Riboswitches Methods and Protocols

METHODS IN MOLECULAR BIOLOGY™ Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfi...

Author: Alexander Serganov

8 downloads 753 Views 4MB Size Report

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

Report copyright / DMCA form

DOWNLOAD PDF

METHODS

IN

MOLECULAR BIOLOGY™

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

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

METHODS

IN

MOLECULAR BIOLOGY™

Riboswitches Methods and Protocols

Edited by

Alexander Serganov Department of Structural Biology, Memorial Sloan-Kettering Cancer Center New York, NY, USA

Editor Alexander Serganov Department of Structural Biology Memorial Sloan-Kettering Cancer Center New York, NY USA [email protected]

ISSN: 1064-3745 e-ISSN: 1940-6029 ISBN: 978-1-934115-88-6 e-ISBN: 978-1-59745-558-9 DOI: 10.1007/978-1-59745-558-9 Library of Congress Control Number: 2008944038 © Humana Press, a part of Springer Science+Business Media, LLC 2009 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified a such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Cover illustrations: illustrations of the three-dimensional crystal structure of the metabolite-sensing domain from the Escherichia coli thiamine pyrophosphate riboswitch (Nature, 2006, 441: 1167–1171). The thiamine pyrophosphate riboswitch, responsive to a phosphorylated derivative of the vitamin B1, was one of the first discovered riboswitches (Nature, 2002, 419: 952–956; Cell, 2002, 111: 747–756) and, to date, the only riboswitch found in both eukaryotes and prokaryotes (RNA, 2003, 9: 644–647). Despite the high divergence of the species, the riboswitches that originated from both kingdoms share a high degree of similarity (Nature, 2006, 441: 1167–1171, Structure, 2006, 14: 1459–1468; Science, 2006, 312: 1208–1121). Printed on acid-free paper springer.com

Preface The transfer of hereditary information from genes to proteins is one of the essential processes in all living organisms on our planet. Some genes are expressed without modulation throughout the life of a cell, while many others require various degrees of control to precisely balance cellular metabolism with environmental conditions. For many years, researchers attributed this regulatory function to protein molecules, which can direct gene expression at multiple levels, in response to various input signals, and with different degrees of selectivity. Even when the control of gene expression was achieved via direct interactions between proteins and mRNAs, the active role was routinely assigned to proteins, while RNAs were considered merely as recipient molecules. The discovery of RNA interference and multiple bacterial regulatory RNAs caused a shift from the perception of proteins as the predominant regulators of gene expression to the acknowledgement of the importance of RNAs in many regulatory circuits. Such a viewpoint received strong support several years ago after the discovery of riboswitches and related RNA sensors – mRNA regions capable of alternating their conformations in response to the presence of cellular metabolites and other physical or chemical cues. These classes of RNA pass on cellular and environmental information directly to transcription or translation machinery without the assistance of proteins. The riboswitches are commonly defined as evolutionarily conserved mRNA regions capable of specific binding to metabolite molecules, and, as a result, adopting a particular RNA conformation that modulates gene expression. This definition restricts riboswitches to mRNAs responding to small organic molecules, and this volume is primarily focused on the techniques used for the identification and characterization of such RNAs. However, the meaning of the term riboswitch can be broadened to incorporate classes of RNA that undergo conformational transitions in response to other stimuli in order to control the expression of genes. Therefore, several contributions to the book are devoted to the mRNAs which adopt complexly folded conformations and directly sense environmental signals or recognize molecules other than metabolites. Among the many RNA-based regulatory systems, there is a special place for mRNAs which sense the regulatory signal delivered by a specific RNA. These RNAs participate in diverse regulatory mechanisms, overviewed in the book, and often require in vivo techniques, described in two chapters, for the elucidation of their function. This volume includes comprehensive and up-to-date coverage of various methods used to study riboswitches and other RNAs involved in gene expression control. Although some protocols utilize the intrinsic properties of metabolite molecules and can be applied exclusively to elucidate the function of metabolite-binding RNAs, a majority of the methods originate from rigorously tested procedures previously used for the characterization of various RNA molecules and RNA–ligand complexes. Several chapters in this book describe classical and emerging biochemical techniques, such as chemical and enzymatic RNA synthesis, RNA structure probing, and footprinting. The latter two techniques, powerful and fast methods of gaining preliminary structural information, have become routine procedures in a number of laboratories. Nevertheless,

