 142 Kirkwood/BBGY hierarchy, 473 L Labeled RNA constructs design assembly of chemical ligations, 58 covalent incorporation, 56 deoxyribozymes, 57 enzymatic ligations, 57 heterogeneity, 57
Watson-Creek base pairing, noncovalent assembly, 56 construct assembly design circular permutations, 51 design strategies, 50 dye labeling, RNA fragments amino and thiol modified oligos, 54 commercial oligos labeling, 53 direct RNA labeling, 55 labeling reaction, 54 oligonucleotide synthesis, 55 dye selection cyanine dyes, 52 photophysical properties, 51–52 single molecule FRET experiments, 51 in vitro transcription polynucleotide kinase (PNK), 56 T7 RNA polymerase, 55 RNA oligos labeling, protocol purification, 60–61 reagent preparation, 59–60 sample labeling reaction, 60 T4 DNA ligase, large RNAs ligation protocol annealing, 62–63 flow-chart, 62 ligation troubleshoot, 64 product purification, 63–64 run ligation reaction, 63 Lac Repressor (LacI) cell growth and induction, 296–297 cell lysis, 297 FPLC purification, 297–298 transformation, 296 Linear RNA products synthesis experimental preparation, 105–106 20 –50 RNA ligation, 104–105 30 –50 RNA ligation, 102–104 M 6-Methylisoxanthopterin (6-MI), 291 Microwave power, 313–314 N Native polyacrylamide gel electrophoresis (Native PAGE) electrophoresis equipment gel boxes, 189 Tetrahymena ribozyme, 188 folded RNA stability casting and prerunning gels, 190–191 data analysis, 192–193 sample preparation, 191–192 ligand induced change, 197–198 PAGE mobility, 194–195 RNA folding kinetics, 193–194
501
Subject Index
RNA measurement, 2-D PAGE, 195–197 theory of chemical exchange, 187–188 macromolecules mobility, 185–186 Native purification, RNAs affinity chromatography method DNA templates preparation, 9–11 HMM protein preparation, 6–9 transcription and purification, RNA, 12–14 anion exchange chromatography cell culture and plasmid purification, 16–19 in vitro transcription, 19–20 plasmid DNA template clone, 16 weak anion-exchange FPLC, 20–24 N-methylisatoic anhydride (NMIA), 129 Nonlinear Poisson–Boltzmann (NLPB) equation vs. anomalous small angle X-ray scattering (ASAXS), 405–407 helical grooves, 429 hydrated ions, 430 ion atmospheres analysis, 427 spatial free energy density, 427–429 Stern layer, 429–430 Tar–Tar* kissing loops, 429 P PAGE mobility, 194–195 Pake pattern, 334–335 Partition function, 475 PCC. See Pearson correlation coefficient Pearson correlation coefficient (PCC), 419–420 PELDOR. See Pulsed electron-electron double resonance Poisson–Boltzmann (PB) theory, 472 Polynucleotide kinase (PNK), 56 pUC18, 16 Pulsed electron–electron double resonance (PELDOR) beyond interval exchange coupling, 345–346 label orientation, 341–345 spin counting, 340–341 data analysis background function, 338–339 distance determination, 339 EPR methods, 330–331 vs. FRET, 346–347 vs. NMR, 346 Pake pattern, 334–335 protocol echo signal optimization, 336 four-pulse sequence, 337–338 inversion pulse, 336 probe pulse frequency, 336 pulses detection, 337 signal-to-noise ratio, 338 pulse sequence, 331–333 RNA structure, 347
sample preparation, EPR, 335 structure generation, 339 time trace, 333–334 R Radial distribution functions (RDFs), 415–416 Refocused echo (RE), 332 RNA crystallography crystal quality improvement, postcrystal analysis analysis, 130 cation additives, 127–128 crystal packing, 128 flowchart, 121 library construction, crystallization trials intermolecular and intramolecular RNA packing, 125–126 peripheral helical lengths variation, 126–127 phylogenetic variants, 124–125 lysine riboswitch regulatory element SAD electron density map, 134 thermotoga maritime, tertiary structure, 133 phasing methods heavy metal cations and nucleotide derivatives, 131–132 molecular replacement, 132 selection and initial characterization folded state characterization, 122–123 native purification, 123–124 RNA families (Rfam) database, 124 RNA folding. See also Native polyacrylamide gel electrophoresis electrostatic forces prediction computational efficiency, 477–478 correlation-corrected PB model, 473–474 counterion condensation (CC) theory, 471 features, 466–467 Kirkwood/BBGY hierarchy, 473 Mg2þ ions, 467 Poisson–Boltzmann (PB) theory, 472 size-modified PB theory, 472 thermal stability, ion-dependence, 467–471 thermodynamic measurement, 467–469 tightly bound ion model, 474–482 steady-state fluorescence group I ribozyme, 274 TPP riboswitch folding, 276–278 T-jump IR measurement advantage, 354 dual fluorophore, 355 features, 357 FRET, 354–355 IR spectral properties, 356–361 methods, 361–366 schematic representation, 363 tetraloop formation, 369–370 thermal equilibrium, 355 tRNA, 366–368
502
Subject Index
