RNA Chaperones: Methods and Protocols [1st ed. 2020] 978-1-0716-0230-0, 978-1-0716-0231-7

Table of contents :
Front Matter ....Pages i-xi
Biochemical Methods for the Study of the FinO Family of Bacterial RNA Chaperones (Hyeong Jin Kim, Steven Chaulk, David Arthur, Ross A. Edwards, J. N. Mark Glover)....Pages 1-18
Quantitative Analysis of RNA Chaperone Activity by Native Gel Electrophoresis and Fluorescence Spectroscopy (Subrata Panja, Ewelina M. Małecka, Andrew Santiago-Frangos, Sarah A. Woodson)....Pages 19-39
Fluorescent Molecular Beacons Mimicking RNA Secondary Structures to Study RNA Chaperone Activity (Pilar Menendez-Gil, Carlos J. Caballero, Cristina Solano, Alejandro Toledo-Arana)....Pages 41-58
Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein Deduced from Hairpin Unfolding (Micah J. McCauley, Ioulia Rouzina, Mark C. Williams)....Pages 59-88
Real-Time Fluorescence-Based Approaches to Disentangle Mechanisms of a Protein’s RNA Chaperone Activity (Tobias Schmidt, Susann Friedrich, Ralph P. Golbik, Sven-Erik Behrens)....Pages 89-106
Use of tRNA-Mediated Suppression to Assess RNA Chaperone Function (Jennifer Porat, Mark A. Bayfield)....Pages 107-120
Salt-Dependent Modulation of the RNA Chaperone Activity of RNA-Binding Protein La (Gunhild Sommer, Christina Sendlmeier, Tilman Heise)....Pages 121-136
Kinetic and Thermodynamic Analyses of RNA–Protein Interactions (Ryo Amano, Taiichi Sakamoto)....Pages 137-150
Detection of MicroRNAs Released from Argonautes (Kyung-Won Min, J. Grayson Evans, Erick C. Won, Je-Hyun Yoon)....Pages 151-159
Mapping the RNA Chaperone Activity of the T. brucei Editosome Using SHAPE Chemical Probing (W.-Matthias Leeder, H. Ulrich Göringer)....Pages 161-178
RNA Remodeling by RNA Chaperones Monitored by RNA Structure Probing (Susann Friedrich, Tobias Schmidt, Sven-Erik Behrens)....Pages 179-192
In Vivo RNA Chemical Footprinting Analysis in Archaea (Robert Knüppel, Martin Fenk, Michael Jüttner, Sébastien Ferreira-Cerca)....Pages 193-208
RNA Structure Analysis by Chemical Probing with DMS and CMCT (José M. Andrade, Ricardo F. dos Santos, Cecília M. Arraiano)....Pages 209-223
Disordered RNA-Binding Region Prediction with DisoRDPbind (Christopher J. Oldfield, Zhenling Peng, Lukasz Kurgan)....Pages 225-239
Delivering Molecular Beacons via an Electroporation-Based Approach Enables Live-Cell Imaging of Single RNA Transcripts and Genomic Loci (Shiqi Mao, Yachen Ying, Xiaotian Wu, Antony K. Chen)....Pages 241-252
Site-Specific Dual-Color Labeling of Long RNAs (Meng Zhao, Richard Börner, Roland K. O. Sigel, Eva Freisinger)....Pages 253-270
Single-Molecule FRET Assay for Studying Cotranscriptional RNA Folding (Heesoo Uhm, Sungchul Hohng)....Pages 271-282
Characterizing Complex Nucleic Acid Interactions of LINE1 ORF1p by Single Molecule Force Spectroscopy (M. Nabuan Naufer, Mark C. Williams)....Pages 283-297
Isolation and Analysis of Bacterial Ribosomes Through Sucrose Gradient Ultracentrifugation (Ricardo F. dos Santos, Cecília M. Arraiano, José M. Andrade)....Pages 299-310
Back Matter ....Pages 311-314

Citation preview

Methods in Molecular Biology 2106

Tilman Heise Editor

RNA Chaperones Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

Series Editor John M. Walker School of Life and Medical Sciences University of Hertfordshire Hatfield, Hertfordshire, UK

For further volumes: http://www.springer.com/series/7651

For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-bystep fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.

RNA Chaperones Methods and Protocols

Edited by

Tilman Heise Department of Pediatric Hematology, Oncology and Stem Cell Transplantation, University Hospital of Regensburg, Regensburg, Germany

Editor Tilman Heise Department of Pediatric Hematology, Oncology and Stem Cell Transplantation University Hospital of Regensburg Regensburg, Germany

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-0230-0 ISBN 978-1-0716-0231-7 (eBook) https://doi.org/10.1007/978-1-0716-0231-7 © Springer Science+Business Media, LLC, part of Springer Nature 2020 Open Access Chapter 3 is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). For further details see license information in the chapter. This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.

Preface The RNA folding problem describes the circumstance that RNA molecules can fold into various secondary and tertiary structures and that only some of those structures are of functional relevance. Hence, the intrinsic nature of RNA to fold into nonfunctional structures implies that mechanisms exist to convert a nonfunctional to a functional RNA structure. RNA might fold into helical regions; might form hairpins, bulges, and loops; and might establish three-dimensional contacts such as pseudoknots. The larger an RNA molecule, the more folding possibilities exist, leading to a compendium of coexisting and interchanging RNA conformations in solution. The discovery of small and long noncoding RNAs annealing in trans to other RNA molecules requires activities facilitating strand annealing and strand separation steps. Software tools and methods exist to predict RNA structures and to confirm the predicted folding of RNA into a specific structure. However, in a cellular context, considering RNA modifications and the presence of RNA-binding proteins which both may significantly influence the RNA fold, the validation of predicted RNA structures is of tremendous importance. The dynamics of RNA folding contribute to the regulation of various cellular processes including transcription, mRNA splicing, translation, mRNA decay, ribosome assembly, ribozyme activity, and viral replication. Hence, it is vital to understand in what way an RNA fold can be converted to an alternative fold, how an RNA hairpin structure can be formed or destabilized, or to what extent two single RNA molecules anneal or separate. RNA is always associated with RNA-binding proteins (RBPs) forming ribonucleoprotein particle (RNPs). Some RBPs are RNA helicase destabilizing, for example, stable RNA hairpins under the consumption of ATP. However, the focus of this book are protocols to study RNA chaperones, a specific class of RBPs proposed to assist structural changes in RNA molecules. RNA chaperones are promoting strand annealing (matchmaker activity) or strand separation, thereby changing structural elements in a given RNA molecule without the consumption of ATP. To assist RNA restructuring, an RNA chaperone has to bind the RNA, resolve the existing structure, and allow the formation of an alternative structure. This chaperoning from an energetically trapped RNA structure to an alternative structure requires activation energy. An RBP can bind RNA in a highly specific and tight way, or they are less specific and form weak interaction with the RNA. It is likely that a highly specific RBP selects for a specific RNA fold due to its specific sequence and structural characteristics required for specific interaction. RBPs with low sequence/structure specificity are likely to bind a variety of RNA folds. RNA chaperones often contain intrinsic disorder regions. This observation led to a model of how RNA chaperones assist structural changes in RNA. It is assumed that upon binding of the RNA, the intrinsic disordered regions of the RNA chaperone are highly organized, allowing additional protein-RNA interactions to occur. During this process of RNA chaperone organization, an “entropy transfer” occurs, meaning that higher organization of the RNA chaperone (loss in entropy) leads to unfolding (increase in entropy) of the RNA. Upon RNA chaperoning, RNA chaperones are releasing the restructured RNA, raising the question of whether the released RNA and RBP fold again into alternative

v

vi

Preface

structures. In contrast to this transient RBP-RNA complex formation, it is also possible that a weak and low specific interaction leads to restructuring of the RNA, evoking a specific interaction and a tight RNP complex formation. In this model, an initial weak RBP-RNA interaction initiates the formation of a strong RBP-RNA interaction. Such a tight complex might be critical in order to keep the RNA sequence/structure stabilized and further provide a nucleation point for assembly of an even larger RNP. RNA chaperones have been described over some decades and are well defined in bacteria such as the FinO domain containing proteins (Chapter 1), the intensively studied protein Hfq (Chapter 2), and the bacterial protein CspA (Chapter 3). Also, some viral nucleocapsid proteins display RNA chaperone activity (Chapter 4). In eukaryotic cells, RNA chaperones are less well characterized; however, growing evidence suggests that RNA chaperones such as NF90, AUF1 (Chapter 5), La (Chapters 6 and 7), Unr, and cold shock domain (CDS) containing proteins, such as YBX1, play important roles in the posttranscriptional control of gene expression. This book contains protocols to study the RNA chaperone activity of FinO (Chapter 1), Hfq (Chapter 2), CspA (Chapter 3), viral nucleocapsid proteins (Chapter 4), NF90 and AUF1 (Chapter 5), and La (Chapters 6 and 7) in vitro. Those basic protocols should be easily adaptable to study the function of other known or potential new RNA chaperones. RNA chaperones bind the RNA substrate; therefore, a protocol to study protein-RNA interactions using surface plasmon resonance (SPR) and isothermal titration calorimetry (ITC) is given (Chapter 8) as well as a protocol to investigate the release of miRNAs from RBPs (Chapter 9). To examine the folding and the structure of RNA, various detailed protocols are included (Chapters 10–13). Instructions for a software program are provided which can be applied to predict the RNA binding of a protein (Chapter 14). Furthermore, protocols for the delivery of molecular beacons into cells (Chapter 15) and the internal labeling of RNA molecules (Chapter 16) are provided. Finally, two protocols describe single molecule approaches (Chapters 17 and 18), and one protocol summarizes the steps for purification of ribosomes using gradient centrifugation (Chapter 19). This book provides a wide spectrum of methods to study RNA chaperones in vitro, at the single molecule level, and protocols useful for cell-based assays. Hence, I hope that the excellent chapters and the topics covered in this volume will be of avail to scientist and students interested in RNA biology and RNA chaperones. Finally, I would like to thank all the authors for their wonderful and comprehensive chapters describing such a variety of refined protocols. It was a compelling time for me to identify the authors and to work with them to create this volume. Furthermore, I would like to thank John Walker, the series editor, for the invitation to edit the volume RNA Chaperones: Methods and Protocols and the many suggestions, tips, and patience. Furthermore, I would like to thank Anna Rakovsky, David C. Casey, and all those involved in the production of the book. Regensburg, Germany

Tilman Heise

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

1 Biochemical Methods for the Study of the FinO Family of Bacterial RNA Chaperones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hyeong Jin Kim, Steven Chaulk, David Arthur, Ross A. Edwards, and J. N. Mark Glover 2 Quantitative Analysis of RNA Chaperone Activity by Native Gel Electrophoresis and Fluorescence Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subrata Panja, Ewelina M. Małecka, Andrew Santiago-Frangos, and Sarah A. Woodson 3 Fluorescent Molecular Beacons Mimicking RNA Secondary Structures to Study RNA Chaperone Activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pilar Menendez-Gil, Carlos J. Caballero, Cristina Solano, and Alejandro Toledo-Arana 4 Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein Deduced from Hairpin Unfolding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Micah J. McCauley, Ioulia Rouzina, and Mark C. Williams 5 Real-Time Fluorescence-Based Approaches to Disentangle Mechanisms of a Protein’s RNA Chaperone Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tobias Schmidt, Susann Friedrich, Ralph P. Golbik, and Sven-Erik Behrens 6 Use of tRNA-Mediated Suppression to Assess RNA Chaperone Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jennifer Porat and Mark A. Bayfield 7 Salt-Dependent Modulation of the RNA Chaperone Activity of RNA-Binding Protein La . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gunhild Sommer, Christina Sendlmeier, and Tilman Heise 8 Kinetic and Thermodynamic Analyses of RNA–Protein Interactions . . . . . . . . . . . Ryo Amano and Taiichi Sakamoto 9 Detection of MicroRNAs Released from Argonautes . . . . . . . . . . . . . . . . . . . . . . . . Kyung-Won Min, J. Grayson Evans, Erick C. Won, and Je-Hyun Yoon 10 Mapping the RNA Chaperone Activity of the T. brucei Editosome Using SHAPE Chemical Probing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . W.-Matthias Leeder and H. Ulrich Go¨ringer 11 RNA Remodeling by RNA Chaperones Monitored by RNA Structure Probing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Susann Friedrich, Tobias Schmidt, and Sven-Erik Behrens 12 In Vivo RNA Chemical Footprinting Analysis in Archaea . . . . . . . . . . . . . . . . . . . . ¨ ppel, Martin Fenk, Michael Ju ¨ ttner, Robert Knu and Se´bastien Ferreira-Cerca

vii

v ix

1

19

41

59

89

107

121 137 151

161

179 193

viii

13 14 15

16 17

18

19

Contents

RNA Structure Analysis by Chemical Probing with DMS and CMCT . . . . . . . . . Jose´ M. Andrade, Ricardo F. dos Santos, and Cecı´lia M. Arraiano Disordered RNA-Binding Region Prediction with DisoRDPbind . . . . . . . . . . . . . Christopher J. Oldfield, Zhenling Peng, and Lukasz Kurgan Delivering Molecular Beacons via an Electroporation-Based Approach Enables Live-Cell Imaging of Single RNA Transcripts and Genomic Loci . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shiqi Mao, Yachen Ying, Xiaotian Wu, and Antony K. Chen Site-Specific Dual-Color Labeling of Long RNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . Meng Zhao, Richard Bo¨rner, Roland K. O. Sigel, and Eva Freisinger Single-Molecule FRET Assay for Studying Cotranscriptional RNA Folding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heesoo Uhm and Sungchul Hohng Characterizing Complex Nucleic Acid Interactions of LINE1 ORF1p by Single Molecule Force Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . M. Nabuan Naufer and Mark C. Williams Isolation and Analysis of Bacterial Ribosomes Through Sucrose Gradient Ultracentrifugation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ricardo F. dos Santos, Cecı´lia M. Arraiano, and Jose´ M. Andrade

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

209 225

241 253

271

283

299 311

Contributors RYO AMANO • The Institute of Medical Science, Project Division of RNA Medical Science, The University of Tokyo, Tokyo, Japan JOSE´ M. ANDRADE • Instituto de Tecnologia Quı´mica e Biologica Antonio Xavier, Universidade Nova de Lisboa, Oeiras, Portugal CECI´LIA M. ARRAIANO • Instituto de Tecnologia Quı´mica e Biologica Antonio Xavier, Universidade Nova de Lisboa, Oeiras, Portugal DAVID ARTHUR • Department of Biochemistry, University of Alberta, Edmonton, AB, Canada MARK A. BAYFIELD • Department of Biology, York University, Toronto, ON, Canada SVEN-ERIK BEHRENS • Charles Tanford Protein Center, Institute of Biochemistry and Biotechnology, Martin Luther University Halle-Wittenberg, Halle, Germany RICHARD BO¨RNER • Department of Chemistry, University of Zurich, Zurich, Switzerland; Laserinstitut Hochschule Mittweida, University of Applied Sciences Mittweida, Mittweida, Germany CARLOS J. CABALLERO • Instituto de Agrobiotecnologı´a, IDAB, CSIC-UPNA-Gobierno de Navarra, Pamplona, Navarra, Spain STEVEN CHAULK • Department of Biochemistry, University of Alberta, Edmonton, AB, Canada ANTONY K. CHEN • Department of Biomedical Engineering, College of Engineering, Peking University, Beijing, China ROSS A. EDWARDS • Department of Biochemistry, University of Alberta, Edmonton, AB, Canada J. GRAYSON EVANS • Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, SC, USA MARTIN FENK • Department of Biochemistry III, Institute for Biochemistry, Genetics and Microbiology, University of Regensburg, Regensburg, Germany SE´BASTIEN FERREIRA-CERCA • Department of Biochemistry III, Institute for Biochemistry, Genetics and Microbiology, University of Regensburg, Regensburg, Germany EVA FREISINGER • Department of Chemistry, University of Zurich, Zurich, Switzerland SUSANN FRIEDRICH • Charles Tanford Protein Center, Institute of Biochemistry and Biotechnology, Martin Luther University Halle-Wittenberg, Halle, Germany J. N. MARK GLOVER • Department of Biochemistry, University of Alberta, Edmonton, AB, Canada RALPH P. GOLBIK • Charles Tanford Protein Center, Institute of Biochemistry and Biotechnology, Martin Luther University Halle-Wittenberg, Halle, Germany H. ULRICH GO¨RINGER • Molecular Genetics, Darmstadt University of Technology, Darmstadt, Germany TILMAN HEISE • Department of Pediatric Hematology, Oncology and Stem Cell Transplantation, University Hospital of Regensburg, Regensburg, Germany SUNGCHUL HOHNG • Department of Physics and Astronomy, Seoul National University, Seoul, Republic of Korea; Institute of Applied Physics, Seoul National University, Seoul, Republic of Korea

ix

x

Contributors

MICHAEL JU¨TTNER • Department of Biochemistry III, Institute for Biochemistry, Genetics and Microbiology, University of Regensburg, Regensburg, Germany HYEONG JIN KIM • Department of Biochemistry, University of Alberta, Edmonton, AB, Canada ROBERT KNU¨PPEL • Department of Biochemistry III, Institute for Biochemistry, Genetics and Microbiology, University of Regensburg, Regensburg, Germany LUKASZ KURGAN • Department of Computer Science, Virginia Commonwealth University, Richmond, VA, USA W.-MATTHIAS LEEDER • Molecular Genetics, Darmstadt University of Technology, Darmstadt, Germany EWELINA M. MAŁECKA • T. C. Jenkins Department of Biophysics, Johns Hopkins University, Baltimore, MD, USA SHIQI MAO • Department of Biomedical Engineering, College of Engineering, Peking University, Beijing, China MICAH J. MCCAULEY • Department of Physics, Northeastern University, Boston, MA, USA PILAR MENENDEZ-GIL • Instituto de Agrobiotecnologı´a, IDAB, CSIC-UPNA-Gobierno de Navarra, Pamplona, Navarra, Spain KYUNG-WON MIN • Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, SC, USA; Department of Biology, College of Natural Sciences, Gangneung-Wonju National University, Gangneung, South Korea M. NABUAN NAUFER • Department of Physics, Northeastern University, Boston, MA, USA CHRISTOPHER J. OLDFIELD • Department of Computer Science, Virginia Commonwealth University, Richmond, VA, USA SUBRATA PANJA • T. C. Jenkins Department of Biophysics, Johns Hopkins University, Baltimore, MD, USA; GeneDx, Gaithersburg, MD, USA ZHENLING PENG • Center for Applied Mathematics, Tianjin University, Tianjin, People’s Republic of China JENNIFER PORAT • Department of Biology, York University, Toronto, ON, Canada IOULIA ROUZINA • Department of Chemistry and Biochemistry, Ohio State University, Columbus, OH, USA TAIICHI SAKAMOTO • Department of Life Science, Faculty of Advanced Engineering, Chiba Institute of Technology, Chiba, Japan ANDREW SANTIAGO-FRANGOS • T. C. Jenkins Department of Biophysics, Johns Hopkins University, Baltimore, MD, USA; Department of Microbiology and Immunology, Montana State University, Bozeman, MT, USA RICARDO F. DOS SANTOS • Instituto de Tecnologia Quı´mica e Biologica Antonio Xavier, Universidade Nova de Lisboa, Oeiras, Portugal TOBIAS SCHMIDT • Charles Tanford Protein Center, Institute of Biochemistry and Biotechnology, Martin Luther University Halle-Wittenberg, Halle, Germany CHRISTINA SENDLMEIER • Department of Pediatric Hematology, Oncology and Stem Cell Transplantation, University Hospital of Regensburg, Regensburg, Germany ROLAND K. O. SIGEL • Department of Chemistry, University of Zurich, Zurich, Switzerland CRISTINA SOLANO • Navarrabiomed-Universidad Pu´blica de Navarra (UPNA)-Complejo Hospitalario de Navarra (CHN), IDISNA, Pamplona, Navarra, Spain GUNHILD SOMMER • Department of Pediatric Hematology, Oncology and Stem Cell Transplantation, University Hospital of Regensburg, Regensburg, Germany ALEJANDRO TOLEDO-ARANA • Instituto de Agrobiotecnologı´a, IDAB, CSIC-UPNA-Gobierno de Navarra, Pamplona, Navarra, Spain

Contributors

xi

HEESOO UHM • Department of Physics and Astronomy, Seoul National University, Seoul, Republic of Korea; Institute of Applied Physics, Seoul National University, Seoul, Republic of Korea MARK C. WILLIAMS • Department of Physics, Northeastern University, Boston, MA, USA ERICK C. WON • Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, SC, USA SARAH A. WOODSON • T. C. Jenkins Department of Biophysics, Johns Hopkins University, Baltimore, MD, USA XIAOTIAN WU • Department of Biomedical Engineering, College of Engineering, Peking University, Beijing, China YACHEN YING • Department of Biomedical Engineering, College of Engineering, Peking University, Beijing, China JE-HYUN YOON • Department of Biochemistry and Molecular Biology, Medical University of South Carolina, Charleston, SC, USA MENG ZHAO • Department of Chemistry, University of Zurich, Zurich, Switzerland; Department of Physics, University of Alberta, Edmonton, AB, Canada

Chapter 1 Biochemical Methods for the Study of the FinO Family of Bacterial RNA Chaperones Hyeong Jin Kim, Steven Chaulk, David Arthur, Ross A. Edwards, and J. N. Mark Glover Abstract The FinO family of proteins constitutes a group of RNA chaperones that interacts with small RNAs (sRNAs) to regulate gene expression in many bacterial species. Here we describe detailed protocols for the biochemical analysis of the RNA chaperone activity of these proteins. Methods are described for preparation of RNA, RNA 50 end labeling with radioisotope and modified EMSA protocols to test the ability of these proteins to catalyze RNA strand exchange and RNA duplex formation. Key words FinO, RocC, ProQ, RNA chaperone, Chaperone activity, Strand displacement assay, Strand exchange assay, Duplexing assay, Annealing assay

1

Introduction RNA chaperones were initially defined by Herschlag as proteins that could overcome kinetic traps along RNA folding pathways to facilitate the stabilization of specific RNA structures to direct a biological outcome [1]. RNA chaperones in general can exhibit a number of functions on their target RNAs. These include protection of the target RNA against enzymatic or chemical degradation, destabilization of misfolded structures, and the facilitation of interactions between RNAs [1–3]. The FinO family of RNA chaperones is widely dispersed throughout prokaryotic species [4–9]. FinO, the first member of this family for whom an RNA chaperone function was identified, is a protein that facilitates sense-antisense RNA interactions to repress gene expression within F plasmids. FinO is highly specific, targeting a single plasmid encoded antisense RNA called FinP. The sense target of FinP is traJ mRNA, which encodes a major activator of F plasmid transcription. FinP-traJ mRNA binding leads to the degradation of traJ [10], and the direct

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_1, © Springer Science+Business Media, LLC, part of Springer Nature 2020

1

2

Hyeong Jin Kim et al.

repression of traJ translation, which together effectively blocks F plasmid gene expression and plasmid conjugation [10, 11]. Proteins containing FinO-like domains are widely dispersed throughout bacterial species [9, 12]. E. coli ProQ was first member of the larger family to be characterized. Early studies revealed ProQ regulated the proline transporter ProP [13, 14]. High-throughput sequencing approaches demonstrated that ProQ is in fact an RNA-binding protein that regulates sRNA-mediated gene regulatory processes by binding highly structured RNAs [15, 16]. Similarly, RNA-seq approaches led to the identification of a set of RNA targets for another FinO-containing protein from Legionella pneumophila, named RocC [7]. These studies have revealed that while ProQ appears to bind and regulate a range of structured sRNAs and mRNAs, RocC and FinO instead appear to be highly specific RNA binding proteins, interacting with only a limited range of sRNA targets. The FinO domain is a mixed α-β fold that serves as the key RNA-binding platform for this family of proteins [4, 17, 18]. There is as yet no high-resolution structural information for how these proteins bind their RNAs. However, protein-RNA crosslinking, FRET, RNase footprinting, and small-angle X-ray scattering studies (SAXS) have been used to develop a low-resolution model for how FinO may recognize minimal RNA targets consisting of a stemloop with a 30 single-stranded tail [19]. More recently, SAXS and hydrogen deuterium exchange mapping of RNA contacting residues have been used to develop a model of full-length ProQ and its interactions with a larger RNA target, SraB [18]. Notably, while this latter study implies that the FinO domain of ProQ acts as a major RNA binding determinant, it also implicates a second Tudorlike domain within ProQ and a flexible linker as also involved in RNA interactions [18]. How FinO chaperones mediate their effects on gene expression is still poorly understood. Early work on FinO revealed that this protein protects FinP against degradation by RNase E [10], and more recent work on ProQ [15] and RocC [7] has shown that these proteins also stabilize their bound RNAs against degradation. A deletion mutant of FinO consisting of the core folded domain lacking the flexible N- and C-terminal tails was shown to be sufficient to tightly and specifically bind FinP and protect FinP from degradation in vivo; however, this fragment did not repress plasmid conjugation [11]. Biochemical studies were used to suggest additional RNA chaperone activities that might explain the ability of FinO to repress traJ expression and conjugation. Early gel-based RNA annealing assays suggested that FinO could directly facilitate FinP interactions with traJ mRNA ~5-fold [20]. These interactions have been proposed to initiate via “kissing” interactions between complementary loops and to proceed to fully duplexed RNA complexes [21]. In the ProQ and RocC systems, the interacting RNAs

FinO RNA Chaperone Activity Assays

3

only display limited base pairing and rely on relatively short complementary seed sequences for RNA-RNA interactions [7, 15]. ProQ is able to bind to its sRNA-mRNA target complex, but it is unclear if this interaction is required to stabilize this complex [15]. The role of FinO chaperones in RNA-RNA pairing, whether between sense-antisense RNAs or partially complementary sRNAmRNA pairs, is unclear. Gel-based assays provide simple but powerful methods to directly test the effect of these proteins on RNA interactions. We have developed two related assays to probe FinO RNA chaperone activity: an RNA duplexing assay and an RNA strand exchange assay. The duplexing assay tests the ability of the chaperone to kinetically facilitate the interaction of two complementary RNAs, while the strand exchange assay monitors the ability of the chaperone to catalyze exchange of RNA strands between a double stranded target RNA and a single-stranded complementary strand. Using these assays, we have been able to show that not only FinO but also ProQ and the FinO ortholog from Neisseria meningitidis can facilitate duplexing between FinP and its complementary RNA [4, 5, 11]. Moreover, these results suggest that these proteins may act, at least in part, by destabilizing internal secondary structures within the individual RNAs that would otherwise kinetically block RNA-RNA interactions. In this chapter, we present detailed methods for the production and purification of in vitro transcribed RNA, and protocols for RNA binding, RNA duplexing and RNA strand exchange assays for RNA chaperone proteins.

2

Materials

2.1 In Vitro Transcription and Purification of RNA 2.1.1 In Vitro Transcription

1. 1.0 M Tris–HCl pH 8.0. 2. 1.0 M MgCl2. 3. 1.0 M NaCl. 4. 1.0 M dithiothreitol (DTT). 5. 0.1 M ATP, UTP, GTP and CTP. 6. 1.0 M spermidine. 7. T7 RNA polymerase. 8. Pyrophosphatase (E. coli). 9. DNA template for SII(A) and SII(B) RNA (see Note 1). 10. Ribonuclease inhibitor. 11. Triton X-100. 12. DEPC (diethyl pyrocarbonate)-treated water (DEPC water).

4

Hyeong Jin Kim et al.

2.1.2 Purification of RNA by Anion Exchange Chromatoraphy

1. Mixture of the transcribed RNA from Subheading 2.1.1. 2. DEPC water. 3. Acrylamide. 4. Bis-acrylamide. 5. Denaturing gel solution: 20% acryl/bis-acrylamide (19:1) in 8 M urea dissolved in 1 TBE. 6. N,N,N0 ,N0 -tetramethylethane-1,2-diamine (TEMED). 7. 10% ammonium persulfate (APS). 8. 10 TBE buffer: Dissolve 108 g Tris base, 55 g boric acid and 7.5 g ethylenediaminetetraacetic acid (EDTA) in 1 L of distilled water. 9. 20% Denaturing gel: Mix 320 μL of 10% APS in 40 mL of denaturing gel solution. The addition of 24 μL of TEMED induces the polymerization. It will be enough to make 6 gels of 0.75 mm thickness, 7.3 cm by 10.2 cm. 10. Denaturing dye: 8 M urea, 0.02% bromophenol blue, 0.02% xylene cyanol in 1 TBE buffer. 11. Low salt buffer: 10 mM HEPES-KOH pH 7.5, 50 mM KCl in DEPC water. 12. High salt buffer: 10 mM HEPES-KOH pH 7.5, 1500 mM KCl in DEPC water. 13. Syringe and needle. 14. Anion exchange chromatography column. 15. Biomolecular imager.

2.1.3 Purification of RNA by Gel Filtration Chromatography

1. Mixture of the transcribed RNA from Subheading 2.1.2. 2. DEPC water. 3. Gel filtration buffer (see Note 2): 10 mM HEPES-KOH pH 7.5, 100 mM KCl, 1 mM TCEP in DEPC water. 4. Denaturing gel solution: 20% acryl/bis-acrylamide (19:1) in 8 M urea dissolved in 1 TBE. 5. TEMED. 6. 10% APS. 7. 10 Tris–glycine buffer: Dissolve 30.0 g of Tris base and 144 g of glycine in 1 L of H2O. The pH of the buffer should be 8.3 and no pH adjustment is required. Store the running buffer at room temperature and dilute to 1 before use. 8. 10% native gel: Dissolve 50 mg of bis-acrylamide, 4 g of acrylamide, 320 μL of 10% APS in 1 Tris–glycine buffer.

FinO RNA Chaperone Activity Assays

5

9. 10 native gel loading dye: 10 mM Tris–HCl pH 8.0, 50% (v/v) glycerol, 0.001% bromophenol blue and 0.001% xylene cyanol. 10. Syringe and needle. 11. Gel filtration chromatography. 12. biomolecular imager. 2.2 Labeling RNA and Purification Using Polyacrylamide Gel Electrophoresis (PAGE)

1. Calf intestinal alkaline phosphatase (CIAP).

2.2.1 RNA Dephosphorylation

4. Transcribed RNA (see Subheading 2.1).

2. Ribonuclease inhibitor. 3. 10 dephosphorylation buffer: 500 mM Tris–HCl (pH 7.9 at 25  C), 1 M NaCl, 100 mM MgCl2, 10 mM DTT. 5. DEPC water. 6. 1.5 mL microcentrifuge tube. 7. Water bath (37 and 65  C). 8. ATP, [γ-32P]-6000 Ci/mmol 150 mCi/mL Lead, 1 mCi. 9. T4 polynucleotide kinase (T4 PNK).

2.2.2 50 End Labeling

1. Dephosphorylated RNA from Subheading 2.2.1. 2. Ribonuclease inhibitor. 3. DEPC water. 4. 1.5 mL microcentrifuge tube. 5. Water bath (37 ). 6. ATP, [γ-32P]-6000 Ci/mmol 150 mCi/mL Lead, 1 mCi. 7. T4 polynucleotide kinase (T4 PNK). 8. 10 reaction buffer for T4 PNK: 500 mM Tris–HCl (pH 7.6 at 25  C), 100 mM MgCl2, 50 mM DTT, 1 mM spermidine. 9. Denaturing dye.

2.2.3 RNA Purification Using PAGE and RNA Extraction Using Phenol-Chloroform Extraction

1. Glass gel plates (17 cm by 28 cm, customized). 2. 0.75 mm spacers and combs. 3. Clamps. 4. Paper towel. 5. Laboratory color tape. 6. Clear tape. 7. Syringe and needle. 8. Metal plate matched to the size of the gel to disperse heat. 9. Electrophoresis apparatus. 10. 15 mL conical tube.

6

Hyeong Jin Kim et al.

11. 20% acryl/bis-acrylamide (19:1) in 8 M urea dissolved in 1 TBE. 12. TEMED. 13. 10% APS. 14. Denaturing dye. 15. ddH2O. 16. 70% ethanol. 17. 95% ethanol. 18. 1 TBE buffer. 19. Plastic wrap. 20. Fluorescent marker. 21. Razor blade. 22. Disposable filter column. 23. X-ray film cassette. 24. Medical X-ray film. 25. X-ray processor. 26. 0.3 M sodium acetate pH 5.2. 27. Phenol (ultrapure buffer-saturated phenol). 28. Chloroform. 29. Phenol: chloroform: isoamyl alcohol (25:24:1, v/v). 30. Tabletop centrifuge. 31. Centrifuge. 32. Vortexer. 33. Vacuum centrifuge concentrator. 2.3 RNA Duplexing Assay

1. 50 nM 32P-labeled RNA. 2. Unlabeled RNA at least 10 more concentrated than the 32 P-labeled RNA. 3. 2 reaction buffer: 50 mM Tris pH 8.1, 40 mM NaCl, 100 μg/mL BSA, 10% glycerol and 0.1% β-mercaptoethanol. 4. Cold stop buffer: 5% glycerol, 0.4% SDS and 20 mM EDTA. 5. 10 μM RNA chaperone protein (FinO). 6. DEPC water. 7. 10 Tris–glycine buffer. 8. 10% non-denaturing gel: 31 mg bis-acrylamide, 2.5 g acrylamide, up to 25 mL in 1 Tris–Glycine. 200 μL 10% APS and 15 μL TEMED are added to initiate polymerization. 9. Electrophoresis apparatus. 10. 10 native gel loading dye.

FinO RNA Chaperone Activity Assays

7

11. Filter paper. 12. Autoradiography cassette. 13. Storage phosphor screen. 14. Gel dryer. 15. Laser scanner for biomolecular imaging applications. 16. Biomolecular imager. 2.4 Strand Exchange Assay

1. 50 nM 32P-labeled RNA RNA (SII(A)). 2. Unlabeled RNA (SII(B)) at least 10 more concentrated than the 32P-labeled RNA (SII(A)). 3. 10 μM RNA chaperone protein. 4. Ribonuclease inhibitor. 5. Annealing buffer: 10 mM Tris pH 8.1, 100 mM NaCl. 6. 2 reaction buffer. 7. DEPC water. 8. 10 Tris–glycine buffer. 9. 15% non-denaturing gel: 46.5 mg bis-acrylamide, 3.75 g acrylamide, made up to 25 mL with 1 Tris–Glycine buffer. 200 μL 10% APS and 15 μL TEMED is added to induce polymerization. 10. Cold stop buffer: 5% glycerol, 0.4% SDS and 20 mM EDTA in distilled water. 11. Electrophoresis apparatus. 12. Filter paper. 13. Autoradiography cassette. 14. Storage phosphor screen. 15. Gel dryer. 16. Laser scanner for biomolecular imaging applications. 17. Biomolecular imager.

3

Methods Key to biochemical studies of RNA structure and function is methodology for the production and purification of RNA in milligram quantities. Here we present our methods for production of RNA molecules using in vitro transcription using recombinant T7 RNA polymerase (T7 RNAP). We use T7 RNAP that is expressed and purified in house by established methods [22]. Denaturing gel electrophoresis has been extensively used for RNA purification; however, the yield of purified RNA can be low due to loss of

8

Hyeong Jin Kim et al.

Table 1 Components for in vitro transcription Stock

Final concentration

1.0 M Tris pH 8.0

40 mM

40

1.0 M MgCl2

10 mM

10

1.0 M NaCl

25 mM

25

1.0 M DTT

4 mM

4

0.1 M ATP

4 mM

40

0.1 M UTP

4 mM

40

0.1 M GTP

4 mM

40

0.1 M CTP

4 mM

40

1.0 M Spermidine

4 mM

4

T7 RNAP (1.2 mg/mL)

0.090 mg/mL

Phosphatase (100 U/mL)

0.25 U/mL

Template (10uM)

0.66uM

RNase OUT (40 U/μL)

0.1 U/μL

Triton x-100

1%

DEPC water Total

Volume (μL)

75 2.5 66 2.5 10 601 1000

product during extraction from the gel. We have developed an alternative RNA purification relying on anion exchange chromatography which enables the rapid purification of milligram quantities of pure RNA in high yield. 3.1 In Vitro Transcription and Purification

1. Mix all the components based on Table 1 for in vitro transcription.

3.1.1 In Vitro Transcription

3. Spin down the white precipitate and take the supernatant (see Note 4).

3.1.2 Anion Exchange Chromatography

1. Equilibrate the column with 2 column volumes (CV) of low salt buffer.

2. Incubate mixture at 37  C for 3 h (see Note 3).

2. Equilibrate the column with 2 CV of high salt buffer. 3. Equilibrate the column with 2 CV of low salt buffer. 4. Inject sample on column (see Note 5) and wash the column with 1 CV of low salt buffer.

FinO RNA Chaperone Activity Assays

9

Fig. 1 RNA purification procedure from in vitro transcription. (a) Chromatogram of anion exchange chromatography. Solid line indicates the absorbance at 280 nm and dotted line indicates the % of buffer B. (b) 6% urea-denaturing gel analysis of fractions from anion exchange chromatography. Grey bar in (a) corresponding to grey bar on (b). Peaks a and b are corresponding to peak from (a). 33 nt is a 33 nucleotide RNA purchased from company for control. (c) Chromatogram of gel filtration (d) 10% native gel from gel filtration. Grey bar indicates the fraction corresponding to grey bar from (c)

5. Elute RNA with a 50 mM to 1500 mM NaCl gradient (see Note 6). 6. Run fractions on a denaturing TBE-urea polyacrylamide gel (see Note 7) to identify pure fractions. 7. The gel is stained with EtBr in TBE buffer (1:10,000) for 30 min. 8. Visualize the gel with biomolecular imager (Fig. 1a, b). 3.1.3 Gel Filtration Chromatography (Optional (See Note 8))

1. Equilibrate the column with the 1 CV of gel filtration buffer. 2. Concentrate collected fraction from anion exchange using centrifugal filter around 1% of bed volume of gel filtration. 3. Inject samples on gel filtration through needle port using syringe. 4. Run fractions on a native PAGE (see Note 7) to identify pure fractions.

10

Hyeong Jin Kim et al.

5. The gel is stained with EtBr in TBE buffer (1:10,000) for 30 min. 6. Visualize the gel with biomolecular imager (Fig. 1c, d), (see Note 9). 3.2 Labeling RNA and Purification Using PAGE

Analysis of RNA interactions using native gel electrophoresis-based assays requires specifically labeled RNAs that can be visualized and quantitated within the gel. We routinely 50 -label our RNA samples with γ-32P-labeled ATP and polynucleotide kinase. The resulting labeled RNA is subsequently purified from unincorporated label and other contaminants by denaturing polyacrylamide electrophoresis.

3.2.1 RNA Dephosphorylation (See Note 10)

1. Mix 10 μL of CIAP (1 U/μL), 1 μL of ribonuclease inhibitor (40 U/μL), 2.5 μL dephosphorylation buffer (10), transcribed RNA (600 pmol) and fill up to 25 μL with DEPC water in 1.5 mL microcentrifuge tube. 2. Incubate mixture in the 37  C water bath for 30 min. 3. Transfer tube to 65  C water bath for 15 min to inactivate phosphatase. 4. This is the CIAP RNA used for labeling.

3.2.2 50 End Labeling

1. Mix 1 μL of T4 polynucleotide kinase (10 U/μL), 1 μL of ATP, [γ-32P], 1 μL of ribonuclease inhibitor (40 U/μL), 2.5 μL of 10 T4 PNK reaction buffer, CIAP RNA (20 pmol) and make up to 25 μL with DEPC water in 1.5 mL microcentrifuge tube. 2. Incubate mixture at 37  C for 30 min in the water bath. 3. Stop the reaction by adding equal volume of denaturing dye.

3.2.3 RNA Purification Using PAGE and RNA Extraction Using Phenol-Chloroform Extraction

1. Rinse the glass spacer plate, short plate, spacer and comb with deionized water followed by ethanol. Remove residual ethanol on the all components with paper towel. 2. Assemble all components and seal the bottom and sides of the glass plates with laboratory tape to prevent leaking. 3. Mix 40 mL of 20% acryl/bis-acrylamide (19:1) in 8 M urea dissolved in 1 TBE in the 50 mL conical tube with 320 μL 10% APS and 24 μL TEMED. Invert several times and try to avoid introduce air bubble in solution (see Note 11). 4. Tilt the plate 45 and gently pour the gel solution between the glass plates. Tap the plates on the bench to remove the residual air bubbles and position the comb between the plates. Polymerization will take around 30 min at room temperature. 5. Remove the sealing tape, comb and spacer on the bottom and assemble the electrophoresis apparatus. Attach the metal plate with clamps to the outside plate of the gel to distribute the heat

FinO RNA Chaperone Activity Assays

11

from high voltage. Fill the upper and bottom reservoir with 1 TBE buffer. 6. Rinse the wells with syringe (see Note 12). Load the denaturing dye that can completely cover the bottom of the wells. Pre-run the gel at 30 W for 30 min. 7. After the pre-run, remove the electrical leads and flush the wells with syringe. Load your samples in the wells. Run the samples at 30 W. Running time can be varied depending on the size of your RNA (see Note 13). 8. Once the electrophoresis is complete, turn off the power and remove the glass plates. Gently separate the glass plates—in general the gel will be stuck one side of the plate. Put the used X-ray film on the gel and flip so that the glass plate is on top, X-ray film on the bottom. Gently remove the remaining glass plate from the gel. Cover the gel with plastic wrap. 9. Place the gel covered with plastic wrap on the X-ray cassette. Attach the fluorescent marker with clear tape on the plastic wrap but don’t put them on the path of lane. After developing the X-ray film, you can find where you need to excise by overlapping the shade of fluorescent marker on the developed film. Close the cassette and bring cassette and X-ray film to dark room to develop X-ray film. 10. Turn on the red safety light and turn off the normal light. Open the cassette and the X-ray film box. X-ray film is light sensitive so it should not be exposed to the regular light. Put a piece of film on the gel in the cassette and lock your cassette. Expose the film for 30 s to 3 min and develop the film with X-ray processor. 11. Once film is developed, turn on the regular light and check the intensity of the band on the film. If it is faint, you can repeat step 10 with a longer exposure time. 12. Place your developed film on the denaturing gel covered with plastic wrap. Carefully overlay the marker on the gel and shade of marker on the developed film. Make a dotted line around your bands using a razor blade. Remove the film and you can still see dot line on the gel. Excise your target bands using the razor blade. 13. Make several holes on the conical part of a 500 μL microcentrifuge tube with a needle. Place the 500 μL microcentrifuge tube in the 1.5 mL microcentrifuge tube. Place the excised gel pieces in the 500 μL tube and close the lid. Place whole assembly in the table top centrifuge and spin gel down with 16,000  g for 5 min. The centrifugation will force the gel through the hole in the 500 μL tube, crushing the gel into

12

Hyeong Jin Kim et al.

small pieces. The crushing will speed diffusion of the RNA from the gel fragments. 14. Remove the 500 μL tube and take 1.5 mL tube with crushed gel. 15. Add 0.3 M sodium acetate pH 5.2400 μL and 10% phenol (v/v). 16. Place it on the 37  C shaker overnight. 17. Next day, spin down the residual liquid from the well and cap. Cut the linker connecting cap and conical tubes. Remove cap. 18. Assemble the 15 mL conical tube and disposable filter column. From the step 18, carefully put together the capless and disposable filter column. 19. Separate the aqueous phase from crushed gel by centrifugation at 2000  g for 5 min. 20. Transfer the liquid to a 1.5 mL tube and add 120 μL of phenol: chloroform: isoamyl alcohol (25:24:1, v/v) and vortex. 21. Separate organic and inorganic layer by spinning for 2 min with 16,000  g. 22. Tilt tube 45 and carefully remove the top aqueous layer containing your RNA and transfer to a fresh tube. Try not to touch the bottom layer. 23. Add 120 μL of chloroform and vortex. 24. Tilt tube 45 and carefully transfer the top layer into new 1.5 mL tube. Add 1 mL 95% ethanol and store at 20  C. 25. Next day, pellet the precipitated RNA by centrifugation at top speed in a pre-chilled table top centrifuge for 20 min. 26. Remove the ethanol with pipette carefully until ~100 μL is left. Do not touch the tube wall with pipette. 27. Rinse the RNA pellet on the wall of the tube with 70% ethanol. 28. Aspirate ethanol and evaporate remaining ethanol using vacuum centrifuge concentrator. 29. Dried RNA can be stored at 20  C and can be resuspended with an appropriate buffer when needed. 3.3 RNA Chaperone Activity Assays 3.3.1 RNA Duplexing Assay

FinO has been shown to bind the antisense RNA FinP and facilitate interaction with its sense RNA partner, traJ mRNA. Biochemical evidence for this was derived from RNA duplexing assays that monitor RNA-RNA interactions via native polyacrylamide gel electrophoresis. Particularly striking results were obtained with minimal substrate RNAs consisting of just the minimal stem-loop and 30 polyU tail from FinP, and the complementary region of traJ mRNA (Fig. 2a). In these assays, one RNA is 50 -32P labeled, and the other RNA is unlabeled. The RNAs are mixed with the unlabeled RNA in

FinO RNA Chaperone Activity Assays

13

Fig. 2 (a) Various FinP RNA constructs used for duplexing assay and strand exchange assay. SLIIx is from FinP and SLIIcx is complementary sequence of FinP SLIIx from TraJ mRNA. (b) Schematic diagram of duplexing assay. Star indicates the radioisotope labeling. (c) Schematic diagram of strand exchange assay

excess and incubated in the presence or absence of FinO. The products of the reaction are then resolved by native gel electrophoresis. To visualize only RNA-RNA interactions within the gel without effects from protein binding, the reactions are quenched with SDS to denature protein while leaving RNA structure intact (Fig. 2b). In this way, we were able to show that full length FinO protein could dramatically enhance the apparent rate of RNA-RNA association over a no protein control. Interestingly, these experiments revealed that the core RNA binding domain of FinO does not catalyze RNA duplexing. A flexible region of the protein N-terminal to the RNA binding domain is required for RNA duplexing (Fig. 3a). We have used this assay to also show that ProQ and the FinO-related protein from Neisseria meningitides, NMB1681, can also catalyze RNA duplexing, suggesting that this activity may be shared by many members of the FinO family of RNA chaperones [4, 5]. Duplexing Assay

1. Prepare unlabeled RNA (see Subheading 3.1) and labeled RNA (see Subheading 3.2). 2. Mix 25 μL of 2 reaction buffer, 5 μL of 10 μM protein, 5 μL of 1 μM unlabeled RNA, and 10 μL of DEPC water. Preincubate the mixture at 37  C for 5 min. 3. Preincubate 32P-labeled RNA at 37  C for 5 min.

14

Hyeong Jin Kim et al.

Fig. 3 Results of RNA chaperone activity assay. (a) Results of duplexing assay. Duplexing ability of FinO was tested by time course in various FinO constructs. (b) Results of strand exchange assay. Strand exchange ability of FinO was tested by time course in various FinO constructs. Results from Arthur and colleagues [11]

4. Start reaction by adding the 5 μL of 50 nM 32P-labeled RNA. 5. Take 5 μL of aliquot from reaction mixture at various time points (see Note 14). 6. Stop each reactions of aliquot by adding cold stop buffer. 7. Samples are separated on 10% non-denaturing PAGE at room temperature and electrophoresed at 180 V, 2 h. 8. After electrophoresis, the plates are removed and separated gel is transferred to Whatman filter paper covered with plastic wrap. Try to avoid any air bubbles and wrinkles. 9. The gel is dried on a gel drier at 80  C, 2 h until the gel is completely dry. 10. Dried gel is transferred to a storage phosphor screen and exposed 30 min to 2 h and scanned with a biomolecular imager. 3.3.2 RNA Strand Exchange Assay

The RNA duplexing assay suggested that pairing between FinP and traJ mRNAs are inhibited by the stable internal stem-loop structures within each RNA. We hypothesized that part of the FinO chaperone activity might be to destabilize these internal structures to facilitate sense-antisense recognition. To test this idea, we used an RNA strand exchange assay. In this assay, we used a two-stranded RNA mimic of the minimal FinO RNA target, SLII from FinP (Fig. 2a). Control experiments showed that this RNA binds FinO with the same affinity as intact SLII, consistent with footprinting experiments that indicated that the loop is not contacted by FinO

FinO RNA Chaperone Activity Assays

15

[19]. For the strand exchange assay, one strand (SII(A)) of the duplex is specifically labeled with 32P, and this duplex is challenged with a molar excess of an unlabeled version of the SII(A) strand. The RNA is incubated with or without FinO, and samples are taken, quenched with SDS to denature protein, and analyzed by native polyacrylamide gel electrophoresis (Fig. 2c). In these experiments the gel is continuously running so that double-stranded and single-stranded RNAs are separated upon loading on the gel. In this way, we could show that the SLII duplex RNA is extremely stable to challenge by the unlabeled SII(A) strand in the absence of FinO, demonstrating the kinetic stability of the base-paired stem. However, addition of FinO facilitated strand exchange so that essentially all the labeled SII(A) was exchanged from duplex to single strand over a 120 min time course. Notably, these experiments also demonstrated that the core FinO RNA-binding domain is not sufficient for strand exchange and the N-terminal flexible region is required for this activity (Fig. 3b). RNA strand exchange experiments using E. coli ProQ and N. meningitides NMB1681 suggest this activity is likely shared by many members of the FinO family of RNA chaperones [4, 5]. Strand Exchange Assay

1. Prepare unlabeled RNA (see Subheading 3.1) and labeled RNA (see Subheading 3.2). 2. Mix 18 μL of annealing buffer with 1 μL of 1 μM the unlabeled strand (SII(B)) and 1 μL of 0.1 μM 32P labeled strand (SII(A)). Incubate at 85  C for 5 min and slowly cool to room temperature in the heating block (see Note 15). 3. Mix 6 μL of 2 reaction buffer, 1 μL of ribonuclease inhibitor (20 U/μL), 1.5 μL of 64 μM FinO protein 1 μL of labeled SII (A)-unlabeled SII(B) duplex (5 nM), 2 μL of the unlabeled SII (A) (1.25 μM) and 0.5 μL of DEPC water. 4. Reaction is initiated by incubation the mixture at 37  C (see Note 16). 2 μL samples are removed at each time point and the reaction is stopped by addition of an equal amount of cold stop buffer. Samples are loaded on a 15% non-denaturing PAGE, continuously running at 180 V, room temperature. 5. Gel is dried and visualized as for the duplexing assay above.

4

Notes 1. Stem loop (SL) II in FinP is known for high binding affinity with FinO. SII(A) and SII(B) are two complementary strands that mimic SLII but lack the connecting loop. This duplex RNA is used to test strand exchange ability of FinO (Fig. 2a).

16

Hyeong Jin Kim et al.

2. In this protocol, a low salt buffer is used for RNA purification since this RNA also used for crystallization. However, the manufacturer recommends the usage of at least 150 mM of salt for standard RNA purification. 3. Incubation time can be varied between 3 and 6 h. 4. The transcription reaction generates pyrophosphate, which precipitates in the reaction mixture. The inclusion of pyrophosphatase in the reaction reduces the amount of precipitate but does not eliminate the precipitate. This by-product can clog your column, so it is important to remove pyrophosphate through centrifugation or filtration before chromatography. 5. We wash the column sample loading loop with 70% ethanol for 30 s and then soak in 1 M NaOH for 5 min, and finally stringently wash the loop with DEPC water to remove trace RNase contaminants. 6. A preliminary test run from 0% to 100% NaCl can be used to determine the elution properties of your RNA. Based on this result, you can fine tune your gradient to optimize purification of your target RNA from abortive RNAs and template DNA. 7. Denaturing gel electrophoresis can identify the presence of RNA contaminants of the wrong size or conformational homogeneity. 8. Gel filtration chromatography can be used as a final step to achieve higher purity RNA, however with some loss of yield. 9. The native gel can be stained with EtBr to visualize nucleic acid Coomassie blue to visualize protein or protein-nucleic acid complexes. 10. If you use chemically synthesized RNA, you can skip this part. Synthetic RNA has 50 -hydroxyl group and can be directly phosphorylated. 11. Acrylamide is classified as a neurotoxin. When you handle these chemicals, wear proper safety equipment such as gloves, safety glasses and a mask. 12. Rinsing the wells is important to remove urea and gel debris. 13. Load a small volume of sample to make a tight (~1 mm) layer in the bottom of the well. This will minimize the smiling effect and give a tight band. The migration of xylene cyanol and bromophenol blue can be used to estimate migration of your RNA sample (see citation [23]). 14. Period of time should be tested to find the right range of reaction. In case of FinO, time was varied depend on constructs we used for experiment from 10 min to 1 h (Fig. 3a).

FinO RNA Chaperone Activity Assays

17

15. If you wrap the microcentrifuge tube with tin foil, it will distribute the heat evenly to prevent water condensation on the lid. 16. Incubation temperature can be adjusted depending on the kinetics of the reaction. More rapid reactions can be studied at lower temperatures.