v

vi

Preface

many other researchers hesitate to utilize these techniques to the fullest extent due to difficulties in the reproducibility and interpretation of the results. The expert contributions to this volume will guide one through all the obstacles in these techniques and will ultimately convert troublesome applications into facile and readily reproducible protocols. Modern biophysical methods also represent a major part of the current volume. Several chapters include the description of cutting-edge technologies used to study riboswitch structures and their folding. These contributions cover a wide range of protocols from X-ray and NMR experiments to single-molecule fluorescent studies and isothermal titration calorimetry. Biophysical techniques are complemented by SELEX and related protocols, which allow for a transition from in vitro to in vivo experiments. In vivo methods, however, are not limited by the elucidation of the function of engineered riboswitches. Microbiologists, cell biologists and geneticists can appreciate the chapters focusing on various aspects of the identification and characterization of riboswitches and regulatory RNAs in living cells. One chapter is devoted to probably the most intriguing and demanding aspects of riboswitch research: computerized searches for riboswitches in genomic sequences, which could be helpful in the prediction of the alternative secondary and tertiary riboswitch structures, and in the identification of the candidate molecules that partner with riboswitches. Research success in any given field is critically dependent upon advances in experimental methodologies. This volume summarizes and illustrates key methods that depict the remarkable recent progress in the riboswitch field and other areas of RNA biology. The wide scope of this book is a feature that will undoubtedly appeal to researchers with different backgrounds, working in various areas of modern biology ranging from evolutionary biology to pharmacology and biotechnology. Moreover, the protocols presented in the book can be successfully utilized by researchers with different levels of expertise, both beginners, who wish only to try the “RNA kitchen”, and experts, who would like to expand their RNA methodology. I thank the authors who contributed to this volume and whose work will hopefully help readers find new stimulating approaches suited for application in their experiments. New York, NY

Alexander Serganov

Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2

3

4 5 6

7 8

9

10 11 12

13

Predicting Riboswitch Regulation on a Genomic Scale . . . . . . . . . . . . . . . . . . . . . Jeffrey E. Barrick Enzymatic Ligation Strategies for the Preparation of Purine Riboswitches with Site-Specific Chemical Modifications . . . . . . . . . . . . . Renate Rieder, Claudia Höbartner, and Ronald Micura Application of Fluorescent Measurements for Characterization of Riboswitch-Ligand Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Benoit Heppell, Jérôme Mulhbacher, J. Carlos Penedo, and Daniel A. Lafontaine Transcriptional Approaches to Riboswitch Studies. . . . . . . . . . . . . . . . . . . . . . . . . Alexander Mironov, Vitaly Epshtein, and Evgeny Nudler Kinetics of Riboswitch Regulation Studied By In Vitro Transcription . . . . . . . . . . J. Kenneth Wickiser Molecular Basis of RNA-Mediated Gene Regulation on the Adenine Riboswitch by Single-Molecule Approaches . . . . . . . . . . . . . . . . . Jean-François Lemay, J. Carlos Penedo, Jérôme Mulhbacher, and Daniel A. Lafontaine Methods for Analysis of Ligand-Induced RNA Conformational Changes . . . . . . . Chad A. Brautigam, Catherine A. Wakeman, and Wade C. Winkler Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sunny D. Gilbert and Robert T. Batey Preparation and Crystallization of Riboswitch–Ligand Complexes. . . . . . . . . . . . . Olga Pikovskaya, Artem A. Serganov, Ann Polonskaia, Alexander Serganov, and Dinshaw J. Patel Crystallization of the glmS Ribozyme-Riboswitch . . . . . . . . . . . . . . . . . . . . . . . . . Daniel J. Klein and Adrian R. Ferré-D’Amaré Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering . . . . . . . . . Jan Lipfert, Daniel Herschlag, and Sebastian Doniach Time-Resolved NMR Spectroscopy: Ligand-Induced Refolding of Riboswitches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Janina Buck, Boris Fürtig, Jonas Noeske, Jens Wöhnert, and Harald Schwalbe Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing and Selective 2′-Hydroxyl Acylation and Primer Extension (SHAPE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Catherine A. Wakeman and Wade C. Winkler

vii

v ix 1

15

25

39 53

65

77

97 115

129 141

161

173

viii

Contents

14

Identification of Metabolite–Riboswitch Interactions Using Nucleotide Analog Interference Mapping and Suppression . . . . . . . . . . . . . Juliane K. Soukup and Garrett A. Soukup 15 RNA-Dependent RNA Switches in Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tina M. Henkin 16 Probing mRNA Structure and sRNA-mRNA Interactions in Bacteria Using Enzymes and Lead(II). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clément Chevalier, Thomas Geissmann, Anne-Catherine Helfer, and Pascale Romby 17 Structural Probing of RNA Thermosensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Claude Chiaruttini, Frédéric Allemand, and Mathias Springer 18 Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting Translational Regulators in Bacteria . . . . . . . . . . . . . . . Pierre Fechter, Clément Chevalier, Gulnara Yusupova, Marat Yusupov, Pascale Romby, and Stefano Marzi 19 Isolation and Characterization of the Heat Shock RNA 1 (HSR1) . . . . . . . . . . . . Ilya Shamovsky and Evgeny Nudler 20 Analysis of tRNA-Directed Transcription Antitermination in the T Box System in vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tina M. Henkin 21 In vitro Selection of Conformational Probes for Riboswitches. . . . . . . . . . . . . . . . Günter Mayer and Michael Famulok 22 A Green Fluorescent Protein (GFP) Based Plasmid System to Study Post-Transcriptional Control of Gene Expression In Vivo . . . . . . . . . . . . Johannes H. Urban and Jörg Vogel 23 High-Throughput Screens to Discover Synthetic Riboswitches . . . . . . . . . . . . . . . Sean A. Lynch, Shana Topp, and Justin P. Gallivan 24 A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laising Yen, Brent R. Stockwell, and Richard C. Mulligan 25 In Vitro Selection of glmS Ribozymes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kristian H. Link and Ronald R. Breaker Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