RNA fragment labeling amino and thiol modified oligos, 54 commercial oligos labeling, 53 direct RNA labeling, 55 dye selection cyanine dyes, 52 photophysical properties, 51–52 single molecule FRET experiments, 51 labeling reaction, 54 oligonucleotide synthesis, 55 RNA interactions simulations, monovalent ions canonical A-RNA and B-DNA helices counterion density, 425–426 features, 421–422 free energy regime., 423–425 melting temperature, 423 Naþ/Kþ accumulation, 426 spatial free energy density, 422–423 Tar–Tar* loop, 423 finite size artifacts coions exclusion, 417 counterions accumulation, 415 cumulative distribution function (CDF), 418–419 ion relaxation, kinetics, 419 macroion, 417–418 molecular dynamics simulations, 414 parameters selection, 415 PCC, 419–420 preferential interaction coefficients, 417 radial distribution functions (RDFs), 415–416 neutralizing counterions, 421 nonlinear Poisson–Boltzmann (P–B) equation atomistic simulations, 426 helical grooves, 429 hydrated ions, 430 ion atmospheres analysis, 427 spatial free energy density, 427–429 Stern layer, 429–430 Tar–Tar* kissing loops, 429 structures, 413 Tar–Tar* kissing-loop system, 412–413 RNA thermal stability, ion-dependence helix–helix assembly, 470 ion binding properties, 470 tertiary structures and RNA–RNA complexes, 471 Rotational correlation time, 319 S Salt dependence data analysis, ion-RNA interactions MgCl2 dependence direct measurement, 458–459 isothermal titration, 452–457 measurement methods, 459–461 melting curves, 457–458
mixed divalent–monovalent cation solution, 448–450 titration relation, 450–452 RNA folding reactions A-riboswitch RNA, 445–447 tar–tar* kissing-loop, 447–448 SAXS. See Small-angle X-ray scattering Single molecule fluorescence resonance energy transfer (smFRET) measurements, 48 Size-modified PB theory, 472 Small-angle X-ray scattering (SAXS), 393 data acquisition, 255–256 data analysis 3D electron density map, 264–266 3-D shape reconstruction, 235–237 kinetic folding pathway, 262–263 Kratky plot, 264 model free analysis, 233–235 radius of gyration, 262 shape generation algorithms, 266 singular value decomposition (SVD), 263–264 low-resolution atomic scale models cis-acting form, 239 VS ribozyme, 237 methods continuous flow mixers, 257–259 stopped flow mixers, 256–257 mixer fabrication flux damage, 261 PDMS–silicon device, 259–260 Tetrahymena ribozyme, 261 X-ray beam, 260 polynucleotide denaturation, 254 radiation damage effect, 233 schematic representation of, 255 small functional RNA change, 232–233 structure determination, low resolution, 232 thermodynamics determination, bead models metal ion core, 241 Poisson–Boltzmann theory, 240 P4-P6 mutant, 242 stereoscopic image, 240 Spectral integration, cw-EPR, 316–318 Splinted ligation, RNA assembly dye-labeled Telomerase RNA preparation, 30 RNA ligation methods FRET-labeled telomerase RNA, 42–43 strategies for, 41 T4 RNA ligase 2, 40 RNA ligation precursor molecules dye-labeling reaction, 39 HPLC purification, 39–40 protocol, 38 reagents, 39 reversed-phase HPLC, 37 RNase H cleavage, 31–33 single-molecule FRET measurements
503
Subject Index
applications, 44 biotin moiety, 45 Tetrahymena thermophila telomerase RNA, 29 unmodified RNA ligation precursor molecules preparation, 31–33 T Tar–Tar* kissing-loop system, 412–413 Temperature jump infrared (T-jump IR) measurement advantage, 354 dual fluorophore, 355 features, 357 FRET, 354–355 IR spectral properties D2O solution, 361 mononucleotide concentrations, 361 spectral assignment, 357–358 structure sensitivity, 356–357 thermodynamics, 358–361 methods FTIR spectroscopy, 362–363 sample cells, 362 sample preparation, 361–362 time-resolved spectroscopy, 363–366 schematic representation, 363 tetraloop formation, 369–370 thermal equilibrium, 355 tRNA parallel pathway folding model, 368 relaxation time, 367 transient absorption, 366 Tetrahymena group I intron ribozyme, 299–301 Thermodynamics analysis ion-RNA interactions equilibrium dialysis experiment, 435–439 interaction coefficients, 435, 439–440 ion activities, 440–441 Mg2þ effect, 434–435 RNA conformational changes, 441–444
salt dependence data analysis, 444–461 T-jump IR measuremens DCp determination, 359 fraction unfolded (fU), 358 Gibbs–Helmholtz equation, 359 melt curve, 360–361 Tightly bound ion model, electrostatic forces prediction application folding stability, 481–482 ion-binding properties, 480–481 partition function, 478–480 computational efficiency, 477–478 correlation parameter, 474 diffusive ions, free energy, 476 free energy (DGd), 475–476 ion correlation effect, 474–475 partition function (Z), 475 polarization energy, 476–477 types, 474 Time-resolved fluorescence intensity decay, 278–284 Time-resolved spectroscopy, T-jump IR measurement cavitation, 366 diode laser, 363–364 D2O solvent, 364 drawbacks, 365 MCT detector, 364–365 Stokes shift, 364 U Ultrafiltration spin columns, BE-AES, 377 W Watson-Creek base pairing, 56 X X-ray beam, SAXS, 260