5

Conclusions Gel-based assays of RNA chaperone activity provide a simple and rapid method to test the ability of putative RNA chaperones to facilitate RNA interactions and also to modulate the stability of internal RNA structures. Careful quantitation of this data can also be used to gain estimates of the apparent rate constants for RNA association. These studies can set the stage for more detailed quantitative biophysical studies of RNA chaperone kinetics, for example, using NMR [24] or FRET methods [25].

Acknowledgments We would like to thank Dr. Alexandru Ghetu for early work establishing the strand exchange assay and duplexing assay in FinOP system. This work was supported by a grant from the Natural Sciences and Engineering Research Council of Canada (NSERC; RGPIN-2016-05163). References 1. Herschlag D (1995) RNA chaperones and the folding problem. J Biol Chem 270:20871–20874. https://doi.org/10. 1074/jbc.270.36.20871 2. Vogel J, Luisi BF (2011) Hfq and its constellation of RNA. Nat Rev Microbiol 9:578–589. https://doi.org/10.1038/nrmicro2615 3. Rajkowitsch L, Chen D, Stampfl S et al (2007) RNA chaperones, RNA annealers and RNA helicases. RNA Biol 4:118–130. https://doi. org/10.4161/rna.4.3.5445 4. Chaulk S, Lu J, Tan K et al (2010) N. meningitidis 1681 is a member of the FinO family of RNA chaperones. RNA Biol 7:812–819. https://doi.org/10.4161/rna.7. 6.13688 5. Chaulk G, Smith-Frieday MN, Arthur DC et al (2011) ProQ is an RNA chaperone that controls ProP levels in escherichia coli. Biochemistry 50:3095–3106. https://doi.org/10. 1021/bi101683a

6. Mark Glover JN, Chaulk SG, Edwards RA et al (2015) The FinO family of bacterial RNA chaperones. Plasmid 78:79–87. https://doi.org/ 10.1016/j.plasmid.2014.07.003 7. Attaiech L, Boughammoura A, BrochierArmanet C et al (2016) Silencing of natural transformation by an RNA chaperone and a multitarget small RNA. Proc Natl Acad Sci U S A 113:8813–8818. https://doi.org/10. 1073/pnas.1601626113 8. Smirnov A, Fo¨rstner KU, Holmqvist E et al (2016) Grad-seq guides the discovery of ProQ as a major small RNA-binding protein. Proc Natl Acad Sci U S A 113:11591–11596. https://doi.org/10.1073/pnas.1609981113 9. Attaiech L, Glover JNM, Charpentier X (2017) RNA chaperones step out of Hfq’s shadow. Trends Microbiol 25:247–249. https://doi. org/10.1016/j.tim.2017.01.006 10. Jerome LJ, Van Biesen T, Frost LS (1999) Degradation of FinP antisense RNA from

18

Hyeong Jin Kim et al.

F-like plasmids: the RNA-binding protein, FinO, protects FinP from ribonuclease E. J Mol Biol 285:1457–1473. https://doi.org/ 10.1006/jmbi.1998.2404 11. Arthur DC, Ghetu AF, Gubbins MJ et al (2003) FinO is an RNA chaperone that facilitates sense-antisense RNA interactions. EMBO J 22:6346–6355. https://doi.org/10. 1093/emboj/cdg607 12. Olejniczak M, Storz G (2017) ProQ/FinOdomain proteins: another ubiquitous family of RNA matchmakers? Mol Microbiol 104:905–915. https://doi.org/10.1111/ mmi.13679 13. Milner JL, Wood JM (1989) Insertion proQ220::Tn5 alters regulation of proline porter II, a transporter of proline and glycine betaine in Escherichia coli. J Bacteriol 171:947–951. https://doi.org/10.1128/jb. 171.2.947-951.1989 14. Kunte HJ, Crane RA, Culham DE et al (1999) Protein ProQ influences osmotic activation of compatible solute transporter ProP in Escherichia coli K-12. J Bacteriol 181:1537–1543 15. Smirnov A, Wang C, Drewry LL, Vogel J (2017) Molecular mechanism of mRNA repression in trans by a ProQ-dependent small RNA. EMBO 36:1029–1045. https://doi. org/10.15252/embj.201696127 16. Holmqvist E, Li L, Bischler T et al (2018) Global maps of ProQ binding in vivo reveal target recognition via rna structure and stability control at mRNA 30 ends. Mol Cell 70:971–982.e6. https://doi.org/10.1016/j. molcel.2018.04.017 17. Ghetu a F, Gubbins MJ, Frost LS, Glover JN (2000) Crystal structure of the bacterial conjugation repressor finO. Nat Struct Biol 7:565–569. https://doi.org/10.1038/76790

18. Gonzalez GM, Hardwick SW, Maslen SL et al (2017) Structure of the Escherichia coli ProQ RNA-binding protein. RNA 23:696–711. https://doi.org/10.1261/rna.060343.116 19. Arthur DC, Edwards RA, Tsutakawa S et al (2011) Mapping interactions between the RNA chaperone FinO and its RNA targets. Nucleic Acids Res 39:4450–4463. https:// doi.org/10.1093/nar/gkr025 20. Van BT, Frost LS (1994) The FinO protein of IncF plasmids binds FinP antisense RNA and its target , traJ mRNA , and promotes dupiex formation. Mol Microbiol 14:427–436 21. Gubbins MJ, Arthur DC, Ghetu AF et al (2003) Characterizing the structural features of RNA/RNA interactions of the F-plasmid FinOP fertility inhibition system. J Biol Chem 278:27663–27671. https://doi.org/10. 1074/jbc.M303186200 22. Grodberg J, Dunn JJ (1988) ompT encodes the Escherichia coli outer membrane protease that cleaves T7 RNA polymerase during purification. J Bacteriol 170:1245–1253. https:// doi.org/10.1128/jb.170.3.1245-1253.1988 23. Ellington A, Pollard JD (2001) Purification of Oligonucleotides Using Denaturing Polyacrylamide Gel Electrophoresis. Curr Protoc Mol Biol 42:2.12.11–2.12.17. https://doi.org/ 10.1002/0471142727.mb0212s42 24. Rennella E, Sa´ra T, Juen M et al (2017) RNA binding and chaperone activity of the E. coli cold-shock protein CspA. Nucleic Acids Res 45:4255–4268. https://doi.org/10.1093/ nar/gkx044 25. Rajkowitsch L, Schroeder R (2007) Dissecting RNA chaperone activity. RNA Biol 13:2053–2060. https://doi.org/10.1261/ rna.671807.tertiary

Chapter 2 Quantitative Analysis of RNA Chaperone Activity by Native Gel Electrophoresis and Fluorescence Spectroscopy Subrata Panja, Ewelina M. Małecka, Andrew Santiago-Frangos, and Sarah A. Woodson Abstract Diverse types of RNA-binding proteins chaperone the interactions of noncoding RNAs by increasing the rate of RNA base pairing and by stabilizing the final RNA duplex. The E. coli protein Hfq facilitates interactions between small noncoding RNAs and their target mRNAs. The chaperone and RNA annealing activity of Hfq and other RNA chaperones can be evaluated by determining the kinetics of RNA base pairing in the presence and absence of the protein. This chapter presents protocols for measuring RNA annealing kinetics using electrophoretic gel mobility shift assays (EMSA), stopped-flow fluorescence, and fluorescence anisotropy. EMSA is low cost and can resolve reaction intermediates of natural small RNAs and mRNA fragments, as long as the complexes are sufficiently long-lived (
10 s) to be trapped during electrophoresis. Stopped-flow fluorescence can detect annealing reactions between 1 ms and 30 s and is best suited for measuring the rapid annealing of oligoribonucleotides. Fluorescence anisotropy reports the physical size of the complex and is well-suited for monitoring the association and dissociation of RNA from Hfq during the chaperone cycle. Key words Stopped-flow spectroscopy, Hfq, Fluorescence anisotropy, Native gel mobility shift, Molecular beacon

1

Introduction Antisense and transacting small RNAs (sRNAs) regulate the expression of other genes by base pairing directly with a complementary site in a target mRNA [1]. These small noncoding RNAs typically act in concert with RNA-binding proteins, such as Hfq, that can facilitate the search for prospective target sites, increase the selectivity of target site selection, and stabilize the resulting sRNAmRNA complex [2–4]. In nearly all RNA-guided gene regulation systems studied to date, including micro-RNAs, CRISPR/Cas, and bacterial sRNAs, target selection depends on kinetic discrimination among prospective sites and not simply on the thermodynamic stability of the noncoding RNA-mRNA duplex [5]. As a result,

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_2, © Springer Science+Business Media, LLC, part of Springer Nature 2020

19

20

Subrata Panja et al.

real-time methods for tracking the kinetics of RNA hybridization have been crucial for dissecting the mechanism of target selection [6–10]. This chapter describes native gel mobility shift and ensemble fluorescence protocols for studying how the bacterial RNA chaperone Hfq facilitates base pairing (or annealing) between bacterial sRNAs and mRNAs. These protocols, however, can be readily adapted to other proteins with RNA annealing activity. They are also useful for defining the parameters of the system that are needed to lay the groundwork for in-cell or single-molecule assays of RNA annealing or sRNA target search kinetics. Bacterial Hfq is a ring-shaped homo-hexamer and a member of the widespread Sm/Lsm protein family that includes eukaryotic proteins involved in mRNA turnover and pre-mRNA splicing [11]. E. coli Hfq was first discovered as a host factor for replication of phage Qβ [12] but has since been widely studied for its role in posttranscriptional regulation by bacterial transacting sRNAs [1]. Hfq can simultaneously bind an sRNA and mRNA and help initiate base pairing between their complementary regions. Once the two RNAs have base paired, the double-stranded region is displaced from the protein, preventing the newly formed duplex from unzippering [13]. Since Hfq can bind each RNA individually (binary complexes) or bind both RNAs at the same time (ternary complex), the annealing reactions are complex. The most frequently used method of determining Hfq-RNA affinity is electrophoretic mobility gel shift assays (EMSA) using radiolabeled RNA. EMSAs have the advantage of separating the free RNA from RNA-protein complexes with different mobility, allowing binary and ternary complexes that form distinct bands in the gel to be quantified individually (Fig. 1, compare ChiX-Hfq and ChiX-chiP binary complexes with ChiX-chiPHfq ternary complex). The ability to quantify each complex greatly simplifies the modeling of the reaction kinetics and can alert the user to problems such as aggregation or unwanted RNA dimerization. Another advantage of EMSA is that it requires small quantities of RNA (10–100 pmol), which can be readily labeled during in vitro transcription. A disadvantage of EMSA is that RNA-protein complexes must be trapped or “caged” in the pores of the polyacrylamide gel during electrophoresis [14]. Complexes that are shortlived or disrupted by the electrophoresis running buffer may appear as long “smears” rather than a discrete band, or may be missed entirely. Considerations for selecting electrophoretic conditions that stabilize the RNA-protein complex and that maximize separation have been described elsewhere [14, 15]. A second disadvantage of EMSA is that the time resolution is typically 10–30 s, depending on the time needed to mix and load the sample into the gel. Therefore, this method is best suited to annealing reactions of full-length sRNAs and mRNAs that typically exhibit slower

RNA Chaperone Kinetics

21

Fig. 1 Kinetics of Hfq-mediated sRNA-mRNA base pairing by EMSA. (a) Native gel showing the annealing of 5 nM 32P-labeled ChiX sRNA (86 nt) to 20 nM chiP mRNA (149 nt) in the presence of 10 nM Hfq. Samples were loaded between 10 s and 60 min. The bands appear to move upwards because the gel is run continuously during the experiment. As a result, samples loaded later in the experiment migrate a shorter distance. To assign the composition of each band loaded at later times, control samples were equilibrated for 30 min, and then loaded at the beginning (30 min) and at the end (90 min) of the time course. (b) The combined fraction of sRNA-mRNA duplex plus sRNA-Hfq-mRNA ternary complex from A was plotted versus time and fit to a double exponential rate equation

reaction kinetics and that form stable complexes that can be easily resolved in a native gel [16–18]. Stopped-flow fluorescence spectroscopy is another method for measuring RNA annealing in real time. We have optimized RNA molecular beacons for use in RNA annealing assays with Hfq [19, 20]. A short hairpin holds a fluorophore on the 50 end of the beacon close to a quencher attached to its 30 end, keeping the fluorescence intensity low until the ends of the hairpin are separated when the loop of the beacon base pairs with a target RNA [21]. The large increase in fluorescence intensity upon annealing makes this assay sensitive (1–5 nM). Moreover, only the strand containing the molecular beacon must be chemically modified. Alternatively, a pair of fluorophores can be attached to separate RNA strands, so that base pairing results in fluorescence resonance energy transfer (FRET) [22]. Although stopped-flow fluorescence provides good time resolution (1 ms), it requires ~1 nmol of material. Because it requires a substantial quantity of fluorescently labeled RNA, stopped-flow spectroscopy is easiest to implement for short RNAs that can be synthesized chemically. Fluorescence detection of annealing has been adapted for use in single molecule assays with Hfq [23]. Fluorescence anisotropy offers an alternative approach to following the progress of the Hfq-mediated RNA annealing reaction [24]. In this method, the anisotropy of the fluorescence A ¼ (IVV  IVH)/(IVV + 2IVH) is determined from the difference between the vertical and horizontal fluorescence intensity, when the sample is excited with vertically polarized light. The anisotropy is related to the lifetime τ and correlation time τc of the fluorophore

22

Subrata Panja et al.

through the Perrin equation, A0/A ¼ 1 + τ/τc. Because the correlation time τc depends on the physical size and hydrodynamic friction of the fluorophore-containing complex, the average fluorescence anisotropy of a fluorescently labeled RNA oligomer generally increases when it binds to Hfq [25] or to a second complementary RNA strand [26]. The advantages of the fluorescence anisotropy method are that it directly reports on the physical size of the complex and that it requires only one component (typically the RNA) to be extrinsically labeled with a fluorophore.

2

Materials For any experiments and sample handling, precautions should be taken to avoid ribonuclease (RNase) contamination. Always use RNase-free pipette tips, tubes, and microfuge tubes. Use highquality, nuclease-free reagents and ultrapure, RNase-free deionized water (with a resistivity of 18.2 MΩ at 25  C). Always wear gloves when handling reagents, samples, or instruments or while doing experiments. Samples containing a fluorophore must be shielded from light. Buffers used for fluorescence anisotropy or stoppedflow spectroscopy should be filtered (0.45 μm) before use.

2.1 Native Polyacrylamide Gel Electrophoresis (PAGE)

1. 40% w/v acrylamide-bisacrylamide (29:1) solution. 2. N,N,N0 ,N0 -Tetramethylethane-1,2-diamine (TEMED). 3. 10% ammonium persulfate (APS). 4. 10 TBE buffer: 890 mM Tris base, 890 mM boric acid, 20 mM EDTA (pH 8.3). 5. Ethanol to clean plates, spacers, and combs. 6. Electrophoresis power supply. 7. Vertical electrophoresis apparatus including glass plates, spacers, well-forming combs, and clamps (see Note 1). 8. Gel dryer and vacuum pump. 9. Plastic food wrap. 10. 3 mm Whatman chromatography paper (sheet).

2.2 Hfq-RNA Binding by Native Gel Mobility Shift (EMSA)

1. 1 TE buffer: 10 mM Tris–HCl (pH 7.5), 1 mM EDTA. 2. 5 Hfq storage buffer: 50 mM Tris–HCl pH 7.5, 250 mM NH4Cl, 1 mM EDTA, 10% (v/v) glycerol. 3. 3 μL 100 nM Note 2).

32

P-labeled RNA in water or TE buffer, (see

4. 20 μL 500 nM unlabeled RNA stock.

RNA Chaperone Kinetics

23

5. 4 μL 60 μM monomer Hfq protein stock in 5 Hfq storage buffer. 6. 10 TNK buffer: 100 mM Tris–HCl (pH 7.5), 500 mM NaCl, 500 mM KCl. 2.3 sRNA-mRNA Annealing by EMSA

1. 3 μL 100 nM 32P-labeled RNA in water or TE buffer. 2. 20 μL 500 nM unlabeled RNA stock. 3. 1 μL 60 μM Hfq protein monomer stock. 4. 10 TNK buffer (see item 6 in Subheading 2.2).

2.4 RNA Annealing Kinetics by StoppedFlow Spectrometry

1. TE buffer (see item 1 in Subheading 2.2). 2. 10 μM working stock of custom designed, reversed-phase HPLC-purified molecular beacon with a fluorophore (FAM) at the 50 end and a quencher (DABCYL) at the 30 end in 1 TE buffer (see Notes 3 and 4). 3. 20 μM working solution of custom designed target RNAs, purified by reversed-phase HPLC or denaturing PAGE in 1 TE buffer (see Note 5). 4. Hfq storage buffer (see item 2 in Subheading 2.2). 5. 0.2–0.5 mL of 25 μM and 300 μM Hfq protein monomer in 5 Hfq storage buffer. 6. 10 TNK buffer (see item 6 in Subheading 2.2). 7. 3 mL BD Luer-Lok syringe with 18 G PrecisionGlide needle or similar. 8. Stopped-flow spectrofluorometer with small internal mixing volume (e.g., SX 18MV from Applied Photophysics or similar). 9. Data analysis and scientific plotting software (e.g., Graphpad Prism, Igor, KaleidaGraph, MatLab, Origin, Sigma Plot or similar).

2.5 Fluorescence Anisotropy RNA Annealing Experiments

1. 100 μM custom designed, reversed-phase HPLC-purified RNA oligomer (D16-FAM) with a fluorophore (FAM) at the 50 end in TE buffer (see Note 6). 2. 5 μM D16-FAM in TE buffer (working solution). 3. 100 μM custom designed, reversed-phase HPLC-purified RNA oligomer (R16) complementary to D16-FAM in TE buffer. 4. 5 μM R16 in TE buffer (working solution). 5. 30 μM Hfq monomer in 5 Hfq storage buffer. 6. 10 TNK buffer (see item 6 in Subheading 2.2). 7. Horiba Fluorolog-3 spectrofluorometer in L format, with excitation and emission polarizers, or similar.

24

Subrata Panja et al.

8. 500 μL quartz cuvette (four windows). 9. Plastic tips for a 200 μL pipettor with a long neck and round tip (e.g., for gel loading). 10. Data analysis and plotting software (see item 9 in Subheading 2.4).

3

Methods The first three protocols describe the use of native PAGE to measure RNA-binding constants and sRNA-mRNA annealing kinetics in the presence of Hfq protein. Protocols four and five describe how to measure binding and annealing kinetics using fluorescence spectroscopy.

3.1

Native PAGE

This protocol is for preparing a native 8% polyacrylamide gel suitable for separating free RNA (50–300 nt) from RNA that is bound to Hfq protein or to a second RNA. The gel composition and electrophoresis running buffer can be varied to optimize the separation for a given protein-RNA complex of interest. 1. Thoroughly clean plates, spacers, and comb with ethanol. Ensure that all components are dry before continuing. Lie one plate flat on the bench, place a spacer along each vertical edge of the plate, and sandwich the spacers with a second plate. Slide the second glass plate up slightly, to create an open space on the bottom of the plates. In this way, the gel plug can be prepared first, which will prevent leaking during gel casting. 2. Prepare 40 mL 8% gel solution (for a gel with dimensions 0.1  20  18 cm) by mixing 4 mL 10 TBE buffer, 8 mL 40% acrylamide:bisacrylamide solution and 28 mL deionized RNase-free water. Add 400 μL 10% APS. 3. Place 5 mL gel solution in a disposable 15 mL tube and add 15 μL TEMED. Pour this mixture immediately between the space on the bottom of the two plates. Ensure that gel solution contacts both spacers. Wait 5–10 min until the gel polymerizes into a solid plug. 4. Add 14 μL TEMED to the remaining 35 mL of gel solution, and pour it between the plates, avoiding bubbles. Insert the comb immediately. Allow the gel to polymerize in a horizontal position for at least 1.5 h. 5. Remove the comb from the gel and mount the gel on the gel apparatus. Add 1 TBE buffer to the upper and lower reservoirs. Make sure the wells are covered by running buffer. Flush the wells with running buffer and pre-run the gel at 10 W for at least 15 min (see Note 1).

RNA Chaperone Kinetics

25

6. When the samples are ready to load, rinse the wells of the gel with running buffer. Load 5 μL sample in each well (see Note 7). 7. Since the sample does not contain tracking dyes, load 5 μL 6 RNA loading dye in an unused well to monitor the progress of electrophoresis. 8. Run the gel at 10 W for at least 1.5–2 h (see Note 8). 9. Turn off the power, pour off the running buffer, and unclamp the gel(s) from the apparatus. If the running buffer is contaminated with 32P, dispose it carefully as regulated by your institution. 10. Remove one plate and cover the gel with 3 mm Whatman paper. Remove the other plate and cover the exposed gel surface with plastic wrap. Put the gel on top of the gel dryer and turn on the pump to create a vacuum. Simultaneously heat to 80  C. Dry the gel for 30 min or until it is completely dry. 11. Place the dried gel (still covered with plastic wrap) in a lighttight cassette with a phosphor storage screen and expose the screen overnight. 12. When electrophoresis is complete, use image analysis software to determine the intensity of each band in the gel. One method is to define the area of each band and then calculate a volume integral which measures the total signal intensity (see Note 9). A similar area in a region of the gel not containing RNA should be integrated to establish the background intensity. 3.2 Hfq-RNA Binding by Native Gel Mobility Shift (EMSA)

This protocol is for determining the equilibrium dissociation constant (Kd) of Hfq with 32P-labeled RNA by EMSA. 1. Prepare 150 μL of 32P-labeled RNA-labeled RNA to 2 nM final concentration (see Note 10) in 1 TNK buffer. 2. Denature the RNA by heating it at 90  C for 1 min (see Note 11). Refold the RNA at room temperature for 20 min. 3. Prepare 40 μL of 6 μM monomer Hfq stock in 1 TNK buffer. Take 20 μL of protein and mix with 20 μL of buffer. In this way, prepare ten more twofold protein dilutions for a total of 12 reactions (see Note 12). 4. Mix refolded RNA and protein dilutions in 1:1 ratio (10 μL each). Prepare control sample without protein by mixing RNA with buffer. Incubate at room temperature for 30 min. 5. Load 5 μL each sample in the well of a native polyacrylamide gel and separate the bound and free RNA as described in Subheading 3.1. 6. After the band intensities are quantified and the background subtracted as described in Subheading 3.1, the fraction of

26

Subrata Panja et al.

bound RNA in each lane of the gel is calculated from ƒB ¼ [cpm bound]/[cpm total], in which cpm total is the total counts in the lane. 7. To obtaining binding constants, the fraction bound should be fit to a binding polynomial (partition function) representing an appropriate binding model for the system of interest (see Note 13). 3.3 sRNA-mRNA Annealing by EMSA

This protocol is for measuring the observed kinetics of sRNAmRNA annealing in the presence or absence of Hfq protein, using 32P-labeled RNA and native PAGE. 1. Prepare 100 μL each of 32P-labeled sRNA, unlabeled mRNA, and Hfq as 3 stocks: 3 nM, 90 nM, 540 nM (concentration of Hfq monomer), respectively, in 1 TNK buffer (see Note 14). In this way, the reaction components can be mixed in a 1:1:1 ratio which simplifies setting up the reaction. 2. Both sRNA and mRNA should be denatured and refolded before the experiment (see step 2 in Subheading 3.2). 3. Prepare four control reactions (see Note 15): (1) 10 μL of labeled sRNA and 20 μL of 1 binding buffer, (2) 10 μL of labeled sRNA +10 μL of unlabeled mRNA and 10 μL of buffer, (3) 10 μL of labeled sRNA +10 μL of Hfq and 10 μL of buffer, and (4) 10 μL of labeled sRNA +10 μL unlabeled RNA + 10 μL Hfq. Incubate the control reactions for 30 min at room temperature. 4. During this time, begin pre-running the native gel. 5. Load 5 μL of each control sample on the running gel. 6. Immediately start the annealing reaction by mixing 30 μL of each component at room temperature (see Note 16). Load a 5 μL aliquot of the reaction in the next lane of the native gel at specific times after mixing (e.g., 10 s, 30 s, 1 min, 1.5 min, 2 min, 3 min, 5 min, 10 min, 15 min, 20 min, 30 min, 60 min). The current should remain on during the experiment (see Note 17). 7. Right after the last time point has been added to the gel, in the last four lanes load 5 μL of the control reactions that were set up at the beginning of the experiment. 8. Continue to run the gel until the desired separation is achieved, taking care not to let the first samples run off the bottom of the gel. Image the gel as described in Subheading 3.2. 9. From the image of the gel, use the control samples to assign the bands corresponding to the free RNA, sRNA-mRNA (binary complex), sRNA-Hfq (binary complex), and sRNA-mRNAHfq ternary complexes.

RNA Chaperone Kinetics

27

Fig. 2 Applied Photophysics SX 18 MV stopped-flow system. (a) A representative Applied Photophysics SX 18 MV stopped-flow spectrometer (reprinted with permission). PSU power supply unit, LH lamp housing, Mo monochromator, SHU sample handling unit, ECU electronic control unit. (b) Scheme of the stopped-flow experiment. Liquids from two drive syringes are rapidly mixed and reach the flow cell in the continuous flow phase. Once the flow of liquid is stopped by the stop syringe, the mixture ages in the flow cell. The solution in the flow cell is excited with a lamp (with monochromator) and the fluorescence intensity is detected by a photomultiplier tube perpendicular to the incident beam

10. Calculate the fraction of sRNA annealed as the sum of counts from sRNA-mRNA and sRNA-mRNA-Hfq complexes relative to the total counts in each lane, ƒA ¼ ([cpm sRNAmRNA] + [cpm sRNA-mRNA-Hfq])/[cpm total]. 11. To calculate the observed rate constant for annealing, the fraction annealed should be plotted versus time and the data should be fit to a single- or double-exponential rate equation (see Note 18), as illustrated in Fig. 1. 3.4 RNA Annealing Kinetics by StoppedFlow Spectrometry

This protocol is designed based on the workflow of the Applied Photophysics SX 18 MV stopped-flow system (Fig. 2), but is also applicable to other stopped-flow systems (i.e., KinTek). We have studied the effect of the RNA chaperone Hfq on RNA annealing, but any other RNA-binding protein can be studied with this protocol if provided with appropriate beacon and target RNA sequences. 1. Before the start of the experiment (30–60 min), turn on the power switch of the “lamp power supply” (“a” in Fig. 2) and then press and hold the “start” switch (“b” in Fig. 2) for ~5 s until a “click” sound is heard (see Note 19). This will turn on the lamp inside the lamp housing (see Note 20). Turn on the Pro Data Electronics Unit (“C” in Fig. 2) and the computer. 2. Turn on the power switch of the circulating water bath and set the temperature to 30  C. 3. Open the regulator of a nitrogen tank (high purity) that is attached to the stopped flow spectrometer and adjust the

28

Subrata Panja et al.

output pressure to 125 psi. The pressure should remain constant throughout the experiment. 4. Thaw the stock solutions of molecular beacon, target RNA and Hfq on ice. Once thawed, mixed the solutions gently (see Note 21) and spin the tubes in a microcentrifuge for a few seconds. Hold thawed solution on ice until use. 5. Install a 25-mm high-quality 515 nm long pass filter between the flow cells and photomultiplier tubes, using three screws. 6. Set the excitation monochromator pathlength to 4 mm. 7. Configure the SX20 Pro-Data acquisition software as follows: (a) Open the SX20 Pro-Data software within the computer’s operating system. (b) Select “Fluorescence” from the drop-down list of the “Signal” panel. (c) On the “Sequencer” panel select “Kinetics” mode. (d) Select “External” in the “Trigger” panel. (e) On the “Timebase” panel, set the time to 200 s and “Points” to 5000. This configuration will enable data acquisition on linear time scale (see Note 22 for other acquisition modes). (f) On the “Monochromator” panel, select the wavelength as 490 nm and press “Set.” 8. Dilute 10 TNK buffer to 1 with water. 9. Fill the instrument drive syringes (2.5 mL) with 1 TNK using 2.5 mL disposable Luer-Lok syringes. Direct the end of the waste line into a suitable container. Put the valves in “drive” position and click the ‘Drive’ button 6 to 8 times (see Note 23). 10. Open “Pro-Data Viewer.” Create a new folder with a unique name and the date of the experiments. Right click on the folder and select “Set experimental directory here.” All of the data will be saved in this folder until a new experimental directory is designated. 11. Prepare samples as listed in Table 1. 12. Load 1 mL of molecular beacon-Hfq solution into the left drive syringe (Position “L”) and 1 mL of target RNA into the right drive syringe (Position “R”). Put the drive valves in the “drive” position. 13. Wait for 5 min to equilibrate the samples to 30  C (see step 2 in Subheading 3.4). 14. Click the “Drive” button three times so that the samples purge the tubing and reach the flow cell.

L2

L3

L4

L5

L6

L7

L8

L9

50

100

200

300

600

900

1200

b

a

1500

10 μM stock 20 μM stock

L1

0

10

10

10

10

10

10

10

10

10

Tube Molecular number beacona (μL)

[Hfq] after mixing (nM)

Samples for syringe L

3000

2400

1800

1200

600

400

200

100

0

300

300

300

300

300

25

25

25

–

Stock [Hfq] in tube concentration ‘L’ (nM) (μM)

Hfq

Table 1 Preparation of samples for stopped flow kinetics experiments

10

8

6

4

2

8

4

2

0

Vol. stock to add (μL)

100

100

100

100

100

100

100

100

100

880

882

884

886

888

882

886

888

890

R9

R8

R7

R6

R5

R4

R3

R2

R1

10

10

10

10

10

10

10

10

10

100

100

100

100

100

100

100

100

100

890

890

890

890

890

890

890

890

890

10 Water Tube Target 10 Water TNK (μL) (μL) number RNAb (μL) TNK (μL) (μL)

Samples for syringe R

RNA Chaperone Kinetics 29

30

Subrata Panja et al.

15. On the “Detector High Voltage” panel click the “. . .” button next to the “Fluorescence” box which will open the “Set HV” dialog box. Click “AutoPM” to set up the detector voltage, do not change it during the experiment (see Note 24). 16. Click “Acquire” to collect data. The green LED on the Pro Data Electronics Unit will turn on. 17. During data acquisition, a live window will pop up to show the progress curve of the experiment. Once the experiment is done, two red LEDs on the Pro Data Electronics Unit will start blinking. 18. Continue acquiring data until all of the prepared sample in the drive syringes has been used (see Note 25). 19. Wash the drive syringes with 1 TNK buffer to start the next set of experiments (see Note 26). 20. After finishing all the experiments, wash the drive syringes and flow cell with 3 mL of buffer and 3 mL of water as described in step 5 (see Subheading 3.4). 21. Once all the experiments are done, export the data files for analysis by dragging individual kinetic trace files from the Pro-Data Viewer to the SX20 Pro-Data window to open the kinetic traces (see Note 27). Use the “Save as” drop-down menu to save individual kinetic traces to “experiment_name. csv” format. 22. Exit the SX20 Pro-Data and Pro-Data Viewer software. Turn off the power switch of the Pro Data Electronics Unit (“c” in Fig. 2). Turn off the power switch of the “lamp power supply” (“a” in Fig. 2). 23. Analyze the data in the following way: (a) Open the “experiment_name.csv” files using Microsoft Excel or another compatible software program. Copy the time column and paste it into the “x” column of Origin or any comparable plotting or graphing software workbook. Then copy the fluorescence intensity column and paste it into the “y” column of the workbook. Average the intensities from all the traces from an individual experiment if not performed earlier. (b) Plot the data as an “x–y scatter plot” (Fig. 3a). Fit the data using the double exponential rate equation ΔF (t) ¼ F0  Afast exp.(kfastt)  Aslow exp.(kslowt) (see Note 28). (c) Repeat this analysis for the data collected at different Hfq concentrations.

RNA Chaperone Kinetics

31

Fig. 3 Representative results for the stopped-flow RNA annealing assay. (a) The molecular beacon is attached to a fluorophore at the 50 end and a quencher at the 30 end. The 50 and 30 ends base pair, so the fluorescence intensity is low. When the loop of the beacon anneals to the target RNA, the fluorophore and quencher are separated and the fluorescence intensity increases substantially. (b) Representative stopped flow kinetic traces of molecular beacon rapidly mixed with target-U6 RNA in presence of different concentrations of Hfq. (c) Rate constants of RNA annealing as a function of Hfq concentration. Different Hfq binding sites (A18 or U6) on the target RNA have different effects on the rate of RNA annealing. (Adapted from Ref. 27 with permission)

(d) Plot kfast and kslow as a function of Hfq concentration to determine how Hfq affects the RNA annealing kinetics (Fig. 3b). 3.5 Fluorescence Anisotropy to Observe Different Stages of RNA Annealing

This protocol is designed for the workflow of the Fluorolog-3 L format spectrofluorometer (Horiba). This protocol can be adopted to any sensitive spectrofluorometer that has an anisotropy measurement module. The time resolution of this procedure is poor (~20 s) due to the manual mixing and the inherent slowness of the anisotropy measurements in an L format. For better time resolution, it is advisable to use an Applied Photophysics SX 18 MV stopped-flow system with the polarizer accessory and dual detectors. 1. Thaw 5 μM D16-FAM, 5 μM R16 RNA and 30 μM Hfq on ice. Once thawed, mix the solutions gently and spin the tubes for a few seconds in a microcentrifuge before returning samples to the ice bath. 2. Turn on the spectrofluorometer as follows: (a) Turn on the main power switch. (b) Turn on the lamp power switch. (c) Turn on the accessory power switch that supplies power to the water bath, spectrofluorometer control unit, and computer.

32

Subrata Panja et al.

(d) Turn on the computer. (e) Before starting the experiment, wait at least 30 min for the instrument to stabilize. (f) Once the lamp has stabilized, open the “FluorEssence” software within the computer’s operating system. (g) Wait for a minute and then click on the M (main) tab on the menu bar of the software. (h) Wait for few minutes and allow the instrument to selfcalibrate (see Note 29). (i) From the main menu, choose “Anisotropy” and then “vs Time” from the “Experiment Type” menu. This will open the “Experiment Setup” window. 3. Configure your experiment (see Note 30): (a) Monochromator: Set the excitation wavelength to 490 nm and the emission wavelength to 515 nm (see Note 31). Set the excitation and emission slits to 5 nm width (see Note 32). (b) Detector: Choose an “integration time” of 1 s. Set the “data interval” to 20 s and set the “total time” to 900 s. In the “Signal” submenu, enable “S1” and under “Formulas,” choose “Anisotropy” (see Note 33). (c) In the accessory tab, set “temperature” to 30  C and “tolerance” to 0.5  C. 4. Thaw Hfq and R16 working stocks and keep on hand. 5. In a 1.5 mL tube, add 445 μL of water, 50 μL of 10 TNK and 5 μL of D16-FAM from the 5 μM working solution. Mix gently by pipetting up and down and spin for a few seconds. 6. Transfer the entire mixture to a 500 μL quartz cuvette and put the cap on the cuvette. Put the cuvette inside the cuvette holder in the sample chamber of the spectrofluorometer. Close the lid of the sample chamber. Wait 5 min for the sample to equilibrate to 30  C. 7. Start recording by pressing the tab ‘Start’ and keep recording for 180–200 s. Press the tab “pause” and then “close shutter.” 8. Open the sample chamber and remove the cuvette cap. Add 5 μL of 30 μM Hfq working solution and mix rapidly with a 200 μL pipettor equipped with a long disposable tip. Close the cuvette and the sample holder, and press “Resume.” Record the anisotropy for another 5 min. 9. Press the tab “pause” again and then “close shutter” as in step 5.

RNA Chaperone Kinetics

33

10. Add 5 μL of 5 μM R16 working solution, and mixed rapidly as described in step 6. Close the sample chamber, press “Resume” and record the anisotropy for another 6–7 min. 11. Once the experiment is done, the data will pop up in a new window. 12. Repeat steps 3–8 at least three times. 13. To transfer the data to a format for export, go to “File” and then “batch transfer.” Select the output files as “.csv.” 14. When the experiment is done, turn off the instrument: (a) Quit the software after saving your data. (b) Shut down the computer. (c) Turn off the lamp power using the switch on the spectrofluorometer. (d) Turn off the main power using the switch on the instrument (see Note 34). (e) Turn off the accessory power switch. 15. Clean the cuvettes. 16. Open all the data (.csv) files using Microsoft Excel or another compatible software program. Average the anisotropy values from three trials. Copy the time column and paste it into the “x” column of Origin or any comparable plotting or graphing software workbook. Then copy the average anisotropy column and paste it into the “y” column of the workbook. 17. Plot the data as a line plot or fit the readings to an appropriate reaction mechanism model (Fig. 4).

4

Notes 1. To prevent the dissociation of complexes during electrophoresis, a water-cooled gel electrophoresis unit should be used to maintain the internal temperature of the gel below 15  C. Alternatively, electrophoresis can be performed in the cold room. The choice of thickness of the spacers depends on the sample volume—for very thin gels (0.05 cm) which dry very quickly, it is advised to load 2 μL of the sample to obtain sharp bands. Thicker gels (0.1 cm) also can be used. They allow a larger volume of sample (5–10 μL) to be loaded in each well, which can be useful for reactions containing a very low concentration of 32P-labeled RNA. For protocols presented in this chapter, use a comb that forms 20 or more sample wells. 2. Use 10 μCi γ-32P-ATP (1 μL 10 mCi/mL ATP) to phosphorylate 20 pmol RNA with T4 polynucleotide kinase in a total volume of 20 μL (1 μM RNA final concentration).

34

Subrata Panja et al.

Fig. 4 Representative results for RNA annealing by fluorescence anisotropy. (a) Experimental work flow and principles behind the anisotropy experiment [26]. The FAM attached to the D16 RNA oligonucleotide has low anisotropy, as the RNA can rotate freely in the solution due to its relatively small size. Once the RNA oligonucleotide is bound by either Hfq or the complementary RNA, the anisotropy increases, due to the increased size, and slower rotational diffusion, of the complex. (b) Fluorescence anisotropy values for D16-FAM recorded every 20 s in a standard cuvette. Error bars represent the S.D. of three trials. Arrows and shaded regions indicate the points when Hfq and complementary R16 RNA are added to the cuvette. The anisotropy of D16-FAM increases when it binds to Hfq hexamer. When the complementary R16 RNA was added, it base-pairs with D16-FAM RNA. The majority of the RNA duplex dissociates from Hfq, resulting in an intermediate anisotropy. (Data replotted from Ref. 28)

3. One can use other fluorophore and quencher pairs. A fluorophore with a high quantum yield should be chosen. When designing a molecular beacon, one should make sure that the molecular beacon has a single secondary structure. The heterogeneity of the structure can be checked by native 12% PAGE, in which all of the molecular beacon molecules should migrate as a single band. 4. Dissolve the molecular beacon in TE buffer to make a 100 μM stock solution. Use the extinction coefficient provided by the vendor. Dilute the molecular beacon to 10 μM working solution in TE buffer. Fluorescently labeled samples should be handled in low light at all times to minimize photo-bleaching of the fluorophores. 5. When storing RNA solutions, one should make multiple aliquots of the stock solution and keep it at 80  C for long-term storage and 20  C for short-term storage. Make multiple aliquots of the working solution in TE buffer. It is strongly

RNA Chaperone Kinetics

35

advisable to avoid multiple freeze thaw cycles of the same aliquot. 6. The “target” region of the FAM-labeled oligoribonucleotide should be unstructured and should be able to interact with Hfq protein or with the protein of interest [6]. An RNA strand complementary to the “target” region is also needed for annealing assays. 7. Make sure not to expel air bubbles which can disturb sample loading. 8. The time of electrophoresis depends on the length and structure of the RNAs being studied, the number of different complexes to be resolved and the difference in mass and conformation between them. The conditions provided here work well for resolving complexes of 50–200 nt RNA with an E. coli Hfq hexamer (67 kDa). 9. When defining the area of a band, keep in mind that the peak shapes are typically Lorentzian, with a broad base above and below the middle of the band. Be sure to define the peak area in a consistent way and to properly account for background intensity. Weak Hfq-sRNA complexes may dissociate during electrophoresis and appear as a broad smear of intensity within the lane; these complexes should be also considered in the binding model. 10. Ideally, the final concentration of labeled RNA in the reaction should be 10–100 times lower than the equilibrium dissociation constant. Under these conditions, the amount of bound protein is a very small percentage of the total protein, and therefore the free protein concentration is approximately equal to the total protein concentration, simplifying the calculation of binding affinities. 11. Different RNAs may require different renaturation conditions for optimal activity, and various methods of refolding should be tested. For example, RNAs with more complex structures might refold better when cooled slowly (e.g., 2  C/min), when held for 20–30 min at 50  C, or when magnesium ions are present (e.g., 2 mM). 12. To calculate the equilibrium dissociation constant, it is necessary to use several protein concentrations that cover the whole range of the binding reaction. A general rule is to use 10–15 protein concentrations that span three orders of magnitude around the equilibrium dissociation constant with twofold protein dilutions. If binding is cooperative, one must use a narrower range of protein concentrations within the steep portion of the binding curve.

36

Subrata Panja et al.

13. Two hexamers of Hfq can bind an sRNA, and binding is often mildly cooperative with respect to Hfq concentration. Therefore, the results of a typical Hfq binding reaction can often be fit to a partition function P containing terms for binding of one or two hexamers [18], n
2 !n 
Hfq6 Hfq6 þ and f H1 P ¼1þ K1 K 1K 2  n
2 !n 
Hfq6 Hfq6 ¼ =P and f H2 ¼ =P K 1K 2 K1 in which ƒH1 is the fraction of RNA bound to one Hfq hexamer, ƒH2 is the fraction bound to two hexamers, and K1 and K2 are the apparent binding constants for each binding reaction. 14. Either the sRNA or the mRNA may be labeled with 32P. Hfq will distribute between the labeled and unlabeled RNA according to their relative affinities for the protein. Consequently, the observed annealing kinetics can differ depending on which RNA is present in excess. In sRNA-mRNA annealing studies, it is crucial to use an Hfq concentration that is above the equilibrium dissociation constant for binding of both RNAs. However, the concentration of Hfq used should not be too high, since RNAs will preferably form binary complexes with Hfq rather than a ternary complex. 15. These control reactions are important for defining the composition of the observed bands in the annealing reaction. 16. It is important to mix the components at the same time to avoid favoring one type of complex. Simultaneous mixing can be achieved by adding 32P-labeled RNA to the tube first and then unlabeled RNA as a droplet on the wall of the tube, which will be mixed together while adding the Hfq with a pipette. Alternatively, two pipettes can be used simultaneously. 17. The current should remain on while samples from the annealing reaction are added to the gel, to ensure that the complexes become quickly caged within the polyacrylamide gel matrix and are resolved from one another. 18. The double-exponential rate equation is fA(t) ¼ 1  A1 exp (k1t)  A2 exp (k2t), in which ƒA(t) is the fraction annealed at time t, A1 and A2 and k1 and k2 are the amplitudes and rate constants of the observed kinetic phases. 19. The instrument components must be powered on and off in the sequence given in the protocol. The lamp should be turned on at least 30 min before the experiments. An unstable lamp will cause a change in excitation intensity between experiments,

RNA Chaperone Kinetics

37

making it difficult to compare results between different experiments. Additionally, an unstable lamp will increase the amount of noise in the data. 20. It is difficult to see if the lamp is turned on or not. One can attach an LED to the hole of the lamp housing that will emit a visible light when it receives current from the lamp. 21. RNA and protein solutions can be mixed by tapping the tubes gently. Do not vortex protein solutions as this can denature the protein. 22. Do not select “Logarithmic” or “Split Timebase” unless these options are needed. The “Logarithmic” mode distributes data points over a logarithmic time base. In “Split Timebase” mode, the data points are divided among two user defined time windows. This is useful when the kinetics has fast and slow phases. 23. One can wash the system with water first before washing with the experimental buffer. For washing, one can start with a manual wash which will use higher volume of liquid than the wash using the “Drive” command. However, several wash steps using the “Drive” command are recommended, as this wash procedure will generate a higher force, ensuring that the fluid path is thoroughly flushed. 24. A detector voltage gain that provides an appropriate signal change for all of the kinetics runs can be determined after doing few experiments. Future experiments with same beacon and fluorophore combination can be performed with the same detector voltage. This provides consistent results and allows the one to compare the start and endpoints between experiments. 25. Each shot consumes 80 μL per syringe, or 160 μL final. With 1 mL sample, one can usually obtain 8–10 kinetic traces. The first three traces are usually mixed with aged sample and should be discarded. 26. One can begin a series of experiments with samples containing no Hfq, and then move to reactions with increasing concentrations of Hfq, to avoid intermediate washing steps. 27. Congruent trajectories from a series of shots may also be averaged before export using the Pro-Data software. 28. RNA annealing kinetics in the presence of Hfq typically contains two or three observed exponential phases. Simpler reactions may be fit with a single exponential decay. Alternatively, a set of progress curves at different ligand concentrations may be globally fit to an appropriate kinetic reaction mechanism. 29. While the instrument is self-calibrating after it is turned on, it will make different noises that come from the movements of the monochromators and the polarizers. Do not disturb the

38

Subrata Panja et al.

instrument until this part of the calibration procedure has finished. 30. Before starting the experiment, one should check the lamp absorption spectrum. To do this, set the emission wavelength at 350 nm and scan the lamp absorption between 200 and 600 nm. The absorption maximum should fall at 467 nm. Next, measure the intensity of the water Raman signal. To do this, set the excitation wavelength at 350 nm and take an emission scan between 365 and 450 nm. The water Raman peak should be at 397 nm. Verify the locations of these peaks to be sure that the monochromators are properly aligned. The peak intensity of the water Raman signal should be >450,000 for the Horiba Fluorolog-3 if the instrument is operating at peak sensitivity. 31. These values are appropriate for fluorescein or 6-FAM and should be adjusted depending on the fluorophore used for the experiment. 32. Narrower slits reduce background from scattered light or direct detection of the excitation beam, but also reduce the signal-to-noise ratio. 33. When anisotropy is selected as the acquisition mode, the Horiba Fluorolog will automatically record the fluorescence intensity with four different polarizer settings and calculate the anisotropy. 34. Wait 2–3 min before turning off the spectrofluorometer main power switch, to allow the lamp to cool down. References 1. Wagner EG, Romby P (2015) Small RNAs in bacteria and archaea: who they are, what they do, and how they do it. Adv Genet 90:133–208 2. Updegrove TB, Zhang A, Storz G (2016) Hfq: the flexible RNA matchmaker. Curr Opin Microbiol 30:133–138 3. Holmqvist E, Vogel J (2018) RNA-binding proteins in bacteria. Nat Rev Microbiol 16:601–615 4. Woodson SA, Panja S, Santiago-Frangos A (2018) Proteins that chaperone RNA regulation. Microbiol Spectr 6:PMID: 30051798 5. Gorski SA, Vogel J, Doudna JA (2017) RNA-based recognition and targeting: sowing the seeds of specificity. Nat Rev Mol Cell Biol 18:215–228 6. Panja S, Woodson SA (2012) Hfq proximity and orientation controls RNA annealing. Nucleic Acids Res 40:8690–8697

7. Chandradoss SD, Schirle NT, Szczepaniak M, MacRae IJ, Joo C (2015) A dynamic search process underlies microRNA targeting. Cell 162:96–107 8. Singh D, Sternberg SH, Fei J, Doudna JA, Ha T (2016) Real-time observation of DNA recognition and rejection by the RNA-guided endonuclease Cas9. Nat Commun 7:12778 9. Fei J, Singh D, Zhang Q, Park S, Balasubramanian D, Golding I, Vanderpool CK, Ha T (2015) RNA biochemistry. Determination of in vivo target search kinetics of regulatory noncoding RNA. Science 347:1371–1374 10. Persson F, Linden M, Unoson C, Elf J (2013) Extracting intracellular diffusive states and transition rates from single-molecule tracking data. Nat Methods 10:265–269 11. Weichenrieder O (2014) RNA binding by Hfq and ring-forming (L)Sm proteins: a trade-off

RNA Chaperone Kinetics between optimal sequence readout and RNA backbone conformation. RNA Biol 11:537–549 12. Franze de Fernandez MT, Hayward WS, August JT (1972) Bacterial proteins required for replication of phage Q ribonucleic acid. Purification and properties of host factor I, a ribonucleic acid-binding protein. J Biol Chem 247:824–831 13. Santiago-Frangos A, Woodson SA (2018) Hfq chaperone brings speed dating to bacterial sRNA. Wiley Interdiscip Rev RNA 9:e1475 14. Hellman LM, Fried MG (2007) Electrophoretic mobility shift assay (EMSA) for detecting protein-nucleic acid interactions. Nat Protoc 2:1849–1861 15. Ryder SP, Recht MI, Williamson JR (2008) Quantitative analysis of protein-RNA interactions by gel mobility shift. Methods Mol Biol 488:99–115 16. Zhang A, Altuvia S, Tiwari A, Argaman L, Hengge-Aronis R, Storz G (1998) The OxyS regulatory RNA represses rpoS translation and binds the Hfq (HF-I) protein. EMBO J 17:6061–6068 17. Brescia CC, Mikulecky PJ, Feig AL, Sledjeski DD (2003) Identification of the Hfq-binding site on DsrA RNA: Hfq binds without altering DsrA secondary structure. RNA 9:33–43 18. Lease RA, Woodson SA (2004) Cycling of the Sm-like protein Hfq on the DsrA small regulatory RNA. J Mol Biol 344:1211–1223 19. Hopkins JF, Panja S, McNeil SA, Woodson SA (2009) Effect of salt and RNA structure on annealing and strand displacement by Hfq. Nucleic Acids Res 37:6205–6213 20. Hopkins JF, Panja S, Woodson SA (2011) Rapid binding and release of Hfq from ternary

39

complexes during RNA annealing. Nucleic Acids Res 39:5193–5202 21. Tyagi S, Kramer FR (1996) Molecular beacons: probes that fluoresce upon hybridization. Nat Biotechnol 14:303–308 22. Rajkowitsch L, Semrad K, Mayer O, Schroeder R (2005) Assays for the RNA chaperone activity of proteins. Biochem Soc Trans 33:450–456 23. Arluison V, Hohng S, Roy R, Pellegrini O, Regnier P, Ha T (2007) Spectroscopic observation of RNA chaperone activities of Hfq in post-transcriptional regulation by a small non-coding RNA. Nucleic Acids Res 35:999–1006 24. LeTilly V, Royer CA (1993) Fluorescence anisotropy assays implicate protein-protein interactions in regulating trp repressor DNA binding. Biochemistry 32:7753–7758 25. Sun X, Wartell RM (2006) Escherichia coli Hfq binds A18 and DsrA domain II with similar 2:1 Hfq6/RNA stoichiometry using different surface sites. Biochemistry 45:4875–4887 26. Panja S, Schu DJ, Woodson SA (2013) Conserved arginines on the rim of Hfq catalyze base pair formation and exchange. Nucleic Acids Res 41:7536–7546 27. Panja S, Santiago-Frangos A, Schu DJ, Gottesman S, Woodson SA (2015) Acidic residues in the Hfq chaperone increase the selectivity of sRNA binding and annealing. J Mol Biol 427(22):3491–3500 28. Santiago-Frangos A, Jeliazkov JR, Gray JJ, Woodson SA (2017) Acidic C-terminal domains autoregulate the RNA chaperone Hfq. eLife 6:pii 27049

Chapter 3 Fluorescent Molecular Beacons Mimicking RNA Secondary Structures to Study RNA Chaperone Activity Pilar Menendez-Gil, Carlos J. Caballero, Cristina Solano, and Alejandro Toledo-Arana Abstract Molecular beacons (MBs) are oligonucleotide probes with a hairpin-like structure that are typically labelled at the 50 and 30 ends with a fluorophore and a quencher dye, respectively. The conformation of the MB acts as a switch for fluorescence emission. When the fluorophore is in close proximity to the quencher, fluorescence emission cannot be detected, meaning that the switch is in an OFF state. However, if the MB structure is modified, separating the fluorophore from the quencher, the switch turns ON allowing fluorescence emission. This property has been extensively used for a wide variety of applications including real-time PCR reactions, study of protein-DNA interactions, and identification of conformational changes in RNA structures. Here, we describe a protocol based on the MB technology to measure the RNA unfolding capacities of the CspA RNA chaperone from Staphylococcus aureus. This method, with slight variations, may also be applied for testing the activity of other RNA chaperones, RNA helicases, or ribonucleases. Key words RNA, Chaperone, RNA-binding protein, Hairpin, Stem loop, Molecular beacon, Fluorescein, Quencher, FAM

1

Introduction Molecular beacons (MBs) are oligonucleotide probes commonly used to target DNA for real-time monitoring of polymerase chain reactions (RT-PCRs). The central nucleotides of the MB are complementary to a specific DNA (or RNA) target and do not base pair with one another, while the five to seven nucleotides at each terminus are complementary to each other, creating a hairpin-like conformation (Fig. 1a). Since the 50 and 30 ends are labelled with a fluorophore and a quencher, respectively, the MB acts as a switch. In their native conformation, the extremes are close enough for the quencher to prevent fluorescent emission from the fluorophore (switch OFF). When the MB hybridizes with its specific target, its native structure is disrupted, and both dye molecules fall apart from

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_3, © The Author(s) 2020

41

42

Pilar Menendez-Gil et al.