193 207

215

233

247

265

281 291

301 321

335 349 365

Contributors FRÉDÉRIC ALLEMAND • UPR9073 du CNRS associée à l’Université de Paris VII, Institut de Biologie Physico-Chimique, Paris, France JEFFREY E. BARRICK • Microbiology and Molecular Genetics, Michigan State University, East Lansing, MI, USA ROBERT T. BATEY • Department of Chemistry and Biochemistry, University of Colorado, Boulder, Boulder, CO, USA CHAD A. BRAUTIGAM • Structural Biology Core Facility and Department of Biochemistry, The University of Texas Southwestern Medical Center, Dallas, TX, USA RONALD R. BREAKER • Department of Molecular, Cellular and Developmental Biology, Department of Molecular Biophysics and Biochemistry, Howard Hughes Medical Institute, Yale University, New Haven, CT, USA JANINA BUCK • Institute of Organic Chemistry and Chemical Biology, Center for Biomolecular Magnetic Resonance, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany CLÉMENT CHEVALIER • Architecture et Réactivité de l’ARN, Université de Strasbourg, CNRS, IBMC, Strasbourg, France CLAUDE CHIARUTTINI • UPR9073 du CNRS associée à l’Université de Paris VII, Institut de Biologie Physico-Chimique, Paris, France SEBASTIAN DONIACH • Departments of Physics, Applied Physics, and Biophysics Program, Stanford University, Stanford, CA, USA VITALY EPSHTEIN • Department of Biochemistry, New York University Medical Center, New York, NY, USA MICHAEL FAMULOK • Rheinische Friedrich-Wilhelms-Universität Bonn Life & Medical Sciences (LIMES) Institute, Chemical Biology & Medicinal Chemistry Laboratory, c/o Kekulé-Institut für Organische Chemie und Biochemie, Bonn, Germany PIERRE FECHTER • Architecture et Réactivité de l’ARN, Université de Strasbourg, CNRS, IBMC, Strasbourg, France ADRIAN R. FERRÉ-D’AMARÉ • Division of Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA BORIS FÜRTIG • Institute of Organic Chemistry and Chemical Biology, Center for Biomolecular Magnetic Resonance, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany JUSTIN P. GALLIVAN • Department of Chemistry and Center for Fundamental and Applied Molecular Evolution, Emory University, Atlanta, GA, USA THOMAS GEISSMANN • Architecture et Réactivité de l’ARN, Université de Strasbourg, CNRS, IBMC, Strasbourg, France

ix

x

Contributors

SUNNY GILBERT • Department of Chemistry and Biochemistry, University of Colorado, Boulder, Boulder, CO, USA ANNE-CATHERINE HELFER • Architecture et Réactivité de l’ARN, Université de Strasbourg, CNRS, IBMC, Strasbourg, France TINA M. HENKIN • Department of Microbiology, The Ohio State University, Columbus, OH, USA BENOIT HEPPELL • RNA Group, Département de Biologie, Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada DANIEL HERSCHLAG • Department of Biochemistry and Biophysics Program, Stanford University, Stanford, CA, USA CLAUDIA HÖBARTNER • Institute of Organic Chemistry, Leopold Franzens University, Innsbruck, Austria DANIEL J. KLEIN • Division of Basic Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA, USA DANIEL A. LAFONTAINE • RNA Group, Département de Biologie, Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada JEAN-FRANÇOIS LEMAY • RNA Group, Département de Biologie, Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada KRISTIAN H. LINK • Howard Hughes Medical Institute, Yale University, New Haven, CT, USA JAN LIPFERT • Department of Physics, Stanford University, Stanford, CA, USA Kavli Institute of Nanoscience, Faculty of Applied Sciences, Delft University of Technology, Delft, The Netherlands SEAN A. LYNCH • Department of Chemistry and Center for Fundamental and Applied Molecular Evolution, Emory University, Atlanta, GA, USA STEFANO MARZI • Architecture et Réactivité de l’ARN, Université de Strasbourg, CNRS, IBMC, Strasbourg, France GÜNTER MAYER • Rheinische Friedrich-Wilhelms-Universität Bonn Life & Medical Sciences (LIMES) Institute, Chemical Biology & Medicinal Chemistry Laboratory, c/o Kekulé-Institut für Organische Chemie und Biochemie, Bonn, Germany RONALD MICURA • Institute of Organic Chemistry, Leopold Franzens University, Innsbruck, Austria ALEXANDER MIRONOV • Department of Biochemistry, New York University Medical Center, New York, NY, USA JÉRÔME MULHBACHER • RNA Group, Département de Biologie, Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada RICHARD C. MULLIGAN • Department of Genetics, Harvard Medical School, Children’s Hospital, Harvard Institutes of Medicine, Boston, MA, USA JONAS NOESKE • Institute of Organic Chemistry and Chemical Biology, Center for Biomolecular Magnetic Resonance, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany EVGENY NUDLER • Department of Biochemistry, New York University Medical Center, New York, NY, USA DINSHAW J. PATEL • Department of Structural Biology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA

Contributors

xi

J. CARLOS PENEDO • RNA Group, Département de Biologie, Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada OLGA PIKOVSKAYA • Department of Structural Biology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA ANN POLONSKAIA • Department of Structural Biology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA RENATE RIEDER • Institute of Organic Chemistry, Leopold Franzens University, Innsbruck, Austria PASCALE ROMBY • Architecture et Réactivité de l’ARN, Université de Strasbourg, CNRS, IBMC, Strasbourg, France HARALD SCHWALBE • Institute of Organic Chemistry and Chemical Biology, Center for Biomolecular Magnetic Resonance, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany ALEXANDER SERGANOV • Department of Structural Biology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA ARTEM A. SERGANOV • Department of Structural Biology, Memorial Sloan-Kettering Cancer Center, New York, NY, USA GARRETT A. SOUKUP • Department of Biomedical Sciences, Creighton University School of Medicine, Omaha, NE, USA JULIANE K. SOUKUP • Department of Chemistry, Creighton University, Omaha, NE, USA MATHIAS SPRINGER • UPR9073 du CNRS associée à l’Université de Paris VII, Institut de Biologie Physico-Chimique, Paris, France BRENT R. STOCKWELL • Department of Biological Sciences and Department of Chemistry, Columbia University, New York, NY, USA SHANA TOPP • Department of Chemistry and Center for Fundamental and Applied Molecular Evolution, Emory University, Atlanta, GA, USA JOHANNES H. URBAN • RNA Biology Group, Max Planck Institute for Infection Biology, Berlin, Germany JÖRG VOGEL • RNA Biology Group, Max Planck Institute for Infection Biology, Berlin, Germany CATHERINE A. WAKEMAN • Department of Biochemistry, The University of Texas Southwestern Medical Center, Dallas, TX, USA J. KENNETH WICKISER • Department of Chemistry and Life Science, The United States Military Academy, West Point, NY, USA WADE C. WINKLER • Department of Biochemistry, The University of Texas Southwestern Medical Center, Dallas, TX, USA JENS WÖHNERT • Institute for Molecular Biosciences, Center for Biomolecular Magnetic Resonance, Johann Wolfgang Goethe-University, Frankfurt am Main, Germany LAISING YEN • Department of Pathology, Department of Molecular & Cellular Biology, Baylor College of Medicine, Houston, TX, USA MARAT YUSUPOV • IGBMC, Department of Structural Biology and Genomics, INSERM U596, CNRS UMR7104, Illkirch, France GULNARA YUSUPOVA • IGBMC, Department of Structural Biology and Genomics, INSERM U596, CNRS UMR7104, Illkirch, France

Chapter 1 Predicting Riboswitch Regulation on a Genomic Scale Jeffrey E. Barrick Summary Riboswitches are vital components of many genomes. Covariance model searches for the characteristic architectures of riboswitch aptamer domains can be used to predict new examples of these structured RNAs. Since riboswitches generally function as cis-regulatory elements, examining the genomic contexts of these hits is critical for evaluating their biological relevance. With these two sources of comparative support, it is possible to identify riboswitches accurately from sequence information alone. Annotating riboswitches on a genomic scale enables more precise functions to be assigned to the proteins that they regulate, better defines their conserved aptamer structures by identifying diverged variants, and provides insight into how the genetic regulation of fundamental metabolic processes varies among species. Key words: Noncoding RNA, Covariance model, Regulon, Genome annotation, Comparative genomics, Secondary structure

1. Introduction Riboswitches are noncoding RNA structures that sense changes in the cellular environment and directly mediate appropriate gene control responses (1). Most riboswitch classes that have been characterized bind to small molecule metabolites – like coenzymes, amino acids, and nucleobases – and regulate genes responsible for transporting, synthesizing, or recycling these compounds in bacteria. These riboswitches are primarily found in the 5¢ untranslated regions of messenger RNAs. They typically leverage a shape change induced during metabolite binding by an aptamer domain to alter the structure of downstream sequences in a way

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_1 © Humana Press, a part of Springer Science + Business Media, LLC 2009