Fig. 1 Examples of different MB designs dedicated to (a) quantifying specific DNA or RNA molecules, (b) analyzing the single-stranded DNA cleavage by specific nucleases [4], (c) studying the structural changes on ribozymes and riboswitches [5, 6], or (d) determining the RNA chaperone activity on hairpin-like structures by cold shock proteins (CSPs) [7–9]. In all cases, the MB switch turns on when the fluorophore (FAM) folds away of the quencher (BHQ_1) due to the base-pairing of the MB with its specific target (a, c) or to the MB cleavage or unfolding by the activity of an RNA-binding protein (b, d)

each other, allowing fluorescence emission (switch ON) (Fig. 1a). Since MBs tolerate very versatile designs, they have been used for various applications [1, 2]. Molecular biologists have taken advantage of their potential for studying different mechanisms such as protein-DNA interactions [3], single-stranded DNA cleavage by specific nucleases (Fig. 1b) [4], and structural changes on ribozymes and riboswitches (Fig. 1c) [5, 6]. In this last case, RNA conformational changes have been determined by the use of MBs that target specific RNA regions that become free for hybridization. This usually occurs after binding of the metabolite, which induces the subsequent RNA structural change on the ribozyme or riboswitch (Fig. 1c) [5, 6]. Thus, only when the MB is bound to its RNA target, the probe structure unfolds and becomes fluorescent. On the other hand, in order to study RNA chaperone activity, a more direct approach by using a MB that mimics the regulatory RNA hairpin targeted by a specific RNA-binding protein (RBP) has been adopted [7, 8]. This strategy assumes that binding of the RNA chaperone to the MB may cause a similar RNA conformational rearrangement to the one occurring on the native RNA. Therefore, the MB may act as a direct reporter of its own structural rearrangement (Fig. 1d). We choose this approach to demonstrate that the

Molecular Beacons to Assess RNA Chaperone Activity

43

Fig. 2 Molecular beacon (MB) design to study the CspA RNA chaperone activity on the RNA hairpin structure of the cspA mRNA [9]. (a) The proposed RNA structure for the 50 UTR of cspA is shown [10]. The U-rich motif required for CspA interaction is highlighted in red. (b) The MB consisted of a 49-mer ssDNA oligonucleotide labelled with the FAM fluorophore and the BHQ_1 quencher at its 50 and 30 ends, respectively [9]

RNA chaperone CspA of Staphylococcus aureus unfolds the RNA hairpin present in the 50 UTR of its own mRNA [9] (Fig. 2a). This hairpin (ΔG ¼ 24.60 kcal/mol) is cleaved by endoribonuclease III (RNase III) mainly at position G-53, generating a shorter cspA mRNA version that is more efficiently translated than the unprocessed mRNA [10]. CspA would repress its own expression by unfolding the hairpin and thus antagonize the function of RNase III [9]. Specifically, we designed a MB that comprised a 49-mer ssDNA oligonucleotide, which included the central functional sequence of the cspA 50 UTR hairpin (ΔG ¼ 13.70 kcal/mol). A molecule of fluorescein (FAM) and a Black Hole Quencher (BHQ_1) were attached to the 50 and 30 ends, respectively (Fig. 2b). In the native MB conformation, BHQ_1 efficiently quenched the fluorescence from FAM, indicating that the designed MB accurately mimicked the cspA 50 UTR hairpin. In contrast, when the MB structure was disrupted (separating FAM from BHQ_1) either by the presence of the RNA chaperone CspA or by an increase in the temperature of incubation, fluorescence emission was registered. The folded

44

Pilar Menendez-Gil et al.

Fig. 3 Schematic representation of the putative auto-regulatory mechanism modulating CspA expression as previously described by Caballero and colleagues [9]. The 50 UTR of the cspA mRNA forms a hairpin structure that is cleaved by RNase III to enhance CspA translation when CspA levels are low [10]. When the concentration of CspA inside the cell is high, the protein is able to interact with the hairpin structure through a U-rich motif and unfold it. As a consequence, the cspA mRNA is not processed by RNase III and CspA translation is decreased

conformation of the MB could be efficiently restored (indicated by the ceasing of fluorescence emission) either by adding Proteinase K, which eliminated the chaperone activity by degrading CspA, or by decreasing the temperature of incubation. The specificity of CspA on the designed MB system was verified by the incubation of the MB with an unrelated protein. This strategy allowed us to demonstrate that CspA unfolded the regulatory hairpin located at the cspA 50 UTR and, thus, interfered with cspA mRNA processing by RNase III. When CspA levels were low, the cspA 50 UTR RNA hairpin was targeted and cleaved by RNase III. The resulting processed mRNA suffered a conformational change that favored CspA translation [10]. When CspA levels rose, CspA decreased its own expression by unfolding the cspA 50 UTR RNA hairpin to avoid RNase III cleavage [9] (Fig. 3).

Molecular Beacons to Assess RNA Chaperone Activity

45

Here, we describe in detail the different steps that should be followed to determine the RNA folding rearrangements caused by the binding of any RBP by using a MB that mimics a natural target (whose synthesis could be ordered from a regular oligonucleotide supplier company). The protocol requires commonly available equipment at a molecular biology research center. It is noteworthy that, with slight modifications, this protocol may be adapted to test (1) any DNA or RNA folding structure that allows close proximity of BHQ_1 to FAM and that provides enough separation between them when disrupted; (2) the activity of RBPs such as RNA chaperones, RNA helicases and ribonucleases that target and/or process hairpin-like structures; and (3) the function of small regulatory RNAs that produce conformational changes on hairpin-like structures of their mRNA targets.

2

Materials Prepare all solutions using ultrapure water (prepared by purifying deionized water to reach a sensitivity of at least 18 MΩ at 25  C) and analytical grade reagents for use in molecular biology. Store solutions at room temperature unless stated otherwise. Follow safety and waste disposal regulations when handling harmful products accordingly.

2.1 Purification of the Recombinant RNA Chaperone CspA 2.1.1 Growth of Bacteria Expressing GST-CspA Fusion Protein

1. E. coli BL21 (DE3) harboring pGEX-6P-2::cspA (see Note 1). 2. Sterile material for bacterial growth: 10-, 100- and 1000-μL pipette tips, test tubes, 2-L Erlenmeyer flasks, graduated cylinders, 250-mL centrifuge tubes, petri dishes, 1.5-mL Eppendorf tubes. 3. 100 mg/mL ampicillin stock solution sterilized by filtration. Store at 20  C. 4. 40% Glucose solution sterilized by filtration. 5. Luria Bertani (LB) agar plates supplemented with 100 μg/mL ampicillin. 6. Microbiological incubator at 37  C. 7. Luria Bertani (LB) medium sterilized by autoclave and supplemented with glucose and ampicillin to a final concentration of 1% and 100 μg/mL, respectively. 8. Shaking incubator at 37  C and 200 r.p.m. 9. Spectrophotometer. 10. Centrifuge with rotor for Eppendorf tubes. 11. 200 mg/mL Isopropyl-β-D-1-thiogalactopyranoside (IPTG) stock solution, sterilized by filtration. Store at 20  C.

46

Pilar Menendez-Gil et al.

12. Refrigerated centrifuge with rotor for 250-mL centrifuge flasks. 13. Phosphate buffered saline (PBS): pH 7.3, 140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4 sterilized by autoclave. 2.1.2 Bacterial Cell Lysis and Recovery of Total Protein Crude Extract

1. Sterile 15-mL and 50-mL conical tubes. 2. Sterile PBS pH 7.3 (see item 13 in Subheading 2.1.1). 3. 50 mg/mL Lysozyme stock solution sterilized by filtration. Store at 20  C. 4. 10 mg/mL RNase A stock solution. Store at 20  C. 5. 100 mM Phenylmethanesulfonyl fluoride (PMSF) stock solution prepared in isopropanol. Store at 20  C (see Note 2). 6. Shaking incubator at 30  C and 200 r.p.m. 7. Branson sonifier 250 with microtip. 8. Centrifuge with a rotor for 50-mL tubes, which allows centrifugation at 16,000  g. 9. High speed centrifuge tubes. 10. 5 mg/mL DNase I stock solution prepared by dissolving DNase I powder in 0.15 M NaCl. 11. 0.45 μm filters and 1.5-mL Eppendorf tubes. 12. 12% SDS-polyacrylamide gels. 13. 6 Sample buffer: 375 mM Tris–HCl pH 6.8, 9% SDS, 50% glycerol, 9% β-mercaptoethanol and 0.03% bromophenol blue. Store at 20  C. 14. Tris–glycine running buffer: 25 mM Tris, 192 mM glycine, 0.1% SDS. 15. Protein molecular weight marker. Store at 20  C. 16. Heating block. 17. Electrophoresis chamber for polyacrylamide gels. 18. Power supply. 19. Coomassie brilliant blue R250 solution. 20. Orbital shaker. 21. Destaining solution: 40% ethanol and 10% acetic acid in water. 22. Fixation solution: 10% ethanol and 3% glycerol in water.

2.1.3 Purification of Recombinant CspA from Total Protein Crude Extracts

1. AKTAprime plus chromatography system. 2. GSTrap FF 5-mL column. 3. GSTrap FF 1-mL column. 4. HiPrep 16/60 Sephacryl S-100 HR column.

Molecular Beacons to Assess RNA Chaperone Activity

47

5. Ultrapure water, sterilized by autoclave and degassed. 6. 20% ethanol solution sterilized by autoclave and degassed. 7. Binding Buffer: degassed sterile PBS pH 7.3 (see item 13 in Subheading 2.1.1). 8. PreScission Protease buffer: 50 mM Tris–HCl pH 7, 150 mM NaCl, 1 mM EDTA, 1 mM DTT sterilized by autoclave and degassed. 9. PreScission Protease. 10. PreScission Protease mix: mix 100 μL (200 units) of PreScission Protease with 4.9 mL of PreScission Protease buffer at 4  C. 11. 5-mL syringe with Luer tip. 12. 1.5-mL Eppendorf tubes and 15-mL conical tubes. 13. Elution buffer: 50 mM Tris–HCl pH 8, 10 mM reduced glutathione sterilized by autoclave and degassed. 14. Gel Filtration 500 mM NaCl.

buffer:

20

mM

Tris–HCl

pH

7.4,

15. Slide-A-Lyzer Dialysis Cassettes. 16. 0.22 μm filters. 17. CspA Storage buffer: 10 mM Tris–HCl pH 8, 1 mM EDTA, 50 mM potassium chloride and 10% glycerol. 18. Bio-Rad protein assay. 19. 96-well standard plates. 20. MultiSkan EX (Labsystems) or any other equivalent microplate photometer. 2.2 Assessment of the RNA Chaperone Activity with a Molecular Beacon

1. Molecular beacon mimicking the hairpin structure under study (see Note 3). 2. Spectrophotometer NanoDrop).

equipped

with

a

UV

lamp

(e.g.,

3. Recombinant RNA chaperone CspA (or protein of interest). 4. Bovine Serum Albumin (BSA) dissolved in CspA storage buffer at the same concentration as the protein of interest (see Note 4). 5. TE buffer: 10 mM Tris–HCl pH 7.5 and 1 mM EDTA. 6. 96-well PCR plates suitable for the Real-Time PCR System available. Plate is sealed with optically clear adhesive film (see Note 5). 7. AriaMx Real-Time PCR System (Agilent Technologies) or any other equivalent thermal cycler including an optical system able to excite the FAM fluorophore and register the fluorescence emission at different temperature incubation times.

48

Pilar Menendez-Gil et al.

8. CspA storage buffer (see item 17 in Subheading 2.1.3). 9. 10 reaction buffer: 100 mM Tris–HCl pH 7.5, 300 mM KCl, 200 mM NH4Cl, 15 mM DTT, 50 mM MgCl2 (see Note 6). 10. 4 U/μL Ribolock (see Note 7). 11. Proteinase K (Sigma) stock solution. Dissolved Proteinase K powder in water to a final concentration of 20 mg/mL (see Note 8). Store at 20  C.

3

Methods

3.1 Purification of the Recombinant RNA Chaperone CspA

1. Streak the E. coli BL21 (DE3) pGEX-6P-2::cspA strain in an LB agar plate supplemented with 100 μg/mL ampicillin and incubate at 37  C overnight.

3.1.1 Growth of Bacteria Expressing the GST-CspA Fusion Protein

2. Inoculate a colony of the previous culture into a sterile test tube containing LB medium supplemented with 100 μg/mL ampicillin and 1% glucose. Grow culture at 37  C and 200 r.p.m. overnight. 3. Inoculate 500 μL of the bacterial preculture (1/1000 dilution factor) into two sterile pre-warmed 2-L Erlenmeyer flasks containing 500 mL of LB medium supplemented with 100 μg/mL ampicillin and 1% glucose. Mix and incubate the cultures at 37  C and 200 r.p.m. until an optical density (OD600nm) of 0.5 is reached. 4. Induce the expression of CspA by addition of IPTG to a final concentration of 0.4 mM. Save 1 mL of culture of one of the flasks and centrifuge it for 3 min at 18,000  g. Store the bacterial pellet at 20  C. This aliquot sample corresponds to the pre-induction control (see Note 9). Resume bacterial growth for another 5 h at 37  C and 200 r.p.m. 5. Save 1 mL of culture of one of the flasks and centrifuge it for 3 min at 18,000  g. Store the bacterial pellet at 20  C (postinduction control) (see Note 9). Harvest the rest of the cultures in 250-mL tubes and centrifuge for 10 min at 5000  g (see Note 10). Discard the supernatant and resuspend the pellets in 1 volume of PBS pH 7.3. Repeat the centrifugation step, discard the supernatant and store the bacterial pellets at 80  C (see Note 11).

3.1.2 Bacterial Cell Lysis and Recovery of Total Protein Crude Extract

1. Thaw the bacterial pellets, resuspend them in 25 mL PBS pH 7.3 in 50-mL conical tubes (per pellet) and add lysozyme, RNase and PSMF to a final concentration of 1 mg/mL, 10 μg/ mL, and 1 mM, respectively. Incubate the samples for 30 min at 30  C and 200 r.p.m.

Molecular Beacons to Assess RNA Chaperone Activity

49

2. Sonicate the samples on ice as follows: 3 cycles of 30 s power 4, 2 cycles of 30 s power 5. Leave the samples on ice for 1 min in between cycles. 3. Centrifuge the samples at 16,000  g for 30 min at 4  C (see Note 10). Transfer the supernatant (soluble fraction) to new tubes and store the pellet at 20  C. Pellets (insoluble fraction) contain inclusion bodies (IB), and constitute the IB control (see Note 9). 4. Supplement the soluble fraction with DNase I and RNase A to a final concentration of 10 μg/mL and 5 μg/mL, respectively, and incubate on ice for 30 min. Store 50 μL of the sample at 20  C (pre-filtered soluble fraction control) (see Note 9). 5. Filter the soluble fraction using a 0.45 μm filter whilst on ice (see Note 12). Store 50 μL of the sample at 20  C (postfiltered soluble fraction control) and the rest of the soluble fraction at 20  C (see Note 9). 6. Mix aliquots of the different control samples (pre-induction control, post-induction control, IB control, pre-filtered soluble fraction and post-filtered soluble fraction), collected in the previous steps (see Note 9), with 6 sample buffer to a final concentration of 1. Denature mixtures at 95  C for 5 min and load them in a polyacrylamide gel (a Molecular Weight Marker should be included) (see Note 13). Run the gel with 1 running buffer at 130 V until the front reaches the bottom of the gel (see Note 14). 7. Stain the gel with Coomassie blue for at least 4 h at room temperature on an orbital shaker. Destain the gel with several washes of destaining solution at room temperature and shaking. Once protein bands are visible and the background level is low, incubate the gel with fixing solution for 15 min at room temperature and shaking. Optimal results are reached when most of the GST-CspA fusion protein appears in the soluble fraction and not in the inclusion bodies fraction (see Note 15). 3.1.3 Purification of the Recombinant CspA Protein

1. Thaw the post-filtered soluble fraction and purify the GST-CspA fusion protein with the use of a GSTrap FF 5-mL column and an AKTAprime plus chromatography system, following the recommendations of the manufacturer. 2. Clean the system with 20% ethanol and ultrapure water. 3. Connect the column to the AKTAprime plus system “drop to drop” to avoid introducing air into the column. Equilibrate the column with 25 mL of binding buffer at a flow rate of 5 mL/ min. 4. Apply the sample at a flow rate of 0.2 mL/min (see Note 16).

50

Pilar Menendez-Gil et al.

5. Wash the column with 50 mL of binding buffer at a flow rate of 5 mL/min. 6. Equilibrate the column with 50 mL of PreScission Protease buffer at a flow rate of 5 mL/min and disconnect the column from the AKTAprime plus chromatography system. 7. Prepare the PreScission Protease mix at 4  C and load it manually onto the column using a syringe at a flow rate of 1 mL/ min. Seal the column with the top and bottom stop plugs and incubate overnight at 4  C. 8. Connect a GSTrap FF 1-mL column to the AKTAprime plus system and equilibrate it with 5 mL of PreScission Protease buffer at a flow rate of 1 mL/min. 9. Place the GSTrap FF 5-mL column on top of the GSTrap FF 1-mL column. This tandem column scheme acts as a filter to capture any released cleaved GST proteins, uncleaved GST-tagged proteins and unbound PreScission Protease. Elute CspA with 15 mL of PreScission Protease buffer at a flow rate of 1 mL/min. Collect 1 mL fractions containing the CspA protein and place them on ice. 10. Elute the GST and GST-PreScission Protease from the columns with 30 mL of elution buffer at a flow rate of 1 mL/ min. Clean the system and columns with ultrapure water and 20% ethanol and remove columns from the system. 11. Dialyze the CspA fractions against Gel Filtration buffer using a Slide-A-Lyzer Dialysis Cassette overnight at 4  C. Collect CspA from the Dialysis Cassette and filter the solution using a 0.22 μm filter. Keep the CspA sample on ice until its purification by size exclusion chromatography. 12. Connect a HiPrep 16/60 Sephacryl S-100 HR Column (see Note 17) to the AKTAprime plus system “drop to drop” to avoid introducing air into the column. Equilibrate the column with 60 mL of ultrapure water at a flow rate of 0.5 mL/min and then with 240 mL of Gel Filtration buffer at a flow rate of 1 mL/min. 13. Inject the CspA sample into the column and run it with 120 mL of Gel Filtration buffer at a flow rate of 0.5 mL/ min. Collect 3 mL fractions and place them on ice. 14. Clean the column with 480 mL of ultrapure water and 480 mL of 20% ethanol at a flow rate of 1 mL/min. Remove the column from the system and clean the system with ultrapure water and 20% ethanol. 15. To select fractions containing CspA, mix an aliquot of each peak fraction with sample buffer 6 and perform a 12% PAGE as described above.

Molecular Beacons to Assess RNA Chaperone Activity

51

16. Load the CspA selected fractions into a Slide-A-Lyzer Dialysis Cassette and dialyze against CspA Storage buffer overnight at 4  C. 17. To assess protein purity, mix an aliquot of the recombinant CspA chaperone with sample buffer 6 and perform a 12% SDS-polyacrylamide gel electrophoresis (PAGE) as described above. 18. Determine the recombinant protein concentration by the Bio-Rad protein assay. 3.2 Assessment of the RNA Chaperone Activity with a Molecular Beacon 3.2.1 Molecular Beacon Design

3.2.2 Testing the Effectiveness of the Designed MB and Setup of the Working Conditions

The success of this assay lies in an adequate MB design, which is based on two main principles: (1) the presence of an RNA structure targeted by the RNA chaperone under study and (2) fluorescence quenching exerted by a quencher dye (e.g., BHQ_1) on a fluorophore (e.g., FAM), which occurs when both molecules are in close proximity to one another. Additionally, the selected RNA structure must keep the quencher close enough to the fluorophore at the working temperature (switch OFF). MB mimicking hairpin-like structures have been shown to comply these criteria before [7–9]. Likewise, alternative MB conformations can be tested according to the characteristics of the protein of interest. Having decided the MB configuration, the synthesis of the labelled probe may be ordered to any oligonucleotide supplier (see Note 18). Based on the binding capacity of the CspA protein and the stability and cost of the probe, we decided to use single-stranded DNA instead of RNA oligonucleotides (see Note 19). Before assessing if the RNA chaperone is able to unwind the MB structure, the effectiveness of the designed MB must be tested following two main criteria. On the one hand, no fluorescence emission should be detected when the MB is in an OFF state. On the other hand, maximum fluorescence levels should be registered when the MB is completely unfolded (ON state). In other words, a MB design will be appropriate when a large ON/OFF fluorescence ratio is detected. Incubating different MB concentrations at increasing temperatures (that denature the oligonucleotide and open the structure leading to FAM fluorescence emission) helps determining both the background fluorescence (see Note 20) and the lowest quantity of oligonucleotide needed to obtain good fluorescence levels when the MB is in an ON state (see Note 21). To test if the MB design was successful, proceed as follows: 1. Dissolve the MB in TE buffer to obtain a concentration of 100 μM, following the manufacturer recommendations (see Note 22). Concentration of the MB should be corroborated with a spectrophotometer (e.g., NanoDrop).

52

Pilar Menendez-Gil et al.

Table 1 Preparation of dilution mixes for testing MB effectivenessa MB concentration (pmol) 0

0.5

1

2

5

10

15

20

MB 10 μM

–

–

–

–

0.5

1

1.5

2

MB 1 μM

–

0.5

1

2

–

–

–

–

CspA storage bufferb

12.5

12

11.5

10.5

12

11.5

11

10.5

10 reaction buffer

2.5

2.5

2.5

2.5

2.5

2.5

2.5

2.5

Ultrapure water

10

10

10

10

10

10

10

10

Final volume

25 μL

Volumes of each reactive are indicated in μL Since the CspA protein is diluted in CspA storage buffer, the MB effectiveness test is performed including this buffer

a

b

2. Program the AriaMx thermal cycler to incubate the MB samples as follows: 37  C, 5 min; 45  C, 5 min; 55  C, 5 min; and 65  C during 5 min (see Note 23). Register the emission of FAM fluorescence every minute. 3. Make serial dilutions of the MB in an optical 96-well plate as indicated in Table 1, which shows mixtures of the components to analyze different concentrations of the MB. Triplicates are highly recommended. 4. Seal the plate with an optically clear adhesive film (see Note 5) and load it into the thermal cycler. Start the incubation program. 5. Once the incubation time is finished, plot the obtained fluorescence signals in function of the MB concentration at the different temperatures. If replicates are used, plot the means of the fluorescence signals. The instrument background signal should be previously subtracted. Figure 4 shows an example of the results obtained with the MB designed for the analysis of S. aureus CspA activity [9] (Fig. 4). In this example, when the MB was incubated at 55  C and 65  C, fluorescence emission was registered, indicating that the MB was in an ON state. These fluorescence levels were directly proportional to the MB concentration. In contrast, when the MB was incubated at 37  C and 45  C, the fluorescence values were close to those of the background confirming that the MB was in an OFF configuration. This experiment validated the functionality of the designed MB (see Note 24).

Molecular Beacons to Assess RNA Chaperone Activity

53

Fig. 4 Test of the molecular beacon functionality. Different concentrations of the MB mimicking the hairpin structure located at the 50 UTR of the cspA mRNA were incubated at different temperatures and fluorescence emission was registered. The experiment was carried out using the AriaMx thermal cycler 3.2.3 Determination of the RNA Chaperone Activity Using the Designed MB

1. Based on the data obtained from the MB effectiveness test, select the lowest MB concentration that gives a good ratio between the fluorescence and background signals (see Note 21). 2. Program the AriaMx thermal cycler to incubate the MB samples as follows: 37  C, 5 min; PAUSE, 37  C, 15 min; PAUSE, 37  C, 30 min; 65  C, 10 min; STOP (Table 2). Register the fluorescence emission every minute (see Note 25). 3. Prepare an optical 96-well plate including the reaction mixes as indicated in Table 2 (see Note 26). Note that the CspA and BSA proteins should be added later. 4. Seal the plate with adhesive film (see Note 5) and load it into the thermal cycler. Start the incubation program. 5. At the first pause of the incubation program, pull out the 96-well plate from the thermal cycler, remove the adhesive film and add the appropriate quantity of CspA and BSA. Re-seal the plate with a new adhesive film. This step must be performed swiftly. 6. Reintroduce the plate into the thermal cycler and continue the incubation at 37  C during 15 min. Register the fluorescence emission every minute. 7. During the second incubation pause, pull out the plate, remove the adhesive film and add 10 μL of proteinase K (20 mg/mL). Re-seal the plate with a new adhesive film. This step must be performed swiftly. 8. Reintroduce the plate into the thermal cycler and continue the incubation for 30 min at 37  C and then increase the temperature up to 65  C during 10 min. Register fluorescence emission

54

Pilar Menendez-Gil et al.

Table 2 Determination of RNA chaperone activity: preparation of reaction mixesa Samplesb 1

2

3

4

5

6

7

8

MB tube labelling



+

+

+

+

+

+

+

CspA tube labelling





+

+

+







BSA tube labelling











+

+

+

Water (Vf: 100 μL)

39

38

38

38

38

38

38

38

CspA storage buffer

50

50

30

15



30

15



10 reaction buffer

10

10

10

10

10

10

10

10

MB 1 μM



1

1

1

1

1

1

1

Ribolock 4 U/μL

1

1

1

1

1

1

1

1

Seal the plate with adhesive film Incubate 37  C—5 min Register fluorescence emission every minute PAUSE incubation programc CspA stock (~200 μM)





20

35

50







BSA stock (~200 μM)











20

35

50

10

10

10

Re-seal the plate with adhesive film Incubate 37  C—15 min Register fluorescence emission every minute PAUSE incubation program Proteinase K 20 mg/mL

10

10

10

10

10

Re-seal the plate with adhesive film Incubate 37  C—30 min Incubate 65  C—10 min Register fluorescence emission every minute Collect the fluorescence data from AriaMx thermal cycler Plot the data accordingly Volumes of each reactive are indicated in μL Replicates of samples should be included c If the thermal cycler software allows it, the entire incubation protocol can be pre-programed including the corresponding PAUSE times a

b

Molecular Beacons to Assess RNA Chaperone Activity

55

every minute. Once the incubation program is finished, collect the result data sheet. 9. Plot the obtained data subtracting the background fluorescence levels. If the experiments work as expected, fluorescence emission should be registered after addition of the RNA chaperone. This fluorescence should disappear after treatment with Proteinase K, showing the specificity of the reaction. Finally, increasing the temperature at 65  C should lead to maximum levels of fluorescence, indicating that the MB remains functional through the course of the experiment. Logically, in the negative controls, no fluorescence emission should be detected until the last step, when temperature is raised (e.g., see ref. 9).

4

Notes 1. Although we purified the RNA chaperone CspA from E. coli using the glutathione S-transferase (GST) gene fusion system, any other recombinant purification alternative can be used for your protein of interest. We fused the cspA coding sequence to the GST gene in the pGEX-6P-2 plasmid. This allowed the convenient site-specific cleavage by the PreScission Protease between the GST domain and CspA at low temperature, minimizing the degradation of the protein of interest. Additionally, it provided the recovery of the recombinant CspA protein with only a few extra amino acids in its sequence (for details about the construction of this strain see ref. 9). The fact that the PreScission Protease was engineered with a GST tag, permitted an on-column cleavage so that the GST moiety of the tagged protein and the PreScission Protease itself remained bound to the Glutathione Sepharose column. Thus, at the end of the procedure CspA was not contaminated with the protease. 2. PMSF is unstable in the presence of water. A stock solution should be prepared in anhydrous isopropanol or anhydrous absolute ethanol. 3. Our designed MB consisted of a 49-mer single-stranded DNA oligonucleotide, which was synthetized and labelled at its extremes with the 6-FAM molecule and Black Hole Quencher (BHQ_1) by the Integrated DNA technologies company. This length was enough to include the functional part of the RNA hairpin (Fig. 2). 4. We used bovine serum albumin (BSA), a protein without capacity to bind nucleic acids, as a negative control. Any alternative protein lacking DNA/RNA binding domains can also be used.

56

Pilar Menendez-Gil et al.

5. We preferred to seal the 96-well plates with adhesive film because removing it and re-sealing the plates is faster than using flat caps. A quick sealing helps registering fluorescence emission sooner, after the RNA chaperone is added to the MB solution. 6. Storage of the 10 reaction buffer will require it to be prepared without DTT. DTT should be added just before use. 7. Any alternative RNase inhibitor can be used. If ssDNA is used as the backbone of the MB, RNase inhibitors are not required. 8. To demonstrate the specificity of chaperone activity, proteinase K (or any alternative protease) may be included to degrade the protein under study. This should eliminate the RNA chaperone activity and restore the MB folding. If fluorescence is not quenched again after Proteinase K treatment (in other words, the MB cannot be refolded), it might indicate a contamination of the RNA chaperone solution with nucleases that affect the MB integrity. 9. It is important to collect samples at various steps during the purification procedure to monitor the yield of the recombinant protein. Comparison of these control samples helps evaluating if (1) the induction of the recombinant protein expression is adequate (pre-induction vs post-induction control), (2) the recombinant protein is present in the soluble and/or the insoluble fraction (IB control vs soluble fractions), and (3) the recombinant protein is lost during the step of sample clearance by filtration (pre-filtered vs post-filtered soluble fractions). 10. Centrifuge should be pre-cooled before use. 11. Bacterial pellets can be stored at 80  C for several days. 12. We recommend the use of filters with a pore size of 0.45 μm instead of 0.2 μm to avoid filter saturation. 13. Pre-cast or custom-made gels may be used with the appropriate percentage of acrylamide according to the protein of interest (we used 12% PAGE). 14. Adjust voltage of the electrophoresis system accordingly. 15. If the protein of interest is not in the soluble fraction, bacterial growing conditions should be modified to force its solubilization. Alternatively, protein purification methods from inclusion bodies may be applied. 16. Due to the slow binding kinetics between GST and glutathione, it is very important to keep the flow rate as low as possible during sample application for maximum binding capacity. 17. The column used in this protocol is specific for separating proteins with a small size. If the RNA chaperone of interest has a bigger size, the column should be changed accordingly.

Molecular Beacons to Assess RNA Chaperone Activity

57

18. Some oligonucleotide supplier companies limit the synthesis of labelled oligonucleotide probes to 50 nucleotides (nt). In our design, the functional RNA hairpin region could be included in an oligonucleotide probe smaller than 50 nt. For larger regulatory structures, the synthesis of a MB may prove more challenging. This problem could be solved by dividing the MB synthesis into two shorter oligomers that can afterwards be ligated as previously described [7]. 19. The reason for using a labelled DNA oligonucleotide as a MB is that it has been proven that CSPs can bind ssDNA as efficiently as RNA molecules [11]. Nevertheless, testing other RNA-binding proteins may require synthesis of RNA-based MBs. 20. If the region of the RNA structure under study is not strong enough to maintain the MB beacon in an OFF state, the basal level of fluorescence might be too high to obtain reliable results once the RNA chaperone is added. 21. Sometimes the quantity or the concentration of the chaperone under study can be limited. We recommend using the lowest concentration of the MB that gives good fluorescent levels in an ON state. This will help saving RNA chaperone sample. 22. The fluorophores of the MB are sensitive to the light; therefore, keep the stock and any other dilutions wrapped in aluminum foil and protect them from exposure to light to maintain their integrity. 23. The selected temperature might vary depending on the melting temperature of the MB structure. 24. If the control of the MB functionality does not show clear differences on the fluorescence signals between ON and OFF states, and/or the fluorescence background is too high, the MB should be redesigned. 25. The entire incubation protocol to be carried out with the AriaMx thermal cycler (or any equivalent equipment) can be programed from the beginning, including the corresponding pause times required to add the different components of the reactions. 26. Volumes of each reactive should be adjusted according to the concentration of the RNA chaperone.

Acknowledgments ˜ igo Lasa for critical reading of the manuscript. We thank Prof. In This work was supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 646869) and the Spanish

58

Pilar Menendez-Gil et al.

Ministry of Economy and Competitiveness (BFU2014-56698-P). C.J.C. was supported by a predoctoral contract from the Public University of Navarre (UPNA), Spain. References 1. Tan W, Wang K, Drake TJ (2004) Molecular beacons. Curr Opin Chem Biol 8:547–553 2. Marx A, Seitz O (2010) Molecular beacons: signalling nucleic acid probes, methods, and protocols. Humana Press, Totowa 3. Li J, Cao ZC, Tang Z et al (2008) Molecular beacons for protein-DNA interaction studies. Methods Mol Biol 429:209–224 4. Li JJ, Geyer R, Tan W (2000) Using molecular beacons as a sensitive fluorescence assay for enzymatic cleavage of single-stranded DNA. Nucleic Acids Res 28:E52 5. Hopkins JF, Woodson SA (2005) Molecular beacons as probes of RNA unfolding under native conditions. Nucleic Acids Res 33:5763–5770 6. Chinnappan R, Dube´ A, Lemay J-F et al (2013) Fluorescence monitoring of riboswitch transcription regulation using a dual molecular beacon assay. Nucleic Acids Res 41:e106 7. Phadtare S, Inouye M, Severinov K (2002) The nucleic acid melting activity of Escherichia coli

CspE is critical for transcription antitermination and cold acclimation of cells. J Biol Chem 277:7239–7245 8. Kuehnert J, Sommer G, Zierk AW et al (2015) Novel RNA chaperone domain of RNA-binding protein La is regulated by AKT phosphorylation. Nucleic Acids Res 43:581–594 9. Caballero CJ, Menendez-Gil P, CatalanMoreno A et al (2018) The regulon of the RNA chaperone CspA and its auto-regulation in Staphylococcus aureus. Nucleic Acids Res 46:1345–1361 10. Lioliou E, Sharma CM, Caldelari I et al (2012) Global regulatory functions of the Staphylococcus aureus endoribonuclease III in gene expression. PLoS Genet 8:e1002782 11. Phadtare S, Inouye M (1999) Sequenceselective interactions with RNA by CspB, CspC and CspE, members of the CspA family of Escherichia coli. Mol Microbiol 33:1004–1014

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Chapter 4 Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein Deduced from Hairpin Unfolding Micah J. McCauley, Ioulia Rouzina, and Mark C. Williams Abstract RNA and DNA hairpin formation and disruption play key regulatory roles in a variety of cellular processes. The 59-nucleotide transactivation response (TAR) RNA hairpin facilitates the production of full-length transcripts of the HIV-1 genome. Yet the stability of this long, irregular hairpin becomes a liability during reverse transcription as 24 base pairs must be disrupted for strand transfer. Retroviral nucleocapsid (NC) proteins serve as nucleic acid chaperones that have been shown to both destabilize the TAR hairpin and facilitate strand annealing with its complementary DNA sequence. Yet it has remained difficult to elucidate the way NC targets and dramatically destabilizes this hairpin while only weakly affecting the annealed product. In this work, we used optical tweezers to measure the stability of TAR and found that adding NC destabilized the hairpin and simultaneously caused a distinct change in both the height and location of the energy barrier. This data was matched to an energy landscape predicted from a simple theory of definite base pair destabilization. Comparisons revealed the specific binding sites found by NC along the irregular TAR hairpin. Furthermore, specific binding explained both the unusual shift in the transition state and the much weaker effect on the annealed product. These experiments illustrate a general method of energy landscape transformation that exposes important physical insights. Key words Energy landscape, Transition state, Nucleic acid hairpin, Reverse transcription, Optical tweezers, Nucleic acid-protein interactions, Mfold

1

Introduction RNA and DNA hairpin folding and unfolding represent critical, often rate-limiting steps in several processes [1]. Folding and unfolding for a variety of hairpins have been investigated, often including the effects of secondary and tertiary contacts [2–7]. Furthermore, as hairpins often interact in concert with other ligands, these interactions have been under increasing focus. Some ligands have been shown to stabilize hairpin structures, while others weaken them as they serve to facilitate key cellular processes [5, 8]. Central to these studies is an important database of base pairing and stacking energies that form a predictive model of hairpin structure for a variety of sequences [9, 10]. Mfold allows vital

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_4, © Springer Science+Business Media, LLC, part of Springer Nature 2020

59

60

Micah J. McCauley et al.

Fig. 1 The nucleocapsid (NC) protein facilitates the rate limiting step of strand transfer during reverse transcription. (a) The sequence of the transactivation response (TAR) RNA hairpin of HIV-1 folds into an irregular hairpin punctuated by bubbles and mismatches. (b) An abbreviated schematic of reverse transcription begins with a single RNA transcript containing a TAR hairpin near each end. Conversion to doublestranded DNA begins with hybridization and synthesis but is limited by (
) strand transfer [12, 15, 20, 21]. Both TAR unfolding and strand annealing are facilitated by the presence of NC which increases the rate of this step by 104 [19]. (c, d) Two zinc “fingers” coordinate aromatic residue stacking with exposed bases, while a basic tail facilities chaperone activity [16–18]. (d) NC binds to the apical loop of SL3 RNA in this structure, which highlights aromatic stacking with unpaired bases (pdb code: 1a1t) [56]

comparisons to experimental unfolding data. Closer examination of base pair and stacking energies have even revealed a full transition landscape, expanding the potential for insight into experimental results [5–7]. HIV-1 RNA contains two TAR hairpins, each located near opposite ends of the transcript. The 59-base hairpin (see Fig. 1a) owes its stability (>40 kBT) to its length, though its structure is highly irregular and punctuated by both bubbles and mismatches with only 24 base pairs. It serves a critical role in the creation of fulllength transcripts while bound to the trans-activator of transcription (TAT) protein [11, 12]. Yet this stability becomes a liability once the virion infects a new host cell. During the process of reverse transcription, both TAR and the complementary cTAR DNA hairpin must unfold and anneal during strand transfer to continue the creation of double-stranded DNA. It is binding by NC that both

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

61

destabilizes TAR and facilitates nucleic acid annealing (see Fig. 1b for a brief outline and see Fig. 1c, d, for sequence and structure) [13–18]. This step has been shown to be the rate-limiting step of reverse transcription, and the absence of NC reduces this rate by a factor of 104 [19]. Yet while NC destabilizes the TAR hairpin, it has no similar effect on the annealed construct [12, 20, 21]. In this work we outline a general technique comparing singlemolecule data to energy landscape predictions from Mfold, with a special emphasis on ligand destabilization. In a previous review, we connected experimental data from single-molecule experiments to these landscapes [22]. Though we will briefly reiterate this technique, here we will focus on changes in TAR hairpin stability in saturating conditions of NC binding. Force unfolding data is used to alter the theoretical landscape, with protein binding directed to alter hairpin stability at a combination of possible individual sites. Comparisons to the fitted data allow us to discern the key NC binding sites on the hairpin that drive changes to the landscape. Specifically, NC is shown to induce an unusual change in the transition state extension, which explains the dramatic shift in the opening rate of the hairpin in its presence [23]. More generally, this method illustrates this technique of both qualitative and quantitative assessment of ligand binding, which applies to a variety of destabilizing and stabilizing ligands.

2

Materials Several studies have detailed the step-by-step creation of hairpin and labeled handle constructs for single-molecule study [23, 24], and the preparation of nucleocapsid protein p7 [25]. Furthermore, previous reviews have outlined landscape construction [5, 7]. The protocol that follows details only the essential steps to create both the theoretical and experimental data for comparison both with and without ligand binding.

2.1 Instruments, Reagents, and Buffers

1. Plasmid pBR322 cloning vector, 4361 base pairs, stored in 10 mM Tris–HCl, 1 mM EDTA, and pH 8.0 at 25  C. 2. Biotin-DNA handle sequences: DNA primer 1: 50 -Biotin-GGCCTTCCCCATTATGATTCTT CTCGC-30 . DNA primer 2: 50 -CCACAGGACGGGTGTGGTCGCCAT GATCG-30 . 3. Restriction endonuclease EcoR1 and reaction buffer: 100 mM Tris–HCl, 50 mM NaCl, 10 mM MgCl2, 0.025% Triton X-100, and pH 7.5 at 25  C. 4. Digoxygenin-DNA handle sequences:

62

Micah J. McCauley et al.

DNA primer 3: 50 -GGCCTTCCCCATTATGATTCTTCTC GC-30 . DNA primer 4: 50 -Digoxigenin-CCACAGGACGGGTGTGG TCGCCATGATCG-30 . 5. Restriction endonuclease BspE1 and reaction buffer: 100 mM NaCl, 50 mM Tris–HCl, 10 mM MgCl2, 100 μg/mL BSA, and pH 7.9 at 25  C. 6. TAR DNA 101 nucleotide template for PCR (T-7 promoter site is in bold, and followed by 12 nucleotide flanking site in italics, and then the TAR sequence and a final 12-nt flanking sequence in italics): 50 -GTAATACGACTCACTATAGGGTGACAGACCGGGT CTCTCTGGTTAGACCAGATCTGAG. CCTGGGAGCTCTCTGGCTAACTAGGGAACCCCCAGA CAGTGCC-30 . TAR primer 1: 50 -GGCACTGTCTGGGGGTTCCC-30 . TAR primer 2: 50 -CTGTAATACGACTCACTATAGG-30 . 7. Generated DNA 103 nucleotide template for in vitro transcription: 50 -CTGTAATACGACTCACTATAGGGTGACAGACCGG GTCTCTCTGGTTAGACCAGATCTGAG. CCTGGGAGCTCTCTGGCTAACTAGGGAACCCCCAGA CAGTGCC-30 . 8. T7 RNA polymerase and reaction buffer: 40 mM Tris–HCl, 6 mM MgCl2, 1 mM DTT, 2 mM spermidine and pH 7.9 at 25  C. Supplemented with 0.5 mM each ATP, UTP, GTP, and CTP. 9. Final TAR RNA sequence (59-nt flanked by 12-nt sequences in italics—see also Fig. 1a): 50 -GGGUGACAGACCGGGUCUCUCUGGUUAGACCAG AUCUGAG. CCUGGGAGCUCUCUGGCUAACUAGGGAACCCCCA GACAGUGCC-30 . 10. Ligation reaction DNA oligos (matching flanking sequences are italicized): 25-nt flanking: 50 -GGTCTGTCACCCCACACTATCTACT-30 . 17-nt flanking: 50 -P-CCGGAGTAGATAGTrGrUrG-30 (contains 3 ribonucleotides). 46-nt flanking: 50 -P-AATTGATACTAGCCTGATCAAGCT GCGGCATCTGGGCACTGTCTGG-30 . 30-nt flanking: 50 -P-CAGATGCCGCAGCTTGATCA GGCTAGTATC-30 .

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

2.2 Mfold (UNAfold) Structures and Energies

63

1. The UNAFold website (version 3.9 as of this writing) is a combined nucleic acid program that finds optimally folded structures for either DNA or RNA (see Note 1) [9]. For a given nucleic acid sequence, servers return an array of stably folded states with the energies of the folded state for a variety of experimental conditions and folding constraints (see Note 2). The main site may be found: http://unafold.rna.albany.edu/ Separate programs for DNA and RNA also exist, as well as programs describing hybridization. The RNA hairpins in this study were folded using the single-stranded RNA web server known as Mfold: http://unafold.rna.albany.edu/?q¼mfold/RNA-FoldingForm This server utilized version 3.6 of Mfold, which returned the same values as UNAfold for the hairpins discussed below [9]. 2. The sequence was folded under varying conditions and folding constraints. For DNA hairpins, the energy of folding was found under different monovalent and divalent salt concentrations as well as varying temperature (see Note 3). However, version 3.6 of Mfold would only fold RNA in 1 M Na+ and at 37  C (see Note 4), and what follows was obtained under those conditions (see Note 5) [10]. 3. Mfold returned all optimal folded structures with the total energy of folding (ΔGf). In the case of TAR and the fully matched hairpins studied below, only one folded structure was deduced (see Fig. 1a for the folded TAR structure). If more than one structure had been determined, then each would have been considered as part of an ensemble (see Notes 6 and 7). 4. In addition to the energy (ΔGf) required to unfold this sequence (length Δxf), Mfold solved a sequence of individual energies for each base pair and for each loop element, G(ni, F ¼ 0) (see Notes 8 and 9). While matched bases contributed to stability, loops and mismatches generally destabilized the overall hairpin. Finally, Σi G(ni, 0) ¼ ΔGf for the total folding hairpin energy.

2.3 RNA Hairpins with Labeled Handles for Single-Molecule Experiments

1. The 50 -biotinylated handle, 3400 base pairs in length, resulted from PCR amplified plasmid pBR322 and the biotin-DNA primers (50 -biotinylated primer 1 and primer 2). Once digested by EcoR1 (see Note 10), the result was purified on a 0.8% agarose gel. 2. The 30 -digoxygenein handle, 3100 base pairs in length, resulted from PCR amplified plasmid pBR322 and the

64

Micah J. McCauley et al.

digoxygenin-DNA primers (primer 3 and 30 -digoxygenein primer 4, see Note 11). After digestion by BSPE1, the result was purified on a 0.8% agarose gel. 3. The 59-base TAR hairpin was prepared from a PCR-amplified 103-base synthetic DNA template, which included additional flanking sites and the T7 promoter site (Note 12). The TAR RNA hairpin, with flanking sequences, was prepared with T7 RNA polymerase, and then purified with denaturing 10% PAGE gel [23, 24]. 4. The RNA hairpin was ligated to short DNA oligos on both the 50 and 30 side. For the 50 side, T4 RNA ligase 1 was used with a 30-base and 46-base DNA oligomer. On the 30 side, T4 RNA ligase 2 was used with a 25-base and a 17-base DNA oligomer, where three bases of the 17-mer (on the 30 side) were ribonucleotides (see Note 13). Each ligation was purified with denaturing 10% PAGE gel. 5. The resulting hairpin and ligated oligomer construct was then ligated to each labeled handle with T4 DNA ligase and purified on a 0.8% agarose gel after each step [23] (see Note 14). 2.4

Purified NCp7

1. DNA containing NC coding elements were generated by PCR and purified and digested with Kpn1 and Sal1 [25]. 2. Bacterial clones were grown and pelleted. Cycles of sonication, rinsing, centrifugation, and resuspension were followed by incubation in factor Xa buffer. 3. Final purification of NCp7 utilized high-performance liquid chromatography, followed by storage in
80  C in storage buffer (see Note 15) [25].

3

Methods Several excellent reviews summarize both the construction and utility of single-molecule optical tweezers, including the dualbeam design used here [26–31]. Other single-molecule techniques have studied hairpin unfolding and generated landscape information (most commonly probing the unfolding energies only) [32, 33]. In a previous review, we detailed the construction of a theoretical energy landscape from Mfold data and the process of relating the measured variables from an OT experiment to an experimental energy landscape [22–24]. Here, we will detail a minimum protocol necessary to recover landscapes from experimental OT data and then focus on changes wrought by ligand binding. We introduce that analysis with a fully matched RNA hairpin, as the TAR hairpin is highly irregular and includes many nonstandard features. The effects of ligand destabilization on a

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

65

uniform hairpin are shown for a few simple cases. Finally, destabilization of the TAR hairpin is quantified, revealing the role of NC in the altered energy landscape. The technique is general and may apply to a variety of other hairpin-binding ligands. 3.1 Optical Tweezers (OT) Experiments Quantify Hairpin Unfolding

1. The OT instrument used here is a dual beam single trap that holds a single micron-sized bead between counterpropagating lasers focused by a pair of water-dipping 1.2 NA objectives. Forces on the trapped bead may be detected to sub piconewton resolution, using a lateral effect detector to measure trapping beam deflection (see Note 16) [26]. 2. A cover glass-faced, custom flow cell with a fixed micropipette in the center of the chamber was held on a piezoelectric stage, capable of movement with sub-nanometer precision. 3. Stock solutions of the labeled beads are diluted 1000:1 into the experimental buffer (see Note 17), at an ambient temperature of 22  C. 4. Single control or hairpin constructs were secured between the 5-μm diameter streptavidin coated bead held in the trap and a 2-μm diameter anti-digoxigenin-coated bead pulled onto the micropipette (see Fig. 2a) (see Notes 18 and 19). 5. Beginning at low tension, the piezo increased the construct extension stepwise in 4 nm increments, with force measurements after each step (see Notes 20 and 21). At ~30 pN (see Note 22), the construct was gradually released, also in 4 nm increments. Increasing/decreasing signal averaging decreased/ increased the speed of extension and release, known as the loading rate (r) (see Note 23), which varied across 30, 10, and 0.8 pN/s (see Note 24). 6. Data for the hairpin-free construct showed the smooth increase and decrease in the force due to polymer elasticity during cycles of extension and release (see Fig. 2b) (see Note 25). 7. Data for constructs incorporating a TAR RNA hairpin revealed a sudden increase in length due to hairpin unfolding (Δxop, see Fig. 2b and see Note 26)), at a single force (Fop). A closing force (Fcl) was evident during release (see Note 27). A total of n ¼ 250 events were observed. 8. The work delivered by the instrument during unfolding (Wop, see Note 28) was found from the area between the extension of the handles with a folded and unfolded hairpin (see Fig. 2b and see Note 29). A similar measurement was made for the work of folding (Wcl). 9. These cycles and measurements were repeated in 50 nM NC, which is believed to saturate the TAR hairpin at these conditions (see Fig. 2b and see Note 30). Here, a total of n ¼ 162 events is observed over the three loading rates.

66

Micah J. McCauley et al.

Fig. 2 Optical tweezers experiments characterize unfolding. (a) A single RNA hairpin construct is fixed between two beads via distinct attachment chemistry. (Inset) Image of the experimental plane where functionalized beads are 5.1 and 2.2 μm in diameter. (b) Sequences of hairpin unfolding and folding cycles at a fixed loading rate of r ¼ 10 pN/s, according to the scale shown. On the left are three cycles of extension (solid line) and release (dotted line) for handles with no hairpin (lime, emerald, and forest green). Offset just to the left are three cycles incorporating the TAR RNA hairpin (navy, blue, and cyan), which show hairpin unfolding as a sudden change in length (Δxop) at a force identified as the opening force (Fop). The work done during unfolding (Wop) is the shaded region. Finally, both the closing force (Fcl) and the work done during closing (Wcl) are also found (the work during closing is not shown here). On the right are two cycles of TAR unfolding and folding in 50 nM NC p7 (red and maroon). Polymer models of construct elasticity (thick green, cyan, and pink lines) fit the data as described in the text and facilitate the identification of unfolding and folding

3.2 Experimental Hairpin Unfolding Energies and Lengths with NC

1. The experimentally measured opening lengths Δxop for TAR and TAR in the presence of 50 nM NC p7 (from Subheading 2.4, item 7 and measured in nm) were each determined at a distinct opening force (Fop). To correct for the elasticity of single-stranded RNA, a polymer model was used to deduce the force-independent length (in base pairs, see Note 31). These values did not vary significantly per changes in the loading rate (see Note 32), and the length did not change in the presence of NC (see Fig. 3a, b for corrected distributions and Table 2 for averaged values).

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

a

c 0.10

P ((kBT)-1)

0.06

d

0.05

50

z (kBT)

0.04

25

0.02

0

0.00 0

20

40 60 N (bases)

0.00

80

e

b 0.10

TAR +NC 30 pN/s 10 pN/s 0.8 pN/s

0

20

40 W (kBT)

60

0.06 TAR +NC

f

25 50 x (kBT) r 30 pN/s 10 pN/s 0.8 pN/s

50

z (kBT)

0.04

0.05

0

80 Pop (W) Pcl (W)

0.08

P ((kBT)-1)

Pop (base-1)

r 30 pN/s 10 pN/s 0.8 pN/s

TAR

TAR 30 pN/s 10 pN/s 0.8 pN/s

Pop (base-1)

Pop (W) Pcl (W)

0.08

67

25

0.02

0

0.00 0

20

40 60 N (bases)

80

0.00

0 0

20

40 W (kBT)

60

80

25 50 x (kBT)

Fig. 3 Determining the length and energy of hairpin unfolding. (a) Distributions of measured unfolding lengths (Δxop) for TAR at loading rates of 30, 10, and 0.8 pN/s (navy diamonds, blue circles, and cyan squares). (b) Unfolding lengths for TAR in 50 nM NC p7 over the same rates (maroon diamonds, red circles, and pink squares). Corrected for polymer elasticity, measured lengths are the same across all rates. (c) Distributions of the measured work during unfolding and unfolding (Wop—filled symbols and Wcl—open symbols) for the TAR hairpin across varying loading rates of 30, 10, and 0.8 pN/s (navy diamonds, blue circles, and cyan squares). The crossing point where P(Wop) ¼ P(Wcl) indicates the equilibrium unfolding free energy ΔGo and is marked by an arrow for each rate. (d) The free energy estimate for each rate, using the unfolding work distribution (z). Data crosses the line (orange) at the free energy of unfolding as described in the text. (e) Distribution of the work of hairpin unfolding and folding in the presence of NC, over the same rates (maroon diamonds, red circles, and pink squares). Estimates of the equilibrium free energy of unfolding are indicated by the arrows and the estimate (f) as above. The measured values of the unfolding length and energy are summarized in Table 2

68

Micah J. McCauley et al.

2. Distributions of the measured work of unfolding and refolding (Wop and Wcl, see Fig. 3c, d) are nonequilibrium energy measurements (as generally Pop 6¼ Pcl everywhere in the distribution). Equilibrium free energies of unfolding (ΔGo) were recovered through several methods (see Notes 33 and 34). As with the unfolded lengths, these were found to be steady with changes in loading rate. Addition of NC lowered the energy of hairpin unfolding by ~16 kBT (see Table 2 and Fig. 3e, f) [22, 23]. 3.3 Measuring the Transition State of TAR with NC

1. Probability distributions of the measured unfolding force (Fop, see Note 35) assembled for all loading rates (Fig. 4a, b) showed that increasing force was required to unfold with increasing rates (see Note 36). 2. A model of force-affected transition state theory fit the data across all loading rates simultaneously (see Note 37). These fits returned best fit values of the barrier height (G{op ) and distance from the folded state to the transition state (x{op ) for TAR and TAR saturated with NC (see Note 38). Both values changed with the addition of NC (see Table 2 for values and see Note 39). 3. These fits also showed the dependence of the rate of hairpin opening (kop) on the loading rate (see Fig. 4c and see Note 40). The natural rate of hairpin unfolding (koop , in the absence of external force) was also found to increase by 104 with the addition of NC (see Table 2) [22, 23].

3.4 Modeling the Hairpin Unfolding Energy

1. The individual elements of the Mfold folding energy G(ni, F ¼ 0) were progressively summed at each element, beginning at the terminal stem, Gi(ni, F ¼ 0). A matched RNA hairpin (see Fig. 5a) sees an increase in the overall energy, culminating in the total folded hairpin energy Σi G(ni, 0) ¼ ΔGf and the total hairpin length (Δxf, see Fig. 5b). The landscape describes the energy required to unfold the hairpin from the lower stem up to each individual element (ni, see Note 41). 2. Increasing force “tilts” this landscape by the stretching work: Gi(xi, F) ¼ Gi(ni, 0)
FΔxini increasingly favoring the unfolded state. The probability density across this landscape of observing unfolding up to each element is found ( p ~ e
Gi(ni, F) ) and normalized. At F½ the probability of the folded state and the unfolded state are equal as measured by the integrated probability density (Pop ¼ Pcl) (see Note 42), and this gives the “tilted” landscape (Gi(xi, F½), see Fig. 5c and Note 43). 3. While the unfolded state is centered with its corresponding probability density, the folded state is shifted due to “fraying” of the lowest bases at low forces. This effect varies with

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

a

c

0.5

TAR 30 pN/s 10 pN/s 0.8 pN/s

Pop (pN-1)

0.4 0.3

69

100

10

0.2

0.0 0

b

0.5

20

TAR +NC 30 pN/s 10 pN/s 0.8 pN/s

0.4

Pop (pN-1)

5 10 15 Opening Force (pN)

0.3

kop (s-1)

0.1

1

0.1 TAR TAR +NC 30 pN/s 30 pN/s 10 pN/s 10 pN/s 0.8 pN/s 0.8 pN/s

0.2 0.1 0.01

0.0 0

5 10 15 Opening Force (pN)

20

4

6

8 10 12 14 Opening Force (pN)

16

18

Fig. 4 Measurements of the transition state from unfolding experiments. (a) Distributions of the unfolding force for the TAR hairpin vary across the loading rates of 30, 10, and 0.8 pN/s (navy diamonds, blue circles, and cyan squares). Lines are best fits across all rates simultaneously to the dynamic force spectroscopy model as discussed in the text. (b) Unfolding force probability distributions for TAR with NC for the loading rates of 30, 10, and 0.8 pN/s (maroon diamonds, red circles, and pink squares). Lines are global best fits. (c) The force-dependent opening rate (kop(F)) for all pulling rates for TAR and TAR in 50 nM NC calculated from the remaining unfolded fraction at each given force. Solid lines are the fits shown in (a) and (b), while dotted lines are direct fits to the rates. These fits return the same values within uncertainty. Extrapolation to zero force gives the natural rate of hairpin opening in the absence of any force as k oop . Fitted values of transition state parameters and the natural opening rate are summarized in Table 2

sequence and leads to predicted values of the hairpin length (Δxop) and the energy of unfolding (ΔGo) that are slightly less than the values directly from Mfold, which gives energy (ΔGf) and length (Δxf), shown on Gi(xi, F). Values are shown in Table 1 (see Fig. 5b and see Note 44). 4. The transition state barrier is identified from this landscape, and the value of the barrier height (G{op ) and distance from the folded state (x{op ) are found from the graph of the tilted landscape (Gi(xi, F½), see Note 45). Values are shown in Table 1 (see Fig. 5c).

40

DGo

20

D xop

0 0 20

20 40 60 Extension (bases) 0.1

Energy (kBT)

15 10 † DGop

5

x

† op

0

0.0 0

10 20 30 Hairpin Extension (nm)

j

Energy (kBT)

60 40

D xop

20

D Go

0 0

20

20 40 60 Extension (bases)

Energy (kBT)

10

DG † x op

5 0

0

† op

10 20 30 Hairpin Extension (nm)

0.0

DGo

20

D xop

0 0

f 20

20 40 60 Extension (bases) F1/2 = 10.9 pN

15

0.1

10 5 0

x 0

† DGop

† op

10 20 30 Hairpin Extension (nm)

0.0

k 80 60 40

D Go

20

D xop

0 0 20

20 40 60 Extension (bases) F1/2 = 10.8 pN

15 10

† D Gop

5 0

0.1

† xop

0

10 20 30 Hairpin Extension (nm)

Probability Density

0.1

15

40

U U U U 30 U U U –A G– C A –U A –U G – C40 U –A U –A C–G A –U 20 A–U G –C U –A U –A G– C l A – U50 A –U G –C U –A U –A 10 C–G A –U A –U G –C U –A U – A60 G– C A –U A –U 1 G – C64 Upper Stem

Probability Density

F1/2 = 7.8 pN

60

Probability Density

F1/2 = 13.4 pN

e 80 Energy (kBT)

Energy (kBT)

60

h 80

U U U U 30 U U U –A G– C A –U A –U G – C40 U –A U –A C–G A–U 20 A–U G –C U –A U –A G– C i A – U50 A –U G –C U –A U –A 10 C–G A –U A –U G –C U –A U – A60 G– C A –U A –U 1 G – C64 Lower Stem

U U 30 U U U –A G– C A –U A –U G – C40 U –A U –A C–G A –U 20 A–U G –C U –A U –A G– C A – U50 A –U G –C U –A U –A 10 C–G A –U A –U G –C U –A U – A60 G– C A –U A –U 1 G – C64 All Sites

Probability Density

g

d UU

Energy (kBT)

b 80

U U U U 30 U U U –A G– C A –U A –U G – C40 U –A U –A C–G A –U 20 A–U G –C U –A U –A G– C c A – U50 A –U G –C U –A U –A 10 C–G A –U A –U G –C U –A U – A60 G– C A –U A –U 1 G – C64 RNA Hairpin

Energy (kBT)

a

Micah J. McCauley et al.

Energy (kBT)

70

0.0

Fig. 5 Modeling the free energy landscape of unfolding under destabilization. (a) A fully matched 64-base RNA hairpin forms 29-base pairs (+6 unpaired bases in the apical loop) in an optimally folded structure a predicted by Mfold. (b) The progressive sum of base pair energies, starting from the lowest pair on the stem, gives the

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

3.5 Characterizing the Effects of a Destabilizing Ligand

71

1. Destabilization due to ligand binding is simulated by simply subtracting energy from individual elements (ni) (see Note 46). In this example, the G-bases are selected and destabilized (see Note 47). For simplicity, they are first destabilized by an equal amount of 2 kBT, for a total of 20 kBT over the starting hairpin (see Fig. 5d for the sites and see Note 48). 2. The landscape is tilted to find F½ as above, and while both the energy of unfolding and the barrier height are affected, the unfolding length and the distance to the transition state are not changed within uncertainty (see Fig. 5e, f and Table 1). 3. If the 20 kBT is localized on the 5 G-bases on the lower stem (as specified in Fig. 5g), the unfolding pathway is changed significantly. Now the lowest stem frays up to nearly ½ of the length (see Note 49) affecting all estimated parameters of the landscape (see Fig. 5h, i for visual comparisons, Table 1 for values and see Note 50). 4. Conversely, if the 20 kBT is localized to the 5 G-bases on the upper stem (see Fig. 5j), then the all unfolding parameters are affected, except for the overall hairpin unfolding length (see Fig. 5k, l for visual comparisons and Table 1 for values). 5. From this example, measured changes in Δxop and x{op due to ligand binding should indicate whether binding is predominately located at the stem, the upper loop, or is more uniformly distributed along the length of the hairpin (see Note 51). This information will give insight into measured changes in the TAR unfolding landscape in the presence of TAR described below (see Notes 52 and 53).

ä Fig. 5 (continued) full unfolding energy (ΔGo) and hairpin length (Δxop). (c) Increasing force applied to the lower stem “tilts” the landscape to favor unfolding and reveals a transition barrier to unfolding (of height G {op and distance from the folded state x {op ). Dotted lines indicate probability density at each extension under an external force (F). At F½ the integrated density gives a 50:50 probability of finding either the folded or unfolded state. (d) Ligand binding to all G bases along hairpin from (a) are circled (green) and assumed to be destabilized equally by 2 kBT (for a 20 kBT total). (e, f) Relative to the free hairpin, energies and landscapes at F½ predict changes in measured unfolding and transition energies, but not the unfolding length or the distance to the transition state. (g) Ligand binding to only the lowest G bases in the stem (yellow). (h) Though the total energy of destabilization is the same as above (d), the lower stem unfolds at a much lower force (i), and all observed landscape variables for the remaining hairpin are affected. (j) Binding to G bases in the upper stem only (red), with the same total destabilization as above (d). (k, l) Destabilization leads to detectable changes in both the unfolded and the transition state, though the overall length of unfolding remains the same. Table 1 summarizes values predicted by these models, highlighting those that vary outside uncertainty. Preferential binding to either the upper or lower stem is distinguishable experimentally

72

Micah J. McCauley et al.

Table 1 Ligand destabilization affects the energy landscape of hairpin unfolding Ligand binding

Dxop (bases)

DGo (kBT)

F½ (pN)

† xop (nm)

† DGop (kBT)

None

60 ± 1

71 ± 2

13.4 ± 0.2

23.4 ± 0.5

9.9 ± 1.0

All sites

60 ± 1

53 ± 2

10.9 ± 0.2

22.4 ± 0.5

9.3 ± 1.0

Lower stem

26 ± 2

35 ± 1

7.8 ± 0.3

11.9 ± 1.0

7.9 ± 1.2

Upper stem

60 ± 1

51 ± 2

10.8 ± 0.2

13.6 ± 1.0

12.5 ± 1.6

Values of the energy landscape predicted for the hairpin destabilized by ligand binding shown in Fig. 5. The length of the hairpin (Δxop) and the energy of unfolding (ΔGo) found from the simple sum of the Mfold base pair lengths and energies Σi G(ni, F ¼ 0) for the hairpin shown in Fig. 5a to give the landscape of Fig. 5b. This hairpin is destabilized by a total of 20 kBT according to each geometry shown (Fig. 5d, g, and j) to give modified landscapes (Fig. 5e, h, and k). The transition barrier height (G{op ) and distance from the folded state (x{op ) are found from the tilted landscape (Σi G(ni, F½), see Fig. 5c, f, i and l), where Pop ¼ Pcl. While parameters describing the unfolded state (Δxop and ΔGo) are independent of applied force, the values characterizing the transition state (G{op and x{op ) are found at F½ and are expressed at that value. Values of this simple landscape that differ significantly from the free hairpin are italicized

Table 2 Comparisons of energy landscapes reveal that NC binding to the TAR hairpin sharply changes the transition state Experiment

D xop (bases)

DGo (kBT)

F½ (pN)

† xop (nm)

† DGop (kBT)

o kop (s-1)

TAR (OT)

47.8 ± 1.1

43.3 ± 0.9

10.4 ± 0.2

9.9 ± 1.1

27.0 ± 2.2

(0.8 ± 0.5)×10-8

TAR (Mfold)

48 ± 1

41.5 ± 1.5

10.1 ± 0.1

11.3 ± 0.9

30.7 ± 2.0

-

TAR +NC (OT)

48.4 ± 0.5

28.3 ± 0.9

7.7 ± 0.2

4.8 ± 0.6

14.3 ± 1.3

(1.2 ± 0.8)×10-4

TAR +NC (Mfold)

48 ± 1

25.6 ± 1.5

7.9 ± 0.1

4.7 ± 0.9

14.8 ± 1.2

-

Experimental OT values and theoretical Mfold values are compared for TAR alone (blue, cyan) and in the presence of NC (red, pink). Experimental landscape values for unfolding the irregular TAR hairpin match theoretical values well. While the measured length (Δxop) was unaffected in saturating concentrations of NC, the energy of unfolding (ΔGo) was decreased by ~16 kBT. Destabilization was spread out evenly over the four sites shown in Fig. 6 to give modified theoretical landscapes. These gave both the barrier height (G{op ) and distance (x{op ) that best match experimental data at F½ and reveal the key NC binding sites to TAR. The natural rate of hairpin unfolding (koop ) was experimentally found to increase 104 with the addition of NC

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

G G G G U G 30U G 30 C A C A C–G C–G G– C G– C A –U A –U G–C G–C U U C C U U A – U40 A – U40 G – C G –C 20 20 A–U A–U C –G C –G C –G C –G A A G–C G–C A –U A –U U –A U–A U –A U–A G – C 50 G – C 50 G U G U 10 10 U –A U–A C –G C –G U G U G C –G C –G U –A U–A C C U –A U–A G –C G –C G – C59 1G – C59 1 G –C G –C 5 sites 6 sites

c 70

60 50 40

D xop

10 0 d 20

0 20 40 60 Hairpin Extension (bases) F1/2 = 10.1 pN

15

0.1

10 5

† xop

0

e 20

8 4 0

4 8 x†op (nm)

12

10 20 30 Hairpin Extension (nm)

0.1

10 5 0

0.0

F1/2 = 7.9 pN

15

Energy (kBT)

3 sites 4 sites 5 sites 6 sites OT data

12

† D Gop

x

0

† op

† DGop

10 20 30 Hairpin Extension (nm)

0.0

Probability Density

DG†op (kBT)

D Go

20

0

16

0

4 sites

30

b 24

20

TAR RNA TAR +NC

Energy (kBT)

G G U G 30 C A C–G G– C A –U G–C U C U A – U40 G –C 20 A–U C –G C –G A G–C A –U U –A U –A G – C 50 G U 10 U –A C –G U G C –G U –A C U –A G –C G – C59 1 G –C 4 sites

Probability Density

G G U G 30 C A C–G G– C A –U G–C U C U A – U40 G –C 20 A–U C –G C –G A G–C A –U U –A U –A G – C 50 G U 10 U –A C –G U G C –G U –A C U –A G –C G – C59 1 G –C 3 sites

Energy (kBT)

a

73

Fig. 6 Using the unfolding landscape to map protein binding to the hairpin. (a) NC binding induces a reduction in the free energy of opening but no change in the opening length. Four models of upper stem destabilization are compared by specifically destabilizing exposed G bases in the irregular TAR structure. Solid ovals mark destabilized sites in each model, while dotted ovals are exposed G bases that are not destabilized. (b) Predicted values of the barrier height G {op and the distance to the transition state x {op for the four cases from (a), with each site destabilized by an equal amount of the measured 16 kBT total. The four-site model reproduces the force unfolding data most closely, indicating these bases are the essential sites for TAR

74

Micah J. McCauley et al.

3.6 Modeling Changes in the Transition State Due to NC Binding

1. Table 2 summarizes OT data describing the unfolded and transition states for hairpin unfolding. Within the uncertainty of the experiments, all the measured landscape parameters are altered by NC binding, except for the overall unfolding length (see Note 54). From the discussion of Subheading 3.5, this indicates that the destabilization is focused on the upper stem (see Note 55). 2. From OT experiments, the total change in hairpin energy due to hairpin binding is ~16 kBT. If NC specifically binds to G bases exposed by irregularities in the TAR sequence (see Note 56), this energy should be used to specifically destabilize these sites. However, not all the available sites may contribute to destabilization. 3. For simplicity, we consider a simple model where destabilization is either fully apparent or not evident at all. Exposed G bases may be destabilized along the stem starting from the apical loop (see Note 57) and progressively including more of the sites in the four most likely cases (see Fig. 6a). 4. For each case, the measured energy difference (16 kBT) is equally distributed across the active sites to obtain GiNC(xi, 0) (see Note 58). Then this landscape is “tilted” with applied force until Pop ¼ Pcl, giving GiNC(xi, F½). As above, all landscape parameters are determined (Δxop, ΔGo, x{op and G{op ), though only the transition state parameters change significantly between the four cases (see Fig. 6c–e for landscape examples for the case of four active binding sites compared to TAR alone and see Note 59) [23]. 5. The values for x{op and ΔG{op that correspond to the closest match to the experimental data (see Fig. 6b) are those for the four site model, indicating that it is destabilization of these four states by NC that both reduces and shifts the transition state of TAR (see Note 60).

3.7 TAR Hairpin’s Irregular Structure Facilitates NC Activity

1. A summary of the energy landscape data in Table 2 may be used to create a simplified energy landscape (see Fig. 7 and see Note 61). This landscape shows that NC not only destabilizes the overall hairpin but drives and unusual shift in the transition state, which drives the 104 increase in the natural rate of

ä Fig. 6 (continued) destabilization. (c) The summed energies Gi(ni, F ¼ 0) for the TAR hairpin (blue). The presence of NC is simulated by discrete destabilization at four sites (arrows) in this example to give GiNC(ni, F ¼ 0) (red). (d) The “tilted” landscape for TAR at F½; Gi(ni, F½) (blue), where the probability of finding the unfolded state is ½ (dotted line). (e) The landscape for TAR with NC, GiNC(ni, F½) (red), with calculated probability density (dotted line). The unfolding force is reduced by the presence of NC. The transition state may be modeled in both cases, and these values are compared to experiment in Table 2

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

a

75

b 5 kBT

F = 8 pN

DG (kBT)

D G (kBT)

5 nm

5 kBT

F = 8 pN 5 nm

Hairpin Extension (nm)

Hairpin Extension (nm)

Fig. 7 Free energy landscapes reveal the function of NC during TAR destabilization. (a) The smoothed landscape consists of the unfolded state and the transition barrier, relative to the unfolded state. Experimental values (blue) compare well with Mfold deduced theoretical values (cyan), and both theory and experiment compare well within uncertainty. (b) The TAR unfolding landscape in the presence of NC, for experimental values (red) as compared to theoretical values (pink). This match was achieved for the four-site model shown in Fig. 6, where destabilization due to binding is concentrated in the upper stem. This leads to a shift in the transition state (shown as ovals in the inset figures) which drives the 104 change in the unfolding rate while preserving the measured unfolding length. Theoretical and experimental values are compared explicitly in Table 2

unfolding (see Note 62). Furthermore, it is the irregular structure of TAR that allows the upper stem to be specifically destabilized. 2. The irregular structure or TAR is not a feature of the final annealed TAR RNA + cTAR DNA product, and so formation of this annealed product is favored in the presence of NCp7 (see Note 63). 3. This example illustrates the technique of landscape construction and variation. Comparisons to experimental data reveal specific binding responsible for the nucleic acid chaperone activity of TAR RNA. However, this technique of landscape variation may apply to a variety of ligand + hairpin activities.

4

Notes 1. Standalone versions are available for download for each program version. 2. Not all constraints and solution conditions are available for each version of the program. Furthermore, different versions may be comprised of different sets of base pairing and stacking energies [9, 10].

76

Micah J. McCauley et al.

3. Salt and temperature dependencies of base pairing and stacking energies have been studied extensively both theoretically and experimentally [9, 10, 34–37], much less is known about solution variations for the stability of bubbles, mismatches, and noncanonical bases [22]. 4. A previous version of Mfold (2.3) does allow variation in these parameters, though it is comprised of a somewhat older data set of base pairing and stacking energies [10]. 5. The effect of decreasing temperature and the effect of decreasing monovalent salt concentration on hairpin folding energies offset to some extent. Decreasing temperature stabilizes the overall hairpin while decreasing salt destabilizes the hairpin. The experimental conditions that follow were 22  C and 100 mM Na+, which was previously shown to be effectively similar to the conditions imposed by Mfold of 37  C and 100 mM Na+ [23]. 6. Multiple structures are likely evident in the data, with distinct unfolding energies and unfolding lengths. Folding and unfolding may undergo multiple pathways, and comparisons to experiment should explicitly consider this [38, 39]. 7. It is possible to use folding constrains to favor a structure that might be known to form (perhaps utilizing evidence from other types of experiments). However, constraints will potentially introduce an unknown energy into the landscape, and so folding constraints were not used here. 8. Elements are energies of base pairing and stacking with nearest neighbor. Data also exists for some noncanonical pairing, including mismatches and nonstandard DNA which may contribute less to the stability. Other bases, including those formed by oxidative damage, were not included [24]. Bases were numbered along the backbone starting from the presumed 50 end (avoiding double counting for the 30 side). 9. The external force on the hairpin was explicitly equal to zero. 10. All enzymes were incubated at 37  C. 11. The single digoxygenin was positioned on the opposing strand of the biotin, so that the labels appear on opposite ends of the construct. 12. Constructs that featured DNA hairpins did not require in vitro transcription, nor the RNA polymerases described below. Though not the focus of this work, construction of labeled DNA hairpin constructs is otherwise similar to the outline that follows [24]. 13. The oligomer pairs were annealed to each other, then to the overhangs on the TAR template and followed by the appropriate ligase.

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

77

14. Constructs omitting the hairpin were also constructed and ligated together, to confirm the functionality of the labeled ends [23]. 15. Once thawed, aliquots appeared active for nearly 1 week. 16. In addition to the dual trap/micropipette design shown here [26–28], other force extension techniques have been used to study unfolding including a two trap OT design (eliminating the micropipette tip) [29], magnetic tweezers [30], atomic force microscopy [31], and centrifuge-based force extension [40]. 17. This buffer includes 100 mM Na+, which was chosen to facilitate catching and retention of individual constructs between the two beads, which becomes progressively more difficult in lower salt. However, the binding affinity of NC for the hairpin is reduced in higher salt (see below). 18. Varying sizes of polystyrene beads may be used (dependent upon the laser trap diameter). But it is preferable that the differently labeled beads be of easily distinguishable sizes. Furthermore, within each stock solution, the beads should not vary strongly (>10%) in diameter, to prevent error in either instrument calibration or stiffness as described below. 19. In addition to commercially labeled beads, protein G-labeled 2-μm diameter beads were also custom coated with antidigoxygenin to facilitate binding to the digoxygenin label. 20. Each experiment returned a measured force for each extension changing step. Together, this data formed an array of force and extension (along with force and release). The force measurement was calibrated in several ways [26, 27], and here we overstretched a long (phage-λ, 48,500 base pairs), which provides a clear force plateau for calibration of 60–65 pN (including known variations due to experimental conditions) [41]. The extension offset is deduced from the known extension per base pair of double-stranded DNA with the models below. The finite stiffness of the trap requires a correction for movement of the trapped bead with increasing force. For the 5-μm beads used here, this stiffness was measured to be ~100 pN/μm. 21. If two or more constructs are inadvertently caught between the beads, the measured force response will be integer multiples of the single construct force response. This data cannot be calibrated and is easily distinguished from the force data from a single construct. The construct/beads should be immediately discarded, and the catching protocol restarted (possibly with lower solution concentrations of constructs).

78

Micah J. McCauley et al.

22. A single digoxygenin—anti-digoxygenin attachment becomes increasingly likely to separate above 30 pN. If a sequence of extension/release cycles is desired, a force limit of 30 pN is necessary. 23. This loading rate was specifically measured over a force range just below unfolding for each measurement. Though the loading rate may be affected by hairpin and handle stiffness, in this instrument, the stiffness of the laser trap dominated [22, 23, 42]. 24. Similar loading rates are found in other optical tweezer studies. While faster loading rates are typically limited by detector noise, slower loading rates are constrained by drift due to position fluctuations in either the trapping laser or the micropipette tip. 25. The elasticity of the double-stranded DNA (dsDNA) comprising the long handles was characterized by the high-force solution to the worm-like chain (WLC) model of elasticity [26, 43]: " #
1=2 1 kB T F b ds ðF Þ ¼ B ds 1
þ : ð1Þ S ds 2 P ds F The force-dependent end-to-end length (bds(F)) was determined for a polymer of contour (backbone) length (Bds), elastic modulus (Sds) and persistence length (Pds). These parameters have been fit to dsDNA under a variety of conditions and contour lengths and for the hairpin free handles, Bds ¼ 0.340  0.001 nm/bp, Pds ¼ 30  4 nm and Sds ¼ 1200  200 pN. There are known deviations in Pds due to finite construct length (Pds is typically 50 nm for long DNA, such as phage-λ) [44]. This extension (xfolded(F) ¼ Nds  bds(F), with Nds ¼ 6600 bp) is the green line in Fig. 2b. 26. The unfolded hairpin, composed of single-stranded RNA, was described by another polymer model, the freely jointed chain (FJC) [23, 26, 43]: " #  
1=2 2P ss F 1 kB T F b ss ðF Þ ¼ B ss coth  1þ
: ð2Þ S ss kB T 2 P ss F These parameters are less well characterized, though typical values are used here for the contour length (Bss ¼ 0.570  0.001 nm/bp), persistence length (Pss ¼ 0.8  0.2 nm), and an elastic modulus (Sss ¼ 800  200 pN). Some variability will affect the measured parameters, so here we remained consistent between experiments with TAR and those with TAR and NC and focused on differences between the two specifically during hairpin

Specific Nucleic Acid Chaperone Activity of HIV-1 Nucleocapsid Protein. . .

79

unfolding [7, 23]. This was justified by the choice of NC solution concentration (the handles are unaffected at c ¼ 50 nM, see below). The total extension (xunfolded(F) ¼ Nds  bds(F) + Nss  bss(F), with Nds ¼ 6600 base pairs and Nss ¼ 64 bases), is the blue and red line in Fig. 2b. 27. This work characterized the opening length, as complete unfolding clearly occurs in a single event (unfolding is two state). Hairpin folding, in contrast, may occur over several paths [28], so only the final, folded state was determined here, as it should yield most closely work of folding (and errors in the folding work from this assumption should be fairly small). The folding force was not analyzed for information. 28. This work done by the instrument was not generally equal to the free energy of hairpin unfolding, even after correcting for the stretching elasticity of the hairpin (as done here). Hairpin unfolding was generally nonequilibrium in these forceextension experiments, especially for distributions where Pop and Pcl are not everywhere equal (equivalently, the averages of Fop and Fcl are not equal). Only the slowest loading rate for the TAR hairpin (r ¼ 0.8 pN/s) appeared to be near equilibrium. 29. The work was numerically equivalent to the area between the two models for xfolded(F) and xunfolded(F), from zero force to Fop, with a small correction for work done during unfolding (due to trap stiffness) [6, 7]. 30. Binding of NCp7 to TAR appeared to be saturated at the solution concentration chosen here of 50 nM. Experiments were also conducted at 200 nM, with no detectable difference in result. It should be noted that this number is smaller than the values seen for binding to polyA and to the tRNALys3 minihelix (Kd ~200 nM for both, corrected for solution conditions) [45, 46]. This implies that nonspecific NC binding contributes very little to the destabilization of the TAR hairpin, while binding to specific sites, with an affinity Kd 100 Da for SPR and an advantageous of ITC is that ITC has no molecular weight limitation. On the other hand, ITC requires large heat change on binding, and it cannot be estimated without measuring ITC. Although ITC exhibits some good features compared with SPR, SPR is the most widely used method by which to analyze RNA–protein interactions for the following reasons: SPR assays require less protein and RNA than ITC (50 μg of each sample for ITC and 10 ng of each sample for SPR may be needed in general, although the amount of the sample depends on the affinity of the interaction and the molecular weight of the sample.), SPR has higher throughput than ITC, and the range of Kd values that can be measured in SPR is wider than that in ITC (from millimolar to subnanomolar).

142

2 2.1

Ryo Amano and Taiichi Sakamoto

Materials SPR Experiment

1. Instrument Biacore X. 2. BIAevaluation 3.0. 3. Heating block. 4. Purified protein (see Note 1). In the case of the high affinity RNA-protein interaction [7], approximately 300 μL of 20 nM protein was needed to determine the Kd value. 5. Purified RNA with 30 -poly(A16) tail. Approximately 250 μL of 10 nM RNA was needed to determine the Kd value [7] (see Notes 1 and 2). 6. 50 -biotinylated poly(dT16) oligonucleotide. Approximately 50 μL of 1 μM 50 -biotinylated poly(dT16) was needed to immobilize it on sensor chip SA [7] (see Note 2). 7. Sensor chip SA (GE Healthcare). 8. Nuclease-free water. 9. 1 SPR buffer: 20 mM sodium phosphate (pH 6.5), 2 mM magnesium acetate, 300 mM potassium acetate, 0.1% Tween 20, and 1 mM dithiothreitol (DTT) (see Note 3). 10. 1 M NaCl in 50 mM NaOH that was passed through 0.22 μm pore filter. 11. Regeneration buffer: 4 M urea solution that was passed through 0.22-μm-pore filter.

2.2

ITC Experiment

1. iTC200 isothermal titration calorimeter (Malvern Panalytical, UK). 2. Microcal Origin7. 3. Heating block. 4. Dialysis membrane with an appropriate molecular weight cutoff for the protein. 5. Centrifugal ultrafiltration unit Vivaspin 2 (cutoff molecular weight 3000–5000). 6. Purified protein (see Note 1). Approximately 280 μL of 10 μM protein was needed in single measurement [7]. 7. Purified RNA (see Note 1). Approximately 60 μL of 100 μM RNA was needed in single measurement [7]. 8. Nuclease-free water (e.g., MilliQ water). 9. 1 ITC buffer: 20 mM sodium phosphate (pH 6.5), 2 mM magnesium acetate, 300 mM potassium acetate, 10% glycerol, and 1 mM DTT (see Note 3).

Kinetic and Thermodynamic Analyses

3 3.1

143

Methods SPR Experiment

All experiments should be performed under Ribonuclease-free conditions. Experimental parameters are shown in Table 1.

3.1.1 Sample Preparation

1. Prepare a twofold dilution series in 1 SPR buffer of the protein (e.g., 0.625, 1.25, 2.5, 5, 10, and 20 nM). 2. Prepare 1 μM RNA with 30 -poly(A16) tail in nuclease-free water. 3. Heat the RNA at 95  C for 5 min and snap cool on ice. 4. Dilute 1 μM RNA with 30 -poly(A16) in 1 SPR buffer to prepare 10 nM 30 -poly(A16) tail RNA in 1 SPR buffer. 5. Incubate at room temperature for 10 min.

3.1.2 Pre-immobilization of 50 -Biotinylated Poly (dT16) on the Sensor Chip SA

1. Equilibrate the new sensor chip SA at room temperature for 30 min. 2. Degas and filter the nuclease free water and SPR buffer through a membrane with a pore size of 0.22 μm just before use. Table 1 Parameters for surface plasmon resonance measurement (see Note 4) Experimental parameters Temperature ( C)

25

Flow path

Flow cell 1, 2

Reference

Flow cell 2

Flow rate (μL/min)

90

Parameters of RNA injection (see Note 5) Injection time (s)

20

Injection volume (μL)

30

Parameters of protein injection Injection time (s)

60

Injection volume (μL)

90

Dissociation time (s)

180

Parameters of regeneration Injection time (s)

60

Injection volume (μL)

60

144

Ryo Amano and Taiichi Sakamoto

3. Turn on Biacore X and set the operating temperature to 25  C. 4. Dock the sensor chip in Biacore X. 5. Execute “PRIME” twice with nuclease free water. PRIME is the procedure for washing the Biacore system including pumps, tubes, and the sensor chip with nuclease free water or buffer. This procedure is selected from the “working tools” menu and used at start up and when the buffer is changed and also to remove small air bubbles. 6. Start new sensorgram. 7. Select “Multichannel” in the “Detection Mode” menu. 8. Select “Fc 1–2” in the “Flow Path” menu, and set flow rate to 10 μL/min. 9. Inject 1 M NaCl in 50 mM NaOH three times for 1 min, and then execute “PRIME” once with nuclease free water. 10. After stopping the sensorgram, execute “PRIME” twice with SPR buffer. 11. Start new sensorgram. 12. Select “Single” in the “Detection Mode” menu, and select “Fc 2” in the “Flow Path” menu, and then set flow rate to 5 μL/ min. 13. Immobilize 50 -biotinylated poly(dT16) on sensor chip SA in the flow cell 2 at a level of approximately 800 response unit (RU) using the manual inject command (see Notes 6 and 7). 14. After stopping the sensorgram, execute “PRIME” with SPR buffer to remove all unbound (or weakly bound) 50 -biotinylated poly(dT16). 3.1.3 SPR Measurement

1. Start new sensorgram. 2. Select “Multichannel” in the “Detection Mode” menu, and select “Fc 1–2” in the “Flow Path” menu, and then set the reference cell to flow cell 1. 3. Set flow rate to 90 μL/min (see Note 8). 4. Inject 10 nM RNA with 30 -poly(A16) tail dissolved in SPR buffer selecting “None” in the “wash after injection” menu, and immobilize it approximately 100 RU on flow cell 2 which 50 -biotinylated poly(dT16) was immobilized (see Notes 6–8). 5. Wait for at least 120 s to stabilize the baseline (see Note 9). 6. Inject the protein dissolved in SPR buffer for 60 s with the dissociation time set to 180 s (see Note 8). 7. After observing the dissociation step, inject the regeneration buffer over flow cells 1 and 2 three times for 60 s with “Normal” selected in the “wash after injection” menu to remove the bound materials.

Kinetic and Thermodynamic Analyses

145

8. Repeat steps 4–7 for a dilution series of the proteins prepared above (see Subheading 3.1.1) and the buffer as a control. 3.1.4 Data Analysis

The following describes the steps in SPR data analysis using the BIAevaluation 3.0. 1. With the software turned on, choose “File: Open” or press “Open” button on toolbar and open the result file (extension.blr). 2. Select the curves in the curve list. 3. Select the curves in the project and click “Overlay Plot” to display the curves in an overlay plot. 4. Remove the unwanted section of the curves, such as RNA immobilization and regeneration parts. 5. Select a flat section of the curves to be treated as the baseline. 6. Press “Y-Transform.” 7. Choose “Zero at Average of Selection” and then press “Replace Original.” 8. Press “X-Transform” and choose “Curve Alignment.” 9. Move the arrows on each curve to the injection start point and press “Finish.” 10. Choose “Fit:Kinetics Simultaneous ka/kd” and click “Next>.” 11. Move the thin vertical black lines to adjust the injection start and stop points. 12. Move the thicker line to select association and dissociation data ranges to be used in the fitting and click “Next>.” 13. Select the binding model to be used for the fitting, and enter the analyte concentration values in the cells of parameters. 14. Click “Fit.” 15. After the fitting process is terminated, click “Accept Fit.”

3.2

ITC Experiment

3.2.1 Sample Preparation

1. Rinse the dialysis membrane with nuclease-free water to remove sodium azide preservative. Secure a clamp to one end of the membrane. 2. Fill the membrane with the purified protein and clamp the open end. Again, check the sealability of the membrane and clamps. 3. Immerse the membrane in a beaker containing 1 L ITC buffer. Dialyze for 12 h at 4  C with gentle stirring. 4. After purifying the RNA, desalt it using new the Vivaspin 2 ultrafiltration unit. 5. Heat the RNA at 95  C for 5 min and snap cool on ice.

146

Ryo Amano and Taiichi Sakamoto

6. Exchange the buffer with the ITC buffer used for protein dialysis using new Vivaspin 2 (see Note 10). 3.2.2 ITC Measurement

The shape of the titration curve obtained by ITC depends on ΔH upon binding and Kd; therefore, to analyze the thermodynamic parameters of the interaction, large ΔH is required. With a ΔSdriven interaction and small ΔH, it would be not easy to analyze the interaction using ITC. Furthermore, to accurately measure Kd, the C value should be within the range of from 10 to 3000; a value within the range of from 50 to 500 is ideal. The unitless parameter C is obtained using Eq. 4: C ¼ Ka  M  N,

ð4Þ

where Ka is the association constant, M is the concentration of the sample in the cell, and N is the stoichiometry. In cases of very tight binding, the shape of the curve is nearly rectangular at the stoichiometric equivalence point N. To accurately and uniquely fit the curve to the titration data, the C value should be 3000 and preferably 500. In cases of high-affinity binding, the competitive ligand binding experiment is necessary to accurately measure Kd [17]. In cases of low-affinity binding, the shape of the curve is shallow but not sigmoidal; therefore, it is not easy to interpret the accurate ΔH using the fitting process when C is 16 mer when you use buffer with a low concentration of salt. Hybridization of poly(A) and poly(dT) becomes unstable under low concentrations of salt, which results in baseline drift. 3. Buffers for SPR and ITC should be optimized for your samples. For SPR, to prevent nonspecific binding of the protein to the RNA and the sensor chip, change salt or detergent concentration. As the first-choice buffer, the following buffer [10 mM HEPES (pH 7.4), 150 mM NaCl, 0.005% TWEEN20] is recommended. For ITC, buffer with a low ionization enthalpy is best such as phosphate or acetate, Good’s buffer such as Tris is not recommended. As the first-choice buffer, phosphate buffer containing 50–100 mM salt is recommended. If you need a reducing agent, we suggest TCEP (below 2 mM) rather than DTT. The use of DTT may cause aberrant baselines. Glycerol and detergents at low concentrations where they don’t foam by stirring are acceptable. If you use DMSO, you need the same DMSO concentration in the syringe and the sample cell because heat of dilution of DMSO is very large. 4. These parameters, except for injection time (s), are set at the Biacore X instrument. 5. These parameters depend on flow rate, the type of RNA sample, and the condition of the sensor chip. Thus, a pilot experiment is recommended in which the appropriate values to obtain the desired immobilization level are assessed. 6. The 50 -biotinylated poly(dT16) is immobilized on the SA chip on which streptavidin is pre-immobilized. Then, the sensor chip is ready for the immobilization of the RNA sample by the hybridization of poly(dT16) and poly(A16). 7. The response unit (RU) shows how much the 50 -biotinylated poly(dT16) is immobilized. Less immobilization of the poly (dT16) and RNA reduces protein binding response when the protein is injected; however, too much immobilization causes rebinding of the protein which results in misinterpretation of kinetic data. Therefore, 800 RU is recommended. 8. Depending on the binding affinity of your samples, it is necessary to optimize flow rate, the amount of RNA immobilized, and the binding and dissociation times. When conducting the first SPR experiment, we usually set the flow rate to 20–50 μL/ min and immobilize the RNA with a poly(A) tail ~500–1000

Kinetic and Thermodynamic Analyses

149

RU into flow cell 2. The binding and dissociation times are 60 and 180 s, respectively. 9. By waiting more than 120 s, weakly unbound and unbound RNAs are washed away. Thereby the baseline is stabilized. 10. At least four wash cycles will remove the initial buffer (or water). 11. Dirt in the sample cell or titration syringe causes noise in the raw data. Keep them clean, and titration of water with water is recommended to check the noise level before the experiments. 12. A buffer mismatch causes noise. To prepare the RNA samples and to wash the sample cell, use the buffer that is identical to that used for protein dialysis. 13. To prevent air bubbles during the measurement, the protein and RNA dissolved in ITC buffer, nuclease-free water, and ITC buffer should be returned to room temperature before use. 14. Depending on the binding affinity of your samples, the concentrations of the protein and RNA should be optimized. 15. Chi-square is a quantity commonly used to evaluate whether any given data are well fitted by some curve of hypothesized function.

Acknowledgments This study was supported by JSPS KAKENHI Grant Number JP15K06982, JP18K11536 from The Ministry of Education, Sports, Culture, Science and Technology (MEXT) of Japan. References 1. Mitchell SF, Parker R (2014) Principles and properties of eukaryotic mRNPs. Mol Cell 54:547–558 2. Morris KV, Mattick JS (2014) The rise of regulatory RNA. Nat Rev Genet 15:423–437 3. Renaud JP, Chung CW, Danielson UH, Egner U, Hennig M, Hubbard RE, Nar H (2016) Biophysics in drug discovery: impact, challenges and opportunities. Nat Rev Drug Discov 15:679–698 4. Cle´ry A, Sohier TJM, Welte T, Langer A, Allain FHT (2017) switchSENSE: a new technology to study protein-RNA interactions. Methods 118–119:137–145 5. Faner MA, Feig AL (2013) Identifying and characterizing Hfq-RNA interactions. Methods 63:144–159

6. Jing M, Bowser MT (2011) Methods for measuring aptamer-protein equilibria: a review. Anal Chim Acta 686:9–18 7. Amano R, Takada K, Tanaka Y, Nakamura Y, Kawai G, Kozu T, Sakamoto T (2016) Kinetic and thermodynamic analyses of interaction between a high-affinity RNA aptamer and its target protein. Biochemistry 55:6221–6229 8. Szabo A, Stolz L, Granzow R (1995) Surface plasmon resonance and its use in biomolecular interaction analysis (BIA). Curr Opin StructBiol 5:699–705 9. Katsamba PS, Park S, Laird-Offringa IA (2002) Kinetic studies of RNA–protein interactions using surface plasmon resonance. Methods 26:95–104

150

Ryo Amano and Taiichi Sakamoto

10. Sakamoto T, Ennifar E, Nakamura Y (2018) Thermodynamic study of aptamers binding to their target proteins. Biochimie 145:91–97 11. Oda M, Nakamura H (2000) Thermodynamic and kinetic analyses for understanding sequence-specific DNA recognition. Genes Cells 5:319–326 12. Prabhu NV, Sharp KA (2005) Heat capacity in proteins. Annu Rev Phys Chem 56:521–548 13. Salim NN, Feig AL (2009) Isothermal titration calorimetry of RNA. Methods 47:198–205 14. Jayaraman B, Mavor D, Gross JD, Frankel AD (2015) Thermodynamics of Rev–RNA interactions in HIV-1 Rev–RRE assembly. Biochemistry 54:6545–6554 15. Burnouf D, Ennifar E, Guedich S, Puffer B, Hoffmann G, Bec G, Disdier F, Baltzinger M,

Dumas P (2012) kinITC: a new method for obtaining joint thermodynamic and kinetic data by isothermal titration calorimetry. J Am Chem Soc 134:559–565 16. Dumas P, Ennifar E, Da Veiga C, Bec G, ˜eiro A, Sabin J, Palau W, Di Primo C, Pin ˜ oz E, Rial J (2016) Extending ITC to Mun kinetics with kinITC. Methods Enzymol 567:157–180 17. Velazquez-Campoy A, Freire E (2006) Isothermal titration calorimetry to determine association constants for high-affinity ligands. Nat Protoc 1:186–191 18. Turnbull WB, Daranas AH (2003) On the value of c: can low affinity systems be studied by isothermal titration calorimetry. J Am ChemSoc 125:14859–14866

Chapter 9 Detection of MicroRNAs Released from Argonautes Kyung-Won Min, J. Grayson Evans, Erick C. Won, and Je-Hyun Yoon Abstract The Argonaute (AGO) family of proteins plays an essential role in the process of microRNA (miRNA)mediated gene silencing. More specifically, they are the only known proteins to associate directly with miRNAs within the RNA-induced silencing complex (RISC). Given the importance of miRNA regulation of the transcriptome and its vast implications for human disease, it is essential to understand the molecular underpinnings of miRNA-AGO interactions. Although there are methods available to investigate mature miRNA decay and loading onto AGO2, no feasible method exists to detail the opposite process: release of miRNA from associated AGO proteins. In this chapter, we describe in detail a methodology derived from biochemical approaches, which can be used to quantify the release of any given miRNA from AGOs. Key words miRNA, Argonautes, RBP, Ribonucleoprotein complexes, qPCR, miRNA release

1

Introduction In mammalian cells, posttranscriptional gene regulation mainly occurs in the interplay between noncoding RNAs (ncRNAs) and RNA-binding proteins (RBPs) [1, 2]. In particular, the interactions between miRNAs and the Argonaute (AGO) family of RBPs are crucial in order to understand the role of miRNAs in gene silencing, as AGO proteins constitute the catalytic and RNA-binding portion of the RNA-induced silencing complex (RISC) [3–5]. The biogenesis and processing of miRNAs have been thoroughly studied [3, 6]. Briefly, biogenesis begins with the transcription of primary miRNAs (pri-miRNAs) [7, 8]. These hairpins are processed by several steps facilitated by the microprocessor composed of Drosha [9] and DGCR8 [10] in order to produce precursor miRNAs (pre-miRNAs), which are then exported from the nucleus via exportin 5 [11–13]. Processing continues via interactions with DICER, cleaving the pre-miRNA into sequences of around ~22 nucleotide base pairs in length [14–16]. The miRNA duplex is then loaded onto AGO proteins with assistance of the chaperone proteins (HSC70/HSP90) [17, 18]. The choice of dominant mature

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_9, © Springer Science+Business Media, LLC, part of Springer Nature 2020

151

152

Kyung-Won Min et al.

miRNA is determined by thermodynamic and structural properties of the processed duplex [19–24], which as described previously, comprise the catalytic component of the RISC [3, 25]. Ultimately, miRNA expression, processing, and association with RISC lead to target mRNA translation repression, deadenylation, and subsequent degradation [3, 4, 26]. A critical step in miRNA function relies on the assembly of RISC complex [27]. During this key step, duplex miRNAs are loaded onto AGO proteins in an ATP-dependent manner [17]; thus, one strand remains in the AGO proteins while discarding the complementary strand. There is the growing evidence supporting the existence of protein-free cytoplasmic miRNAs [28] and new types of miRNA-binding proteins (miRBPs) that can bind directly to miRNAs without the aid of chaperone proteins. Researchers are increasingly interested in the fate of protein-free miRNAs [28–32], as well as the molecular processes that could be involved (e.g., recycling to target mRNAs, alternative miRNA decay, or packaging of miRNA into exosomes) [33–35]. Furthermore, RBP AUF1 preferentially binds to single-strand miRNAs rather than to duplex miRNAs [30], with likely involvement in single-strand miRNA metabolism such as miRNA decay, transfer, and release from AGO proteins. Since miRNAs potentially target 30–60% of the protein-coding genes understanding the sophisticated mechanism for miRNA regulation by miRBP is necessary and, to a greater extent the regulation of miRNA levels in exosomes by miRBP which functions as potent signaling molecules for cell-cell communication. Consequently, to further our understanding of miRNA metabolism, it is necessary to investigate the role of miRNA release from AGO proteins. In this chapter, we describe the biochemical method to quantify single-stranded miRNAs released from AGO2 in vitro (in a test tube) and in vivo (in mammalian cells). The in vitro assay utilizes a recombinant miRBP incubated with miRNA-bound AGO2 to measure the non-protein-associated miRNA level present in the reaction buffer. The in vivo (or cellular) assay utilizes a recombinant miRBP incubated with AGO2 enriched with antibodies from mammalian cell lysate. In each experimental setting, the protein-free miRNAs/or non-AGO2-associated miRNAs are quantified via RT-qPCR. Detailed step-by-step instructions, the reagents used, and the additional notes concerning the experimental protocol are provided to ensure maximal understanding and ease of use.

2

Materials

2.1 Inducing AGO2 and MiRNA Complex

1. Recombinant His-tagged AGO2 protein (Sino Biological Inc.).

miRNA Release from Argonaute2 Protein

153

2. Synthetic hsa-let-7b miRNA with phosphoryl group at the 50 -end /5Phos/UGAGGUAGUAGGUUGUGUGGUU-30 . 3. NT2 Reaction buffer: 50 mM Tris–HCl at pH 7.5, 150 mM NaCl, 1 mM MgCl2 and 0.05% NP-40. 2.2 Enrichment of AGO2/MiRNA Complex

1. Nickel NTA agarose beads.

2.3 Incubation of AGO2/MiRNA with RB

1. Recombinant RBPs and control protein (see Note 1).

2.4 SK-N-SH Cell Culture

2. Ice-cold NT2 buffer (see item 3 in Subheading 2.1).

2. NT2 buffer (see item 3 in Subheading 2.1). 1. SK-N-SH cell. 2. Dulbecco’s Modified Eagle Medium (DMEM), 10% Fetal bovine serum (FBS), 100 U/mL Penicillin/streptomycin (see Note 2). 3. Ice-cold phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, and 1.5 mM KH2PO4. 4. 1 mL of ice-cold NT2 buffer: NT2 buffer (see item 3 in Subheading 2.1) supplemented with 1 μL of RiboLock RNase inhibitor (40 U/μL), 20 μL of protease inhibitor cocktail (50) and 1 μL of DTT (100 mM). 5. 100 mm culture dish.

2.5 Immunoprecipitation

1. Protein A or G Sepharose beads. 2. AGO2 antibody. 3. Recombinant RBPs and control proteins (see Note 1). 4. Ice-cold NT2 buffer (see item 3 in Subheading 2.1). 5. BCA protein assay.

2.6 Acidic Phenol Extraction

1. Acidic phenol:chloroform, pH 4.5 (with Isoamylalcohol, 25:24:1). 2. 15 mg/mL GlycoBlue™. 3. 3 M Sodium Acetate, pH 5.2. 4. 100% Ethanol.

2.7 Reverse Transcription and qPCR

1. dNTP mix. 2. Poly(A) Polymerase. 3. Poly(A) tailing buffer. 4. Reverse transcriptase. 5. 10 μM of universal 30 miRNA reverse primer.

154

Kyung-Won Min et al.

6. Forward primer for 2.5 μM of hsa-let-7b qPCR: 50 -TGAGGTA GTAGGTTGTGTGGTT-30 . 7. KAPA SYBR qPCR reagent.

3

Methods

3.1 Inducing AGO2/ MiRNA Complex Formation In Vitro

1. Mix 20 pM of hsa-let-7b miRNA wth 20 nM of His-AGO2 recombinant protein (1:1000 molar ratio) in 100 μL of NT2 buffer (see Note 3). 2. Incubate reaction mixture at 37  C for 1 h.

3.2 Enrichment of AGO2/MiRNA Complex In Vitro

1. Wash 30 μL of Ni-NTA beads twice with 1 mL of ice-cold NT2 buffer. 2. Discard supernatant without disturbing the beads. 3. Add 100 μL (see Subheading 3.1) of the reaction mixture to the washed Ni-NTA beads. 4. Incubate the reaction mixture at 4  C for 2 h with rotation.

3.3 Incubation of AGO2/MiRNA with RBP In Vitro

1. After centrifugation at 2000  g for 1 min at 4  C, wash beads three times with 1 mL of ice-cold NT2 buffer, aspirate supernatant and add 200 μL of NT2 buffer to beads (see Note 4). 2. Add 20 nM of recombinant RBPs (AUF1 [36], HuR [29] and La [37]), then incubate 100 μL of reaction mixture at 37  C for 1 h (see Note 5). 3. Collect supernatant containing miRNA released from AGO2 (see Fig. 1). 4. Continue with Subheading 3.6.

3.4 Mammalian Cell Culture and Cell Harvesting

(In vivo cell assay) 1. Expand human cell lines such as SK-N-SH (see Note 2) in a 100 mm culture dish, using DMEM supplemented with 10% (v/v) fetal bovine serum and 100 U/mL Penicillin/streptomycin. Grow cells up to 80% confluence. 2. Aspirate the culture medium, wash cells twice with 5 mL of ice-cold PBS, and aspirate PBS completely. Keep the dishes on ice during this procedure. 3. Lyse cells with 1 mL of ice cold NT2 buffer. 4. Collect lysate and transfer lysate to a 1.5 mL tube. 5. Incubate on ice for 10 min and centrifuge at 10,000  g for 15 min at 4  C. 6. Transfer supernatant to a 1.5 mL tube and then measure protein concentration with BCA protein assay (see Note 6).

miRNA Release from Argonaute2 Protein

155

Fig. 1 Schematic representation of the detection of miRNA released from AGO2. The in vitro miRNA release assay (using miRNA and recombinant AGO2 protein) and in vivo cell-based assay (using cell lysate and AGO2 antibody) are shown. Addition of miRBP might lead to the release of the miRNA from AGO2. After AGO2 pulldown (immunoprecipitation) using the appropriate conjugated beads, the miRNAs that dissociated from AGO2 (present in supernatant) can be identified using RT-qPCR

3.5 Immunoprecipitation and Incubation with a Desired RBP

(In vivo cell assay) 1. To prepare Sepharose beads, wash 30 μL of beads with ice-cold lysis buffer twice. 2. Remove supernatant and add 1 μg of AGO2 antibody to the beads in 1 mL of NT2 buffer. Incubate the beads and AGO2 antibody for 1 h at 4  C. 3. Add 1 mg of cell lysate to the Sepharose beads coated with antibody against AGO2 for 2 h at 4  C. 4. Centrifuge the beads at 2000  g for 1 min at 4  C. 5. Wash the pellet three times with lysis buffer (see Note 4).