1

2

Barrick

that causes premature transcription termination or prevents ribosome binding. Riboswitches are important features of many genomes. Most bacteria seem to harbor a complex menagerie of different riboswitches (2), and thiamin pyrophosphate riboswitches also occur in plant and fungal genomes (3). Covariance model (CM) searches are a flexible and powerful way to identify a conserved RNA when an alignment of representative sequences with secondary structure markup is available (4, 5). The Rfam database maintains high-quality, hand-curated alignments of hundreds of RNA families, including riboswitches (6). However, riboswitches, like many other families of noncoding RNAs, are not yet consistently annotated in genomic sequences (7). This shortcoming is particularly unfortunate because riboswitch identification can abet the prediction of precise functions for the genes that they regulate. Thus, the presence of a riboswitch can disambiguate the function of a downstream gene or even lead to the prediction of new protein families associated with its cognate ligand, e.g., (8). Riboswitches are outstanding targets for covariance model searches. Their aptamer domains typically form complex RNA structures that are rich in the combination of base conservation and base-pairing constraints that CMs were designed to exploit. Furthermore, one can validate putative riboswitch hits by examining their genomic contexts. Combining these two sources of information allows one to iteratively refine the descriptions of both a riboswitch’s conserved structure and its genetic regulon with great confidence (2). Other families of noncoding RNAs may not be as amenable to predictions based on comparative information alone because they either have relatively small conserved structures, as is the case for many RNA structures bound by regulatory proteins (9), or are not consistently associated with the same neighboring genes, as is the case with most ribozymes (10). This chapter describes how to exhaustively identify riboswitches in an intergenic region, genome, or larger sequence database of interest. The techniques are arranged in increasing order of the technical expertise they require, ranging from using a web-based form to running command-line tools on a computer cluster. The procedures can be used with any RNA family in the Rfam database or (for the last two) any RNA structure for which the user has a sequence alignment suitable for training a covariance model. However, they are designed to be especially useful for analyzing bacterial cis-regulatory elements such as riboswitches, riboswitch-like RNA structures of unknown function (11–13), and new candidate regulatory RNA motifs such as those that might be identified by the cmfinder computational pipeline (14, 15).

Predicting Riboswitch Regulation on a Genomic Scale

3

2. Materials 2.1. Identifying Riboswitches in an Intergenic Region of Interest

1. The Rfam database form for submitting a sequence search [http://www.sanger.ac.uk/Software/Rfam/search.shtml]. The website interface from version 8.1 (October 2007) was used in this example. 2. The nucleotide sequence of the intergenic region of interest. The intergenic region upstream of the thiD gene from Arthrobacter aurescens TC1, including 100 nucleotides of each adjacent reading frame, was used here. This corresponds to positions 2604078-2604530 from GenBank record NC_008711.1. Currently, the query sequence can be no longer than 2,000 nucleotides.

2.2. Annotating Riboswitches in a Bacterial Genome

1. The Infernal software package [http://infernal.janelia.org/] (5), which is distributed as ANSI C source code that can be compiled on a variety of computer platforms. Version 0.81 was run on Mac OS X in this example. 2. A multiple sequence alignment in Stockholm format (see Note 1) for the riboswitch of interest. The thiamine pyrophosphate (TPP) riboswitch seed alignment from the Rfam database was used here [http://www.sanger.ac.uk/cgi-bin/ Rfam/getacc?RF00059]. 3. The bacterial genome sequence to be searched in both FASTA and GenBank formats. This example uses the 5.4 megabase Serratia proteamaculans 568 genome, which is available as GenBank record NC_009832.1. 4. Perl scripts for automating various analysis procedures and drawing genomic contexts [http://www.barricklab.org/]. As described in their documentation, some of these scripts require BioPerl (16), additional Perl modules, and accessory code libraries to be installed. 5. A text editor for viewing and manipulating RNA alignments. The RNA Alignment Editor RALEE [http://www.sanger. ac.uk/Users/sgj/ralee/] (17), which functions as an extension to Emacs, is recommended. However, any editor that allows wrapping of long lines to be disabled can be used.

2.3. Exhaustively Identifying Riboswitches in a Sequence Database

1. Software programs, scripts, and a riboswitch alignment as in Subheading 2.2. 2. The sequence database to be searched in FASTA and GenBank formats. Records from the RefSeq database [http://www.ncbi. nlm.nih.gov/RefSeq/] (18) are preferred because they include annotation from the Conserved Domain Database (CDD) (19). The microbial subdivision of RefSeq Release 24 was searched here.

4

Barrick

3. A computer cluster able to access a shared directory of the sequences to be searched. Many research institutions now have central facilities suitable for parallelizing covariance model searches. Searches were conducted on a cluster of Intel computers running the Linux OS using an in-house job management system to create the example results.