156

Kyung-Won Min et al.

6. Incubate beads (AGO2:miRNA) with 20 nM of recombinant RBPs (AUF1 [36], HuR [29]and La [37]) in 100 μL of NT2 buffer (The ratio of AGO2 to AUF1 is 1:1) (see Note 7). 7. After centrifugation at 2000  g for 1 min at 4  C, collect supernatant containing miRNA released from AGO2 (see Note 8). 3.6 RNA Purification Using Acidic Phenol Extraction

1. Add 500 μL acidic phenol and 400 μL of DEPC-treated water to the collected supernatant. 2. Vortex for 1 min, then spin at 12,000  g for 15 min. 3. Transfer 400 μL of supernatant to a 1.5 mL tube, then add 800 μL of 100% ethanol, 40 μL of 3 M sodium Acetate (pH 5.2), and 1 μL of Glycoblue. 4. Mix gently and keep at 4  C for overnight. 5. Spin at 12,000  g for 30 min and discard supernatant. 6. Wash with 1 mL of 70% ethanol. 7. Dry pellet at room temperature for 5 min. 8. Resuspend pellet in 20 μL of DEPC-treated water.

3.7 Reverse Transcription and qPCR

1. For Poly(A) tailing reaction, mix 2 μL of Poly(A) tailing buffer (5) and 1 μL of Poly(A) polymerase with 7 μL of purified RNAs. 2. Incubate 1 h at 37  C followed by 5 min at 70  C. Briefly centrifuge to collect the contents. Keep on ice before cDNA synthesis. 3. Set up the cDNA synthesis reaction by mixing 10 μL of the mixture (from step 2) with 9 μL of reaction buffer and 1 μL of Reverse transcriptase. 4. Incubate 20 min at 42  C followed by 5 min at 85  C. 5. cDNA is ready to use for RT-qPCR. 6. For detecting hsa-let-7b miRNA, mix 2.5 μL of cDNAs with 2.5 μL of hsa-let-7b specific forward primer (2.5 μM), universal 30 miRNA reverse primer (10 μM from QuantiMir Kit), and SYBR green master mix. 7. Perform PCR and calculate Ct values of each sample. GST or MBP can be used as a negative control to normalize AGO2 pulldown (see Fig. 2).

4

Notes 1. Compare the result of GST protein control with the result of no protein control. If there is a significant variation in the level

miRNA Release from Argonaute2 Protein

157

Fig. 2 Let-7b release from AGO2 is regulated by RBPs. (a) His-tagged recombinant AGO2/let-7b (1000:1 molar ratio) complex was incubated and enriched by Ni-NTA beads. The indicated miRBP was added to Ni-NTA beads then the let-7b released from AGO2 present in the supernatant of the reaction mixture was analyzed by RT-qPCR. (b) SK-N-SH cells were lysed with NT2 buffer supplemented with RNase inhibitor, protease inhibitors, and 1 mM DTT. 1 mg of cell lysates were incubated with the Sepharose beads coated with AGO2 specific antibody. After washing the beads, 20 nM of the indicated proteins (either GST, AUF1, HuR or La) were added to beads and the let-7b that was released from AGO2 (from supernatants) were analyzed via RT-qPCR. Note that the different result between in vitro and in vivo assay may have arisen from posttranslational modification of AGO2 which occurred in cells [38]. (c) Image of protein stains on a 4–15% gradient polyacrylamide gel. The images presented in (c) were cropped to improve clarity

of target miRNA in the supernatant, consider the use of a different non-RBP as a control protein. 2. Compositions of cell culture medium and supplementary materials are dependent on cell lines utilized. 3. A concentration of target miRNA and AGO2 protein can be adjusted but keep it at a 1:1000 molar ratio of miRNA to AGO2 to minimize the quantity of non-AGO2-associated miRNA.

158

Kyung-Won Min et al.

4. Collect the supernatant from each washing step. Check the miRNA levels in the collected supernatants by RT-qPCR. Optimize the number of washing steps to the point until non-AGO2-associated miRNAs are not present in the supernatant. Western blot analysis can be utilized to assess if AGO2 is not present in the supernatant. 5. No ATP is required in the reaction mixture for the RBPs tested in this chapter. However, the effect of ATP on miRNA release should be tested in other RBPs. 6. Presence of AGO2 protein in either the supernatant or the pellet can be monitored by western blot analysis. If AGO2 protein is detected in the pellet consider using sonication for cell lysis. 7. The incubation time and concentration for the desired RBP can be adjusted empirically. Because non-AGO2-associated miRNA could arise from Koff [miRNA:AGO2], to minimize that possibility, non-RBP such as GST or MBP can be used as a control protein for normalization. 8. Adjust the incubation time of AGO2/miRNA complex with a choice of RBP empirically. Note that non-RBP can be used as a control protein for normalization.

Acknowledgments Medical University of South Carolina and Hollings Cancer Center to J.H.Y. 2019 Academic Research Support Program in GangneungWonju National University to K.W.M References 1. Wu E et al (2017) A continuum of mRNP complexes in embryonic microRNA-mediated silencing. Nucleic Acids Res 45(4):2081–2098 2. Gehring NH, Wahle E, Fischer U (2017) Deciphering the mRNP Code: RNA-bound determinants of post-transcriptional gene regulation. Trends Biochem Sci 42:369–382 3. Bartel DP (2018) Metazoan microRNAs. Cell 173:20–51 4. Jonas S, Izaurralde E (2015) Towards a molecular understanding of microRNA-mediated gene silencing. Nat Rev Genet 16:421–433 5. Pasquinelli AE (2012) MicroRNAs and their targets: recognition, regulation and an emerging reciprocal relationship. Nat Rev Genet 13:271

6. Ha M, Kim VN (2014) Regulation of microRNA biogenesis. Nat Rev Mol Cell Biol 15:509–524 7. Kwon SC et al (2016) Structure of human DROSHA. Cell 164:81–90 8. Nguyen TA, Park J, Dang TL, Choi Y-G, Kim VN (2018) Microprocessor depends on hemin to recognize the apical loop of primary microRNA. Nucleic Acids Res 46(11):5726–5736 9. Lee Y et al (2003) The nuclear RNase III Drosha initiates microRNA processing. Nature 425:415 10. Han J et al (2004) The Drosha-DGCR8 complex in primary microRNA processing. Genes Dev 18:3016–3027

miRNA Release from Argonaute2 Protein 11. Yi R, Qin Y, Macara IG, Cullen BR (2003) Exportin-5 mediates the nuclear export of pre-microRNAs and short hairpin RNAs. Genes Dev 17:3011–3016 12. Bohnsack MT, Czaplinski K, Gorlich D (2004) Exportin 5 is a RanGTP-dependent dsRNAbinding protein that mediates nuclear export of pre-miRNAs. RNA 10:185–191 13. Zeng Y, Cullen BR (2004) Structural requirements for pre-microRNA binding and nuclear export by Exportin 5. Nucleic Acids Res 32:4776–4785 14. Hutva´gner G et al (2001) A cellular function for the RNA-interference enzyme dicer in the maturation of the let-7 small temporal RNA. Science 293:834–838 15. Park J-E et al (2011) Dicer recognizes the 50 end of RNA for efficient and accurate processing. Nature 475:201 16. Tian Y et al (2014) A phosphate-binding pocket within the platform-PAZ-connector helix cassette of human dicer. Mol Cell 53:606–616 17. Iwasaki S et al (2010) Hsc70/Hsp90 chaperone machinery mediates ATP-dependent RISC loading of small RNA duplexes. Mol Cell 39:292–299 18. Naruse K, Matsuura-Suzuki E, Watanabe M, Iwasaki S, Tomari Y (2018) In vitro reconstitution of chaperone-mediated human RISC assembly. RNA 24:6–11 19. Matranga C, Tomari Y, Shin C, Bartel DP, Zamore PD (2005) Passenger-strand cleavage facilitates assembly of siRNA into Ago2containing RNAi enzyme complexes. Cell 123:607–620 20. Okamura K, Liu N, Lai EC (2009) Distinct mechanisms for microRNA strand selection by Drosophila Argonautes. Mol Cell 36:431–444 21. Jo MH et al (2015) Human Argonaute 2 has diverse reaction pathways on target RNAs. Mol Cell 59:117–124 22. Yao C, Sasaki HM, Ueda T, Tomari Y, Tadakuma H (2015) Single-molecule analysis of the target cleavage reaction by the Drosophila RNAi enzyme complex. Mol Cell 59:125–132 23. Salomon WE, Jolly SM, Moore MJ, Zamore PD, Serebrov V (2015) Single-molecule imaging reveals that Argonaute reshapes the binding properties of its nucleic acid guides. Cell 162:84–95 24. Chandradoss SD, Schirle NT, Szczepaniak M, MacRae IJ, Joo CA (2015) Dynamic search

159

process underlies microRNA targeting. Cell 162:96–107 25. Fabian MR, Sonenberg N (2012) The mechanics of miRNA-mediated gene silencing: a look under the hood of miRISC. Nat Struct Mol Biol 19:586–593 26. Huntzinger E, Izaurralde E (2011) Gene silencing by microRNAs: contributions of translational repression and mRNA decay. Nat Rev Genet 12:99–110 27. Nakanishi K (2016) Anatomy of RISC: how do small RNAs and chaperones activate Argonaute proteins? Wiley Interdiscip Rev RNA 7:637–660 28. Janas MM et al (2012) Alternative RISC assembly: binding and repression of microRNA–mRNA duplexes by human Ago proteins. RNA 18:2041–2055 29. Yoon J-H et al (2013) Scaffold function of long non-coding RNA HOTAIR in protein ubiquitination. Nat Commun 4:2939 30. Yoon JH et al (2015) AUF1 promotes let-7b loading on Argonaute 2. Genes Dev 29:1599–1604 31. Min KW et al (2017) AUF1 facilitates microRNA-mediated gene silencing. Nucleic Acids Res 45(10):6064–6073 32. Zealy RW, Wrenn SP, Davila S, Min K-W, Yoon J-H (2017) microRNA-binding proteins: specificity and function. Wiley Interdiscip Rev RNA 8:e1414 33. Baccarini A et al (2011) Kinetic analysis reveals the fate of a microRNA following target regulation in mammalian cells. Curr Biol 21:369–376 34. Golden RJ et al (2017) An Argonaute phosphorylation cycle promotes microRNAmediated silencing. Nature 542:197–202 35. Shurtleff MJ, Temoche-Diaz MM, Karfilis KV, Ri S, Schekman R (2016) Y-box protein 1 is required to sort microRNAs into exosomes in cells and in a cell-free reaction. eLife 5:e19276 36. Choi YJ, Yoon J-H, Chang JH (2016) Crystal structure of the N-terminal RNA recognition motif of mRNA decay regulator AUF1. Biomed Res Int 2016:9 37. Kota V et al (2016) SUMO-modification of the La protein facilitates binding to mRNA in vitro and in cells. PLoS One 11:e0156365 38. Jee D, Lai EC (2014) Alteration of miRNA activity via context-specific modifications of Argonaute proteins. Trends Cell Biol 24:546–553

Chapter 10 Mapping the RNA Chaperone Activity of the T. brucei Editosome Using SHAPE Chemical Probing W.-Matthias Leeder and H. Ulrich Go¨ringer Abstract Mitochondrial pre-mRNAs in African trypanosomes adopt intricately folded, highly stable 2D and 3D structures. The RNA molecules are substrates of a U-nucleotide-specific insertion/deletion-type RNA editing reaction, which is catalyzed by a 0.8 MDa protein complex known as the editosome. RNA binding to the editosome is followed by a chaperone-mediated RNA remodeling reaction. The reaction increases the dynamic of specifically U-nucleotides to lower their base-pairing probability and as a consequence generates a simplified RNA folding landscape that is critical for the progression of the editing reaction cycle. Here we describe a chemical mapping method to quantitatively monitor the chaperone-driven structural changes of pre-edited mRNAs upon editosome binding. The method is known as selective 20 -hydroxyl acylation analyzed by primer extension (SHAPE). SHAPE is based on the differential electrophilic modification of ribose 20 -hydroxyl groups in structurally constraint (double-stranded) versus structurally unconstrained (single-stranded) nucleotides. Electrophilic anhydrides such as 1-methyl-7-nitroisatoic anhydride are used as probing reagents, and the ribose 20 -modified nucleotides are mapped as abortive cDNA synthesis products. As a result, SHAPE allows the identification of all single-stranded and base-paired regions in a given RNA, and the data are used to compute experimentally derived RNA 2D structures. A side-by-side comparison of the RNA 2D folds in the pre- and post-chaperone states finally maps the chaperone-induced dynamic of the different pre-mRNAs with single-nucleotide resolution. Key words African trypanosomes, U-indel RNA editing, Editosome, Pre-edited mRNA, RNA chaperone, G-quadruplex, SHAPE chemical probing, RNA dynamic

1

Introduction The maturation of mitochondrial transcripts in the parasitic protozoan organism Trypanosoma brucei requires an insertion/deletion (indel)-type RNA editing reaction [1, 2]. The process converts cryptic, i.e., sequence-deficient primary transcripts into translatable messenger (m)RNAs, which ultimately code for essential proteins of the oxidative phosphorylation system such as the NADH-ubiqinone oxidoreductase (complex I), cytochrome bc1 (complex III), cytochrome oxidase (complex IV), and the ATPase synthase (complex V). Importantly, the editing reaction has a strict nucleotide

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_10, © Springer Science+Business Media, LLC, part of Springer Nature 2020

161

162

W.-Matthias Leeder and H. Ulrich Go¨ringer

specificity: only U-nucleotides (nt) are site-specifically inserted into and/or deleted from the different pre-mRNAs, which for some of the transcripts can amass to more than 50% of the sequence information of the final, fully edited mRNA sequence. This overwhelming form of RNA editing has been termed pan-editing. The process is catalyzed by a high molecular mass (0.8 MDa) multienzyme complex known as the editosome [3, 4]. Editosomes act as reaction surfaces for the binding and subsequent processing of the different pre-edited substrate RNAs and execute a multistep catalytic reaction cycle. Key molecules in the process are trypanosome-specific, small noncoding RNAs, so-called guide (g) RNAs. Through the formation of gRNA/pre-mRNA hybrid structures, they perform a template-like function that guides the insertion and deletion of U-nucleotides by virtue of their primary sequences. Editing proceeds with a general 30 to 50 directionality on the pre-edited mRNA and in the majority of cases multiple (10) gRNAs are required to fully edit a single pre-mRNA. This emphasizes the structural dynamic of the editing reaction: successive pre-mRNA/gRNA hybrid structures have to be formed, which can only hybridize if the interacting RNA sequences are structurally accessible for hydrogen bonding. This argues for a high degree of single strandedness in the different pre-mRNA molecules; however, the contrary is the case. Pre-edited mRNAs adopt thermodynamically highly stable secondary (2D) and perhaps tertiary (3D) folds [5]. This is in part due to the substantial sequence deficiencies of the different RNAs (on average 45% of the nt-information is missing), which causes an extreme nt-bias. Pan-edited pre-mRNAs are purine (R)-rich and pyrimidine (Y)-deprived (average R/Y ¼ 2). In addition, they contain high numbers of G-nucleotide runs. Two-thirds of all G-residues are embedded in G-clusters of 2  G  8. Runs of G-nt have been shown to fold into thermodynamically highly stable G-quadruplex (GQ) folds and with the exception of the pre-mRNA of RPS12 all pan-edited T. brucei mRNAs contain minimally two up to five GQ-elements [5]. As a result of the highly folded nature of the different primary transcripts, it is not surprising that the editing reaction cycle involves RNA helicases and RNA annealing factors as auxiliary activities [6, 7]. In addition, the editosome itself has been identified to execute a complex-inherent RNA chaperone activity. Binding of the pre-edited substrate mRNAs to the catalytic complex triggers a structural refolding reaction that generates a simplified RNA folding landscape [8, 9]. This involves about one-third of the nucleotides of every pre-mRNA and specifically involves uridylate residues. The flexibility increase of the affected U’s lowers their base-pairing probability and as a consequence reduces the energy barrier to promote the formation of gRNA/pre-mRNA hybrid structures. The editosomal RNA chaperone activity is a surfacedriven process. It is executed by the intrinsically disordered protein

SHAPE-Chemical Probing of RNA Chaperone-Mediated RNA Dynamic

163

(IDP) domains of the oligonucleotide/oligosaccharide-binding (OB) fold proteins of the editing complex and likely follows a fly-casting type mechanism [9]. In the following we describe an experimental protocol to precisely examine the chaperone-driven RNA unfolding reaction that is executed by the T. brucei editosome. The procedure is known as SHAPE chemical probing or mapping [10–13]. SHAPE is an acronym for “selective 20 -hydroxyl acylation analyzed by primer extension.” It was developed by Kevin Weeks and colleagues at the University of North Carolina, Chapel Hill, USA. The method enables the quantitative monitoring of the structural dynamic of all nucleotide positions in a given RNA molecule in a single experiment. SHAPE chemical probing relies on the fact that the nucleophilicity of every ribose 20 -hydroxyl (OH) group in a polynucleotide is affected by the nearby 30 -phosphodiester moiety. As a consequence, nucleotides in a single-stranded, i.e., flexible sequence context show an increased nucleophilicity and nucleotides in a double-stranded, i.e., structurally constrained conformation show a decreased nucleophilicity. This chemical difference is monitored with the help of OH-selective, electrophilic compounds such as N-methylisatoic anhydride (NMIA), 1-methyl-7-nitroisatoic anhydride (1M7), benzoyl cyanide (BzCN), or 2-methylnicotinic acid imidazolide (NAI) (for a comparison of all available reagents see Lee et al. [14]). All SHAPE compounds form stable, bulky 20 -Oadducts preferentially with single-stranded nucleotides (Fig. 1), which can be identified as abortive cDNA synthesis products in a follow-up reverse transcription reaction (cDNA synthesis stops precisely one nucleotide before the 20 -O-modified ribonucleotide). The generated cDNA fragments are electrophoretically separated, and the signal intensity at a given nt position represents a quantitative read-out of the SHAPE reactivity, i.e., the structural flexibility of that nucleotide (Fig. 2). Normalization allows the calculation of SHAPE reactivity units (SU), which can be converted into pseudo Gibbs-free energies (ΔGSHAPE). These values in turn can be used to aide 2D RNA structure prediction algorithms to calculate experimentally derived RNA 2D structures [15, 16] (Fig. 2). Importantly, the formation of the 20 -O-adducts is fast (seconds to minutes), it is self-limiting due to hydrolysis of the electrophilic compounds, and the reaction is not restricted by the size of the probed RNA. SHAPE can be performed with short and long RNA molecules alike [17, 18]. Lastly, SHAPE probing is insensitive to solvent accessibility [13, 19], and as such it is perfectly suited to trace the RNA chaperone-induced dynamic changes of pre-edited mRNAs by the editosomal complex.

W.-Matthias Leeder and H. Ulrich Go¨ringer

164

A

5‘

pre-edited mRNA

B

Base Base

O O

O

O

O

+ O O

P

N

OH

flexible nucleotide

+

O

O

pH 7.5 - 8 O O

O-

O

O

O

P

O

O

O-

H N

O NO2

1M7

NO2

Fig. 1 (a) Sketch of the 2D structure of a small SHAPE target RNA. The ribose-phosphodiester backbone of the molecule is depicted as a continuous grey line, Watson/Crick base pairs as short grey strokes and GU base pairs as grey dots. (b) Acylation of a 20 -OH group of a non-base-paired nucleotide by 1-methyl-7-nitroisatoic anhydride (1M7). The covalent modification is catalyzed at mild alkaline pH (7.5–8) and is restricted to flexible nucleotides. The formation of the 20 -O adduct is accompanied by the release of CO2

2

Materials In the following we focus on 1-methyl-7-nitroisatoic anhydride (1M7) as SHAPE reagent. The general workflow however is comparable for all other 20 OH-acylating electrophiles [14].

2.1 Buffers and Reagents

All chemicals should be of pro analysi (p.a.) grade. 1. TE buffer: 10 mM Tris–HCl pH 7.5, 1 mM Na2EDTA. 2. T. brucei pre-edited mRNAs (see Note 1) in TE buffer: A6 (401nt), CO3 (463nt), CR3 (164nt), CR4 (283nt), Cyb (632nt), MURF2 (1091nt), ND3 (268nt), ND7 (783nt), ND8 (361nt), ND9 (322nt), RPS12 (221). RNA concentration 1 μM. 3. T. brucei 20S editosomes (see Note 2) in 1 RNA folding buffer. A typical concentration is in the range of 0.1 μg/μL (¼0.13 pmol/μL). 4. 1-Methyl-7-isatoic anhydride (1M7); IUPAC: 1-methyl-5nitro-2H-benzo[d][1,3]oxazine-2,4(1H)-dione [20].

SHAPE-Chemical Probing of RNA Chaperone-Mediated RNA Dynamic

A

165

B 1.0

norm. reactivity

1M7

RNA

RT

0.85 0.6 0.4 0.2 0.0

GCUUCUUUUGAAUAAAAGGGAGGC

pseudo free energy (kcal/mol)

ΔGSHAPE=1.8*ln[SHAPE+1]+(-0.6)

cDNAfragments

1

0 1

norm. SHAPE-react. (SU)

-0

.2 -0 2 .5 -0 6 .5 -0 8 .5 -0 8 .5 -0 8 .5 -0 8 .5 -0 3 .3 -0 9 .1 8 +0 .1 9

CE

+0

.5

3

.0

5

2

elution time

.0 +0 9 .2 -0 4 .3 -0 1 .3 -0 4 .3 -0 0 .3 -0 4 .3 -0 2 .3 -0 3 .1 -0 8 .2 -0

fluorescence intensity

5‘ G C U U C U U U U G A 0.55 A+ U +0.52 3‘ C G G A G G G A A A A+ 0

Fig. 2 (a) Graphical flowchart of the workflow of a SHAPE experiment using the example of a simple RNA hairpin. The RNA is modified with 1M7 at single-hit conditions and reversed transcribed to identify the 20 -Omodified nucleotides as abortive cDNA synthesis products. cDNA synthesis stops one nucleotide before the 20 -O-modification. The generated cDNAs are separated by capillary electrophoresis (CE) and the quantity of each cDNA-fragment can be retrieved from the corresponding peak-integral in the electropherogram. (b) Graphical representation of the background-corrected and normalized SHAPE reactivity profile. The SHAPE reactivity data are converted into pseudo Gibbs free energies (ΔGSHAPE), which are used to calculate a minimal free energy RNA 2D structure. The pseudo-free energy contributions for each nucleotide (in kcal/mol) are shown next to the hairpin structure. The conversion is based on the heuristic equation: ΔGSHAPE ¼ 1.8  ln [SHAPE  react. + 1] + (0.6) [12]. A detailed video tutorial of the SHAPE technology can be found at http:// www.jove.com/video/50243 [35]

166

W.-Matthias Leeder and H. Ulrich Go¨ringer

5. 3 RNA folding buffer: 60 mM HEPES pH 8.0, 90 mM KCl, 30 mM MgCl2, 7.5 mM DTT. 6. Dimethylsulfoxid (DMSO), anhydrous. 7. Phenol, H2O-saturated. 8. Phenol/chloroform/isoamylalcohol (PCI) (25:24:1 v/v/v). 9. Chloroform/isoamylalcohol (CI) (24:1 v/v). 10. Linear polyacrylamide 1 μg/μL. 11. 100% (v/v) EtOH. 12. 70% (v/v) EtOH. 13. Spin chromatography columns. 14. Size exclusion gel matrix for spin chromatography. 15. Superscript® IV reverse transcriptase or a comparable enzyme. 16. 5 Reverse transcriptase (RT) buffer: 250 mM Tris–HCl pH 8.3, 375 mM LiCl, 15 mM MgCl2. 17. 0.1 M Dithiothreitol (DTT). 18. 10 mM each dNTP-mix (Deoxyadenosine-, deoxyguanosine-, deoxycytidine- and deoxythymidine-triphosphate). 19. 2.5 mM Dideoxyadenosine-, dideoxycytidine-, dideoxythymidine- and dideoxyguanosine-triphosphate (ddATP, ddCTP, ddTTP, ddGTP). 20. 0.5 M Na2EDTA, pH 8. 21. 125 mM Na2EDTA, pH 8. 22. 1 μM 50 -fluorophor (FAM, JOE, TAMRA, ROX)-labeled DNA-oligonucleotide primer in TE buffer. 23. 3 M NaOH. 24. 100% (v/v) HOAc. 25. HiDi™-formamide. 2.2 Laboratory Equipment and Software

1. 20  C Freezer. 2. Refrigerated laboratory microcentrifuge. 3. PCR-thermocycler. 4. Centrifugal vacuum concentrator. 5. Capillary electrophoresis (CE) system (e.g., ABI PRISM™ 310 Genetic Analyzer). 6. Software: ShapeFinder [21]: https://github.com/drsuuzzz/ ShapeFinder (Mac OSX only. Make sure the software has read and write privileges for /Library/BaseFinder and all subdirectories.). Software alternatives: QuShape [22]: https://weeks.chem.unc.edu/qushape/ Requires Python, platform-independent.

SHAPE-Chemical Probing of RNA Chaperone-Mediated RNA Dynamic

167

RiboCAT [23]: https://research.cbc.osu.edu/musierforsyth.1/tools/ (Requires Microsoft Excel). HiTRACE [24, 25]: https://ribokit.github.io/ HiTRACE/ (stand-alone) or www.hitrace.org (online version) (Stand-alone requires MatLab).

3

Methods

3.1 RNA Folding and Editosome Binding

First step in the protocol is the generation of an RNA conformational landscape that is competent to bind to the editosome complex. This is achieved by the heat-induced unfolding of the RNA at low ionic strength followed by a refolding reaction in the presence of RNA-stabilizing monovalent and divalent cations. 1. Mix 4 pmol of pre-mRNA in a 0.2 mL PCR reaction tube with 6 μL 1 TE and add ddH2O to a final volume of 24 μL. 2. Place the tube in a PCR thermocycler and denature the RNA by quickly heating to 95  C. Incubate for 1.5 min and snap cool on ice for 1 min. 3. Add 12 μL of 3 RNA folding buffer. Mix and incubate for 30 min at 27  C (see Note 3). 4. Split the sample into two 18 μL samples using fresh 0.2 mL PCR reaction tubes. Add an equimolar amount (2 pmol) of editing-active 20S editosomes to one of the vials [(+) editosomes ¼ post-chaperone state] and add the same volume of 1 RNA folding buffer to the second vial [() editosomes ¼ prechaperone state] (see Note 4). Incubate at 27  C for 40 min (see Note 5). 5. After the incubation each sample is split in half again representing the (+) 1M7- and () 1M7-samples.

3.2

RNA Modification

1. Add 35 mM 1M7 in DMSO to a final concentration of 3.5 mM to all (+) 1M7-samples and the same volume of neat DMSO to all () 1M7-samples. 2. Incubate all vials for 70 s (5 the half-life of 1M7) at 27  C (see Note 6). 3. Transfer all samples to fresh 1.5 mL reaction tubes. 4. Add 1 vol. H2O-saturated phenol to each tube. Mix and centrifuge (21,000 rcf) for 10 min at 4  C. 5. Transfer the aqueous phases to fresh 1.5 mL reaction tubes and add 1 vol. of PCI (25:24:1 (v/v/v)). Mix and spin at 21,000 rcf for 5 min at 4  C.

168

W.-Matthias Leeder and H. Ulrich Go¨ringer

6. Again, transfer the aqueous phases to fresh reaction vials and add 1 vol. of CI (24:1 (v/v)). Mix and centrifuge at 21,000 rcf for 1 min at 4  C. 7. Transfer the aqueous phases to fresh tubes. Add 0.1 vol. 3 M NaOAc pH 4.8 and 1 μg linear polyacrylamide as a co-precipitant. Add 2.5 vol. of 100% (v/v) EtOH, mix and incubate at 20  C for 30 min. 8. Spin samples at 21,000 rcf for 45 min at 4  C. Remove supernatants and dry the RNA pellets in a vacuum concentrator for 2 min. 9. Dissolve the RNA pellets in 30 μL 0.25 TE buffer and desalt the RNA by size exclusion chromatography (see Note 7). 3.3

cDNA Synthesis

1. Transfer 26 μL of each pre-mRNA sample (~1 pmol) into a fresh 0.2 mL PCR reaction tube. Add 1 μL of a 1 μM solution of the appropriate fluorescently labeled DNA-oligonucleotide primer. We recommend FAM-labeled primers for the (+) 1M7-samples and JOE-labeled primers for the () 1M7-samples (see Note 8). 2. Set up two 27 μL DNA-sequencing reactions in 0.25 TE for the pre- and the post-chaperone state in 0.2 mL PCR reaction tubes. Each sequencing reaction is performed with 1 pmol untreated RNA using an equimolar amount of a fluorescently labeled (e.g., 50 -TAMRA) DNA-oligonucleotide primer. Adjust the final concentration of the ddNTP’s to 0.125 mM in 40 μL. 3. Denature all samples by rapidly heating to 95  C. Incubate for 1.5 min and cool down quickly to a temperature 2  C below the Tm of the DNA-oligonucleotide primer. Anneal for 5–10 min. 4. Prepare a RT reaction mix by combining 8 μL 5 RT buffer, 2 μL 0.1 M DTT, and 2 μL 10 mM dNTP-mix. 5. Snap cool all samples on ice for 1 min. 6. Add 12 μL of the RT-reaction mix to each sample, mix, and incubate at 52  C for 1 min. 7. Add 1 μL SuperScript IV (200 U/μL) to each sample and incubate for 20 min at 52  C (see Note 9). 8. Stop cDNA synthesis by snap cooling on ice. 9. Set-up one SHAPE-sample for the pre-chaperone state and one sample for the post-chaperone state by combining the cDNA of the (+) 1M7-sample and the () 1M7-sample plus one sequencing reaction in a 1.5 mL reaction tube.

SHAPE-Chemical Probing of RNA Chaperone-Mediated RNA Dynamic

169

10. Add 0.1 vol. of 3 M NaOH and incubate samples for 5 min at 95  C to completely hydrolyze the RNA. Quickly place samples on ice for 1 min. 11. Add 0.1 vol. 100% (v/v) HOAc and 0.1 vol. 125 mM Na2EDTA pH 8. Mix and precipitate the cDNA by adding 100% (v/v) EtOH at room temperature to a final concentration of 69% (v/v). Mix thoroughly and place the tubes in a microcentrifuge without further incubation. 12. Centrifuge at 21,000 rcf for 30 min at 4  C. Aspirate the supernatant. 13. Wash the pellet with 750 μL 70% (v/v) EtOH at room temperature. Spin samples at 21,000 rcf for 5–10 min at 4  C. Aspirate the supernatant. 14. Repeat step 13 (see Note 10). 15. Aspirate the supernatant and dry the cDNA pellets in a vacuum concentrator for 2 min. 16. Dissolve the cDNA-pellets in 15 μL HiDi™-formamide. Mix thoroughly. Incubate at 95  C for 2–5 min and place the tubes on ice for 2 min. 3.4 Capillary Electrophoresis

1. Upload the samples into the capillary electrophoresis (CE) system (see Note 11). 2. Electrokinetic injection of the samples is performed at 2.5 kV for 30 s. Electrophoresis is performed at 50  C at a voltage of 12.2 kV (see Note 12). 3. Output of the CE system is an electropherogram in which the fluorescence intensity (FI) of the separated cDNA fragments (in relative fluorescence units (RFU)) is plotted as a function of the electrophoretic elution time (FIcDNA ¼ f(ET)). The plots usually contain the information for all fluorescence channels in the form of overlapping traces (Fig. 3a).

3.5 SHAPE Data Analysis

1. The data are retrieved from the CE system and imported into ShapeFinder for all subsequent data manipulation and analysis steps [21]. ShapeFinder performs an alignment and integration of the different peaks and tabulates the results next to the primary sequence of the RNA. This is followed by subtracting the () 1M7-data from the (+) 1M7-data and by determining statistical outliers as >1.5-fold the interquartile range in a boxplot-based analysis [15]. All values are then normalized by dividing the individual reactivities by the maximal reactivity, which is defined as the mean of the top 10% of peaks (see Note 13). 2. The normalized SHAPE reactivities are finally uploaded into RNAstructure [15] or ViennaRNA [16]. Both programs allow

W.-Matthias Leeder and H. Ulrich Go¨ringer

170

primer peak

A

full-length peak

8000

RFU

6000 4000 2000 0 0

1000

2000

3000 elution time/s

4000

5000

baseline adjust matrixing mobility shift signal decay correction scaling

B

(+) 1M7 (-) 1M7

12000 8000

RFU

4000 0 12000 8000 4000 0

ddGTP seq. lane

2000

3000

4000

5000

norm. SHAPE-react. (SU)

C

peak integration background subtraction normalization

1.6 1.2 0.8 0.4 0.0 20

40

60

80

100

120

140 (nt)

160

180

200

220

240

260

280

norm. SHAPE-reactivity (SU)

elution time/s

1.6

0.85 0.4 0

Fig. 3 (a) Raw electropherograms of a 1M7-SHAPE experiment of T. brucei RPS12 pre-mRNA. Blue trace: (+) 1M7-sample. Green trace: () 1M7-sample. Black trace: ddGTP-sequencing profile. Red trace: empty chanel. (b) Processed profiles after baseline adjustment, matrixing, signal decay correction and scaling. (c) Peak integration, background subtraction, and normalization generate the final SHAPE reactivity profile (adapted from [8])

for a straight forward implementation of SHAPE-reactivity data to calculate experimentally guided RNA 2D structures (see Note 14). 3. Both, technical and biological replicates should be performed to analyze the statistical significance of the probing data (see Note 15).

4

Notes 1. Pre-edited mRNAs are generated by standard run-off in vitro transcription [26] from linearized plasmid DNA templates using T7-RNA polymerase. Full-length synthesis of the different pre-mRNAs is verified by electrophoresis in denaturing (8 M urea-containing) polyacrylamide gels (Fig. 4a). RNA preparations containing 10% of truncated transcription products should be purified by preparative, denaturing

SHAPE-Chemical Probing of RNA Chaperone-Mediated RNA Dynamic

A

RPS12

nt

A6

ND9

CO3

171

ND7

1000 800 600

400 300

B primer peak 8000

+ 1M7-modification

full length peak

6000

RFU

4000 2000 unmodified control

8000 6000 4000 2000 1000

2000

3000 4000 5000 elution time/s

6000

7000

Fig. 4 (a) Denaturing polyacrylamide gel electrophoresis of in vitro transcribed T. brucei pre-mRNAs RPS12, A6, ND9, CO3, and ND7. The different RNA preparations are 95% pure. (b) Capillary electrophoresis (CE) profiles of cDNA fragments derived from 1M7-modified (top panel) and unmodified (bottom panel) ND7 pre-mRNA. Comparison of the two electropherograms verifies the successful modification of individual nucleotides in the (+) 1M7 sample and the dominant full-length peak attests single-hit reaction conditions

polyacrylamide gel electrophoresis (PAGE). Additionally, a primer extension reaction should be performed to test the binding of the different DNA primer oligonucleotides and to confirm full-length cDNA synthesis (Fig. 4b). RNA concentrations are calculated from UV-absorbance measurements at 260 nm. Molar extinction coefficients (ε260) are derived from the sum of the nucleotide absorbancies of the different pre-mRNAs based on the nearest-neighbor model and its published parameter [27, 28]. 2. T. brucei editosomes are enriched from whole cell trypanosome detergent lysates by tandem affinity purification [3, 29]. The protein composition of the 0.8 MDa (20S) complexes is

172

W.-Matthias Leeder and H. Ulrich Go¨ringer

analyzed by mass spectrometry and the protein concentration is determined by Bradford dye-binding. The RNA editing activity (EA) of the isolates is tested in vitro using the established pre-cleaved U-insertion and U-deletion RNA editing assays [30, 31]. Visualization of the complexes can be achieved by atomic force microscopy (AFM) as in [32]. Editosomes are stored frozen in 1 RNA folding buffer containing 30% (v/v) glycerol. Prior to usage, the glycerol is removed by drop dialysis. 3. The temperature of 27  C represents the optimal growth temperature of insect-stage African trypanosomes. 4. Editosomes have one substrate pre-mRNA binding site [32]. To assure equimolar editosome/pre-mRNA reaction conditions, the fraction of editosomes that is competent to bind substrate pre-mRNA must be determined and adjusted accordingly for the incubation with the pre-mRNA substrate molecules. This can be done by nitrocellulose (NC)-filter binding using (32P)-labeled pre-mRNA preparations [33] or by surface plasmon resonance (SPR) measurements using nonradioactive pre-mRNA isolates [34]. The different pre-mRNAs bind to editosomes with Kd-values between 3 and 10 nM. 5. The half-lives of the different RNA/editosome complexes range between 2 and 10 min [32]. Thus, the incubation time is 4-fold the half-life of the complexes. 6. A quenching step to stop the acylation reaction is not necessary because of hydrolysis of the reagent. Furthermore, the reaction is by and large insensitive to ionic strength, to redox and crowding conditions. However, it requires (mild) alkaline reaction conditions (pH 7.5–8) [13]. The recommended 1M7 concentration is 3.5 mM, although it can vary if long target RNAs are probed or if the acylation is performed below or above the ambient temperature window. The most critical constraint of the method is that the probing reaction must be conducted with single-hit kinetic: Individual target RNAs in the sample must be modified only once in the region being analyzed. For RNAs with a molecular length of 300–1700 nt a 3.5 mM 1M7 solution yields single-hit conditions in a roughly 350 nt window. If the optimal concentration has to be determined, it is prudent to vary the 1M7 concentration and to determine the ratio of the peak integral of the full-length (FL) peak over the sum of all other peaks in the electropherogram. Single hit conditions are usually met at ratios 80/20 and if the signal intensity differences between long and short cDNA fragments is 2–3 (Fig. 4b). 7. Use for instance Bio-Rad Mirco Bio-Spin® columns with Bio-Gel® P-30.

SHAPE-Chemical Probing of RNA Chaperone-Mediated RNA Dynamic

173

8. All DNA-primer oligonucleotides should be approximately 18–25 nt in length with melting temperatures (Tm) between 55 and 65  C. The FAM, JOE, TAMRA, and ROX fluorophores are covalently linked to the 50 -phosphates of the primer oligonucleotides using a carbon-6-linker. The different fluorophores and the linkage are resistant to alkaline hydrolysis at 95  C. Other fluorophores such as Cy5, Cy5.5, WellRedD2, and IRDye800 have been demonstrated to work equally well [35]. 9. The required synthesis of full-length cDNA molecules can represent a prime obstacle for the SHAPE protocol. This is the case for all target RNAs that contain highly stable 2D structures, which do not melt out even at the elevated reaction temperature of 52  C. In that case an optimization of the cDNA synthesis step is required in which the concentration of structure-stabilizing mono- and divalent cations is lowered/ adjusted in combination with a variation of the synthesis temperature. A special case applies to the pan-edited pre-mRNAs of T. brucei. As mentioned above the different pre-mRNAs are extremely G-rich and contain multiple, up to five G-quadruplex (GQ) folds. Due to their high thermodynamic stability, GQ elements have been shown to stall cDNA synthesis, and as a consequence the reaction has to be performed at GQ-disfavouring conditions. GQ folds are stabilized by monovalent cations (K+ >> Rb+ > Na+ > Li+ ¼ Cs+) [36], and thus the reaction is performed in the presence of “suboptimal” monovalent cations such as Li+ or Cs+ (75 mM). In the case of the CR4 pre-mRNA 0.75 mM Ni2+- or Co2+-ions further suppress GQ-induced RT-stops. Note: If Ni2+-/Co2+-ions are used, do not use DTT in the RT-reaction mix. As a rule of thumb, optimized RT conditions result in signal intensities no greater than fivefold of the median of all peaks in the electropherogram (Figs. 3a and 4b). 10. The complete removal of RNA and the stringent EtOH precipitation of the generated cDNA fragments is of the essence. Residual complementary RNA and/or remaining traces of salt strongly impact the read length as well as the resolution and signal intensity of the capillary electrophoresis. 11. We recommend a 61 cm (50 cm to the detector) capillary with a diameter of 52 μm and POP-6™ as the gel matrix. 12. This setup leads to the detection of unincorporated DNA primer oligonucleotides after roughly 30 min. The electrophoretic “elution time” (ET) in minutes of a cDNA fragment can be estimated by the following approximation: ET (min) ¼ ntlength  0.15 + 39. In the case of a 350 nt long RNA about 90 min should be considered for the detection of the full-

174

W.-Matthias Leeder and H. Ulrich Go¨ringer

length cDNA and an additional 5 min for resolving the fulllength peak and the subsequent baseline. A read length of roughly 350 nt is generally achievable. The signal intensity can be adjusted by a prolonged or shortened injection time (15–45 s) and a voltage range for the electrokinetic injection between 2.5–3 kV. 13. In the case that an ABI PRISM™ 310 Genetic Analyzer is used for the CE separation of the cDNA fragments followed by a manipulation of the data with the help of ShapeFinder [21], a mobility shift calibration and a matrixing step are required. For that prepare a ddNTP-sequencing reaction from 2 pmol of RNA starting material in a final volume of 80 μL for each fluorescently labeled DNA primer (see step 2 in Subheading 3.3). After termination of cDNA synthesis (see step 8 in Subheading 3.3) pool 40 μL of each reaction to generate one mobility shift calibration sample. Treat the remaining 40 μL of each of the samples individually and continue as described (see step 11 in Subheading 3.3). A detailed description of all ShapeFinder tools is available in the ShapeFinder “Help menu” or in Vasa et al. [21]. Figure 3 visualizes all processing steps starting from the raw electropherograms of a typical 1M7-SHAPE probing experiment using the T. brucei RPS12 pre-mRNA as an example (Fig. 3a). This includes a baseline adjust, matrixing, and a signal decay correction to offset the imperfect processivity of the reverse transcriptase. Fluorophore-induced variations of the cDNA-elution time are adjusted by a mobility shift and are calibrated for each fluorophore individually. After alignment of the traces of the (+) 1M7 samples and the () 1M7 samples with the ddGTP-sequencing profile, the smallest peaks in the modified and the unmodified profile are superimposed and scaled (Fig. 3b). This assumes that every folded RNA contains a few nucleotides that are unreactive toward 1M7. The nucleotide identity of the different peaks is assigned with the help of the ddGTP-sequencing profile followed by peak integration. The final SHAPE-reactivity profile is shown in Fig. 3c. The profile is the result of subtracting the unmodified () 1M7 background from the (+) 1M7 reactivity profile followed by a normalization to the top 10% of values. 14. A representative result of a SHAPE-based analysis of the chaperone activity of the T. brucei editosome is summarized in Fig. 5. The analysis shows the side-by-side comparison of the SHAPE-derived 2D structures of CYb pre-mRNA in its preand post-chaperone folding state. i.e., in the absence and presence of bound editosomes. The figure details the entire workflow of the experiment starting from the two normalized SHAPE reactivity profiles to the derived 2D minimal free

SHAPE-Chemical Probing of RNA Chaperone-Mediated RNA Dynamic post-chaperone state, (+) editosomes

B

norm. SHAPE-reactivity (SU)

pre-chaperone state, (-) editosomes

A

175

2

0.8 0.35 0

100nt

C

D

norm. SHAPEreactivity

E

99.8% DMS (dimethyl sulfate). 2. 100% ethanol. 3. DMS dilutions (1:6, 1:8, 1:12) in 100% ethanol. Always prepare fresh. 4. Milli-Q water. 5. DMS quenching solution: 4.29 M β-mercaptoethanol, 0.3 M NaOAc pH 5.2. 6. Heating block or water bath set at 37  C.

DMS and CMCT Modifications to Determine the RNA Fold

2.3 Modification with CMCT

213

1. 100 mg/ml CMCT (1-cyclohexyl-3-(2-morpholinoethyl) carbodiimide metho-p-toluene sulfonate). 2. Milli-Q water. 3. CMCT dilutions (1:6, 1:8, 1:12) in water. Always prepare fresh. 4. 100% ethanol. 5. 3 M NaOAc, pH 5.2. 6. Heating block or water bath set at 37  C.

2.4 RNA Precipitation

1. Refrigerated centrifuge adapted for 1.5 ml microtubes. 2. 100% ethanol. 3. 75% ethanol. 4. Milli-Q water.

2.5 Primer Extension Analysis

1. 10 pmol/μl primer. 2. 10 U/μl T4 polynucleotide kinase. 3. 10 PNK buffer A. 4. [γ-32P]-ATP (3000 Ci/mmol). 5. Milli-Q water. 6. Heating blocks or water baths. 7. MicroSpin G-25 column. 8. 10 mM dNTPs. 9. 1 mM ddNTPs. 10. SuperScript III Reverse Transcriptase. 11. 5 SuperScript III Reverse Transcriptase buffer. 12. RNasin Ribonuclease Inhibitor. 13. 1 M DTT. 14. 2 N NaOH. 15. 2 N HCl. 16. 100% ethanol. 17. 75% ethanol. 18. 10 mg/ml glycogen. 19. Gel loading buffer: 8 M urea, 1 mM EDTA pH 8, 0.03% bromophenol blue.

2.6 Gel Fractionation and Exposure

1. Sequencing System (2400 V, 90 W). 2. Glass plates (45  20  0.32 cm). 3. Gel spacers and combs (0.4 mm). 4. High-voltage power supply.

214

Jose´ M. Andrade et al.

5. 40% acrylamide-bisacrylamide (19:1). 6. 10% (w/v) ammonium persulfate (freshly prepared). 7. TEMED. 8. 1 TBE buffer: 89 mM Tris–HCl, 80 mM boric acid, 2 mM EDTA pH 8. 9. Gel loading buffer: 8 M urea, 1 mM EDTA pH 8, 0.03% bromophenol blue. 10. Saran wrap. 11. PhosphorImager equipment.

3

Methods

3.1 General Considerations

1. It is important to work under RNase-free conditions to avoid RNA degradation. To minimize ribonuclease contamination, we always suggest wearing gloves, to bake glassware at 200  C and to use plastic disposable material. Solutions are made with autoclaved Milli-Q water, when appropriate. 2. DMS is highly toxic and corrosive. Consult the security guidelines followed at your institution regarding disposal of DMS residues.

3.2

Folding of RNA

1. Dilute 10 μg of total RNA (see Note 2) of each strain to be analyzed in 25 μl of the appropriate chemical reagent buffer (DMS buffer or CMCT buffer) without magnesium (see Note 3). 2. Denature the reaction mix for 3 min at 90  C in a heat block. 3. Switch off the heat block and let the sample cool down to room temperature for 10–15 min. 4. Spin down samples. 5. Add 0.25 μl 1 M MgCl2 to each sample (see Note 3). 6. Incubate for 30 min at 50  C to equilibrate and then for 60 min at 37  C (or other appropriate folding temperature) (see Note 4).

3.3 Chemical Modification

1. To 25 μl of the folded RNA, add 1 μl of freshly prepared DMS solution (diluted in EtOH) (see Notes 5 and 6).

3.3.1 Modification with DMS

2. Replace DMS with H2O (diluted in EtOH) when performing the negative control sample. 3. Incubate for 10 min at 37  C. 4. Add 475 μl of DMS quench solution to all samples (see Note 7). 5. Add 2.5 vol of 100% EtOH. Mix thoroughly. 6. Incubate overnight at 20  C.

DMS and CMCT Modifications to Determine the RNA Fold 3.3.2 Modification with CMCT

215

1. To 25 μl of the folded RNA, add 1 μl of freshly prepared CMCT solution (prepared in H2O) (see Notes 5 and 6). 2. Replace CMCT with H2O when performing the negative control sample. 3. Incubate for 10 min at 37  C. 4. Add 2.5 vol of EtOH and 0.3 vol of NaOAc. Mix thoroughly. 5. Incubate overnight at 20  C.

3.4 RNA Precipitation

1. Centrifuge sample at 15,000  g for 30 min at 4  C. 2. Wash pellets with 75% ethanol. 3. Centrifuge at 15,000  g for 15 min at 4  C. 4. Dry the RNA pellets at room temperature (avoid overdrying). 5. Resuspend each pellet in 16 μl Milli-Q water. 6. Quantify the RNA sample by Nanodrop measurement.

3.5 Primer Extension Analysis 3.5.1 Primer Labeling

1. Label the primer of choice (10 pmol) by phosphorylation at the 50 terminus in a mixture containing 2 μl of [γ-32P]-ATP, 1 μl of T4 polynucleotide kinase, and 2 μl of 10 T4 PNK buffer in a 20 μl total reaction. Incubate for 60 min at 37  C. 2. Add 20 μl Milli-Q H2O and separate the labeled primer from the unincorporated [γ-32P]-ATP by gel filtration on a MicroSpin G-25 column according to the manufacturer’s instructions.

3.5.2 Annealing

1. Mix 250 ng of total RNA chemically modified (or the untreated sample for control) with 0.25 pmol of the 50 -end labeled primer and 1 μl 10 mM dNTPs in a 13.5 μl total reaction. Perform a similar reaction without adding RNA, to later use as control in gel fractionation (see Subheading 3.6). 2. Incubate the mixture for 1 min at 90  C. 3. Let the mixture cool down to room temperature (see Note 8). 4. Save the mixture on ice.

3.5.3 Primer Extension

1. Add 4 μl of 5 SuperScript III Reverse Transcriptase buffer, 0.5 μl of SuperScript III Reverse Transcriptase (200 μ/μl), 1 μl of RNasin, and 1 μl 100 mM DTT to the 13.5 μl reaction mixture (see step 4 in Subheading 3.5.2) which performs a 20 μl total reaction. 2. Incubate for 30 min at 55  C for the elongation step by the MMLV SuperScript III RT (see Note 9). 3. Incubate for 10 min at 80  C to inactivate the SuperScript III RT and stop the reaction.

216

Jose´ M. Andrade et al.

3.5.4 RNA Removal

1. Add 6.7 μl of 2 N NaOH to the reaction mixture (see step 3 from Subheading 3.5.3) (see Note 10). 2. Incubate for 3 min at 95  C. 3. Allow to cool down to room temperature. 4. Add 3.3 μl of 2 N HCl.

3.5.5 RNA Precipitation

1. Add 75 μl of 100% ethanol and 1 μg of glycogen (see Note 11). 2. Incubate overnight at 20  C. 3. Centrifuge at 15,000  g for 15 min at 4  C. 4. Wash pellets with 75% ethanol. 5. Centrifuge at 15,000  g for 5 min at 4  C. 6. Resuspend each pellet in 4 μl Milli-Q water and 8 μl gel loading buffer. Store on ice.

3.5.6 Sequencing Ladder Preparation

1. Prepare four sequencing tubes, labeled A, C, G, and T (see Note 12). 2. To each tube, mix 250 ng of total RNA sample (not treated with chemical probes) with 0.25 pmol of 50 -end [γ-32P]-ATP labeled primer (~50,000–100,000 cpm) and 3.5 μl Tris 20 mM pH 8, in a total volume of 7 μl. 3. Incubate the mixture for 2 min at 95  C. 4. Quickly chill on ice. 5. Add 5 μl of the appropriate ddNTP (ddTTP, ddGTP, ddCTP, and ddATP) at 1 mM concentration to the corresponding sequencing tube. 6. Add 1 μl of 10 mM dNTPs, 4 μl of 5 SuperScript III RT buffer, 1 μl of SuperScript III RT (200 μ/μl), 1 μl of RNasin, and 1 μl 100 mM DTT in total volume of 20 μl. 7. Mix gently and briefly spin the samples. 8. Incubate for 30 min at 55  C and then incubate for 10 min at 80  C. 9. Proceed as mentioned above (see Subheading 3.5.4). 10. Proceed as mentioned above (see Subheading 3.5.5).

3.6 Gel Fractionation, Exposure, and Analysis 3.6.1 Gel Fractionation

1. Pour an 8% polyacrylamide and 8 M urea sequencing gel in 1 TBE buffer (see Note 13). 2. Pre-run the gel at 1500 V for 1–2 h (see Note 14). 3. Denature the samples (the synthesized cDNAs, including controls, and the sequencing ladder reactions) for 3 min at 95  C. 4. Quickly chill on ice. 5. Load 3 μl of each sample (~10,000 cpm) on the pre-warmed sequencing gel.

DMS and CMCT Modifications to Determine the RNA Fold

217

6. Run the gel at 2000 V until the bromophenol blue is 5–10 cm above the end of the gel (see Note 15). 3.6.2 Gel Exposure

1. Disassemble the gel. Remove one of the glasses, allowing the gel to rest on the other glass (see Note 16). 2. Cover the gel with Saran wrap and remove any air bubbles from the interface between the gel and the plastic (see Note 17). 3. Apply on top of the gel a phosphorimager screen and let exposure to proceed overnight, typically for 12–16 h, at room temperature. Be cautious to work safely, protecting yourself and others from radioactivity exposure.