3. Methods 3.1. Identifying Riboswitches in an Intergenic Region of Interest

1. Paste the sequence of the intergenic region of interest into the Rfam search page and submit the form. Include 50–200 nucleotides overlapping each flanking gene (see Note 2). 2. Figure 1 is an example search result. It reports a bit score of 55.54 for a putative TPP riboswitch aptamer in the intergenic region upstream of the thiD gene from Arthrobacter aurescens TC1. Matches with bit scores above the gathering cutoff for the RNA family are shown (see Note 3). Higher bit scores indicate more confidence in a prediction. It is possible, though relatively unlikely, that a valid match might be missed because a heuristic filtering step is applied before evaluating the CM (see Note 4). 3. Examine the RNA structure predicted for each putative riboswitch aptamer in the query sequence to determine whether it is biologically plausible (see Note 5). The alignment on the Rfam results page is generated by Infernal’s cmsearch tool. Figure 2 is a key to the symbols used in this alignment.

Fig. 1. Example Rfam sequence search results page for the intergenic region upstream of the thiD gene from Arthrobacter aurescens TC1.

Predicting Riboswitch Regulation on a Genomic Scale

5

Fig. 2. Key to cmsearch alignment symbols adapted from the Infernal user guide.

3.2. Annotating Riboswitches in a Bacterial Genome

1. Navigate to the directory containing the Stockholm alignment of the riboswitch of interest (TPP.sto) and the genome sequences to be searched in FASTA (NC_009832.1.fna) and GenBank (NC_009832.1.gbk) formats. 2. Run Infernal’s cmbuild program to generate a covariance model specification file (TPP.cm) from the multiple sequence alignment: > cmbuild TPP.cm TPP.sto 3. Search this CM against the genome using Infernal’s cmsearch program: > cmsearch --hmmfilter -T 27.2 TPP.cm NC_009832.1.fna > hits.txt The threshold option (−T) tells the program to only report matches with bit scores equal to or greater than the following value. When using Rfam seed alignments, this option should generally be set to the gathering cutoff of the RNA family (see Note 3). This search takes 10 min on the author’s computer. With these options, Infernal uses filtering techniques to accelerate the covariance model search (see Note 4). Therefore, it is possible that it will occasionally fail to report a match with a bit score above the specified threshold because it did not survive these filtering steps. In this case, the search takes 29 h when no filtering options are enabled, and it reports the same four matches.

6

Barrick

4. If you have installed the Perl scripts and necessary prerequisites, generate a multiple sequence alignment of all the hits with the following commands: > cmsearch_reformat.pl -s scores.tab hits.txt hits.fna > cmalign -o hits.sto TPP.cm hits.fna The output file (hits.sto) can be opened in RALEE or a text editor. The SS_cons and RF annotation lines show the covariance model’s secondary structure and consensus sequence, respectively, with the symbols described in Fig. 2. 5. Optionally, create a colorized HTML version of the alignment: > stockholm_to_html.pl hits.sto hits.html Open the output file (hits.html) in a web browser. 6. If you have installed the Perl scripts and necessary prerequisites, generate drawings of the genomic context of each hit: > genomic_context.pl -b scores.tab -s NC_009832.1.gbk hits. sto context Open the output (context/index.html) in a web browser. 7. Using the Stockholm or HTML multiple sequence alignment or by examining the alignments of hits to the covariance model in the original cmsearch output, verify that each putative riboswitch has a plausible RNA structure (see Note 5). Using the genome context drawings or by examining what genes are near each match by looking at the GenBank file, verify that each putative riboswitch is appropriately positioned for regulation (see Note 2). 3.3. Exhaustively Identifying Riboswitches in a Sequence Database

1. Download all the FASTA and GenBank files of the database to be searched into two separate directories (/db/refseq24/ fasta and /db/refseq24/genbank). 2. Divide the FASTA sequences into files that will take a reasonable amount of time for a single processor on the computer cluster to process. In this case, approximately 40 MB of sequence is placed in each fragmented database file: > fragment_db.pl -s 40 /db/refseq24/fasta /db/refseq24/ fragment This script creates 92 sequence files from the microbial portion of the RefSeq release 24 database. Copy the directory of fragmented databases files (/db/refseq24/fragment) to the cluster. 3. The details of this step will depend on the computer cluster. For help, contact your system administrator. Compile and install Infernal. Create a covariance model for the riboswitch of interest from a Stockholm alignment file (TPP.sto).

Predicting Riboswitch Regulation on a Genomic Scale

7

> cmbuild TPP.cm TPP.sto With whatever job management interface provided, submit cmsearch runs for each of the database fragments. The command for searching the first database fragment might look like this: > cmsearch --hmmfilter TPP.cm /db/refseq24/fragment/1. fna > hits.1.txt After the jobs are complete, gather all the resulting output files into one directory and concatenate them into a single file. For example, with this command: > cat hits.* > hits.txt Copy this file back to your local computer to continue the analysis. 4. Generate a multiple sequence alignment of the hits. In total, this search identifies 10,007 matches with positive bit scores for the example search. The distribution of these scores is shown in Fig. 3. In a large sequence database, there generally will not be a single bit score cutoff that can separate true riboswitches from false positives. In order to exhaustively identify all riboswitches, the genomic contexts of matches with bit scores below the usual gathering cutoff (27.20 in this case) should be examined. In this case a conservative bit score threshold of 10 identifies 2,684 hits and is likely to discover low-scoring aptamer variants if they exist: > cmsearch_reformat.pl -T 10 -b scores.tab hits.txt hits.fna > cmalign -o hits.sto TPP.cm hits.fna Note that putative riboswitch aptamers in the output Stockholm alignment file (hits.sto) are sorted by score, such that questionable hits will be nearest the end.