3.6.3 Gel Analysis

DMS and CMCT provide information on the pairing status of the nucleotides, as these chemical probes react preferably to unpaired (single-stranded) nucleotides rather than paired (double-stranded) nucleotides. The main steps involved in the RNA structure analysis by DMS and CMCT are shown in Fig. 1. Some points must be taken into consideration when performing the RNA structural mapping: 1. Reaction of DMS or CMCT with nucleotides results in bands which are not present in the unmodified sample reactions. The intensities of these bands are correlated with the accessibility of the nucleotides to the chemical reagents: higher band intensities correspond to exposed nucleotides (which are singlestranded), while lower band intensities indicate a minor accessibility of the chemicals to those nucleotides (which preferably are double-stranded). 2. Controls include cDNA reactions from unmodified RNA samples, which allows identification of bands corresponding to spontaneous reverse transcriptase stops. Such bands should have equivalent intensity between samples analyzed in the same conditions, as this indicate a similar activity of the RT on the different samples. An additional control is performed without RNA template but with the primer, and it is used to identify bands due to possible primer multimers. Altogether, bands arising from control samples correspond to background and should not be considered when performing the mapping reading. 3. Dideoxy sequencing lanes from untreated RNA samples are fractionated in parallel with the sample reactions and are used as reference to identify the position of nucleotide modifications within the RNA chain. 4. Note that the bands from the chemically modified samples run faster by one nucleotide on the gel than the unmodified bands. This is because RT is blocked one position upstream from the modified position. Therefore, modified nucleotides appear one position lower than the corresponding bands in sequencing lanes.

218

Jose´ M. Andrade et al.

Fig. 2 DMS and CMCT accessibility probing of the 16S rRNA. (a) Reverse-transcribed cDNA was fractionated on a 10% polyacrylamide/7 M urea gel. The reverse transcriptase stops, and the primer multimer bands are indicated by arrows. The position of some nucleotides is given on the right of each panel (b). Residues with altered reactivities in the Δhfq mutant are indicated. The inset depicts the analyzed region of the 50 -end of the 16S rRNA. (c) A scheme of the overall secondary structure of the 50 -end of the 16S rRNA from RNA isolated from WT and Δhfq is shown. Most nucleotides in the WT are unpaired and thus accessible to the chemical reagents; on the other hand, the minor reactivities found in the Δhfq indicate that the same nucleotides are preferably found in a double-stranded conformation. This figure is adapted from [14]

DMS and CMCT Modifications to Determine the RNA Fold

219

5. When performing the reading analysis, remember that lower bands migrate faster and are closer to the primer sequence while upper bands migrate slower as they are longer (because they are extended toward the 50 -end of the RNA). 6. Estimating differences in band intensities between strains and/or conditions can be done by eye, but it might be a difficult and time-consuming task, particularly when differences are not obvious. This requires optimization and repetition to validate the results [15]. However, the intensity of the bands can be quantified using image-processing tools, such as the software Semi-Automated Footprinting Analysis (SAFA) [16]. This is particularly useful when large sequences are analyzed. A practical example of this technique is illustrated in Fig. 2, where the 50 -end of 16S rRNA from the wild-type strain and its isogenic Δhfq mutant was analyzed by DMS and CMCT chemical probing. Both probes reacted much better with nucleotides from RNA isolated from the WT, originating bands of higher intensity when compared to the mutant Δhfq sample (Fig. 2a, b). These results indicate that the structure of the 50 -end of the 16S rRNA is different between these two strains. The less reactive nucleotides found in the Δhfq mutant suggest that these nucleotides are mostly found in double-stranded regions and thus adopt a closed conformation that prevents reaction with the chemical probes (Fig. 2c). We have used this method to study the effect of the RNA chaperone Hfq on the structure of the 16S rRNA [14]. This strategy may, however, be applied for the study of many other RNA-binding proteins and RNAs.

4

Notes 1. Total RNA was isolated from exponential phase cultures (OD600 ~ 0.35–0.40) of the wild-type strain and the mutant of interest. However, this condition is variable and may be adjusted to your experimental needs. For example, if it is known that the levels of the RNA chaperone of interest are induced at a different stage (e.g., stationary phase), this is probably the best condition to start your analysis. RNA extraction was performed with phenol-chloroform following standardized protocols (see, e.g., [17]). Otherwise, TRIzol or column-based kits for RNA isolation can be used following the manufacturer’s instructions. After a precipitation step in ethanol and 300 mM sodium acetate, RNA is resuspended in MilliQ-water. The integrity of RNA samples was evaluated by agarose gel electrophoresis. When necessary, DNase treatment following a new phenol-chloroform step is used to remove

220

Jose´ M. Andrade et al.

contaminant DNA. Total RNA is suspended in Milli-Q water and quantified spectrophotometrically. RNA samples are kept at 80  C until use. 2. The amount of sample RNA to use in this experiment is variable, depending on the number of reactions to perform after chemical-induced modification and on the relative amount of the substrate of interest, in the case of using total RNA. To analyze the 16S rRNA, we used 10 μg of total RNA for chemical challenge with DMS and CMCT. When using a purified in vitro transcribed RNA (e.g., when performing footprinting experiments), we recommend starting with 1 μM, instead. 3. Magnesium is often required for the correct folding of RNA. However, RNA can be degraded when heated in the presence of divalent ions (like Mg2+). To avoid this, denaturation of the RNA is performed in buffers without magnesium. 4. Adjust the folding temperature according to your experimental conditions. Typically, a 25  C or 37  C is used. 5. Modification reactions are typically performed under limiting chemical reagent concentrations, to label only a fraction of the RNAs in a sample. Ideally, one modification per RNA molecule should be obtained, to lower the chance that the first modification causes an RNA structural rearrangement that alters access for the following modifications. This depends on the substrate and should be empirically determined by testing different dilutions of the chemical probe in the appropriate solvent. The best dilution is a balance between having a low level of modification per molecule which is at the same time high enough to produce readily detected signals. We suggest starting this analysis by preparing the following dilutions from stock solutions: 1:6, 1:8, and 1:12. These will be further diluted (1:25) in the experimental reaction (see step 1 in Subheadings 3.3.1 and 3.3.2). 6. Avoid buffers containing amines (such as Tris), as these can often react efficiently with the probing reagents. HEPES or cacodylate buffers are generally preferable. 7. To effectively stop the DMS modification reaction on RNA, we perform a quenching reaction with β-mercaptoethanol and NaOAc (following the protocol described in [11]). We found that the use of a quenching solution increased reproducibility. However, we note that other published methods do not include a quench step and rely on the termination of the DMS reaction upon ethanol precipitation. 8. Alternatively, the annealing step can be performed in a PCR machine using the gradient tool, lowering the temperature from 65 to 40  C for 30 min.

DMS and CMCT Modifications to Determine the RNA Fold

221

9. Different reverse transcriptases are commercially available and may be suitable for use in this protocol. We routinely use SuperScript III enzyme, as this enzyme performs well on highly structured RNAs. Depending on your substrate, you may consider to select an even more robust enzyme, able to perform at higher temperatures, like the SuperScript IV enzyme that has been engineered for higher thermostability, processivity, and cDNA yields. 10. Elimination of the template RNA from the synthesized cDNAs improves visualization of the products. The alkaline treatment with NaOH followed by neutralization with HCl destroys the RNA but not the cDNAs. Alternatively, RNA template can be digested with RNase H. However, it is advisable to perform a phenol-chloroform extraction of the reaction mixture after digestion. 11. Following ethanol precipitation, cDNA pellets are not very visible. Place tubes in the centrifuge rotor so that the tabs face out, to never lose the orientation. The addition of glycogen improves nucleic acid precipitation and helps to better visualize the pellet. 12. A sequencing ladder is fractionated in parallel with the sample reactions in order to identify the modifications on the RNA. Note that the bands from the chemically modified samples run one position lower on the gel than the corresponding bands in sequencing lanes. This is because RT is blocked one position upstream from the modified position [11]. 13. The resolution range of a sequencing gel is correlated to the concentration of polyacrylamide used. Higher concentrations of polyacrylamide result in a gel mesh of reduced-sized pores that increases separation resolution for smaller fragments but greatly slows migration. In an 8% polyacrylamide gel, a good resolution is generally obtained for reading up to 100 nts fragments. If the RNA template under analysis is much longer or if the run did not provide a good resolution, the use of different primers should be considered. 14. Pre-run prevents hyperfocusing and distortion of the gel and should not be omitted. 15. The bromophenol blue present in the gel loading buffer might indirectly inform on the separation of the run. In an 8% polyacrylamide gel, the bromophenol blue co-migrates with a 26 nts primer. 16. It is very important to use clean glass plates, and this usually is enough to make the gel stick to only one of the plates during disassembling. We usually do not treat the glass with a silanization agent. However, if you plan to dry the gel after running is complete, you may treat one of the glasses with a silicone-based

222

Jose´ M. Andrade et al.

product that will increase its hydrophobicity; this will prevent adsorption of the gel to the treated glass surface and make disassemble of the gel easier. 17. Alternatively, the gel can be transferred from the glass to a 3MM paper and covered with Saran wrap. The gel is then dried under vacuum at 80  C for 30 min and exposed to X-ray film or in a PhosphorImager cassette. We found that for an overnight exposure, it is not critical to dry the gels. However, if the radioactivity signal is too weak or if a longer exposure is planned, it would be advisable to dry the gel as this eliminates sample diffusion and increases the sharpness of bands.

Acknowledgments This work was financially supported by Project LISBOA-01-0145FEDER-007660 (Microbiologia Molecular, Estrutural e Celular) funded by FEDER through COMPETE2020-Programa Operacional Competitividade e Internacionalizac¸˜ao (POCI) and by FCT-Fundac¸˜ao para a Cieˆncia e a Tecnologia (Portugal), including Program IF (IF/00961/2014) and Grants PTDC/BIA-MIC/ 32525/2017 to J.M.A. and PTDC/BIA-MIC/1399/2014 to CMA; R.F.dS. is recipient of an FCT Doctoral fellowship (PD/BD/105733/2014). We also acknowledge the European Union Horizon 2020 Research and Innovation Programme grant agreement no. 635536 to CMA. References 1. dos Santos RF, Quendera AP, Boavida S et al (2018) Major 30 –50 exoribonucleases in the metabolism of coding and non-coding RNA. In: Teplow DB (ed) Progress in molecular biology and translational science. Academic Press, Cambridge, pp 101–155 2. Weeks KM (2010) Advances in RNA structure analysis by chemical probing. Curr Opin Struct Biol 20:295–304 3. Andrade JM, Pobre V, Arraiano CM (2013) Small RNA modules confer different stabilities and interact differently with multiple targets. PLoS One 8:e52866 4. Brunel C, Romby P (2000) Probing RNA structure and RNA-ligand complexes with chemical probes. Methods Enzymol 318:3–21 5. Ehresmann C, Baudin F, Mougel M et al (1987) Probing the structure of RNAs in solution. Nucleic Acids Res 15:9109–9128

6. Marryman C, Noller HF (1998) Footprinting and modification-interference analysis of bindind sites on RNA. In: Smith CWJ (ed) RNA-protein interactions: a practical approach. Oxford University Press, Oxford, pp 237–254 7. Ziehler WA, Engelke DR (2001) Probing RNA structure with chemical reagents and enzymes. Curr Protoc Nucleic Acid Chem Chapter 6: Unit 6.1 8. Sachsenmaier N, Handl S, Debeljak F et al (2014) Mapping RNA structure in vitro using nucleobase-specific probes. Methods Mol Biol 1086:79–94 9. Philippe J-V, Ayadi L, Branlant C et al (2015) Probing small non-coding RNAs structures. In: Rederstorff M (ed) Small non-coding RNAs. Methods in molecular biology. Humana Press, Totowa, pp 119–136

DMS and CMCT Modifications to Determine the RNA Fold 10. Behm-Ansmant I, Helm M, Motorin Y (2011) Use of specific chemical reagents for detection of modified nucleotides in RNA. J Nucleic Acids 2011:1–17 11. Tijerina P, Mohr S, Russell R (2007) DMS footprinting of structured RNAs and RNA–protein complexes. Nat Protoc 2:2608–2623 12. Caprara M (2011) RNA structure determination using chemical and nuclease digestion methods. In: Rio D, Hannon G, Ares M et al (eds) RNA: a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, pp 269–275 13. Yu AM, Evans ME, Lucks JB (2018) Estimating RNA structure chemical probing reactivities from reverse transcriptase stops and mutations. bioRXiv. https://doi.org/10. 1101/292532

223

14. Andrade JM, Dos Santos RF, Chelysheva I et al (2018) The RNA-binding protein Hfq is important for ribosome biogenesis and affects translation fidelity. EMBO J 37:e97631 15. Wan Y, Kertesz M, Spitale RC et al (2011) Understanding the transcriptome through RNA structure. Nat Rev Genet 12:641–655 16. Das R, Laederach A, Pearlman SM et al (2005) SAFA: semi-automated footprinting analysis software for high-throughput quantification of nucleic acid footprinting experiments. RNA 11:344–354 17. Andrade JM, Pobre V, Matos AM et al (2012) The crucial role of PNPase in the degradation of small RNAs that are not associated with Hfq. RNA 18:844–855

Chapter 14 Disordered RNA-Binding Region Prediction with DisoRDPbind Christopher J. Oldfield, Zhenling Peng, and Lukasz Kurgan Abstract RNA chaperone activity is one of the many functions of intrinsically disordered regions (IDRs). IDRs function without the prerequisite of a stable structure. Instead, their functions arise from structural ensembles. A common theme in IDR function is molecular recognition; IDRs mediate interactions with other proteins, RNA, and DNA. Many computational methods are available to predict IDRs from protein sequence, but relatively few are available for predicting IDR functions. Available methods primarily focus on protein-protein interactions. DisoRDPbind was developed to predict several protein functions including interactions with RNA. This method is available as a user-friendly web interface, located at http://biomine. cs.vcu.edu/servers/DisoRDPbind/. The development and architecture of DisoRDPbind is briefly presented, and its accuracy relative to other RNA-binding residue predictors is discussed. We explain usage of the web interface in detail and provide an example of prediction results and interpretation. While DisoRDPbind does not identify RNA chaperones directly, we provide a case study of an RNA chaperone, HCV core protein, as an example of the method’s utility in the study of RNA chaperones. Key words Intrinsic disorder, Protein-RNA interactions, Intrinsically disordered regions, Molecular recognition

1

Introduction RNA chaperone activity is one of the many functions of intrinsically disordered proteins (IDPs) and intrinsically disordered regions (IDRs) [1]. The sequences of IDPs and IDRs are self-insufficient to form a stably folded structure in isolation and instead exist as structural ensembles, where structures vary over time and over population [2–4]. Available sequence analysis methods accurately predict the locations of IDRs from protein sequence [5–15]. These methods estimate that IDPs and IDRs are prevalent in proteomes [16–23], particularly in eukaryotes where 25–40% of their proteins contain significant IDRs [16, 21]. Despite the lack of stable structure, IDPs and IDRs perform many and varied biological functions [21, 24–26]. Among their functions, IDPs play an important role

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_14, © Springer Science+Business Media, LLC, part of Springer Nature 2020

225

226

Christopher J. Oldfield et al.

in mediating molecular interactions by binding to proteins and nucleic acids [21, 27–37]. In particular, many IDPs are known to function as RNA chaperones [34, 38]. While the potential for IDRs to recognize RNA is well known, predicting novel RNA-binding IDPs from sequence using many of the existing tools [39–47] is problematic; the majority of these tools have been developed for structured proteins interacting with RNA. While no specific IDP chaperone prediction method is available to date, many studies have demonstrated that IDP function can be predicted from protein sequence [8, 15, 48–54]. Until recently, most studies focused on interactions between IDPs and other proteins, and several methods are available for prediction of protein recognition regions within IDRs [8]. Less attention has been paid to other types of IDP functions, including interaction with DNA and RNA. To fill this gap, we developed DisoRDPbind, a method that simultaneously predicts protein, DNA, and RNA interacting regions within IDRs [55, 56]. This method predicts each type of function separately, allowing identification of the type of interaction partner for each predicted region. Additionally, DisoRDPbind predicts interactions at residue-level resolution, allowing identification of the protein regions responsible for each prediction type. Performance of the method is significantly better than other available methods for RNA interaction region predicted when applied to IDPs [55]. While not specific to RNA chaperones, DisoRDPbind is a useful tool for identification of novel intrinsically disordered RNA chaperones when combined with additional function information. As an example, we examine the hepatitis C virus (HCV) core protein. This protein is multifunctional, playing a structural role in capsid formation and RNA organization [57], as well as serving as an RNA chaperone [58, 59]. The RNA chaperone activity is located in the intrinsically disordered N-terminal region of HCV core protein [58] (Fig. 1). The main idea behind methods such as DisoRDPbind is to predict these disordered RNA-binding regions directly from the amino acid sequence. We demonstrate that the residue-level DisoRDPbind predictions correctly identify residues in this region of the input protein chain as intrinsically disordered and RNA binding (see Case Study). The DisoRDPbind method is publicly available as a userfriendly webserver. Further, it was designed as a high-throughput method that can predict entire proteomes in a matter of hours. Here we briefly review the DisoRDPbind method and provide detailed instructions on usage of its web interface. Finally, we discuss the case study of DisoRDPbind applied to an HCV core protein, an intrinsically disordered RNA chaperone.

Prediction of Disordered RNA-Binding Regions

227

Fig. 1 Sequence of HCV core protein and annotated IDR, RNA chaperone region, and RNA-binding region

2

Materials DisoRDPbind infers the function of novel sequences through the use of historical protein functional data. Available historical data were broken into two sequence dissimilar subsets: the training dataset and the testing dataset. The training dataset is used to build the models DisoRDPbind uses to infer function in novel sequences. The testing data is used to assess the performance of the final model on novel data. Estimation of model accuracy aids in interpreting prediction results for novel proteins; it informs how many correct and incorrect predictions are typically expected for an average protein sequence.

2.1

Datasets

Training and testing datasets for development of DisoRDPbind [55] were extracted from the DisProt database [60], a database of IDPs including function annotations. IDPs from DisProt were clustered at 30% sequence identity, and clusters were assigned to either the training or testing set. This procedure is intended to avoid overestimation of method performance by ensuring that orthologous proteins do not appear in both the training and testing sets. It also demonstrates whether DisoRDPbind can make correct predictions in the absence of sequence similarity, i.e., when sequence alignment typically would not produce accurate results. The training dataset included 315 proteins. Two test sets were used, one with 114 proteins and another of 36 proteins that consisted of only recent additions to the DisProt database. These datasets are available at http://biomine.ece.ualberta.ca/Dis oRDPbind/.

2.2

Architecture

The architecture of DisoRDPbind was designed to allow for highthroughput predictions, enabling practical whole proteome IDP function prediction. The architecture has five primary stages (Fig. 2), where here we focus on the RNA-binding prediction portion of DisoRDPbind; see Peng et al. for full architecture details [55]. Stage 1 develops a profile representation of the protein sequence that covers several relevant sequence properties. Stage

228

Christopher J. Oldfield et al.

Fig. 2 Partial architecture of DisoRDPbind, focused on the RNA-binding portion of method. Each stage referenced in the text is indicated with a number in a yellow circle

2 extracts a processed set of features that are computed from the profile. These features numerically quantify information that is relevant for prediction, where independent feature sets are selected for each type of functional region. Stage 3 uses these features as input to a trained logistic regression model to produce modelbased predictions. Stage 4 is done in parallel to stages 1–3, where the input sequence is compared to a database of proteins with annotated RNA-binding function. If a sufficiently similar sequence is found in the database, annotations are transferred to the input protein at aligned positions in the sequence. Finally, stage 5 merges model-based predictions with similarity-based predictions to give the final DisoRDPbind prediction. The feature representation of each sequence (Fig. 2, stage 1) involves several calculated features. Intrinsic disorder predictions are made using the IUPred algorithm [5], and low-complexity regions—a sequence property correlated with IDRs—are identified using the SEG algorithm [61]. Additional features include residue compositions, secondary structure predictions made with the PSIPRED algorithm [62], and 17 selected amino acid scales from

Prediction of Disordered RNA-Binding Regions

229

the AAIndex database [63]. Amino acid scales quantify physiochemical properties such as hydrophobicity, net charge, and folding free energy. Sliding windows were applied to average each input features, which transforms residue predictions into local sequence averages, where a window size of 55 residues was used for RNA-binding prediction. Windowed averages are used to make predictions for the center residue of the window. Empirical feature selection (Fig. 2, stage 2) was used to remove uninformative and redundant features prior to model training or prediction. A separate set of features was selected empirically for each function prediction method. A small set of 11 features was found to give good results for RNA-binding prediction. Each selected feature set is used in a separate logistic regression model for each function type (Fig. 2, stage 3). Logistic regression models are robust to overfitting, are extremely fast, and provide a propensity in the range of 0–1 for each residue in a protein. The overall method provides three separate propensity scores, one of which indicates propensity of a residue to be intrinsically disordered and interact with RNA. The final stage of DisoRDPbind merges these predicted propensities with functional annotations found through sequence similarity with the training dataset (Fig. 2, stage 4). Input sequences are compared with the training dataset using BLAST [64]. The alignments produced by BLAST are used to transfer functional annotations from training set protein sequences to input protein sequences. DisoRDPbind output consists of the RNA-, DNA-, and protein-binding propensity scores, as well as binary classification of each residue as RNA, DNA, and protein binding based on a model specific threshold. The thresholds were selected to produce predictions with a low (10%) false-positive rate on the training dataset [55]. Residues with propensity scores that exceed the model-specific threshold are then classified as either RNA, DNA, or protein binding. Greater propensity scores are indicative of a higher likelihood of binding to a particular molecule type (see Note 1). 2.3 Predictive Quality and Runtime

Prediction quality of DisoRDPbind for RNA-binding residues of IDPs is significantly better than other computational predictors of RNA-binding residues that were not specifically designed for IDPs, including BindN+ [40] and RNABindR [45]. The area-under-thecurve (AUC) metric is a threshold agnostic measure of predictive performance. DisoRDPbind RNA-binding predictions produced AUC values around 0.67, depending on the specific test set used, which was significantly better than other methods tested, with AUC values between 0.54 and 0.64 [55]. The other methods tested were developed from structured RNA-binding proteins, whereas DisoRDPbind was developed form intrinsically disordered proteins, which suggests that these predictions may be complementary. A comparison of these methods indicates that this is in fact the

230

Christopher J. Oldfield et al.

case; DisoRDPbind is poorly correlated with other methods of RNA-binding prediction with a correlation coefficient less than 0.3. This demonstrates that DisoRDPbind is not only accurate but also complementary to existing RNA-binding prediction methods. This was also recently confirmed in a study of putative RNA-binding protein in the human proteome [37]. The runtime of DisoRDPbind increases quadratically with protein length (see Note 2), ranging between a fraction of a second to several seconds per protein on a modern computer system [55]. This runtime includes predictions of DNA, RNA, and protein interactions. This efficient runtime performance makes proteome scale predictions practical. For example, predictions for the entire human proteome can be obtained in around 40 h on a modern computer system. 2.4

3

Webserver

The user-friendly web interface for DisoRDPbind can be accessed at http://biomine.ece.ualberta.ca/DisoRDPbind/. The only system requirements for submitting sequences for prediction are an Internet connection and a modern web browser. The interface has been tested with Firefox, Internet Explorer, and Chrome. Prediction submissions are made at the main page for DisoRDPbind. This page will accept up to 5000 protein sequences in FASTA format, submitted as either a file upload or with a text entry field. Notifications of completed predictions are provided by email, so an email address is required for submission of sequences for prediction. Notifications provide a link to prediction results and an explanation of result file format. Once sequences are submitted, the webserver runs all programs necessary to make DisoRDPbind predictions. Disorder predictions are made with IUPred [5], secondary structure predictions are made with PSIPRED (in single sequence mode) [62], low complexity regions are identified with the SEG method [61], and annotation transfer is made with BLAST [64]. There are no required options for running DisoRDPbind; simply supply sequences for prediction, enter an email address, and click the “Run DisRDPbind” button. The webserver will then run all required programs and send a notification to the supplied email address when predictions are completed.

Methods

3.1 Running DisoRDPbind

There are three steps to submit sequences for prediction to the DisoRDPbind server (Fig. 3, labels 1a/b, 2, and 3): 1. Provide FASTA formatted sequences (see Note 3) for prediction using 1a or b, depending on the desired submission method. Clicking the “Reset sequence(s)” button below 1b

Prediction of Disordered RNA-Binding Regions

231

Fig. 3 The DisoRDPbind prediction submission form. Red numbers indicate the three necessary steps to submit sequences for predictions, discussed in the text

will clear both submission options. There are limits to both the number of sequences (see Note 4) and maximum length of sequences (see Note 5) submitted for prediction. (a) Upload a file of FASTA formatted sequences. (b) Provide FASTA formatted sequences as text. This can be done using the copy and paste function of your operating system; copy from a local file and paste to the text field. For an example of properly formatted sequences, click the “Example” button located below the text field. 2. Provide an email address (see Note 6). This email address is only used to send notification of completed prediction results; you will receive only one notification email per submission. 3. Click “Run DisoRDPbind” to submit sequences and run predictions. Clicking “Run DisoRDPbind” submit will take the user to a status page, reporting on the current state of the submitted prediction. Submissions are entered into the webservers queue system, and the status page will report the current position in the queue and when predictions on the submission have begun. After predictions have completed, the status page will redirect to the prediction results page, and an email will be sent to the notification email address provided. There is no need to keep the status page open

232

Christopher J. Oldfield et al.

Fig. 4 The DisoRDPbind prediction results page. Red numbers indicate important features of this page, discussed in the text

while predictions are pending; a notification email is always sent on prediction completion. 3.2 Results Generated by DisoRDPbind

The results page can be reached by leaving a browser open to the status page or following the link provided in the results email. The results page includes a link to a text file “results.txt” with prediction results (Fig. 4, label 1) and a description of the result file format (Fig. 4, label 2). The result file contains RNA, DNA, and protein interaction prediction results for each of the submitted protein sequences. Prediction results include both interaction propensities, ranging from 0 for low propensity to 1 for high propensity, and binary interaction predictions (see Note 7), 0 for noninteracting and 1 for interacting, for each of the three interaction types. Each sequence is represented by eight lines in the results file, where the first four are relevant for RNA interaction prediction: 1. The protein name taken from the FASTA header of each sequence. 2. The protein sequence with interacting residues encoded with character case; lower case residues are predicted to be in intrinsically disordered regions that interact with DNA, RNA, and/or protein, and upper case residues are predicted not to be in intrinsically disordered regions or to not interact with DNA, RNA, or protein.

Prediction of Disordered RNA-Binding Regions

233

Fig. 5 The DisoRDPbind notification email. The email provides links, indicated with red numbers, to prediction results, discussed in the text

3. RNA interaction binary predictions, either 1 for interaction or 0 for no interaction. 4. RNA interaction propensity, ranging between 1 for high interaction propensity and 0 for low interaction propensity. When prediction results have been completed by the server, a notification email (Fig. 5) is sent to the email address provided during sequence submission. The email notification contains a link to the results page (Fig. 5, label 1) and a direct link to the results text file (Fig. 5, label 2). Each job has a unique numerical identifier (Fig. 5, “XXXXXXXXXXXXXX”) that is given at the top of the notification email and is used in links to identify each submission (see Note 8). In the case of issues with a prediction submission, the prediction identification number is used to trace the corresponding submission. 3.3

Case Study

Flaviviridae are a group of single-stranded RNA, enveloped viruses that infect mammals, including humans. This group includes HCV, which chronically infects up to 130–170 million people worldwide resulting in over 350,000 deaths annually [65]. The core protein of Flaviviridae serves as both a capsid protein and as an RNA chaperone [58]. The proteomes of Flaviviridae are encoded as a polyprotein, where the core protein is located at or near the extreme N-terminus and is released by proteolysis [57]. The core protein of HCV is characterized by a basic N-terminus and hydrophobic C-terminus. While the domain organization of Flaviviridae may differ, they are generally characterized by the presence of a basic region. These basic regions have been shown to be intrinsically disordered by circular dichroism and to carry RNA-binding activity. A shorter N-terminal region has been shown to be sufficient for RNA chaperone activity by base-pairing assays [58]. The

234

Christopher J. Oldfield et al.

Fig. 6 Annotated regions and prediction results for HCV core protein. Annotated regions are an IDR, known RNA binding region, and sufficient RNA chaperone region. DisoRDPbind’s RNA-binding prediction results are shown both as binary (red regions indicate predicted RNA-binding residues) and propensity per residue. The VSL2B’s prediction of intrinsic disorder is also shown as both a binary prediction (green regions indicate predicted IDRs) and a propensity score, where residues with values greater than 0.5 are predicted to be disordered

intrinsically disordered, RNA binding, and chaperone regions are shown at the top of Fig. 6. We note that this figure was created using a specialized graphical software package. DisoRDPbind’s users have access to the corresponding text-based output that is summarized in Fig. 4. Application of DisoRDPbind to HCV core protein demonstrates good agreement between the HCV core RNA-binding region and predicted RNA-binding residues (middle of the Fig. 6). RNA-binding propensity values for this protein range from 0.006 to 0.372, with nearly all of the highest scores located in the known RNA-binding region. These propensity values are used to obtain a binary prediction—either RNA binding or non-RNA binding—for each residue in the protein by application of a threshold of 0.151. This threshold was selected to balance identification of novel binding residues against spurious predictions (see Note 7). DisoRDPbind predicts residues throughout the known RNA-binding region to interact with RNA. The predicted residues include nearly all of the region known to have RNA chaperone function. Residues not predicted to be RNA binding

Prediction of Disordered RNA-Binding Regions

235

by DisoRDPbind are primarily located at the extremes of the defined RNA-binding region. This suggests the hypothesis that the necessary and sufficient RNA-binding region could be the shorter region suggested by DisoRDPbind, which could be tested experimentally. Disorder predictions for HCV core protein performed with the VSL2b method [7] also agree well with the characterized disordered region (bottom of Fig. 6). Disordered propensity scores are converted to binary disorder predictions—either disorder or structured—for each residue of a protein in the same manner as DisoRDPbind, but with a threshold value of 0.5. While disorder predictions give an accurate estimation of the location of IDRs and structured regions of a protein, they do not carry a direct indication of protein function. For function prediction, specialized prediction methods, like DisoRDPbind, can be used to decompose IDRs into functional regions.

4

Notes 1. In the analysis of individual proteins, it may be useful to examine propensity scores in addition to binary predictions. Elevated propensity scores that do not exceed the prediction threshold (and consequently which do not result in the binary prediction of binding) may be indicative of function when combined with other data. The threshold was originally selected to ensure low (10%) false-positive rate on the training dataset, resulting in a conservative set of binary predictions of binding. Thus, high propensity scores suggest that the corresponding residues have elevated likelihood for binding; however, the user should expect higher levels of false positives among these predictions. 2. A formula for estimating the run time in milliseconds of DisoRDPbind for a given sequence was determined to be [55]: 0:007n2 þ 0:9028n þ 301:06 where n is the number of amino acids in the protein. For n ¼ 200, the estimate is 0.79 s, and for n ¼ 1000, the estimate is 8.9 s. Predictions for proteins for each webserver submission are run serially, so applying the above formula to each sequence in the submission and taking the sum will provide a run time estimate. 3. The FASTA format is described at https://en.wikipedia.org/ wiki/FASTA_format. Briefly, the format consists of a series of sequence label lines, beginning with “>,” followed by the sequence beginning on the next line.

236

Christopher J. Oldfield et al.

4. Up to 5000 FASTA formatted sequences can be submitted at one time to the web interface. Submission sizes exceeding this limit will result in an error notification from the server, and no predictions will be run by the server. For submission of more than 5000 sequences, it will be necessary to break the sequences into multiple submissions each with 5000 or fewer sequences. 5. The programs used to generate predictor inputs limit the maximum length of protein sequences submitted to the webserver. Submitted sequences should be limited to fewer than 10,000 residues. 6. Although single sequence predictions can be made in as little as a fraction of a second, prediction of 5000 sequences will typically require several hours. Rather than requiring an active browser connection, notification of completed predictions is provided via email. The email message will contain instruction on how to access prediction results. 7. Binary predictions are directly related to propensity scores; propensities greater than the predictor-specific threshold are classified as interacting (binary value of 1), and propensities less than the same threshold are classified as noninteracting (binary value of 0). For RNA-binding prediction, the threshold is set at a propensity of 0.151. This threshold was selected to give a 10% false-positive rate on the training dataset. 8. Please save the result notification email or the included links. Predictions will be accessible via these links for at least 3 months after prediction. It is also recommended to save the status page URL, which can be used in the case of a typo in notification email address resulting in no notification email receipt.

Acknowledgments This research was supported in part by the Robert J. Mattauch Endowment funds and the National Science Foundation grant 1617369 to Lukasz Kurgan. References 1. van der Lee R, Buljan M, Lang B, Weatheritt RJ, Daughdrill GW, Dunker AK, Fuxreiter M, Gough J, Gsponer J, Jones DT, Kim PM, Kriwacki RW, Oldfield CJ, Pappu RV, Tompa P, Uversky VN, Wright PE, Babu MM (2014) Classification of intrinsically disordered regions and proteins. Chem Rev 114(13):6589–6631

2. Dunker AK, Obradovic Z (2001) The protein trinity–linking function and disorder. Nat Biotechnol 19(9):805–806 3. Wright PE, Dyson HJ (1999) Intrinsically unstructured proteins: re-assessing the protein structure-function paradigm. J Mol Biol 293 (2):321–331

Prediction of Disordered RNA-Binding Regions 4. Uversky VN, Gillespie JR, Fink AL (2000) Why are “natively unfolded” proteins unstructured under physiologic conditions? Proteins 41(3):415–427 5. Dosztanyi Z, Csizmok V, Tompa P, Simon I (2005) The pairwise energy content estimated from amino acid composition discriminates between folded and intrinsically unstructured proteins. J Mol Biol 347(4):827–839 6. Walsh I, Martin AJ, Di Domenico T, Tosatto SC (2012) ESpritz: accurate and fast prediction of protein disorder. Bioinformatics 28 (4):503–509 7. Peng K, Radivojac P, Vucetic S, Dunker AK, Obradovic Z (2006) Length-dependent prediction of protein intrinsic disorder. BMC Bioinformatics 7:208 8. Meng F, Uversky VN, Kurgan L (2017) Comprehensive review of methods for prediction of intrinsic disorder and its molecular functions. Cell Mol Life Sci 74(17):3069–3090 9. Lieutaud P, Ferron F, Uversky AV, Kurgan L, Uversky VN, Longhi S (2016) How disordered is my protein and what is its disorder for? A guide through the “dark side” of the protein universe. Intrinsically Disord Proteins 4(1): e1259708 10. Monastyrskyy B, Kryshtafovych A, Moult J, Tramontano A, Fidelis K (2014) Assessment of protein disorder region predictions in CASP10. Proteins 82(Suppl 2):127–137 11. Necci M, Piovesan D, Dosztanyi Z, Tompa P, Tosatto SCE (2017) A comprehensive assessment of long intrinsic protein disorder from the DisProt database. Bioinformatics 34 (3):445–452 12. Fan X, Kurgan L (2014) Accurate prediction of disorder in protein chains with a comprehensive and empirically designed consensus. J Biomol Struct Dyn 32(3):448–464 13. Meng F, Uversky V, Kurgan L (2017) Computational prediction of intrinsic disorder in proteins. Curr Protoc Protein Sci 88:2 16 11–2 16 14 14. Mizianty MJ, Stach W, Chen K, Kedarisetti KD, Disfani FM, Kurgan L (2010) Improved sequence-based prediction of disordered regions with multilayer fusion of multiple information sources. Bioinformatics 26(18): i489–i496 15. Jones DT, Cozzetto D (2015) DISOPRED3: precise disordered region predictions with annotated protein-binding activity. Bioinformatics 31(6):857–863 16. Peng Z, Mizianty MJ, Kurgan L (2014) Genome-scale prediction of proteins with

237

long intrinsically disordered regions. Proteins 82(1):145–158 17. Xue B, Dunker AK, Uversky VN (2012) Orderly order in protein intrinsic disorder distribution: disorder in 3500 proteomes from viruses and the three domains of life. J Biomol Struct Dyn 30(2):137–149 18. Pancsa R, Tompa P (2012) Structural disorder in eukaryotes. PLoS One 7(4):e34687 19. Ward JJ, Sodhi JS, McGuffin LJ, Buxton BF, Jones DT (2004) Prediction and functional analysis of native disorder in proteins from the three kingdoms of life. J Mol Biol 337 (3):635–645 20. Tompa P (2012) Intrinsically disordered proteins: a 10-year recap. Trends Biochem Sci 37 (12):509–516 21. Peng Z, Yan J, Fan X, Mizianty MJ, Xue B, Wang K, Hu G, Uversky VN, Kurgan L (2015) Exceptionally abundant exceptions: comprehensive characterization of intrinsic disorder in all domains of life. Cell Mol Life Sci 72 (1):137–151 22. Hu G, Wang K, Song J, Uversky VN, Kurgan L (2018) Taxonomic landscape of the dark proteomes: whole-proteome scale interplay between structural darkness, intrinsic disorder, and crystallization propensity. Proteomics 18: e1800243 23. Yan J, Mizianty MJ, Filipow PL, Uversky VN, Kurgan L (2013) RAPID: fast and accurate sequence-based prediction of intrinsic disorder content on proteomic scale. Biochim Biophys Acta 1834(8):1671–1680 24. Dyson HJ, Wright PE (2005) Intrinsically unstructured proteins and their functions. Nat Rev Mol Cell Biol 6(3):197–208 25. Dunker AK, Brown CJ, Lawson JD, Iakoucheva LM, Obradovic Z (2002) Intrinsic disorder and protein function. Biochemistry 41 (21):6573–6582 26. Xie H, Vucetic S, Iakoucheva LM, Oldfield CJ, Dunker AK, Uversky VN, Obradovic Z (2007) Functional anthology of intrinsic disorder. 1. Biological processes and functions of proteins with long disordered regions. J Proteome Res 6(5):1882–1898 27. Chen JW, Romero P, Uversky VN, Dunker AK (2006) Conservation of intrinsic disorder in protein domains and families: II. Functions of conserved disorder. J Proteome Res 5 (4):888–898 28. Cumberworth A, Lamour G, Babu MM, Gsponer J (2013) Promiscuity as a functional trait: intrinsically disordered regions as central players of interactomes. Biochem J 454:361–369

238

Christopher J. Oldfield et al.

29. Dyson HJ (2012) Roles of intrinsic disorder in protein-nucleic acid interactions. Mol BioSyst 8(1):97–104 30. Fuxreiter M, Toth-Petroczy A, Kraut DA, Matouschek AT, Lim RYH, Xue B, Kurgan L, Uversky VN (2014) Disordered proteinaceous machines. Chem Rev 114(13):6806–6843 31. Haynes C, Oldfield CJ, Ji F, Klitgord N, Cusick ME, Radivojac P, Uversky VN, Vidal M, Iakoucheva LM (2006) Intrinsic disorder is a common feature of hub proteins from four eukaryotic interactomes. PLoS Comput Biol 2 (8):890–901 32. Peng Z, Oldfield CJ, Xue B, Mizianty MJ, Dunker AK, Kurgan L, Uversky VN (2014) A creature with a hundred waggly tails: intrinsically disordered proteins in the ribosome. Cell Mol Life Sci 71(8):1477–1504 33. Peng Z, Mizianty MJ, Xue B, Kurgan L, Uversky VN (2012) More than just tails: intrinsic disorder in histone proteins. Mol BioSyst 8 (7):1886–1901 34. Tompa P, Csermely P (2004) The role of structural disorder in the function of RNA and protein chaperones. FASEB J 18(11):1169–1175 35. Wu Z, Hu G, Yang J, Peng Z, Uversky VN, Kurgan L (2015) In various protein complexes, disordered protomers have large per-residue surface areas and area of protein-, DNA- and RNA-binding interfaces. FEBS Lett 589(19 Pt A):2561–2569 36. Wang C, Uversky VN, Kurgan L (2016) Disordered nucleiome: abundance of intrinsic disorder in the DNA- and RNA-binding proteins in 1121 species from Eukaryota, bacteria and Archaea. Proteomics 16(10):1486–1498 37. Chowdhury S, Zhang J, Kurgan L (2018) In silico prediction and validation of novel RNA binding proteins and residues in the human proteome. Proteomics 18:e1800064 38. Ivanyi-Nagy R, Davidovic L, Khandjian EW, Darlix J-L (2005) Disordered RNA chaperone proteins: from functions to disease. Cell Mol Life Sci 62(13):1409–1417 39. Liu ZP, Wu LY, Wang Y, Zhang XS, Chen LN (2010) Prediction of protein-RNA binding sites by a random forest method with combined features. Bioinformatics 26 (13):1616–1622 40. Wang L, Huang C, Yang MQ, Yang JY (2010) BindN+ for accurate prediction of DNA and RNA-binding residues from protein sequence features. BMC Syst Biol 4(1):S3 41. Walia RR, Xue LC, Wilkins K, El-Manzalawy Y, Dobbs D, Honavar V (2014) RNABindRPlus: a predictor that combines machine learning and sequence homology-based methods to

improve the reliability of predicted RNA-binding residues in proteins. PLoS One 9(5):e97725 42. Wang L, Brown SJ (2006) BindN: a web-based tool for efficient prediction of DNA and RNA binding sites in amino acid sequences. Nucleic Acids Res 34(Web Server):W243–W248 43. Kumar M, Gromiha MM, Raghava GP (2008) Prediction of RNA binding sites in a protein using SVM and PSSM profile. Proteins 71 (1):189–194 44. Yang X, Wang J, Sun J, Liu R (2015) SNBRFinder: a sequence-based hybrid algorithm for enhanced prediction of nucleic acidbinding residues. PLoS One 10(7):e0133260 45. Walia RR, Caragea C, Lewis BA, Towfic F, Terribilini M, El-Manzalawy Y, Dobbs D, Honavar V (2012) Protein-RNA interface residue prediction using machine learning: an assessment of the state of the art. BMC Bioinformatics 13:89 46. Yan J, Kurgan L (2017) DRNApred, fast sequence-based method that accurately predicts and discriminates DNAand RNA-binding residues. Nucleic Acids Res 45 (10):e84 47. Yan J, Friedrich S, Kurgan L (2016) A comprehensive comparative review of sequence-based predictors of DNA- and RNA-binding residues. Brief Bioinform 17(1):88–105 48. Meszaros B, Simon I, Dosztanyi Z (2009) Prediction of protein binding regions in disordered proteins. PLoS Comput Biol 5(5): e1000376 49. Khan W, Duffy F, Pollastri G, Shields DC, Mooney C (2013) Predicting binding within disordered protein regions to structurally characterised peptide-binding domains. PLoS One 8(9):e72838 50. Disfani FM, Hsu WL, Mizianty MJ, Oldfield CJ, Xue B, Dunker AK, Uversky VN, Kurgan L (2012) MoRFpred, a computational tool for sequence-based prediction and characterization of short disorder-to-order transitioning binding regions in proteins. Bioinformatics 28 (12):i75–i83 51. Meng F, Kurgan L (2018) High-throughput prediction of disordered moonlighting regions in protein sequences. Proteins 86 (10):1097–1110 52. Meng F, Kurgan L (2016) DFLpred: highthroughput prediction of disordered flexible linker regions in protein sequences. Bioinformatics 32(12):i341–i350 53. Oldfield CJ, Uversky VN, Kurgan L (2018) Predicting functions of disordered proteins

Prediction of Disordered RNA-Binding Regions with MoRFpred. Methods Mol Biol 1851:337–352 54. Yan J, Dunker AK, Uversky VN, Kurgan L (2016) Molecular recognition features (MoRFs) in three domains of life. Mol BioSyst 12(3):697–710 55. Peng Z, Kurgan L (2015) High-throughput prediction of RNA, DNA and protein binding regions mediated by intrinsic disorder. Nucleic Acids Res 43(18):e121 56. Peng Z, Wang C, Uversky VN, Kurgan L (2017) Prediction of disordered RNA, DNA, and protein binding regions using DisoRDPbind. Methods Mol Biol 1484:187–203 57. Gawlik K, Gallay PA (2014) HCV core protein and virus assembly: what we know without structures. Immunol Res 60(1):1–10 58. Ivanyi-Nagy R, Lavergne J-P, Gabus C, Ficheux D, Darlix J-L (2008) RNA chaperoning and intrinsic disorder in the core proteins of Flaviviridae. Nucleic Acids Res 36(3):712–725 59. Sharma K, Didier P, Darlix JL, de Rocquigny H, Bensikaddour H, Lavergne JP, Penin F, Lessinger JM, Mely Y (2010) Kinetic analysis of the nucleic acid chaperone activity of the hepatitis C virus core protein. Nucleic Acids Res 38(11):3632–3642 60. Piovesan D, Tabaro F, Micˇetic´ I, Necci M, Quaglia F, Oldfield CJ, Aspromonte MC, Davey NE, Davidovic´ R, Doszta´nyi Z,

239

Elofsson A, Gasparini A, Hatos A, Kajava AV, Kalmar L, Leonardi E, Lazar T, MacedoRibeiro S, Macossay-Castillo M, Meszaros A, Minervini G, Murvai N, Pujols J, Roche DB, Salladini E, Schad E, Schramm A, Szabo B, Tantos A, Tonello F, Tsirigos KD, Veljkovic´ N, Ventura S, Vranken W, Warholm P, Uversky VN, Dunker AK, Longhi S, Tompa P, Tosatto SCE (2017) DisProt 7.0: a major update of the database of disordered proteins. Nucleic Acids Res 45 (Database issue):D219–D227 61. Wootton JC, Federhen S (1993) Statistics of local complexity in amino-acid-sequences and sequence databases. Comput Chem 17 (2):149–163 62. McGuffin LJ, Bryson K, Jones DT (2000) The PSIPRED protein structure prediction server. Bioinformatics 16(4):404–405 63. Kawashima S, Ogata H, Kanehisa M (1999) AAindex: amino acid index database. Nucleic Acids Res 27(1):368–369 64. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25(17):3389–3402 65. World Health Assembly (2010) Viral hepatitis: report by the secretariat, vol A63/15. World Health Organization, Geneva

Chapter 15 Delivering Molecular Beacons via an Electroporation-Based Approach Enables Live-Cell Imaging of Single RNA Transcripts and Genomic Loci Shiqi Mao, Yachen Ying, Xiaotian Wu, and Antony K. Chen Abstract Molecular beacons (MBs) are synthetic oligonucleotide probes that are designed to fluoresce upon hybridization to complementary nucleic acid targets. In contrast to genetically encoded probes that can be readily introduced into cells via standard transfection procedures, using MBs to obtain reliable intracellular measurements entails a reliable delivery method that maximizes MB entry while minimizing cell damage. One promising approach is microporation, a microliter volume electroporation-based method that exhibits reduced harmful events as compared with traditional electroporation methods. In this chapter, we describe in detail microporation steps for MB delivery that we have utilized over the past several years, followed by examples demonstrating successful MB-based imaging of specific RNA transcripts and genomic loci at the single-molecule level. Key words Molecular beacons, Microporation, Tandem repeats, Single-molecule

1

Introduction The ability to illuminate specific genomic loci and RNA molecules at the single-molecule level is crucial for deepening our understanding of the relationships between gene expression and disease etiology. One promising approach is to use molecular beacons (MBs), which are small, hybridization-activated antisense oligonucleotidebased probes labeled with a fluorophore and a quencher at the two termini [1]. In the absence of complementary nucleic acid target, the short arm sequences at the two termini self-anneal to cause MBs to assume a stem-loop conformation. In this context, the quencher is placed nearby the fluorophore, causing the fluorophore to emit a low fluorescence signal. Hybridization of a unique MB target sequence (MTS) to the loop region disrupts the duplex stem, causing separation of the dye and the quencher to ultimately restore MB fluorescence.

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_15, © Springer Science+Business Media, LLC, part of Springer Nature 2020

241

242

Shiqi Mao et al.

Currently, methods for efficient delivery of MBs into living cells include streptolysin-O [2, 3], a bacterial exotoxin that reversibly forms pores on the cell surface, and microporation [4], a microliter volume electroporation process that avoids many harmful events such as heat generation, metal ion dissolution, pH variation, and oxide formation that are pertained to conventional electroporation procedures. Over the past few years, our laboratory has extensively optimized the microporation procedures for efficient intracellular delivery of MBs with high viability, demonstrating MB-based imaging of both RNA and genomic loci at the single-molecule level in living cells [5–7]. In brief, our optimized procedure involves first resuspending the cells in a small volume of a specialized buffer containing MBs before microporation, followed by the recovery step in which the cells are incubated in normal cell culture media and plated on surfaces pre-coated with extracellular matrix proteins that facilitate cell adhesion (Fig. 1). In this chapter, we describe in detail the microporation protocol for MB delivery and provide data showing EGFP mRNA labeled by MBs (Figs. 2a and 3) and human telomere loci labeled by the clustered regularly interspersed short palindromic repeats (CRISPR)-based genomic imaging system incorporating MBs (i.e., CRISPR/MB) (Figs. 2b and 4), at the single-molecule level in living cells.

2

Materials

2.1

MB

50 μM anti-MTS MB stock solution, dissolved in nuclease-free water (see Notes 1–4).

2.2

Plasmid

A mammalian expression vector termed sgTelo-MTS/EGFP/ pdCas9-C1, which encodes the sgTelo-MTS (i.e., telomeretargeting sgRNA harboring an MTS), EGFP, and dCas9 derived from Streptococcus pyogenes. The expression of the three genes is driven under the control of separate promoters. This vector is used for CRISPR-/MB-based telomere imaging (see Note 5).

2.3

Cell Culture

1. HeLa cells stably expressing an engineered EGFP mRNA harboring eight tandem repeats of the MTS upstream of the EGFP coding sequence (encoded by pEGFP-N1-8
; see Note 6), denoted as HeLa-EGFP-8
. 2. HEK293 cells. 3. Cell culture media: Dulbecco’s Modified Eagle’s Medium (DMEM) without phenol red and antibiotics, supplemented with 10% (v/v) FBS and 1
Glutamax. 4. Phenol red-free solution of 0.25% trypsin and 1 mM EDTA. 5. 100
penicillin-streptomycin solution. 6. T25-flask.

Delivering Molecular Beacons Using Microporation

243

Fig. 1 Schematic of MB delivery into cells using microporation. Adherent cells are trypsinized and then resuspended in a small volume of a specialized buffer containing MBs before microporation. During microporation, MBs enter the cytoplasm and nucleoplasm due to a transient application of an electric field. After microporation, the cells were plated onto a fibronectin-coated surface and incubated in normal cell culture media for recovery

2.4

Microporation

1. Microporation system (e.g., Thermo Fisher Neon transfection system). 2. 1
phosphate buffered saline (PBS), without Mg2+ and Ca2+. 3. Cell culture media (see item 3 in Subheading 2.3; see Note 7). 4. Resuspension buffer R (Neon transfection system). 5. Electroporation buffer (Neon transfection system). 6. Electroporation gold tips (for 10 μL volume). 7. Electroporation tube. 8. Eight-well chambered coverglass. 9. 10 μg/mL fibronectin. 10. Refrigerated microcentrifuge. 11. Microcentrifuge tube. 12. Cell counter.

2.5 Fluorescence In Situ Hybridization (FISH)

1. Nuclease-free water.

2.5.1 RNA FISH

4. 70% (v/v) ethanol, prepared from anhydrous ethanol.

2. RNase-/DNase-free pipette tips. 3. 4% (w/v) paraformaldehyde diluted in 1
PBS (4% PFA). 5. 2
saline-sodium citrate (SSC) buffer. 6. Wash buffer: 2
SSC, 10% (v/v) formamide. 7. RNA hybridization buffer: 10% (w/v) dextran sulfate, 2
SSC, 10% (v/v) formamide.

244

Shiqi Mao et al.

Fig. 2 Schematic of MB-based detection of RNA and genomic loci at the single-molecule level. (a) Collective hybridization of MBs to an engineered RNA transcript harboring multiple tandem repeats of the MTS can illuminate the target RNA as a single bright spot. (b) In the CRISPR/MB system, dCas9 and sgRNA modified to harbor an MTS (sgRNA-MTS) are first transfected in cells. After sufficient time is given to allow dCas9 and sgRNA-MTS to form a dCas9-sgRNA-MTS complex, which binds to a target DNA sequence, MBs are then introduced into cells. Hybridization of MBs to dCas9-sgRNA-MTS complexes collectively bound to adjacent sequences within a single genomic locus can illuminate the locus as a single bright spot

8. 100 μM EGFP RNA FISH probes, dissolved in nuclease-free water. The probes consisting a pool of oligonucleotides each labeled with a TAMRA at the 30 end and complementary to different regions of the EGFP coding sequence [6]. 9. Parafilm. 2.5.2 DNA FISH

1. Nuclease-free water. 2. RNase-/DNase-free pipette tips. 3. 4% (w/v) paraformaldehyde diluted in 1
PBS (4% PFA). 4. 0.5% (v/v) nonyl phenoxypolyethoxylethanol (NP-40) in 1
PBS. 5. 2
/1
/0.2
SSC buffer.

Delivering Molecular Beacons Using Microporation

245

Fig. 3 Single-molecule detection of engineered EGFP mRNA harboring eight tandem repeats of the MTS in HeLa cells by MBs and RNA FISH. Eight hours post-microporation of the anti-MTS MBs, HeLa-EGFP-8
cells were fixed and permeabilized and then subjected to RNA FISH processing. Representative images of MB and FISH signals are shown. Colocalization between MB and FISH signals was ~88%. This figure is adapted from [6] with permission in accordance with the Creative Commons Attribution 4.0 International (CC BY 4.0) License

Fig. 4 Single-molecule detection of telomere loci in HEK293 cells by CRISPR/ MBs and DNA FISH. Twenty-four hours post-microporation of the anti-MTS MBs, cells were fixed and permeabilized and then subjected to DNA FISH processing. Representative images of MB and FISH signals are shown. Colocalization between MB and FISH signals was ~84%. DAPI (blue) stains the nucleus. This figure is adapted from [7] with permission in accordance with the Creative Commons Attribution Non-Commercial (CC BY-NC 4.0) License

6. Wash buffer: 2
SSC, 10% (v/v) formamide. 7. DNA hybridization buffer: 1% (v/v) Tween® 20, 10% (w/v) dextran sulfate, 2
SSC, 50% (v/v) formamide, 500 ng/mL Salmon sperm DNA. 8. 100 μM telomere leading strand-targeting DNA FISH probes, dissolved in nuclease-free water. The probe is labeled with a

246

Shiqi Mao et al.

TAMRA at the 50 end and complementary to the leading strand of human telomeres. The sequence is 50 -CCCTAACCCTAA CCCTAA-30 [7]. 9. 40 ,6-Diamidino-2-phenylindole (DAPI). 10. Parafilm.

3

Methods

3.1 Cellular Delivery of MBs

3.1.1 Day 0

The protocol below describes the delivery of anti-MTS MBs into HeLa-based stable cell lines (i.e., HeLa-EGFP-8
) and HEK293 cells previously transfected with dCas9 and sgTelo-MTS (for CRISPR-/MB-based telomere imaging). Specific steps are as follows: 1. Seed proper number of cells into a T25-flask with cell culture media such that cells reach ~70% confluency on Day 1 (see Note 8). 2. Add 200 μL fibronectin solution into each well of an eight-well chambered coverglass and incubate the coverglass at 37  C overnight (see Note 9).

3.1.2 Day 1

1. Aspirate the cell culture media and incubate the cells with 5 mL pre-warmed 1
PBS at room temperature for 2 min. 2. Aspirate the PBS and incubate the cells with 1 mL of phenol red-free trypsin/EDTA at room temperature for 1 min. 3. Aspirate the majority of the trypsin and incubate the cells in a trace amount of trypsin at 37  C until all of the cells are detached from the flask surface. 4. Add 5 mL of cell culture media without penicillin-streptomycin into the flask. Pipette gently to resuspend the cells (see Note 10). 5. Transfer 1 mL of the cell suspension to a 1.5 mL microcentrifuge tube. 6. Pellet the cells by centrifugation at 400
g for 5 min at 4  C (see Note 11). 7. Aspirate the supernatant. 8. Add 1 mL of 1
PBS into the tube. Pipette gently to resuspend the cells. 9. Count the cells. 10. Pellet the rest of the cells by centrifugation at 400
g for 5 min at 4  C. 11. Aspirate 1
PBS and resuspend the cell pellet in resuspension buffer R at 5000 cells per μL.

Delivering Molecular Beacons Using Microporation

247

12. Add 1 μL of 50 μM MB stock solution to every 10 μL of cells so that ~50,000 cells are resuspended in the buffer containing ~ 5 μM MBs (see Note 12). 13. Pipette gently to mix the cells with MBs. 14. Add ~ 4 mL electroporation buffer into the electroporation tube. 15. Use an electroporation gold tip to microporate 10 μL of the cell suspension. Set microporation parameters to 1005 V with a 35 ms pulse width and 2 pulses total for HeLa-based cell lines and to 1150 V with a 20 ms pulse width and 2 pulses total for HEK293 cells (see Note 13). 16. Gently pipette out the microporated cells from the tip to a microcentrifuge tube prefilled with 1.5 mL of fresh cell culture media without penicillin-streptomycin (see Note 11). 17. Pellet the cells by centrifugation at 400
g for 5 min at 4  C. 18. Aspirate the supernatant and then gently resuspend the pellet with 1.5 mL of fresh cell culture media without penicillinstreptomycin (see Note 14). 19. Repeat steps 14–15 two more times. 20. After the last wash, gently resuspend the cells with 250 μL of fresh cell culture media without penicillin-streptomycin. 21. Seed the cells into a fibronectin-coated well of an eight-well chambered coverglass (see Note 15), and incubate the coverglass at 37  C. Avoid disturbing the coverglass before cells can sufficiently adhere to the coverglass (~5 h). 3.2 Fluorescence In Situ Hybridization (FISH)

We used RNA FISH to validate the capacity of anti-MTS MBs for labeling the pEGFP-8
transcripts. Care should be taken to avoid disturbing the cells throughout the process.

3.2.1 RNA FISH Day 1

1. Pipette out the media from each well of the eight-well chambered coverglass and wash the cells thrice with 350 μL of 1
PBS. 2. After the last PBS wash, add 250 μL of 4% PFA pre-warmed at 37  C into each well. Incubate the cells in PFA at room temperature for 30 min. 3. Pipette out the PFA. 4. Incubate the cells in 350 μL 1
PBS at room temperature for 5 min and then pipette out the PBS. 5. Repeat the above washing step two more times. 6. After the last wash, add 400 μL of 70% ethanol into each well.

248

Shiqi Mao et al.

7. Close the lid and wrap the chambered coverglass with parafilm to minimize evaporation. 8. Protect the chambered coverglass from light and store it at 4  C overnight (>16 h). Day 2

1. Pipette out the ethanol. 2. Incubate the cells in 350 μL wash buffer at room temperature for 5 min and then pipette out the wash buffer. 3. Repeat the above washing step one more time. 4. Dilute 1 μL of 100 μM TAMRA-labeled EGFP RNA FISH probes in 400 μL of the RNA hybridization buffer to achieve the final probe concentration of 250 nM. 5. After the second wash, add 250 μL of the RNA FISH probes (250 nM final concentration). 6. Close the lid and wrap the chambered coverglass with parafilm to minimize evaporation. 7. Protect the chambered coverglass from light and incubate it at 37  C overnight (>16 h).

Day 3

1. Pipette out the unbound FISH probes. 2. Incubate the cells in 400 μL wash buffer at room temperature for 5 min, and then pipette out the wash buffer. 3. Repeat the above washing step one more time. Do not let samples dry between washes. 4. After the last wash, add 400 μL wash buffer into each well. 5. Close the lid and store the chambered coverglass, protected from light, at 37  C for 30–60 min. 6. Pipette out the wash buffer and add 400 μL 2
SSC into each well. Repeat this step two more times. Do not let the samples dry between washes. 7. Pipette out the 2
SSC and add 300 μL of 1
PBS into each well. 8. Use fluorescence microscopy to image the samples (see Note 16).

3.2.2 DNA FISH

Day 1

We employed DNA FISH to validate the capacity of CRISPR/MB for labeling telomere loci in HEK293 cells (see Note 17). 1. Pipette out the media from each well of the eight-well chambered coverglass. 2. Wash the cells thrice with 350 μL of 1
PBS. 3. After the last PBS wash, add 250 μL of 4% PFA pre-warmed at 37  C into each well.

Delivering Molecular Beacons Using Microporation

249

4. Incubate the cells in PFA at room temperature for 30 min. 5. Pipette out the PFA. 6. Incubate the cells in 350 μL 1
PBS at room temperature for 5 min, and then pipette out the PBS. 7. Repeat the above washing step two more times. 8. After the last wash, incubate the cells in 250 μL of 0.5% (v/v) NP-40 in 1
PBS at room temperature for 10 min. Then pipette out the buffer. 9. Incubate the cells in 350 μL 1
PBS at room temperature for 5 min and then pipette out the PBS. 10. Repeat the above washing step one more time. 11. Dilute 1 μL of 100 μM TAMRA-labeled telomere leading strand-targeting DNA FISH probes in 1 mL of DNA hybridization buffer to achieve the final probe concentration of 100 nM. 12. After the second wash, add 250 μL of the DNA FISH probes (100 nM final concentration). 13. Close the lid and wrap the chambered coverglass with parafilm to minimize evaporation. 14. Protect the chambered coverglass from light and incubate it at 37  C overnight (>16 h). Day 2

1. Pipette out the unbound FISH probes. 2. Incubate the cells in 400 μL wash buffer at room temperature for 5 min. Pipette out the wash buffer. 3. Repeat the above washing step one more time. Do not let samples dry between washes. 4. After the last wash, add 400 μL wash buffer into each well. Close the lid and store the chambered coverglass, protected from light, at 37  C for 30–60 min. 5. Pipette out the wash buffer and add 400 μL 2
SSC into each well. Repeat this step one more time. Do not let the samples dry between washes. 6. Pipette out the 2
SSC and add 400 μL 1
SSC into each well. Repeat this step one more time. Do not let the samples dry between washes. 7. Pipette out the 1
SSC and add 400 μL 0.2
SSC into each well. Repeat this step one more time. Do not let the samples dry between washes. 8. Pipette out the 0.2
SSC and add 300 μL of 1
PBS into each well.

250

Shiqi Mao et al.

9. Add DAPI in 1
PBS to stain the nucleus (working concentration, 100 ng/mL). 10. Use fluorescence microscopy to image the samples (see Note 16).

4

Notes 1. Anti-MTS MB is synthesized with a backbone composed of 20 -O-methyl RNA (2Me) and a fully phosphorothioate (PS)modified loop domain. We have previously demonstrated that MB with this modification is highly biostable and biocompatible [5]. 2. Anti-MTS MB is labeled with an ATTO647N at the 50 end and an Iowa Black® RQ-Sp quencher at the 30 end. Researchers can use MBs incorporating other fluorophore/quencher pairs for their own experiments. 3. Anti-MTS MB has the following sequence: 50 -mCmUmUm CmG∗mU∗mC∗mC∗mA∗mC∗mA∗mA∗mA∗mC∗mA ∗mC∗mA∗mA∗mC∗mU∗mC∗mC∗mU∗mGmAmAmG30 . Underlined letters indicate the MB stem. m represents 2Me modification. ∗ represents PS linkage modification. Typically, we use MB with a 5-nucleotide stem and a 17-nucleotide loop domain. The sequence of anti-MTS MB is designed not to be complementary to endogenous RNAs in mammalian cells and therefore should remain quenched in the cellular environment unless hybridized to its MTS. 4. MBs can be purchased from Integrated DNA Technologies (Coralville, IA, USA). 5. CRISPR/MB consists of three main components: a dCas9, an MB, and an sgRNA scaffold modified to harbor an MTS. To use CRISPR/MB for telomere labeling, the expression vector sgTelo-MTS/EGFP/pdCas9 was first transfected into HEK293 cells. After sufficient time was given to allow for the formation of dCas9-sgTelo-MTS complexes, as well as their binding to the telomere loci (i.e., 24 h post-transfection), antiMTS-MBs were introduced into cells by microporation. Hybridization of the MB to the MTS of the dCas9-sgTelo-MTS complexes bound across the telomere can illuminate the telomere as a bright spot when imaged by standard widefield fluorescence microscopy. 6. The plasmid pEGFP-N1-8
harbors 8 tandem repeats of the 50-base sequence: 50 -CAGGAGTTGTGTTTGTGGACGAA GAGCACCAGCCAGCTGATCGACCTCGA-30 upstream of the EGFP coding sequence. The underlined sequence is the MTS used in our experiment.

Delivering Molecular Beacons Using Microporation

251

7. Cell culture media free of antibiotics (i.e., penicillinstreptomycin) should be used during microporation to minimize cell death. 8. To use CRISPR/MB, HEK293 cells were first transfected with sgTelo-MTS/EGFP/pdCas9 using FuGENE® 6 1 day before microporation. 9. Cells generally adhere and spread more readily on a coverglass pre-coated with fibronectin, an extracellular matrix protein, than on an untreated coverglass. 10. Gentle pipetting minimizes damaging of cells. 11. Do not use microcentrifuge tubes with a low-retention surface as cell pellets are difficult to form in these tubes after centrifugation. 12. The final MB concentration for microporation could be adjusted according to different experimental conditions. In our experience, 5 μM MB concentration was suitable for the detection of EGFP mRNA in HeLa cells, whereas 1 μM MB concentration was suitable for detection of telomere loci in HEK293 cells. 13. The microporation parameter and the number of cells used are cell type specific. Optimized parameters for different cell types are available at https://www.thermofisher.com/cn/zh/ home/life-science/cell-culture/transfection/transfection%2D %2D-selection-misc/neon-transfection-system/neonprotocols-cell-line-data.html. 14. It is recommended to leave the cell pellet and a trace amount of media (~50 μL) in the microcentrifuge tube after every wash to avoid aspirating out the loose cell pellet formed by the small number of cells. 15. To maximize cell attachment, unbound fibronectin in solution should be removed with enough 1
PBS. This is because unbound fibronectin can saturate the binding sites on the cells, preventing the cells to attach onto a fibronectin-coated glass surface. 16. TRITC and Cy5 filter sets were used for imaging. For FISH signals (acquired in the TRITC channel) and MB signals (acquired in the Cy5 channel), image stacks were acquired with an increment of 0.25 μm in the z-direction. Steps for colocalization analysis were described previously [8]. 17. The dCas9-sgTelo-MTS complex can unwind the duplex of telomere loci, making the leading strand accessible by FISH probes. Therefore, harsh treatments including heat and formamide associated with conventional DNA FISH protocols are not required to denature the DNA duplex.

252

Shiqi Mao et al.

Acknowledgments This project was supported by grants from the National Key R&D Program of China (No. 2016YFA0501603), the National Natural Science Foundation of China (No. 31771583), and the Beijing Municipal R&D Key Project (Grant No. Z151100003915081). References 1. Tyagi S, Kramer FR (1996) Molecular beacons: probes that fluoresce upon hybridization. Nat Biotechnol 14(3):303–308. https://doi.org/ 10.1038/nbt0396-303 2. Bhakdi S, Bayley H, Valeva A, Walev I, Walker B, Kehoe M, Palmer M (1996) Staphylococcal alpha-toxin, streptolysin-O, and Escherichia coli hemolysin: prototypes of pore-forming bacterial cytolysins. Arch Microbiol 165(2):73–79. https://doi.org/10.1007/s002030050300 3. Santangelo PJ, Nix B, Tsourkas A, Bao G (2004) Dual FRET molecular beacons for mRNA detection in living cells. Nucleic Acids Res 32(6):e57. https://doi.org/10.1093/nar/gnh062 4. Chen AK, Behlke MA, Tsourkas A (2008) Efficient cytosolic delivery of molecular beacon conjugates and flow cytometric analysis of target RNA. Nucleic Acids Res 36(12):e69. https:// doi.org/10.1093/nar/gkn331 5. Zhao D, Yang Y, Qu N, Chen M, Ma Z, Krueger CJ, Behlke MA, Chen AK (2016) Singlemolecule detection and tracking of RNA

transcripts in living cells using phosphorothioate-optimized 20 -O-methyl RNA molecular beacons. Biomaterials 100:172–183. https://doi.org/10.1016/j.biomaterials.2016. 05.022 6. Chen M, Ma Z, Wu X, Mao S, Yang Y, Tan J, Krueger CJ, Chen AK (2017) A molecular beacon-based approach for live-cell imaging of RNA transcripts with minimal target engineering at the single-molecule level. Sci Rep 7 (1):1550. https://doi.org/10.1038/s41598017-01740-1 7. Wu X, Mao S, Yang Y, Rushdi MN, Krueger CJ, Chen AK (2018) A CRISPR/molecular beacon hybrid system for live-cell genomic imaging. Nucleic Acids Res 46(13):e80. https://doi. org/10.1093/nar/gky304 8. Yang Y, Chen M, Krueger CJ, Tsourkas A, Chen AK (2018) Quantifying gene expression in living cells with ratiometric bimolecular beacons. Methods Mol Biol 1649:231–242. https://doi. org/10.1007/978-1-4939-7213-5_15

Chapter 16 Site-Specific Dual-Color Labeling of Long RNAs Meng Zhao, Richard Bo¨rner, Roland K. O. Sigel, and Eva Freisinger Abstract Labeling of large RNAs with reporting entities, e.g., fluorophores, has significant impact on RNA studies in vitro and in vivo. Here, we describe a minimally invasive RNA labeling method featuring nucleotide and position selectivity, which solves the long-standing challenge of how to achieve accurate site-specific labeling of large RNAs with a least possible influence on folding and/or function. We use a customdesigned reactive DNA strand to hybridize to the RNA and transfer the alkyne group onto the targeted adenine or cytosine. Simultaneously, the 30 -terminus of RNA is converted to a dialdehyde moiety under the experimental condition applied. The incorporated functionalities at the internal and the 30 -terminal sites can then be conjugated with reporting entities via bioorthogonal chemistry. This method is particularly valuable for, but not limited to, single-molecule fluorescence applications. We demonstrate the method on an RNA construct of 275 nucleotides, the btuB riboswitch of Escherichia coli. Key words Long RNAs, Site-specific labeling, Orthogonal chemistry, Riboswitch, Single-molecule fluorescence resonance energy transfer (smFRET)

1

Introduction Labeling of long RNAs remains challenging due to the lack of a universally applicable method that allows the conjugation of reporting entities in an orthogonal yet site-specific manner. Existing methods suffer from either the exponentially reduced labeling efficiency with increasing RNA size or the disruption of the correct fold and thus function of the RNA [1–11]. Here, we describe a posttranscriptional labeling method [12] that employs a customized DNA oligonucleotide (guide DNA, Fig. 1b) as reactive strand (dRS, Fig. 1b) to transfer an alkyne group on any adenine or cytosine residue within an RNA strand of arbitrary length [13, 14]. To increase the yield of this chemical reaction, we make use of DNA oligonucleotides as helper sequences (dHSs, Fig. 2b and c) preventing secondary structure formation and increasing the accessibility of the target site [15]. In this one-pot reaction, the 30 -terminal ribose ring is concomitantly opened to a dialdehyde moiety [16] leading to the functionalized RNA with an internal

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_16, © Springer Science+Business Media, LLC, part of Springer Nature 2020

253

254

Meng Zhao et al.

Fig. 1 (a) Synthesis scheme of the reactive group (RG) and the reactive strand (dRS). Details of the synthesis of compound 1, 2, 3, and 4 were described previously [13]. (b) Scheme of the dRS binding to the RNA, showing the relative position of the labeling site to the guide DNA. (Adapted from [12])

Fig. 2 Scheme of the secondary structure of the btuB riboswitch (a) and the hybridization with the guide DNA of dRS at room temperature assisted by dHSs for labeling at A35 (b) or A213 (c). (Adapted from [12])

alkyne and a reactive 30 -terminal end, which allows for biorthogonal chemistry without cross-reactivity. We use the two carbocyanine dyes Cy3-hydrazide for coupling to the 30 -terminal dialdehyde moiety and Cy5-azide for reaction with the internal alkyne unit, to obtain a dual-color labeled btuB riboswitch RNA of 275 nt [17, 18]. The labeled construct is used for single-molecule fluorescence resonance energy transfer (smFRET) studies [12] as an example, but the presented method can be adapted to any technique that requires the site-specific labeling of RNAs, in particular

Dual Site-Specific RNA Labeling

255

to study RNA folding kinetics and chemical reactions along the desired reaction coordinates. Our method is of particular interest to site-specific and biorthogonal labeled long (>100 nt) RNAs. The posttranscriptional labeling enables the production of large amounts of RNA via in vitro transcription, usually limited by very low reaction yields in solid-phase synthesis. Further, our method allows the labeling of nucleotides located in structured regions of RNAs, temporarily disintegrated with the help of complementary DNA oligonucleotides, thus the labeling of RNAs without restriction in size and fold.

2

Materials If not stated otherwise, chemicals were obtained in highest commercially available grades and used without further purification. All buffers were prepared with ultrapure water and filtered before use. Solutions and chemicals are kept at room temperature unless indicated otherwise. DNA oligonucleotides were purchased. Carbocyanine fluorophores were purchased from Lumiprobe and always protected from light.

2.1 Organic Solvents and Agents

1. DMSO (purum). 2. DMF (purum). 3. 100% ethanol (EtOH). 4. Acetone (purum). 5. Reactive group precursor (Fig. 1a), synthesized as described in [13]. 6. N-Hydroxysuccinimide (NHS). 7. N,N0 -Dicyclohexylcarbodiimide (DCC). 8. 1.6 M 2,20 -Pyridine disulfide (PDS) in DMSO. 9. 1.2 M 4-Dimethylaminopyridine (DMAP) in DMSO. 10. 1.2 M triphenylphosphine (PPh3) in DMF. 11. Ethylenediamine (purum). 12. Formamide (purum). 13. 2% (w/v) lithium perchlorate (LiClO4) in acetone. 14. Ethylenediaminetetraacetic acid (EDTA). 15. Xylene cyanol. 16. Bromophenol blue. 17. Tris(benzyltriazolylmethyl)amine (TBTA). 18. Copper(II) sulfate (CuSO4). 19. 30% acrylamide/bisacrylamide solution. 20. Urea.

256

Meng Zhao et al.

2.2 Buffers and Solutions

1. 100 mM potassium phosphate buffer pH 7.5. 2. 3.0 M sodium chloride (NaCl) in water. 3. 5 sodium acetate (NaOAc) buffer: 1.0 M, pH 5.5. 4. 1 NaOAc buffer: 20 mM, pH 5.5. 5. Triethylammonium acetate (TEAA) buffer: 0.5 M, pH 7.0. 6. Soaking buffer: 10 mM MOPS, 1.0 mM EDTA, and 250 mM NaCl, pH 6.0. 7. 5 folding buffer: 330 mM HEPES, 200 mM Tris–HCl, and 250 mM KCl, pH 7.5. 8. Running buffer: 66 mM HEPES, 34 mM Tris–HCl, and 3.0 mM MgOAc, pH 7.5. 9. 10 Tris-borate-EDTA (TBE) buffer. 10. Formamide loading buffer (FLB): 82% (v/v) formamide, 10 mM EDTA, 0.16% (w/v) xylene cyanol, and 0.16 (w/v) bromophenol blue. 11. 8% (w/v) cetyltrimethylammonium bromide (CTAB) in water. 12. 20 mM sodium periodate (NaIO4) in 1 NaOAc buffer (see item 3 in Subheading 2.2). 13. 15 mM ethylene glycol in 1 NaOAc buffer (see item 3 in Subheading 2.2). 14. 5.0 mM ascorbic acid in water. 15. 30 mM MgCl2 in water. 16. 1.0 M KCl in water. 17. Cu-TBTA solution: 10 mM CuSO4 and 10 mM TBTA in 55% (v/v) DMSO/water.

2.3 Deoxyribonucleotide Triphosphate (dNTP), DNA Oligonucleotides, and RNA

1. 10 mM dNTPmix. 2. Guide-DNA (Fig. 1b) solution: 4.0 mM guide DNA in water. 3. 100 μM helper sequence (dHS) solution in water. 4. 1.0 μM 50 -32P-labeled DNA primer (32P-primer) in water, prepared as described in [19]. 5. 1 μM biotin-tagged DNA oligonucleotides in water. 6. The modified btuB riboswitch RNA was designed for smFRET immobilization studies according to the literature [12]. In vitro RNA transcription was performed with homemade T7 polymerase [20].

2.4

Fluorophores

1. 10 mM Cyanine3 hydrazide (Cy3-hydrazide) in DMSO. 2. 10 mM Sulfo-Cyanine5 azide (Cy5-azide) in DMSO.

Dual Site-Specific RNA Labeling

2.5

Kits and Columns

257

1. AMV reverse transcriptase kit: 2 U/μL enzyme in glycerol, 4 enzyme buffer, and 40 U/μL RNase inhibitor. 2. SuperScript III reverse transcriptase kit: 20 U/μL enzyme in glycerol, 4 enzyme buffer, 0.1 M Dithiothreitol (DTT), and 40 U/μL RNase inhibitor. 3. NAP-5 desalting column.

2.6 Polyacrylamide Gel Electrophoreses (PAGE) Gel

3

1. 5%, 10%, and 18% denaturing PAGE: 7 M urea and 5%, 10%, and 18% (w/v) acrylamide/bisacrylamide in 1 TBE buffer. 2. 6% native PAGE: 6% (w/v) acrylamide/bisacrylamide in running buffer (see item 7 in Subheading 2.2).

Methods

3.1 Preparation of the Reactive Group, Solution I

1. Weigh 3.1 mg of reactive group precursor (Fig. 1a; see Note 1) and 1.5 mg of NHS in a 1.5 mL tube, and dissolve it in 80 μL DMF. Weigh 2.7 mg of DCC in another 1.5 mL tube, and dissolve it in 20 μL DMF. Mix both solutions together, and incubate the mixture at room temperature and 500 rpm overnight (see Note 2). 2. Centrifuge the mixture at 4  C and 16.1  103 rfc for 10 min. Collect the supernatant which contains the activated reactive group in a clean 1.5 mL tube and keep it on ice (see Note 3). Referred to as Solution I.

3.2 Preparation of the EthylenediamineModified Guide DNA, Solution II

1. Prepare 25 μL of the guide DNA (see Note 4) solution in a 1.5 mL tube, and add 6.0 μL of CTAB solution (see Note 5). Mix gently and incubate at room temperature and 500 rpm for 30 min. Centrifuge at room temperature and 16.1  103 rfc for 30 min. Remove and discard the supernatant and dry the precipitate in vacuum for 10 min. Dissolve the remaining pellet in 50 μL DMSO. 2. Mix 50 μL of PDS (see Note 6), 50 μL of DMAP, and 50 μL of PPh3 solution (see Note 7). Combine the mixture in the tube with the 50 μL DMSO solution containing the guide DNA from the last step. Mix well and incubate the mixture at room temperature and 500 rpm for 30 min. Add 8.0 μL of ethylenediamine and incubate for 30 min. Add 1.0 mL of LiClO4 acetone solution. Mix thoroughly and store at 80  C for 3 h. 3. Take the tube from 80  C and centrifuge at 4  C an d 16.1  103 rfc for 30 min. Wash the pellet with acetone. Centrifuge at 4  C and 16.1  103 rfc for 5 min. Remove the supernatant carefully with a pipette. Dry the pellet in vacuum at room temperature for 10 min. Dissolve the dried pellet in 10 μL of potassium phosphate buffer. Referred to as Solution II.

258

Meng Zhao et al.

3.3 Preparation of the Reactive Strand (dRS), Solution III

1. Mix 2.0 μL of Solution I with 10.0 μL of Solution II in a 1.5 mL tube. Mix gently. Incubate the mixture at room temperature and 500 rpm for 1 h. Add 1.0 μL of 3.0 M NaCl solution and 1.0 mL of 100% ethanol. Mix the solution well and store it at 80  C for 3 h. 2. Take the tube from 80  C and centrifuge at 4  C and 16.1  103 rfc for 30 min. Remove the supernatant and dry the precipitate in vacuum at room temperature for 10 min. Dissolve the dried pellet in 25 μL of 5 NaOAc buffer. Place the solution on ice (see Note 8). Referred to as Solution III.

3.4 dRS-Hybridized RNA, Solution IV

1. Prepare 80 μL of the btuB riboswitch RNA (~36.0 μM) in 1 NaOAc buffer, and add 30 μL of dHS solutions (Fig. 2; see Note 9). Mix gently. Heat the mixture at 70  C for 5 min and cool down slowly to room temperature for about 30 min without agitation to allow for hybridization of the dHS. Centrifuge at low speed shortly to recollect any condensated solvent from the lid. 2. Add 25 μL of Solution III. Mix gently. Incubate the solution at room temperature for 10 min without agitation to allow for hybridization of the dRS, and place it on ice before continuing to the next step (see Note 8). Referred to as Solution IV.

3.5 Crude Functionalized RNA, Solution V

1. Add 15 μL of NaIO4 solution to Solution IV of the previous step. Incubate the mixture at room temperature and 500 rpm for 90 min (see Note 10). Add 30 μL of ethylene glycol solution and mix well. Incubate at room temperature and 500 rpm overnight (see Note 11). Stop the reaction by applying the solution to a NAP-5 column and elute the sample with water. 2. Lyophilize the eluted sample solution overnight, giving white spongy precipitate. Spin down the precipitate to the bottom of the tube. Add a minimal volume of water or potassium phosphate buffer to dissolve the spongy precipitate (see Note 12). Store the product at 20  C or keep it on ice if the purification will be carried out the same day. Referred to as Solution V.

3.6 Purification of Functionalized RNA, Solution VI

1. Add one volume equivalent of FLB to Solution V. Mix gently. Load the mixture on a 5% denaturing PAGE (see Notes 13 and 14). Run the PAGE at 10–20 W at 4  C for about 1.5 h or before the lower marker-dye migrates to the bottom edge of the gel. Check carefully not to lose RNA. 2. Use a spatula to pry the glass plates and move the gel body to a plastic wrap. Place the gel body under a UV lamp to visualize RNA bands. Cut out the bands that correspond to the correct size (see Note 15). Collect the gel pieces in clean 1.5 mL tubes. To each tube, add 800 μL soaking buffer. Fix the tubes on a low-speed rotary at 4  C for 2.5 h and gently mix every 20 min.

Dual Site-Specific RNA Labeling

259

3. Centrifuge the tubes containing the soaked mixture at 4  C and 16.1  103 rfc for 30 min. Collect the supernatant to new 1.5 mL tubes (see Note 16) keeping the volume of the supernatant in each tube below 350 μL. Add three volume equivalents ice-cold 100% ethanol to each tube. Mix well and store it at 80  C overnight. 4. Take the tubes from 80  C and centrifuge the precipitated sample at 4  C and 16.1  103 rfc for 30 min. Remove the supernatant and dry the precipitate in vacuum at room temperature for 5 min. Dissolve the pellet in water or buffer as needed. Referred to as Solution VI (see Note 17). 3.7 30 -Terminal Labeled RNA, Solution VII

1. Prepare 25 μL of Solution VI (RNA ~25 μM) in TEAA buffer in a 1.5 mL tube. Mix with 5.0 μL formamide and 5.0 μL DMSO. Flush the solution with argon for 1 min (see Note 18). Add 2.0 μL of argon-flushed Cy3-hydrazide solution (see Note 19) [16]. Mix gently and incubate the solution in the dark at room temperature and 500 rpm overnight. 2. To stop the reaction, add 1.0 mL of ethanol to the tube. Mix gently and store at 80  C for 3 h. Take the tube from 80  C and centrifuge at 4  C and 16.1  103 rfc for 30 min. Remove the supernatant and wash with 200 μL ice-cold 100% ethanol. Centrifuge again for 10 min (see Note 20) and dry the pellet under vacuum at room temperature for 5 min. Dissolve the pellet in 25 μL TEAA buffer. Referred to as Solution VII (see Note 21).

3.8 Dual-Color Labeled RNA, Solution VIII

1. Mix 25 μL of Solution VII with 10 μL formamide, 55 μL DMSO, 2.0 μL Cy5-azide solution (see Note 21), and 10 μL 5.0 mM ascorbic acid solution (see Note 22). Flush the mixture with argon for 1 min (see Note 18) and add 10 μL Cu-TBTA solution. Flush the mixture with argon for 1 min and incubate it in the dark at 25  C and 500 rpm overnight. 2. To stop the reaction, add 1.0 mL of 100% ethanol to the tube. Mix gently and store the mixture at 80  C for 3 h. Take the tube from 80  C and centrifuge at 4  C and 16.1  103 rfc for 30 min. Remove the supernatant, wash with 200 μL ice-cold ethanol, and centrifuge again for 10 min (see Note 20). Dry the pellet in vacuum at room temperature for 5 min. Dissolve the pellet in water and apply it to a NAP-5 desalting column (see Note 23). Elute with water and lyophilize overnight (see Note 12). 3. Purify the lyophilized product in a 5% denaturing PAGE (see Note 24) following the same protocol as in Subheading 3.6 (see Note 25). Dissolve the recovered RNA in 20 μL water or potassium phosphate buffer. Referred to as Solution VIII.

260

Meng Zhao et al.

Characterize the dual-color labeled btuB riboswitch RNA in comparison with the non-labeled one in a 10% denaturing PAGE (Fig. 4; see Notes 26, 36 and 37). 3.9 Reverse Transcription Stop Assay

For a labeling site at the base-paired region, A213 (Fig. 5a; see Note 27) 1. Mix 2.0 μL of Solution VIII (RNA ~ 1.0 μM; see Note 28) with 2.0 μL of the 32P-labeled primer solution (see Note 29) and 4.0 μL of dNTPmix. Mix gently. Heat at 70  C for 5 min and cool on ice for 1 min. 2. Add 1.0 μL of RNAse inhibitor and 2.0 μL of AMV reverse transcriptase (RT) buffer. Mix gently and incubate at 42  C for 1 min. 3. Add 1.0 μL of AMV RT and mix gently. Incubate the sample at 42  C for 1 h without agitation. For a labeling site at the internal loop, A35 (Fig. 5b; see Note 27) 1. Mix 1.0 μL of Solution VIII (RNA ~ 1.0 μM, see Note 28) with 1.0 μL of the 32P-labeled primer (see Note 29) and 4.0 μL of dNTPmix. Mix gently. Heat at 70  C for 5 min. Place on ice for 5 min. 2. Add 1.0 μL of Rnase inhibitor, 1.0 μL of DTT, and 3.0 μL of SuperScript III RT buffer. Mix gently and incubate at 52  C for 1 min. 3. Add 1.0 μL of SuperScript III RT and mix gently. Incubate the sample at 52  C for 1 h without agitation. 4. For both cases, stop the reactions by heating the sample to 90  C and incubate for 5 min. Then, add one equivalent volume of FLB to the sample. Load the sample directly on a 18% denaturing PAGE gel. After drying, the gel was exposed to a phosphor imaging screen overnight and scanned the following day (Fig. 5 and Note 30).

3.10 Sample Annealing and Purification for smFRET Measurements

1. Mix 1.0 μL of Solution VIII (RNA ~ 1.0 μM) with 1.0 μL biotin-DNA oligonucleotides (for surface immobilization, see Note 31), 0.5 μL 1.0 M KCl solution, 2.0 μL 5 folding buffer (see Note 32), and 4.5 μL water and mix gently. 2. RNA annealing/folding: Incubate the sample at 70  C for 5 min; incubate at room temperature for 3 min and add 1.0 μL 30 mM MgCl2 solution. Mix gently (see Note 33). Incubate the sample solution at 37  C for 30 min and, then, at room temperature for another 30 min. Low-speed centrifuge the sample tube to recollect any condensated solvent from the lid.

Dual Site-Specific RNA Labeling

261

As a next step, the single-molecule FRET experiment follows. With regard to methodology [22–24] and data analysis [25–27], we refer the reader to the respective literature. We use prism-based total internal reflection fluorescence (TIRF) microscopy. In brief, the FRET pair labeled RNA molecules are surface immobilized via a biotin-streptavidin linkage, the fluorophores are excited with a laser beam which is totally reflected at the interface between the sample and the quartz prism, and the fluorescence light split in donor and acceptor emission is detected on an EMCCD camera [22, 23, 25]. A typical problem in smFRET is the double-labeling efficiency. The number of double-labeled molecules is increased by applying an additional native PAGE purification step of the pre-annealed RNA sample carrying the biotin-DNA. 3. Native PAGE (6%) purification: Load the sample and run the gel at 4  C and 10–20 W for 100 min (see Note 34). 4. Pry the gel plates with a spatula. Cut out the gel bands with co-localized fluorescence emission of Cy3 and Cy5 (see Note 25). Soak the gel pieces and precipitate the hybridized RNAs following the same protocol (see Subheading 3.6). 5. The precipitated sample is reannealed under the same condition as described (see Subheading 3.10), giving 10 μL of the sample solution diluted with the folding buffer containing the desired Mg(II) concentration for the smFRET measurements (see Note 35).

4

Notes 1. The synthesis of compounds 1, 2, 3, and 4 for the RG (Fig. 1) was described previously [14]. Complementary notes are listed here: (1) Reaction between compound 1 and dry methanol should be strictly controlled in a water-free environment. The incubation should last no longer than 15 min. Both the presence of water and longer incubation times can lead to by-products. To this end, we recommend carrying out this reaction with the subsequent methanol evaporation on a rotary evaporator. (2) To completely remove the methanol, we recommend adding 5 mL of dry CH2Cl2 (distilled over CaH2) to the residue and to co-evaporate both, methanol and CH2Cl2. Repeating the purification step twice yields the pale-yellow oil of compound 2. (3) Compound 2 is very sensitive under ambient condition. Use it immediately for the next reaction step or store it at 20  C after drying under high vacuum. However, long-time storage is not recommended. (4) In general, all compounds should be stored at 20  C to avoid degradation. We particularly recommend drying the RG

262

Meng Zhao et al.

precursor under high vacuum and in an ice/H2O bath to remove water as much as possible. The freshly synthesized RG precursor is usually a viscous oil. Upon storage at 20  C for months, crystals or solids can form. 2. As the reaction proceeds, crystals or precipitates can be observed at the bottom of the tube. 3. Activated RG easily degrades at room temperature and therefore should always be kept on ice. 4. Design criterions for the guide DNA of the dRS (Fig. 1b): (1) A phosphorylated terminal end of the guide DNA is required for the addition reaction of ethylenediamine. (2) The guide DNA has to bind to the RNA in such a way that the modified terminal base pair (position n, Fig. 1b) locates one or two ribonucleotides up- or downstream (depending on the 50 or 30 modification of the guide DNA with the RG) of the labeling site (position n + 2, Fig. 1b). (3) Adenine and cytosine shall not be placed at the phosphorylated terminal end of the guide DNA to prevent its selfmodification. (4) The size of the guide DNA should not exceed 10 nt to allow a rational design of the dHS but should have a GC content greater than 50% to ensure specific and tight binding to the RNA strand. 5. Before use, heat the CTAB solution at 60  C for 30 min or until the solid is completely dissolved. 6. Store at 4  C. Solutions should be prepared freshly before use. 7. A bright-yellow solution should be observed after the three chemicals are mixed. If the solution is only light yellow or colorless, discard it, validate the used chemicals, and repeat the reaction by mixing again the three chemicals. 8. To prevent degradation of dRS, Solution III is recommended to be prepared immediately before use and should always be kept on ice. 9. dHSs are particularly helpful for labeling ribonucleotides which are base-paired and/or shielded by secondary and/or tertiary structural elements. Rationally designed dHSs that bind up- or downstream of the RNA sequence complementary to the dRS can enhance the hybridization of the dRS to the RNA at moderate temperatures and thus increase the overall labeling efficiency. An additional benefit is that hybridization can be performed at moderate temperature, which reduces degradation of the RG. Design of the dHSs mainly considers factors such as the RNA fold, the GC content, and the predicted melting temperature of the respective dHS/RNA hybrid. If the RNA fold is unknown, secondary structure predictions should be consulted

Dual Site-Specific RNA Labeling

263

for dHS and dRS design. Herein, we show two examples in the context of the btuB riboswitch in which implementation of suitable dHS sequences has a pronounced effect. For the modification of A35 (Fig. 2), two helper sequences are designed to mask short segments on both sides of the target nucleotide in order to open up the stem loop temporarily to allow dRS hybridization. In the absence of any dHS, only 60% of the dRS binds to the RNA at room temperature. In contrast, prior incubation with one equivalent of each dHSs increases the binding efficiency of the dRS to over 95% at ambient temperature. By comparison, temperature-induced unfolding in the absence of any dHS shows only about 70% hybridization efficiency. For A213 (Fig. 2), the effect is even more drastic. Only about 20% of the dRS anneal to the RNA in the absence of dHS at room temperature, whereas the addition of one equivalent of each dHS leads to ~75% bound dRS. Again, thermal melting without dHS produces no more than 50% hybridized strands. 10. We validated that 90 min is the optimal incubation time with sodium periodate to convert the reactive group to an aldehyde [13, 14]. Longer times would result in over-oxidation of the reactive group to an inactive ketone. At the same time, sodium periodate oxidizes the 30 -terminal ribose unit of the RNA to a dialdehyde moiety required for biorthogonal labeling. Again, we validated that 90 min of incubation is sufficient for this reaction. 11. Ethylene glycol is used as a reducing agent to quench sodium periodate. Overnight incubation is acceptable. A longer incubation time may induce RNA degradation. 12. If the volume of the eluted sample does not exceed 100 μL, the lyophilization step can be omitted. 13. This step is to purify the functionalized RNA from dRS, dHSs, and their derivatives and potentially degraded RNA fragments with denaturing PAGE. The percentage of acrylamide varies depending on the molecular weight differences between the RNA and the DNA oligonucleotides. Roughly, we recommend using a low percentage of, e.g., 5%, because of the increased RNA-soaking efficiency. Importantly, it is essential to remove the DNA oligonucleotides before proceeding to the subsequent steps. Without PAGE purification, we observed a dramatic degradation of the RNA after the subsequent dye-coupling reactions. We speculate binding of dRS, dHSs, and/or their derivatives to the RNA may result in RNA conformations that are prone for degradation under the conditions of the dye-coupling reactions.

264

Meng Zhao et al.

14. Load the sample into multiple gel wells as needed, since the total volume including FLB is usually larger than the volume of a single well. 15. Load the non-modified RNA construct as a reference. Ensure that the loaded RNA concentration allows for the visualization via UV shadowing. We use 10 μg RNA, corresponding to 0.1 nmol of the 275 nt long btuB RNA to yield a sufficient optical density at 260 nm irradiation. 16. Avoid touching the gel pieces with the tip of the pipette. Otherwise, unintendedly collected gel pieces/substances will coprecipitate with the RNA upon ethanol precipitation. 17. We recommend measuring the fluorescence emission profile of Solution VI before proceeding to the subsequent dye-coupling steps. An emission peak around 415 nm upon excitation at 275 nm (or 308 nm) is indicative for a successful internal RNA modification (Fig. 3). 18. Be attentive not to blow the sample out of the tube. We recommend adjusting the argon flow rate with a blank test sample in a 1.5 mL tube containing water or TEAA buffer of the same volume as the sample.

Fig. 3 Normalized fluorescence emission profiles (background subtracted) of the 1,N6-ethenoadenines at positions A35 (black) and A213 (red) of Solution VI upon excitation at either 275 nm or 308 nm. Reproduced from Fig. S8 in [12], by permission of Oxford University Press

Dual Site-Specific RNA Labeling

265

19. Because the 30 -dialdehyde is generally more reactive than the alkyne, we recommend performing the 30 -terminal dye-coupling prior to the internal dye-coupling. As we observed that Cy3 is prone to photo-bleaching and has a reduced reorientation probability when attached to internal sites of the RNA, we used Cy3 for the 30 -terminal end. In addition, we recommend using sulfonated carbocyanine dyes due to their increased water solubility and reduced interaction with the RNA backbone owing to their negative charge [28]. 20. Ethanol washing can be repeated until the supernatant becomes nearly colorless. 21. Compared to Cy3, Cy5 is less likely to photo-bleach when labeled at internal sites of RNAs. 22. Ascorbic acid solution must be prepared freshly before use as it is easily oxidized if exposed to air. 23. A NAP-5 desalting column removes free dyes efficiently and is fast compared to multiple rounds of ethanol precipitation. 24. For “biophysical” measurements, the denaturing PAGE purification step is mandatory, as it completely removes trace amounts of free dyes. Be sure the sample was purified by ethanol precipitation and NAP desalting column to remove most excess of dyes before loading to the PAGE gel. 25. We used a Typhoon scanner with fluorescence imaging capabilities to locate the target bands. The fluorescence detection is much more sensitive than UV shadowing. 26. We estimate that about 5% of the total RNA after PAGE purification is dual-color labeled based on UV-vis absorption spectra [12]. This amount is sufficient for single-molecule fluorescence spectroscopy. We used higher-percentage PAGE to get denser bands enabling UV shadowing for lower concentrated RNA samples (Fig. 4). 27. Reverse transcription stop assay is employed to probe the internal fluorophore-labeling sites, in which a primer anneals downstream of the putative labeling site and is extended by a reverse transcriptase (Fig. 5). In our experiments, reverse transcriptase exhibits varied susceptibility to be stalled by the fluorophore-labeled ribonucleotides depending on the different positions. We describe the optimized protocols of AMV reverse transcriptase that is best for the base-paired labeling site (A213) and SuperScript III reverse transcriptase that is best for the labeling site (A35) within an internal loop. 28. In order not to waste the precious dual-color labeled RNA (Solution VIII), the single internally labeled sample was used skipping the 30 -terminal coupling step.

266

Meng Zhao et al.

Fig. 4 Probing the double-labeling by a comparative UV shadowing and fluorescence imaging gel. Example denaturing PAGE (10% w/v, 7 M urea) of the dual-color labeled btuB riboswitch with the Cy3 at the 30 -terminal end and the Cy5 at position A35. Bands on the gel are visualized with both, UV shadowing (λex ¼ 254 nm, top) and fluorescence imaging (excitation at 635 nm or 532 nm, respectively, bottom). Color-coded channels show Cy3 (green) and Cy5 (red) emission. Co-localized dyes appear as yellow bands in the superimposed image of both channels. Control lane with the non-modified btuB riboswitch in the presence of both free dyes shows no fluorescence signal. Thus, intercalation of the non-covalently added dyes Cy3 and Cy5 can be excluded. (Reproduced and adapted from Fig. S12 in [12], by permission of Oxford University Press)

29. Importantly, the 32P-marked primer should be shorter than 20 nucleotides and should be designed in such a way that it binds with its 30 -terminus less than ten nucleotides downstream of the putative labeling site of the RNA (Fig. 5). In this way, we were able to separate the differently extended 32 P-marked primers at a one-nucleotide resolution on a highpercentage (18%) denaturing PAGE. 30. Be aware that the reverse transcriptase is not stalled at the presumed labeling site but inserts one additional nucleotide before stopping and releasing the 32P-marked primer (Fig. 5). This effect is known as delayed chain termination (CT) [21]. 31. The biotin-carrying DNA oligonucleotide is used to immobilize the RNA molecule on the quartz surface via biotinstreptavidin interaction. The oligonucleotide is designed to bind to the 50 -elongated region of the RNA and not to interfere with the RNA’s native fold [19]. We recommend adding a twofold excess of biotin-DNA primer to ensure that all dualcolor labeled RNAs (~5%, see Note 26) are hybridized. The non-hybridized biotin-labeled will be removed by the subsequent native PAGE purification which also removes the degraded RNAs and other impurities (see Note 34).

Dual Site-Specific RNA Labeling

267

Fig. 5 Probing the internal labeling of A213 (a) and A35 (b) by reverse transcription. The autoradiograph of the denaturing PAGE footprinting gel (18% w/v, 7 M urea) shows the original 32P-labeled primers (lane 1), the extended 32P-labeled primers in presence of the non-labeled btuB riboswitch (lane 2), and the internally labeled btuB riboswitch (lane 3). Short DNA oligonucleotides, 32P-labeled primer1 + 7 nt (Ref oligo1, a), and 32 P-labeled primer2 + 3 nt (Ref oligo1, b) are shown as references (lane 4). Delayed chain termination (CT) is observed which reflects the insertion of one additional nucleotide past the modification site [21]. (Reproduced and adapted from Fig. 2 in [12], by permission of Oxford University Press)

32. The precise composition of the folding buffer depends on the folding conditions of the particular RNA studied. 33. Addition of Mg(II) is required for formation of the tertiary structure. 34. The native PAGE purification is not mandatory. However, we recommend its application, as otherwise the excess of biotinlabeled primers that is not hybridized to the RNA would block the streptavidin binding sites (Fig. 6), reducing the number of immobilized RNA molecules.

268

Meng Zhao et al.

Fig. 6 Schematic depiction of the surface-immobilized dual-color labeled btuB riboswitch, with the Cy3 fluorophore at the 30 -end and the Cy5 fluorophore at position A213, located in a double-stranded region of the functional riboswitch (see Note 36). (Reproduced and adapted from Fig. S13 in [12], by permission of Oxford University Press)

35. Empirically, we started with the 105-fold dilution (100 pM) of the reannealed double-labeled RNA sample for single-molecule TIRF imaging. Adjustments may be necessary depending on the RNA concentration recovered from the native PAGE purification and the labeling efficiency. 36. Site-specific biorthogonal fluorescent labeling of RNA is of particular interest for FRET studies, as this method demands labeling with two fluorophores. Further details on singlemolecule FRET or FRET gels can be found elsewhere [7, 12, 22, 23, 25, 29]. 37. The labeling scheme is not limited to fluorescent entities but can be adapted to any reporter molecule such as spin labels for EPR studies [30–32].

Acknowledgement We thank Anna Zemann, Nora Grundmann, and Dr. Sofia Gallo for their valuable feedback regarding the synthesis of the RG precursor. Financial support from the Swiss National Science Foundation [to E.F. and R.K.O.S.], the European Research Council [ERC to R.K.O.S.], SystemsX.ch [to R.K.O.S.], the UZH Forschungskredit [FK-14-096, FK-15-095 to R.B.], the UZH Stiftung fu¨r wissenschaftliche Forschung [to R.K.O.S. and R.B.], the University of Zurich, and the SBFI [COST Action CM1105 to E.F. and R.K. O.S] is gratefully acknowledged. Funding for open access charge: University of Zurich.

Dual Site-Specific RNA Labeling

269

References 1. Rao H, Tanpure AA, Sawant AA, Srivatsan SG (2012) Enzymatic incorporation of an azidemodified UTP analog into oligoribonucleotides for post-transcriptional chemical functionalization. Nat Protoc 7:1097–1112 2. Lavergne T, Lamichhane R, Malyshev DA, Li Z, Li L, Sperling E et al (2016) FRET characterization of complex conformational changes in a large 16S ribosomal RNA fragment site-specifically labeled using unnatural base pairs. ACS Chem Biol 11:1347–1353 3. Seidu-Larry S, Krieg B, Hirsch M, Helm M, Domingo O (2012) A modified guanosine phosphoramidite for click functionalization of RNA on the sugar edge. Chem Commun 48:11014–11016 4. Liu Y, Holmstrom E, Zhang J, Yu P, Wang J, Dyba MA et al (2015) Synthesis and applications of RNAs with position-selective labelling and mosaic composition. Nature 522:368–372 5. Lang K, Micura R (2008) The preparation of site-specifically modified riboswitch domains as an example for enzymatic ligation of chemically synthesized RNA fragments. Nat Protoc 3:1457–1466 6. Smith GJ, Sosnick TR, Scherer NF, Pan T (2005) Efficient fluorescence labeling of a large RNA through oligonucleotide hybridization. RNA 11:234–239 7. Steiner M, Karunatilaka KS, Sigel RKO, Rueda D (2008) Single-molecule studies of group II intron ribozymes. Proc Natl Acad Sci U S A 105:13853–13858 8. Schmitz AG, Zelger-Paulus S, Gasser G, Sigel RKO (2015) Strategy for internal labeling of large RNAs with minimal perturbation by using fluorescent PNA. Chembiochem 16:1302–1306 9. Bu¨ttner L, Javadi-Zarnaghi F, Ho¨bartner C (2014) Site-specific labeling of RNA at internal ribose hydroxyl groups: terbium-assisted deoxyribozymes at work. J Am Chem Soc 136:8131–8137 10. Plotnikova A, Osipenko A, Masevicˇius V, Vilkaitis G, Klimasˇauskas S (2014) Selective covalent labeling of miRNA and siRNA duplexes using HEN1 methyltransferase. J Am Chem Soc 136:13550–13553 11. Baum DA, Silverman SK (2007) Deoxyribozyme-catalyzed labeling of RNA. Angew Chem Int Ed 46:3502–3504 12. Zhao M, Steffen FD, Bo¨rner R, Schaffer MF, Sigel RKO, Freisinger E (2018) Site-specific dual-color labeling of long RNAs for singlemolecule spectroscopy. Nucleic Acids Res 46 (3):e13

13. Egloff D, Oleinich IA, Zhao M, Ko¨nig SLB, Sigel RKO, Freisinger E (2016) Sequencespecific post-synthetic oligonucleotide labeling for single-molecule fluorescence applications. ACS Chem Biol 11:2558–2567 14. Egloff D, Oleinich IA, Freisinger E (2015) Sequence-specific generation of 1,N6-ethenoadenine and 3,N4-ethenocytosine in singlestranded unmodified DNA. ACS Chem Biol 10:547–553 15. Fuchs BM, Glockner FO, Wulf J, Amann R (2000) Unlabeled helper oligonucleotides increase the in situ accessibility to 16S rRNA of fluorescently labeled oligonucleotide probes. Appl Environ Microbiol 66:3603–3607 16. Qin PZ, Pyle AM (1999) Site-specific labeling of RNA with fluorophores and other structural probes. Methods 18:60–70 17. Nahvi A, Barrick JE, Breaker RR (2004) Coenzyme B12 riboswitches are widespread genetic control elements in prokaryotes. Nucleic Acids Res 32:143–150 18. Perdrizet GA, Artsimovitch I, Furman R, Sosnick TR, Pan T (2012) Transcriptional pausing coordinates folding of the aptamer domain and the expression platform of a riboswitch. Proc Natl Acad Sci U S A 109:3323–3328 19. Hilario E (2004) End labeling procedures: an overview. Mol Biotechnol 28(1):77–80 20. Gallo S, Furler M, Sigel RKO (2005) In vitro transcription and purification of RNAs of different size. Chimia 59(11):812–816 21. Sarafianos SG, Clark AD, Tuske S, Squire CJ, Das K, Sheng D et al (2003) Trapping HIV-1 reverse transcriptase before and after translocation on DNA. J Biol Chem 278:16280–16288 22. Selvin PR, Ha T (eds) (2008) Single-molecule techniques: a laboratory manual. Cold Spring Harbor Laboratory Press, New York, NY 23. Zhao R, Rueda D (2009) RNA folding dynamics by single-molecule fluorescence resonance energy transfer. Methods 49(2):112–117 24. Zelger-Paulus S, Hadzic MCAS, Sigel RKO, Bo¨rner R (2020) Encapsulation of fluorescently labeled RNAs into surface-tethered vesicles for single-molecule FRET studies in TIRF microscopy. In: Arluison V, Wien F (eds) RNA spectroscopy: methods and protocols. Methods Mol Biol 2113. Springer, New York, NY 25. Bo¨rner R, Kowerko D, Guiset Miserachs H, Schaffer MF, Sigel RKO (2016) Metal ion induced heterogeneity in RNA folding studied by smFRET. Coord Chem Rev 327-328:123–142

270

Meng Zhao et al.