Fig. 3. Distribution of scores for TPP riboswitch matches in RefSeq24.

8

Barrick

Fig. 4. Example HTML alignment of TPP riboswitch matches. The B. subtilis and E. coli thiC riboswitches are shown, followed by the A. aurescens thiD riboswitch from Fig. 1 and the two hits whose genomic contexts are shown in Fig. 5. Symbols in each aligned sequence and the RF (reference annotation) and SS_cons (secondary structure) lines are described in Fig. 2. In each sequence, bases that can pair (including G:U pairs) are shaded according to the secondary structure annotation.

5. Optionally, create a colorized HTML version of the resulting alignment: > stockholm_to_html.pl hits.sto hits.html Selected hits from this output file (hits.html) are shown in Fig. 4. 6. Create genomic context drawings for the putative riboswitches. Since this step requires accessing many GenBank records, index the sequence database first: > index_db.pl -d refseq24_genbank /db/refseq24/genbank /db/index Then, generate the genomic context drawings: > genomic_context.pl -b scores.tab -d /db/index/refseq24_ genbank hits.sto context It may take hours to process a large alignment file for the first time. This process will be considerably quicker the next time the script is run, because the relevant information from each GenBank record is cached in a more efficient format the first time it is accessed. Open the resulting web page (context/index.html). Representative drawings are shown in Fig. 5. 7. Using the Stockholm or HTML multiple sequence alignment, verify that each putative riboswitch has a plausible RNA structure (see Note 5). Using the genome context drawings, verify that each putative riboswitch is appropriately positioned for regulation (see Note 2). Delete hits that fail either of these criteria from the Stockholm alignment file (hits.sto).

Fig. 5. Genomic contexts of selected TPP riboswitch matches. Putative riboswitches are shown in black on the sequence ruler in the center of each panel. Nearby genes are shown above this line with names and descriptions from the GenBank record. The filled regions of each gene overlap predictions of conserved domains from the CDD (19), which are displayed below the sequence ruler with their id numbers and names. The first example is a high-scoring match that overlaps annotation of a probably spurious hypothetical protein. The second example is a hit positioned to regulate TPP-related genes that has a bit score well below the usual gathering cutoff for this family.

Predicting Riboswitch Regulation on a Genomic Scale 9

10

Barrick

8. Optionally, use the Stockholm alignment of validated riboswitch hits to create a new covariance model and repeat this procedure. The bit scores generated by this new model should more cleanly separate true positives from true negatives. Repeating multiple iterations of this cycle where initially lowscoring, but correctly positioned, hits are incorporated into a new training set may bridge to still further diverged RNA structures or discover new gene families that are regulated by this riboswitch class.

4. Notes 1. The specification for the Stockholm file format can be found at [http://www.cgb.ki.se/cgb/groups/sonnhammer/Stockholm.html]. The Infernal user guide provides additional information that is specific to RNA sequence alignments. When downloading alignments from the Rfam website, be sure to save them in plain text format. The first line should be “# STOCKHOLM 1.0.” 2. Almost all bacterial riboswitches are found in the 5¢ untranslated regions of messenger RNAs encoding genes related to the transport, biosynthesis, salvage, or utilization of their target metabolites. Sometimes predicted “noncoding” riboswitch aptamers overlap annotated protein-coding genes. In these cases, the riboswitch may regulate the downstream gene by incorporating the ribosome binding site and/or start codon of the downstream gene into its conserved structures. It is also possible that the protein open reading frame (ORF) may be misannotated. Gene-finding algorithms can be prone to choosing alternate start codons that are upstream of actual translational starts or overpredicting hypothetical reading frames. If an entire ORF or the N-terminal portion of an ORF does not harbor any conserved protein domains, then a prediction of a high-scoring aptamer structure in the same region may indicate that the protein annotation needs to be revised. This is most likely the case for the Ruminococcus gnavus hit shown in Fig. 5. 3. CM matches are assigned raw “bit scores” calculated from the relative probabilities of generating the match sequence with the covariance model versus a null sequence model (assuming that the four RNA bases occur with equal probabilities at each position, for example). The base two logarithm of the ratio of these odds is the bit score, such that there is more confidence that hits with higher bit scores were generated by