26. Ko¨nig SLB, Hadzic MCAS, Fiorini E et al (2013) BOBA FRET: bootstrap-based analysis of single-molecule FRET data. PLoS One 8 (12):e84157 27. Bo¨rner R, Kowerko D, Hadzic MCAS et al (2018) Simulations of camera-based singlemolecule fluorescence experiments. PLoS One 13(4):e0195277 28. Steffen FD, Sigel RKO, Bo¨rner R (2016) An atomistic view on carbocyanine photophysics in the realm of RNA. Phys Chem Chem Phys 18 (42):29045–29055 29. Ramirez-Carrozzi V, Kerppola T (2001) Gel-based fluorescence resonance energy

transfer (gelFRET) analysis of nucleoprotein complex architecture. Methods 25(1):31–43 30. Kim N-K, Murali A, DeRose VJ (2004) A distance ruler for RNA using EPR and sitedirected spin labeling. Chem Biol 11 (7):939–948 31. Esquiaqui JM, Sherman EM, Ye J-D, Fanucci GE (2014) Site-directed spin-labeling strategies and electron paramagnetic resonance spectroscopy for large riboswitches. Methods Enzymol 549:287–311 32. Saha S, Jagtap AP, Sigurdsson ST (2015) Sitedirected spin labeling of RNA by postsynthetic modification of 20 -amino groups. Methods Enzymol 563:397–415

Chapter 17 Single-Molecule FRET Assay for Studying Cotranscriptional RNA Folding Heesoo Uhm and Sungchul Hohng Abstract Cotranscriptional RNA folding plays important roles in gene regulation steps such as splicing, transcription termination, and translation initiation. Progression of our understanding of cotranscriptional RNA folding mechanisms is still retarded by the lacking of experimental tools to study the kinetics of cotranscriptional RNA folding properly. In this chapter, we describe fluorescence resonance energy transfer (FRET) assay that enables the study of RNA cotranscriptional folding at the single-molecule level. Key words Cotranscriptional folding, Single-molecule FRET, Elongation complex reconstitution, Riboswitch

1

Introduction Proper folding of functional RNAs is important for their functioning. It is well known that several factors, such as salts, metabolites, and temperature, influence RNA structure [1–3]. RNAs in the cell, however, fold sequentially from the 50 -end to the 30 -end as they are synthesized. Therefore, transcriptional speed and pauses can significantly influence RNA folding [4], and it is now well accepted that the functional structure of RNAs may be determined mainly by transcription kinetics rather than by the global energy landscape of final RNA products [5]. However, the kinetics of cotranscriptional RNA folding is not easily accessible in conventional experiments at the ensemble level. Over the past two decades, single-molecule experiments such as optical tweezers [6], magnetic tweezers [7], and single-molecule FRET [8] have been developed and used to investigate mechanisms of several biological processes. Single-molecule techniques have a unique character that we can observe dynamic conformational changes directly and in real time, which is usually hidden in ensemble-averaged values. Among a number of single-molecule measurement technique, single-molecule FRET is arguably the

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_17, © Springer Science+Business Media, LLC, part of Springer Nature 2020

271

272

Heesoo Uhm and Sungchul Hohng

Fig. 1 Single-molecule fluorescence cotranscriptional folding assay. Acceptor dye-labeled seed RNA is incubated with a template DNA strand and T7 RNAP in a tube for 50 min at 37  C. To assemble an elongation complex, donor dye-labeled UTP and a nontemplate DNA strand are added to the tube and incubated for 20 min. Elongation complexes are immobilized on a polymer-passivated quartz surface using streptavidinbiotin interactions. Elongation is resumed by injecting NTP while RNA folding is observed using a singlemolecule FRET microscope

easiest and most efficient one and provides the opportunity to observe important aspects of molecular behavior that cannot be studied using other single-molecule techniques. We recently developed single-molecule FRET assay to study cotranscriptional folding and successfully characterized the liganddependent cotranscriptional folding of the Escherichia coli thiM riboswitch that regulates the translation initiation [9] (Fig. 1). The aptamer of the riboswitch and the full riboswitch are 80 nt and 150 nt in length, respectively. We found that the aptamer folds into the gene-off state and the conformation is transformed into the gene-on state during transcriptional pausing near the translational start codon. Importantly, ligand binding maintains the riboswitch in the gene-off state during transcriptional pauses, representing a good example of the cotranscriptional effect. In this chapter, we describe in detail the protocols for the assay.

Single-Molecule FRET Assay for Cotranscriptional RNA Folding

2 2.1

273

Materials Equipment

1. Diamond drill with 0.75 mm bits. 2. Glass cover slips (24  40 mm). 3. Quartz microscope slides (1 mm thick). 4. Heat-resistant slide container. 5. Acid-resistant tray. 6. Polyethylene tubing (0.6 mm inner diameter). 7. Syringe pump (14.5 mm inner diameter). 8. Double-sided tape. 9. Razor. 10. Epoxy glue. 11. Nitrogen gas blower. 12. TIRF microscope (see Note 1).

2.2 RNA and DNA Oligonucleotides

1. Custom RNA oligonucleotides (PAGE or HPLC purified) (see Notes 2 and 3). Seed RNA sequence: 50 -ACGACUCGGGGUGCCCUUC UG-C(amine-modified)-GUGAAGGCUGAGAAAUACCCG UAUCACCUGAUCUGGA-30 . 2. Custom DNA oligonucleotides (PAGE or HPLC purified) (see Note 4). Template DNA sequence: 50 -Biotin-CCGGAATTCC GGGGGGAATGTTGGTGAAAAAGGTGTAACGCGTGCG CAGATTGCGCTGAACCCAGCAGGTCGACTTGCATAG TTTGCTCCTGCCATAACGTGAAGAAGCAATGACCTGG TGGTCCGTGACTTCCCTACGCTGGCATTATCCAGATC ACTTTCTCATTCGTGGC-30 . Nontemplate DNA sequence: 50 -GCCACGAATGAG AAAGTCTAGACCTAAATGCCAGCGTAGGGAAGTCAC GGACCACCAGGTCATTGCTTCTTCACGTTATGGCA GGAGCAAACTATGCAAGTCGACCTGCTGGGTTCAG CGCAATCTGCGCACGCGTTACACCTTTTTCACCAAC ATTCCCCCCGGAATTCCGG-30 .

2.3

Reagents

1. Distilled water. 2. Methanol. 3. Acetic acid. 4. Piranha solution: 3:1 mixture of concentrated sulfuric acid and 30% (v/v) hydrogen peroxide. 5. N-(2-aminoethyl)-3-aminopropyltrimethoxysilane silane).

(amino-

274

Heesoo Uhm and Sungchul Hohng

6. Biotin-PEG-NHS ester (Bio-PEG-SC; MW 5000). 7. PEG-NHS ester (mPEG-SVA; MW 5000). 8. 100 mM sodium bicarbonate (see Note 5). 9. Dye-labeled UTP and/or CTP (see Note 6). 10. Dye NHS ester. 11. Dimethyl sulfoxide (DMSO). 12. RNA labeling buffer: 0.1 M sodium tetraborate (pH 8.5). 13. T50 buffer: 10 mM Tris–HCl (pH 8.0), 50 mM NaCl. 14. Streptavidin. 15. Saturated trolox (see Note 7). 16. Protocatechuic acid (PCA). 17. Protocatechuate 3,4-dioxygenase (PCD). 18. Imaging buffer: 10 mM Tris–HCl (pH 8.0), 20 mM KCl, 5 mM MgCl2, 5 mM DTT, 5 mM PCA, 500 nM PCD, saturated trolox (see Note 8). 19. T7 RNA polymerase (T7 RNAP). 20. T7 RNAP 10 buffer: 400 mM Tris–HCl (pH 7.9), 60 mM MgCl2, 10 mM DTT, 20 mM spermidine. 21. 25 mM NTP: 25 mM ATP, 25 mM UTP, 25 mM CTP, 25 mM GTP.

3

Methods

3.1 Flow Cell Preparation

Our customized flow cell is designed for convenient buffer exchange and allows buffer exchange even during single-molecule imaging.

3.1.1 Drilling and Cleaning

To make a reusable flow cell that allows easy buffer exchange, we make holes in a quartz slide. The quartz slides are first cleaned thoroughly for good silanization and clear single-molecule imaging. 1. Drill holes in a quartz slide using a diamond drill with 0.75 mm bits (Fig. 2) (see Notes 9 and 10). 2. Put quartz slides and cover slips into heat-resistant slide containers. 3. Place the containers on an acid-resistant tray. 4. Fill the containers with piranha solution and incubate them for 20 min (see Note 11). 5. Pour out piranha solution carefully.

Single-Molecule FRET Assay for Cotranscriptional RNA Folding

275

Fig. 2 Flow cell assembly. A procedure to make a four-channel flow cell (top) and a scheme to connect the inlet and outlet tubings to a flow cell (bottom) are shown. During single-molecule FRET imaging, buffer exchange can be applied by using the flow cell

6. Rinse thoroughly the slide glasses and cover slips in the containers with distilled water. 7. Repeat steps 3–5 one more time. 8. Rinse the slide glasses and cover slips in the containers with methanol. 3.1.2 Silanization

Silanization coats the quartz surfaces with amines which will be conjugated with PEG-NHS in the next step. 1. Add 5 mL of acetic acid and 1 mL of amino-silane to 100 mL of methanol and mix the solution well. 2. Pour the solution into the containers and incubate them for 30 min. 3. Pour out the solution. 4. Rinse the slide glasses and cover slips with methanol. 5. Rinse the slide glasses and cover slips with distilled water. 6. Dry the slide glasses and cover slips by blowing nitrogen gas, and keep the dried slide glasses and cover slips on a clean surface.

276

Heesoo Uhm and Sungchul Hohng

3.1.3 PEGylation

PEGylation is a method of surface passivation which reduces the nonspecific protein adsorption on the quartz surface. A small fraction of biotin-PEG allows the immobilization of target molecule. 1. Dissolve 2 mg of bio-PEG-SC and 80 mg of mPEG-SVA in 640 μL of 100 mM sodium bicarbonate solution (see Note 12). 2. Centrifuge the solution at 10,000  g for 1 min to spin down the PEG powder. 3. Mix the solution well and centrifuge the solution at 10,000  g for 1 min to spin down remaining powder. 4. Drop 70 μL of PEG solution on a slide glass, and put a cover slip on it carefully not to make air bubbles (see Note 13). 5. Incubate them for 2 h in a dark room at room temperature (see Note 14). 6. Store them in 20  C freezer.

3.1.4 Flow Cell Assembly

1. Take out a slide glass-cover slip pair from the 20  C freezer, and rinse the slide glass and cover slip with distilled water. 2. Dry the slide glass and cover slip by blowing nitrogen gas. 3. Place the slide glass on a clean surface so that the PEGylated side faces upward. 4. Cut a double-sided tape by a clean razor (Fig. 2) and place it on the slide glass. Be careful not to block the holes. 5. Place the cover slip on the double-sided tape so that the PEGylated side faces downward. 6. Make the flow channels tight by carefully rubbing the region of the cover slip on the double-sided tape using a pipette tip. 7. Remove excess tape from the slide using a clean razor. 8. Seal both ends of the flow channels with epoxy glue. 9. Insert tubings into the holes in the slide glass. Inlet tubings should be long enough so that the other ends of inlet tubings can be dipped into buffers (Fig. 2). 10. Fix the tubings with epoxy glue.

3.2 Fluorescence Labeled Seed RNA Preparation

1. Make 20 mM Cy5 NHS ester using DMSO. 2. Add 5 μL of 1 mM amine-modified RNA and 5 μL of 20 mM Cy5 NHS ester to 25 μL of RNA labeling buffer. 3. Incubate the solution for 3 h at room temperature. 4. Add 4 μL of 3 M NaCl and 90 μL of ethanol to the solution. 5. Incubate the solution for 30 min at 20  C. 6. Centrifuge the solution at 13,000  g for 30 min at 4  C. 7. Remove supernatant carefully.

Single-Molecule FRET Assay for Cotranscriptional RNA Folding

277

8. Rinse the pellet with ethanol two times, and dry the pellet. 9. Dissolve the pellet with T50 buffer, and check the labeling efficiency by measuring the absorbance of RNA and fluorescence. 3.3 Elongation Complex Reconstitution

For the higher efficiency of the elongation complex assembly, the concentration of a seed RNA should be the highest among the concentrations of others. We used 800 nM of seed RNA for 40 nM T7 RNAP and 200–400 nM DNAs (see Note 15). During the elongation complex reconstitution, we used a one-nucleotide elongation with a dye-labeled nucleotide to check the activity of the elongation complex. The dye-labeled nucleotide can be incorporated only into active elongation complex with a right template DNA and a corresponding seed RNA. 1. Add 1 μL of T7 RNAP 10 buffer to 7 μL distilled water. 2. Add 0.8 μL of 10 μM seed RNA and 0.2 μL of 10 μM template DNA strand to the solution and mix it well. 3. Add 1 μL of 0.4 μM T7 RNAP to the solution. 4. Incubate the solution at 37  C for 50 min. 5. Add 0.2 μL of 100 mM Cy3-UTP to the solution. 6. Add 0.4 μL of 10 μM nontemplate DNA strand to the solution. 7. Incubate the elongation complex solution at 37  C for 20 min.

3.4 Single-Molecule Fluorescence Measurement

1. Place a flow cell on a microscope. 2. Maintain the temperature of the flow cell using a temperature controller (see Note 16). 3. Connect the outlet tubing to a syringe pump (see Note 17). 4. Put the end of the inlet tubing in the Eppendorf tube filled with 200 μL T50 buffer (Fig. 2). 5. Flush a flow channel with 170 μL T50 buffer by pulling a syringe (see Note 18). 6. Infuse 85 μL of 0.2 mg/mL streptavidin solution into the channel as described in steps 4 and 5, and incubate for 5 min. 7. Remove unbound streptavidin by infusing 85 μL of T50 buffer into the channel. Repeat this step one more time. 8. Add 1 μL of 40 nM elongation complex solution (see step 7 in Subheading 3.3) to 99 μL of imaging buffer (see Note 5). 9. Infuse 85 μL of 400 pM elongation complex solution into the channel, and incubate them for 2 min. 10. Infuse 85 μL of imaging buffer three times to make sure complete removal of free dyes, and check the number of elongation complex molecules immobilized on the surface using a brief excitation of fluorescence for a few seconds. Single-

278

Heesoo Uhm and Sungchul Hohng

molecule images should be clearly separated for reliable data analysis. In our case, the optimum number of fluorescence spots was around 100–200, but the number can change depending on the image sensor. 11. Add 0.8 μL of 25 mM NTP to 99.2 μL of imaging buffer (see Note 8). 12. Start single-molecule imaging and infuse 85 μL of 200 μM NTP solution with imaging buffer into the detection chamber. Elongation activity is monitored via the decrease of the total intensity which is called as protein-induced fluorescence enhancement (PIFE) phenomenon. Cotranscriptional RNA folding is monitored via FRET between a donor dye (Cy3) and an acceptor dye (Cy5). The imaging can be continued for several hundreds of seconds depending on bleaching property of dyes and time resolution you require. In typical cases in which Cy3 and Cy5 dyes were used, we observed RNA dynamics for 300 s with 50 ms time resolution. 3.5 Drift Correction and Autofocusing

Drift correction and defocusing of the sample chamber hinder long-time FRET imaging. We corrected the lateral drift by using 2D translation image registration algorithm after experiments [10]. However, we should solve the defocusing problem in real time. We used an autofocusing method based on optical astigmatism analysis for the purpose [11].

3.6

Resumption of the elongation activity is monitored via the decrease of the total intensity (disappearance of PIFE). Cotranscriptional RNA folding is monitored via FRET between a donor dye (Cy3) and an acceptor dye (Cy5). FRET efficiency is determined as the calculation of the acceptor signal divided by the sum of the donor and acceptor signals at donor excitation. For instance, we observed the PIFE phenomenon as a signal of elongation and FRET change upon ligand binding in 200 μM thiamine pyrophosphate (TPP) solution (Fig. 3a). TPP binding on the TPP riboswitch induces the stabilizing of the aptamer structure which inhibits the transition into the gene-on state. A comparison of FRET histograms before and after TPP injection reveals a large increase in the high FRET population in the presence of TPP (Fig. 3b). However, full length transcript has the lower portion of the high FRET population in the same condition (Fig. 3c, see Note 19). In the higher concentration of TPP during transcription, we can observe the increase of the high FRET population in the final 200 μM TPP solution. This shows TPP’s ability to maintain the high FRET state cotranscriptionally. To emulate the effect of a transcriptional pause, we used transcripts of varying lengths controlled by template DNA lengths (see Note 19). We observed a large TPP effect when the ribosome binding site (RBS) and translation start codons were exposed at

Data Analysis

Single-Molecule FRET Assay for Cotranscriptional RNA Folding

279

Fig. 3 Cotranscriptional folding of transcript. (a) Representative time traces of the single-molecule cotranscriptional experiment. Fluorescence intensities of a donor and an acceptor are colored in green and red, respectively, while the corresponding FRET appears in black. The total intensity is shown in orange. NTP (200 μM) and TPP (200 μM) were injected at 30 s and 100 s, respectively (dashed lines). (b) FRET histograms before (top) and after 200 μM TPP injection (bottom). (c) The fraction of the high FRET population of the aptamer (R89) and the full riboswitch (R159) with varying TPP concentrations during elongation. TPP binding during elongation keeps the riboswitch in the high FRET state. (d) The fraction of the high FRET population of the riboswitch with varying lengths. The riboswitch without TPP shows the transition into the low FRET state (i.e., gene-on state) at the start region of translation. On the other hand, for the riboswitch with TPP, the transition is inhibited by TPP binding, resulting in the robust high FRET population for all lengths of transcript

transcription stalling (Fig. 3d). Furthermore, the concentration of NTP could be controlled to study the effect of the elongation speed.

4

Notes 1. We used a homebuilt prism-type total internal reflection fluorescence (TIRF) microscope equipped with an electronmultiplying charge-coupled device (EMCCD) for singlemolecule FRET technique. 532 and 635 nm lasers were used for the excitation of donor (Cy3) and acceptor (Cy5) dyes. The Cy3 and Cy5 signals are separated using a 540 nm dichroic mirror. We used a temperature control system (Live Cell

280

Heesoo Uhm and Sungchul Hohng

Instrument) to maintain the sample temperature. The other details of the microscope can be found elsewhere [12–14]. 2. To monitor cotranscriptional folding of RNA using singlemolecule FRET, we need to label RNA transcript with a FRET pair. The dye-labeling positions should be selected so that they can provide different FRET values for critical conformations of RNA folding (see Note 3). Before assembling the elongation complex, we internally labeled the seed RNA with an acceptor (Cy5) using an amine-modified base and aminereactive Cy5 [12]. On the other hand, a donor was added to the 30 -end of the seed RNA using a dye-labeled uracil after the assembly of the core part of the elongation complex [9]. Secondary structures in the seed RNA can lower the efficiency of elongation complex formation. In that case, you should try seed RNAs with different length. For example, the seed RNA that we used is shown. 3. To check the FRET states of the target RNA structure, it is highly recommended to prepare a long synthetic RNA labeled with two dyes and observe the FRET states and dynamics of them before the cotranscriptional folding experiment [8]. 4. To assemble the elongation complex, we need two DNA strands: a template DNA strand and a nontemplate DNA strand. When you design DNA strands, you need to consider the following two things: First, the seed RNA binding region of the template DNA should not be base-paired with the nontemplate DNA for higher efficiency of the elongation complex assembly [15]. Second, the elongation complex stalls at the end of DNA, and therefore the region of RNA transcript involved in cotranscriptional folding is determined and controlled by the DNA length. When you design DNA strands, DNA length adequate for your purpose should be selected. For investigating cotranscriptional folding effects, however, a series of DNA with different lengths and the same seed RNA should be used. 5. 100 mM sodium bicarbonate should be fresh. It is recommended to use the solution in 30 min. 6. For example, we used 100 mM Cy3-UTP. 7. Add 25 mg of Trolox into 50 mL of distilled water, and incubate it on rotator at room temperature under room light for a day. Filter the solution with 0.2 μm membrane filter. 8. It is highly recommended to prepare the imaging buffer just before the injection into the flow cell. 9. It is important to drill right-size holes. If a hole is too large, epoxy glue leaks into the flow channel. If a hole is too small, tubing cannot be inserted into the hole.

Single-Molecule FRET Assay for Cotranscriptional RNA Folding

281

10. Drilled quartz slides can be reused. For reuse, separate quartz slides from cover slips by boiling with some detergent in a microwave oven, and wash impurities from the quartz slide using detergent. 11. The piranha solution should be treated with great care. Hot solution can break the container. Be aware of how to deal with liquid waste. 12. This amount of solution is enough for the PEGylation of nine pairs of slides and cover slips. 13. It is recommended to finish this step within 20 min after fresh PEG solution is prepared. 14. Avoid drying of PEG solution during the reaction. For example, to maintain a proper humidity, we used a pipette tip case with small amount of water at the bottom. Slides and cover slips are placed on top of a rack inside a pipette tip case. 15. For the higher efficiency of the elongation complex assembly, the concentration of T7 RNAP should be the lowest and the concentration of a nontemplate DNA strand should be higher than the concentration of a template DNA strand. 16. We maintained the sample temperature at 37  C. 17. Tubing can be connected to other tubing or a syringe by using a syringe needle, for example, a 23-G needle fitted into our tubing nicely. 18. The end of the inlet tubing should be kept inside buffer to avoid air bubbles. Always prepare more solution than the amount of injection. For example, prepare 100 μL solution and inject 85 μL of the solution. 19. We referred each RNA transcript as “R#,” e.g., R89, where # refers to the position number of the last nucleotide exposed outside RNAP during transcription stalling after the end of elongation. Template DNA length determines the stalling position of RNA elongation (see Note 4). In our experiment, the last sequence of RNA exposed outside RNAP is located at 17th sequence from the end of template DNA [9].

Acknowledgments This work was supported by a Creative Research Initiative program (2009-0081562) to SH.

282

Heesoo Uhm and Sungchul Hohng

References 1. Badelt S, Hammer S, Flamm C, Hofacker IL (2015) Chapter 8–thermodynamic and kinetic folding of riboswitches. In: Chen S-J, BurkeAguero DH (eds) Methods in enzymology. Academic Press: Cambridge, Massachusetts, US, pp 193–213. https://doi.org/10.1016/ bs.mie.2014.10.060 2. Draper DE (2008) RNA folding: thermodynamic and molecular descriptions of the roles of ions. Biophys J 95:5489–5495. https://doi. org/10.1529/biophysj.108.131813 3. Bhaskaran H, Russell R (2007) Kinetic redistribution of native and misfolded RNAs by a DEAD-box chaperone. Nature 449:1014–1018. https://doi.org/10.1038/ nature06235 4. Neuman KC, Abbondanzieri EA, Landick R et al (2003) Ubiquitous transcriptional pausing is independent of RNA polymerase backtracking. Cell 115:437–447. https://doi.org/10. 1016/S0092-8674(03)00845-6 5. Pan T, Sosnick T (2006) Rna folding during transcription. Annu Rev Biophys Biomol Struct 35:161–175. https://doi.org/10. 1146/annurev.biophys.35.040405.102053 6. Anthony PC, Perez CF, Garcı´a-Garcı´a C, Block SM (2012) Folding energy landscape of the thiamine pyrophosphate riboswitch aptamer. Proc Natl Acad Sci 109:1485–1489. https:// doi.org/10.1073/pnas.1115045109 7. Uhm H, Bae S, Lee M, Hohng S (2016) Single-molecule FRET combined with magnetic tweezers at low force regime. Bull Kor Chem Soc 37:408–410. https://doi.org/10. 1002/bkcs.10688 8. Uhm H, Hohng S (2017) Ligand recognition mechanism of thiamine pyrophosphate riboswitch aptamer. Bull Kor Chem Soc

38:1465–1473. https://doi.org/10.1002/ bkcs.11328 9. Uhm H, Kang W, Ha KS et al (2018) Singlemolecule FRET studies on the cotranscriptional folding of a thiamine pyrophosphate riboswitch. Proc Natl Acad Sci 115:331–336. https://doi.org/10.1073/pnas.1712983115 10. Guizar-Sicairos M, Thurman ST, Fienup JR (2008) Efficient subpixel image registration algorithms. Opt Lett 33:156–158. https:// doi.org/10.1364/OL.33.000156 11. Hwang W, Bae S, Hohng S (2012) Autofocusing system based on optical astigmatism analysis of single-molecule images. Opt Express 20:29353–29360. https://doi.org/10.1364/ OE.20.029353 12. Roy R, Hohng S, Ha T (2008) A practical guide to single-molecule FRET. Nat Methods 5:507–516. https://doi.org/10.1038/nmeth. 1208 13. Hohng S, Lee S, Lee J, Jo MH (2014) Maximizing information content of singlemolecule FRET experiments: multi-color FRET and FRET combined with force or torque. Chem Soc Rev 43:1007–1013. https://doi.org/10.1039/C3CS60184F 14. Kapanidis AN, Lee NK, Laurence TA et al (2004) Fluorescence-aided molecule sorting: analysis of structure and interactions by alternating-laser excitation of single molecules. Proc Natl Acad Sci 101:8936–8941. https:// doi.org/10.1073/pnas.0401690101 15. Daube SS, von Hippel P (1992) Functional transcription elongation complexes from synthetic RNA-DNA bubble duplexes. Science 258:1320–1324. https://doi.org/10.1126/ science.1280856

Chapter 18 Characterizing Complex Nucleic Acid Interactions of LINE1 ORF1p by Single Molecule Force Spectroscopy M. Nabuan Naufer and Mark C. Williams Abstract The L1 retrotransposon is the dominant transposable element in mammalian genomes. L1 comprises at least 20% of the human genome. While most L1 regions are inactive, a few still retain the ability to retrotranspose. L1 encodes two proteins, ORF1p and ORF2p, which are required for retrotransposition. During retrotransposition, ORF2p functions as the reverse transcriptase and the endonuclease. ORF1p is a nucleic acid chaperone that binds nucleic acids with high affinity. However, to date, a detailed mechanistic understanding of ORF1p function in L1 retrotransposition is lacking. The single molecule DNA stretching methods described here have been extensively used to understand ORF1p’s complex nucleic acid binding properties. By correlating these properties to ORF1p’s ability to support L1 retrotransposition in in vivo cell-culture based assays, these studies have significantly contributed to advance the understanding of ORF1p function. Although described in the context of ORF1p, these methods provide a general mechanism to study complex protein-DNA interactions. Key words LINE1, ORF1p, Retrotransposon, Single molecule, Nucleic acid chaperone

1

Introduction The long interspersed nuclear element 1 (LINE1, L1) is an autonomously replicating retrotransposon that has been amplifying and evolving since before the mammalian radiation (~100 Myr) and now comprises ~20% of certain mammalian genomes [1–3]. L1 propagates by reverse transcribing its transcript into genomic DNA (Fig. 1a). While this is similar to retroviruses and retrovirallike retrotransposons, L1 replicates by the dramatically different process of target site-primed reverse transcription (TPRT) [4, 5] (Fig. 1b). Because L1 activity can also copy other RNAs into genomic DNA, it has generated more than 40% of the mass of many mammalian genomes [6, 7]. Despite its deleterious [8–10] and catastrophic effects [11, 12], L1 replication and evolution persist in modern humans [13–15], creating genetic diversity and various genetic alterations [16–19]. Therefore, understanding the

Tilman Heise (ed.), RNA Chaperones: Methods and Protocols, Methods in Molecular Biology, vol. 2106, https://doi.org/10.1007/978-1-0716-0231-7_18, © Springer Science+Business Media, LLC, part of Springer Nature 2020

283

284

M. Nabuan Naufer and Mark C. Williams

Genomic DNA L1 RNP

L1 DNA New L1 DNA Target Site

Transcription

TPRT ORF1p

1st L1 cDNA synthesis

ORF2p 2nd L1 DNA synthesis

L1 RNA L1 RNP

Translation

New L1 insertion

Fig. 1 L1 retrotransposition cycle and TPRT. The genomic L1 DNA is transcribed, and the L1 transcript encodes for two proteins, ORF1p and ORF2p. The two proteins associate with their encoding transcript (cis preference) to form a ribonucleoprotein complex (RNP), which mediates the new L1 insertion via TPRT as shown in the inset. During TPRT, the target site is cleaved and then annealed with the L1 RNA to prime the reverse transcription and extend the first L1 cDNA. Then the second genomic site is cleaved and primed with the first L1 cDNA to synthesize the second L1 DNA. During TPRT, ORF2p functions as the endonuclease and the reverse transcriptase. Presumably, ORF1p’s nucleic acid chaperone capabilities mediate strand exchange reactions required to prime the cDNA synthesis and facilitate the nucleic acid rearrangements required during reverse transcription

molecular mechanism by which L1 replicates remains a formidable biological question. Mammalian L1 elements are 6–7 kb in length and encode for two proteins, ORF1p and ORF2p, which are required for retrotransposition [20]. ORF1p and ORF2p associate with their encoding transcript to form a ribonucleoprotein complex (RNP), which mediates retrotransposition [21–24] via TPRT. Highly conserved endonuclease and reverse transcriptase domains in ORF2p indicate its replicase function in TPRT [20, 25, 26], in which the L1 RNA transcript is primed from a 30 hydroxyl and reverse-transcribed at the nicked genomic site [27]. In contrast, ORF1p lacks homology with any protein of known function, rendering its role in retrotransposition less clear. However, it is the major protein component of the L1RNP and present in large molar excess over the presumed single molecule of ORF2p [21, 28]. ORF1p has a unique domain composition with an N-terminal coiled coil motif that mediates

Single Molecule Characterization of ORF1p-Nucleic Acid Interactions

285

trimerization, a noncanonical RNA recognition motif (RRM) domain and a C-terminal domain (CTD) [28–32] . In an attempt to elucidate ORF1p function, single molecule DNA stretching experiments have been utilized in conjunction with bulk biochemical in vitro assays as well as cell culture-based in vivo assays (reviewed in [33]). Biophysical characteristics as observed in single molecule studies are typically correlated to the protein’s ability to support retrotransposition in in vivo assays. Single molecule mutational analyses of mouse ORF1p variants extensively characterized the nucleic acid chaperone mechanism of ORF1p that presumably facilitates L1 activity during TPRT [34–36]. Furthermore, recent single molecule studies on human ORF1p variants showed that ORF1p’s ability to rapidly form stable nucleic acidbound oligomers is critical in L1 retrotransposition [37]. Because single molecule DNA stretching experiments are extremely sensitive to protein binding, they can be used to differentiate protein variants that may appear indistinguishable in bulk biochemical assays. Therefore, the methods described below can be used as a general procedure to accurately dissect nucleic acid interactions for proteins involved in complex nucleic acid rearrangements, such as ORF1p.

2

Materials 1. Clean buffer: 100 or 50 mM NaCl, 10 mM HEPES (pH 7.5). 2. Protein solution: 50 or 100 nM mouse [34–36] and human/ primate [37] ORF1p variants diluted in clean buffer. 3. Optical tweezers: a dual-beam single trap optical tweezers instrument with sub-piconewton force resolution [38] (see Notes 1). 4. DNA labeling buffer: 50 mM NaCl, 10 mM Tris–HCl, 10 mM MgCl2, 1 mM DTT, pH 7.9 (1 NEBuffer2, NEB). 5. λ DNA: 5 μg of 48.5 kbp λ DNA. 6. 0.4 mM Biotin-14-dCTP. 7. 0.4 mM Biotin- 14-dATP. 8. 100 mM dTTP. 9. 100 mM dGTP. 10. DNA polymerase: five units/ul DNA polymerase I, large (Klenow) fragment. 11. Elution buffer: 10 mM Tris–HCl pH 8.5. 12. QIAEX II Gel Extraction Kit. 13. Labeled DNA: 0.1 ng/ml 48.5 kbp λ DNA labeled with multiple biotins at both 30 termini diluted in clean buffer. 14. Bead solution: 0. 001% (solids) 5 μm streptavidin functionalized beads (Bangs Lab) diluted in clean buffer.

286

3 3.1

M. Nabuan Naufer and Mark C. Williams

Methods DNA Labeling

1. Incubate the λ DNA in the DNA labeling buffer with 50 nmol dGTP, 50 nmol dTTP, 5 nmol Biotin-14-dATP, 5 nmol Biotin14-dCTP, and one unit of DNA polymerase at 37  C for 1 hour to fill in the 50 overhangs [39, 40]. 2. Purify the labeled DNA using the QIAEX II Gel Extraction Kit and store in elution buffer.

3.2 Capturing a DNA Molecule in the Optical Tweezers

1. Flow in the bead solution into the flow cell. 2. Immobilize one bead by suction onto the micropipette tip and another bead in the optical trap. 3. Rinse out the free beads in the flow cell with clean buffer. 4. Introduce the labeled DNA into the flow cell and capture a single molecule of DNA between the two beads (see Notes 2). 5. After capturing the DNA, rinse the cell with clean buffer to remove free DNA molecules in the solution. 6. Displace the micropipette to extend the DNA molecule until a set tension is achieved [34–37], depending on the type of experiment (described in Subheading 3), and exchange the surrounding buffer with the protein solution (see Notes 3).

3.3 Quantifying the Equilibrium Binding Affinity and Nucleic Acid Chaperone Activity

The overstretching transition (see Notes 4) of a double-stranded DNA (dsDNA) force extension cycle (FEC) in the presence of nucleic acid chaperone proteins can be used to measure its equilibrium binding affinity (see Notes 5) as well as to directly quantify its nucleic acid chaperone activity. To quantify the overstretching transition, 1. Capture a DNA molecule in the optical tweezers. 2. Exchange the surrounding buffer with the protein solution while holding the DNA at a force of ~10 pN (see Notes 6). 3. Obtain the FEC of the protein-DNA complex (see Notes 7). 4. The overstretching transition width, δF (see Notes 4), is determined by dividing the slope of the tangent drawn at the force midpoint by the change in the DNA length during the transition (Fig. 2a) [41]. This force midpoint corresponds to the force at half the extension change between the double-stranded (ds) and single-stranded (ss) regimes of the FEC. 5. δF is a direct measure that is possitively correlated to the protein’s nucleic acid chaperone activity (see Notes 8). However, excessive DNA aggregation, which may negatively affect the nucleic acid chaperone capabilities, could also result in an increased overstretching transition width. Therefore, it is useful

Single Molecule Characterization of ORF1p-Nucleic Acid Interactions

A

287

B 120

dsDNA aggregation

ssDNA aggregation

ds agg ss agg 1

δF

60 30 0 0.25

DNA only 15 nM WT mouse ORF1p

Normalized Activity

90 Force (pN)

nd nd nd nd nd nd + + + nd nd nd nd nd nd nd

0.75

Retro F

0.5 0.25 0

0.35 0.45 0.55 Extension (nm/bp)

Fig. 2 Single molecule characterization of nucleic acid chaperone activity. (a) FECs (extension, solid lines; return, dashed lines) of DNA with no protein (black) and in the presence of 15 nM mouse ORF1p WT (blue) (Figure adapted from [33]). The transition width, δF, is determined by the intersection of the dashed red lines on the figure, which represent ds- and ss- DNA regimes of the ORF1p-DNA complex. (b) The relative increases in the transition width, ΔF (ΔF ¼ δF–δF0 where δF, δF0 are the measured transition widths in the presence and absence of protein, respectively), for several mouse ORF1p variants (blue) are compared with their retrotransposition activity in an in vivo cell-culture assay (red) [34–36] . The values are normalized with respect to the wild-type values. Extensive aggregation of ss- or ds DNA in comparison with the wild-type protein are denoted with a plus sign. ’nd’ denotes not determined

to quantify DNA aggregation in order to fully evaluate the nucleic acid chaperone properties of the protein (see Notes 9). 6. To determine the binding affinity, the concentration dependence of the transition width can be fit to a simple binding isotherm: δF ðC Þ ¼ δF 0 þ

ðδF max  δF 0 Þ 1 þ KCd

ð1Þ

where C is the protein concentration, δF0 is the measured transition width in the absence of protein, and δFmax and Kd are fitting parameters that represent the maximum transition width and the equilibrium dissociation constant, respectively. 3.4 Quantifying Oligomerization Kinetics on ssDNA

Single molecule DNA stretching assays were also carried out to probe ssDNA binding kinetics of human ORF1p variants in the context of its previously measured oligomerization properties [37, 42]. In the absence of protein, a force-melted dsDNA rapidly reanneals during return, indicated by the almost complete reversal of the dsDNA extension curve, which exhibits minimal hysteresis (Figs. 2 and 3, dotted black line). Incubating the overstretched dsDNA with protein exposes the melted ssDNA regions for protein

M. Nabuan Naufer and Mark C. Williams

A

B

90 2. Incubation with ORF1p

1. Stretch

Force (pN)

60

30 3. Return

0 0.3

Incubation time 0 min 3 min 6 min 12 min 24 min 48 min Sat

0.4 0.5 Extension (nm/bp)

0.6

Fast fraction ssDNA bound

288

1

Modern retro + Ancestral retro + Mosaic retro _

0.5

0

0

1000 2000 3000 Incubation time (s)

Fig. 3 Single molecule characterization of ORF1p oligomerization. (a) Representative return curves of ORF1pssDNA complexes for different incubation times (Figure adapted from [33]). The sequential experimental steps are denoted in numbered gray text. Solid and dashed black lines are the stretch and return of a bare DNA. The colored empty circles are the return curves of the DNA after incubating it with 2 nM modern human wild-type ORF1p (111p) at ~70 pN for different time periods. The ORF1p-ssDNA saturated curve that exhibited the maximum fractional binding is represented in gold. The return curves are fitted to Eq. 2 (detailed description available in [37]) to determine the ssDNA-bound fraction of different ORF1p conformers. (b) The fast ssDNAbound fraction ( ffast) as a function of time for modern human (111p), its resuscitated ancestral primate (555p), and a mosaic ORF1p (151p) in which nine residues in the coiled coil are replaced with the corresponding ancestral residues. As shown in the legend to the figure, only the mosaic ORF1p variant was defective in in vivo retrotransposition assays, although all three protein variants were equally stable in the cell and were indistinguishable in in vitro nucleic acid chaperone and binding assays [37]. The temporal evolution of the ffast is modeled as a sum of increasing and decreasing exponential functions and is shown in solid lines. ffast saturated rapidly for all three protein variants indicating rapid ssDNA association kinetics. However, the fast fraction decreases with incubation time as the protein transforms into more stable conformers ( fslow, fint, data not shown here). Therefore, the decrease rate of the fast fraction is representative of the rate at which ORF1p transforms into more stable conformers and was at least ten-fold slower for the defective mosaic ORF1p variant

binding, thereby preventing subsequent duplex formation. Because ssDNA is much longer than dsDNA at forces between 6 pN and ~60 pN, inhibition of duplex formation results in an increased DNA length in this force regime (Fig. 3a). Therefore, the relative DNA length increase in the return curve after incubation with protein is an accurate measure of the ORF1p-bound ssDNA fraction, which has also been used to characterize the ssDNA binding of the innate immune protein APOBEC3G [43]. 1. Capture a DNA molecule in the optical tweezers. 2. Overstretch the dsDNA up to ~75 pN, converting most of the dsDNA into ssDNA (see Notes 4). 3. Replace clean buffer in flow cell with protein solution.

Single Molecule Characterization of ORF1p-Nucleic Acid Interactions

289

4. Incubate the overstretched DNA with protein via a force feedback loop (see Notes 10). 5. After incubation, retract the DNA and obtain a return curve of the DNA-ORF1p complex. This curve reflects the fraction of protein-bound ssDNA. 6. Obtain a subsequent FEC re-stretching the DNA-ORF1p complex immediately after the initial return. This curve represents the stably bound ORF1p oligomers (see Notes 11). 7. Obtain the protein-saturated DNA curve by incubating the overstretched DNA at a high protein concentration (>>Kd) and for long timescales (see Notes 12). 8. At any given force below the melting plateau, the extension of the DNA attained during either the return after incubation or the subsequent stretch represents the sum of dsDNA and ORF1p-bound ssDNA fractions. Therefore, to find the protein-bound ssDNA fraction ( f ), the data can be modeled as a linear combination of the curves for a bare dsDNA (bds(F), see Notes 13), with no protein bound ( f ¼ 0) and an ORF1psaturated ssDNA (bsat(F), see Notes 12), for which all the possible binding sites of ssDNA are occupied by ORF1p ( f ¼ 1), where: b ðF Þ ¼ ð1  f Þb ds ðF Þ þ fb sat ðF Þ:

ð2Þ

Fitting the subsequent stretch curve obtained in step 6 to Eq. 2 will yield the ssDNA fraction that is bound to stable ORF1p oligomers ( f ¼ fslow). 9. Fitting the return curve after protein incubation will yield the sum of all the ssDNA fractions that are bound to different protein conformers. In this case, the data was best described by a phenomenological model with three protein-bound ssDNA fractions—fslow, fint, ffast—that retain fast, intermediate, and slow dissociation kinetics (Figs. 3b and 4). Detailed information on the data analysis is available in [37]. 3.5 Quantifying Dissociation Kinetics of ssDNA-Bound ORF1p

The following method allows direct measurement of the dissociation timescales of ssDNA-bound ORF1p. Different binding conformers will yield distinct dissociation kinetics. Therefore, this method was also used to confirm the existence of distinct ssDNAbound ORF1p conformers. 1. Capture a DNA molecule in the optical tweezers. 2. Overstretch the dsDNA up to ~75 pN, converting most of the dsDNA into ssDNA. 3. Replace clean buffer in flow cell with protein solution. 4. Incubate the overstretched DNA with protein via a force feedback loop.

M. Nabuan Naufer and Mark C. Williams

B

Force (pN)

90

1. Stretch

60

3. Return

30

4. Force Feedback

0

C 1000

0.44

2. Incubation with ORF1p

Time constant (s)

A

Extension (nm/bp)

290

0.4

0.36 0.3

0.4 0.5 0.6 Extension (nm/bp)

100

10

111p 555p 151p

1 0

200 400 Time (s)

30 40 50 Stopped force (pN)

Fig. 4 Dissociation kinetics of ssDNA-bound ORF1p. (a) Representative data for directly measuring dissociation kinetics of ssDNA-bound ORF1p (figure adapted from [37]). The sequential experimental steps are denoted in numbered gray text. Overstretched dsDNA (solid black) is incubated with ORF1p (2 nM, 111p for 360 s), and then the return is stopped and maintained at a constant force (40 pN) via a force feedback loop. (b) The temporal change in extension during the constant force feedback loop (green circles) is fit to a double exponential function of time (solid black). Two time constants, τint and τfast, represent the characteristic dissociation time constants of the ssDNA-bound 111p populations exhibiting intermediate and fast dissociation kinetics, respectively. (c) Force dependence of the dissociation time constants (τint, circles; τfast, triangles) for the denoted ORF1p variants. Overall variant averages are fit to a simple exponential functions and are shown in dashed gray lines

5. During the return after incubation, stop and maintain the ORF1p-bound ssDNA complex at a given force via a force feedback loop. 6. ORF1p dissociation events will allow additional duplex formation in the ssDNA regions that otherwise were inhibited due to bound protein (Fig. 4). As more DNA bases reanneal, the tension along the template will start to increase. The force feedback mechanism responds to these sudden force increments by releasing the extension of the protein-DNA complex. Therefore, the temporal decrease in the extension directly quantifies the net dissociation of the ssDNA-bound ORF1p as a function of time (see Notes 14). 3.6 Future Experiments: Quantifying ORF1pssDNA Binding Dynamics

The methods described so far probe the properties of ORF1p’s nucleic acid chaperone activity and the ability to rapidly form stable conformers on ssDNA, both of which have been shown to play a critical role in L1 retrotransposition. Although the question of how exactly either of these properties support L1 activity remains untested, it is reasonable to assume that while the chaperone properties play a primary role during TPRT, the oligomerization properties mediate the packaging of the L1 transcript into an ORF2pcontaining RNP. To further investigate ORF1p’s nucleic acid packaging properties, future experiments are required to investigate the conformational aspects of the ORF1p-ssDNA binding dynamics.

Single Molecule Characterization of ORF1p-Nucleic Acid Interactions

291

Single molecule DNA stretching experiments with preformed ssDNA (rather than force-melted dsDNA) will allow direct probing of the binding dynamics and the supramolecular properties of the resulting ssDNA-ORF1p complex. Because terminal labeling of a dsDNA molecule for optical tweezers experiments is relatively easy, in situ single-strand digestion by exonucleolysis is a relatively easy method to generate a single ssDNA molecule for binding experiments. This can be achieved with a dsDNA molecule that contains a single 30 recessed terminal that provides a primer-template substrate to a force-controllable exonuclease such as the commercially available T7 DNA polymerase [44–46].

4

Notes 1. Two counterpropagating laser beams are focused with a pair of water immersion high NA objectives (Nikon), forming an optical trap inside a custom-made flow cell. The flow cell is composed of four inlet tubes on one end and an outlet tube (Warner Instruments) on the other end. A micropipette tip (WPI) extends from the top of the flow cell to allow bead manipulation. Solutions containing clean buffer, beads, DNA, and protein are attached to the four inlets separately. The flow cell is fixed to a piezoelectric stage (nPoint) with sub-nanometer translation resolution. 2. The correlated movement of the immobilized bead on the micropipette tip and the bead in the optical trap indicates a formation of a DNA tether in between these two beads (Fig. 5). 3. Single molecule FEC of a DNA molecule is very sensitive to the bound protein fraction. Therefore, it is important to take extreme care to maintain the effective protein concentration in the reaction buffer for consistent reproducibility of the data. Because ORF1p aggregates in low-salt conditions and also is extremely sensitive to freeze-thaw cycles, it is best to prepare single-use ORF1p aliquots and store them in 80  C. Thaw the protein aliquot right before it is diluted in the reaction buffer and immediately use it in the optical tweezers in DNA binding experiments. 4. The overstretching transition of a bare dsDNA occurs at ~62 pN and is observed as a plateau (Fig. 2a, black line), a rapid increase of the extension in the narrow 60–65 pN force regime in the FEC. Although this plateau may represent transition from B-form dsDNA into S-DNA [47, 48] in high-salt conditions (>0.15 M), at salt conditions used in these studies this primarily represents the helix-coil transition due to forceinduced melting of the dsDNA. The overstretching transition is quantified by defining the transition width (δF) of a FEC to

292

M. Nabuan Naufer and Mark C. Williams

Glass micropipette

Bead in optical trap

Immobilized bead DNA

Laser beam

Microscope objectives Laser Beam

Fig. 5 Optical tweezers. Schematic depiction of an optical tweezers system (bottom) and the camera view inside the flow cell with a tethered DNA molecule (Figure adapted from [33]). A single molecule of λ DNA (48.5 kbp) is attached by its biotinylated ends to streptavidin-coated polystyrene beads. One bead is immobilized by a glass micropipette attached to a flow cell, while the other is held in an optical trap. The optical trap is created by converging two counterpropagating laser beams to overlap in space using microscope objectives. By moving the glass micropipette attached to the flow cell, the tethered DNA molecule is stretched, and the force exerted on the DNA is measured as a function of extension in the presence and absence of protein

be the force interval over which the DNA molecule transitions from an entirely ds to its ss state. 5. The FEC of a DNA molecule is extremely sensitive to the presence of DNA binding proteins such as ORF1p, and significant variation in these FECs is observed when the protein binding results in a change in conformation or a deformation in the DNA at the protein binding site. Although aggregation of DNA is evidence of constriction at the binding site that may occur due to protein-DNA and/or DNA-bound protein-protein interactions, to date, the exact mechanistic nature of dynamic ORF1p-DNA binding conformations is not known. Nevertheless, these variations provide information on the conformation of the DNA-protein complex, allowing us to quantitatively characterize the protein-DNA binding interactions. The most distinct feature of the changes induced by the presence of nucleic acid chaperone proteins in the FEC is the significant increase in the slope of the overstretching transition due to protein binding. Because this is a consequence of ORF1p binding to both ssDNA and dsDNA with comparable

Single Molecule Characterization of ORF1p-Nucleic Acid Interactions

293

affinities, variation in the overstretching transition is a measure of binding to both forms of DNA. Therefore, the overstretching transition can be used to quantify the average composite equilibrium binding affinity of ORF1p to DNA. 6. Because ORF1p binds dsDNA, incubating the DNA with protein at near zero-force may form large protein-mediated DNA loops and introduce large force fluctuations in the FEC in the low force regime. This can be avoided by holding the DNA at a force of ~10 pN during protein incubation. Then retract the DNA down to zero-force and immediately obtain the FEC. A more systematic method is to program the data acquisition software to complete a full-range FEC that starts and ends at ~10 pN. 7. A complete FEC that also contains data beyond the helix-coil transition (>65 pN) is required to quantify the transition width. 8. The transition width is positively correlated to the protein’s ability to function as a nucleic acid chaperone [34–36]. In the absence of protein, the transition width of a DNA molecule is relatively small because the transformation from its ds to ss form occurs cooperatively, in which multiple base pairs melt simultaneously. Therefore, an increase in the transition width is indicative of a less cooperative transition due to protein binding. This is presumably due to the lowering of the energetic barrier to rearrange nucleic acid secondary structure by the protein’s nucleic acid chaperone capabilities, thereby promoting melting of small numbers of base pairs [49, 50]. HIV-1 NC, a model nucleic acid chaperone protein, dramatically increases the transition width [49–52], as also observed with ORF1p. Thus, the transition width is a quantitative measure of the protein’s nucleic acid chaperonee capabilities, which is shown to positively correlate with ORF1p’s ability to support L1 retrotransposition in in vivo cell culture-based assays. 9. Efficient nucleic acid chaperone activity requires the protein to induce nucleic acid attraction, partially destabilize the DNA helix, and exhibit rapid binding kinetics between ss- and dsDNA [53, 54]. Therefore, in the presence of such proteins, in addition to an increased transition width, the FEC will also exhibit alteration at the low force dsDNA and the high force ssDNA regimes (Fig. 2a). The FEC of the WT protein, in which, presumably, all the competing effects are optimally balanced, may serve as a good reference to compare and evaluate different mutant proteins. For instance, relative increases in dsDNA or ssDNA stretching forces are indicative of DNA aggregation and may lead to poor protein binding kinetics and thereby negatively affect nucleic acid chaperone

294

M. Nabuan Naufer and Mark C. Williams

capabilities. In these cases, the transition width may appear to be comparable with the wild-type transition even in the presence of a defective protein. Therefore, it is useful to analyze the complete FEC and quantify DNA aggregation in addition to the transition width to accurately assess the nucleic acid chaperone capabilities of the protein (Fig. 2b). To quantify ssDNA aggregation, one can examine the extensions beyond the overstretching transition where the DNA is mostly single-stranded. An extensive decrease in extension in the ssDNA regime with respect to the wild-type protein is indicative of unfavorable ssDNA aggregation by the bound protein variant. Similarly, high stretching forces due to dsDNA binding can be examined at extension below ~0.3 nm/bp. A significant decrease in the extension in this regime is indicative of dsDNA aggregation due to bound protein. A more systematic method to quantify dsDNA aggregation at low forces is described in [55]. In this method, a compaction force (Fc), which is a measure of the additional force required to stretch the dsDNA at low force regimes due to bound protein, is found. Fc is determined as the average force difference over extensions