Predicting Riboswitch Regulation on a Genomic Scale

11

the model rather than chance. The Rfam curators establish several bit score cutoffs for each RNA family based on the performance of the covariance model built from their seed alignment (1) a “trusted cutoff,” above which all hits to the covariance model are true matches to the RNA family, (2) a “noise cutoff,” which is the highest score of a match believed to not be a member of the RNA family, and (3) the “gathering cutoff,” which is between the trusted and noise cutoffs. The full alignment of an RNA family includes all matches found in the Rfam sequence database with scores above the gathering cutoff. Infernal can also estimate a BLAST-like E-value for each hit that may be useful because it directly reflects the statistical significance of its bit score given the search database (20). 4. Covariance model searches are computationally intensive. For this reason, various filtering techniques have been developed that use faster algorithms to eliminate portions of a database that are unlikely to have high-scoring CM matches. These techniques can be conceptually divided into “rigorous” filters, which guarantee that all CM matches with bit scores above a specified threshold will be recovered, and “heuristic” filters, which may sacrifice some sensitivity for speed. Rfam currently uses BLAST searches with relaxed matching parameters against each sequence in the seed alignment of a family as a heuristic filter. Infernal’s cmsearch program employs a more sophisticated heuristic hidden Markov model (HMM) filter (21) when invoked with the --cmfilter option. For riboswitches, both techniques generally recover nearly all the possible CM hits while allowing dramatic speedups in database scans, but it is possible that any search with a heuristic filtering step will eliminate some valid hits from consideration. The Infernal user guide explains more advanced options for controlling how filtering is applied as well as how to use rigorous HMM filters (22, 23). Much of this filtering functionality was first developed in the RaveNnA software package [http:// bliss.biology.yale.edu/~zasha/ravenna/]. 5. Since covariance models do not consider the thermodynamic stabilities of the structures they predict for RNA sequence matches, they can be prone to assigning high scores to implausibly AU-rich hits, especially for the smaller purine and SAMII riboswitch models. CMs also do not allow pseudoknots and cannot capture all the covariation information that is present in complex RNA structure motifs, like K-turns (24). Many riboswitch classes contain these features (2), and it is important to check that they are present and properly aligned in putative riboswitch structures. CMs may misalign and assign low scores to variant riboswitches that have unexpectedly long

12

Barrick

insertions within their structures. If such a “variable insertion” truly does not disrupt conserved elements and has the potential to fold into a compact base-paired structure, then that sequence is likely to be a genuine riboswitch. Other riboswitch variants that dispense with consensus hairpins or alter entire conserved domains can sometimes be detected better by using cmsearch in --local mode. This option allows a potential match to skip consensus secondary structure elements by exiting and re-entering the covariance model.

Acknowledgments The author would like to thank Zasha Weinberg for helpful comments on this protocol, discussions about filtering techniques, and contributing the script for generating HTML alignments; Chris Fields for his work on the BioPerl modules for parsing Stockholm alignments and interfacing with Infernal; Sean Eddy and colleagues for meticulously documenting and improving Infernal; Alex Bateman and Sam Griffiths-Jones for a generous look at the inner workings of the Rfam database; and especially Ronald Breaker and many other collaborators in his lab for converting bioinformatic fantasies into experimental realities. References 1. Winkler, W. C. and Breaker, R. R. (2005). Regulation of bacterial gene expression by riboswitches. Annu. Rev. Microbiol . 59 , 487– 517 . 2. Barrick, J. E. and Breaker, R. R. (2007). The distributions, mechanisms,and structures of metabolite-binding riboswitches. Genome Biol. 8(11), R239 3. Sudarsan, N., Barrick, J.E. and Breaker, R. R. (2003). Metabolite-binding RNA domains are present in the genes of eukaryotes. RNA 9, 644–647. 4. Eddy, S. R. and Durbin, R. (1994). RNA sequence analysis using covariance models. Nucleic Acids Res. 22, 2079–2088. 5. Nawrocki, E. P. and Eddy, S. R. (2007). Query-dependent banding (QDB) for faster RNA similarity searches. PLoS Comput Biol. 3, e56. 6. Griffiths-Jones, S., Moxon, S., Marshall, M., Khanna, A., Eddy, S. R. and Bateman, A. (2005). Rfam: annotating non-coding RNAs in complete genomes. Nucleic Acids Res. 33 , D121–D124.

7. Griffiths-Jones, S. (2007). Annotating noncoding RNA genes. Annu. Rev. Genomics Hum. Genet. 8, 279–298. 8. Rodionov, D. A., Vitreschak, A. G., Mironov, A. A. and Gelfand, M. S. (2002). Comparative genomics of thiamin biosynthesis in procaryotes: new genes and regulatory mechanisms. J. Biol. Chem. 277, 48949–48959. 9. Stülke, J. (2002). Control of transcription termination in bacteria by RNA-binding proteins that modulate RNA structures. Arch Microbiol 177, 433–440. 10. Hammann, C. and Westhof, E. (2007). Searching genomes for ribozymes and riboswitches. Genome Biol. 8, 210. 11. Barrick, J. E., Corbino, K. A., Winkler, W. C., Nahvi, A., Mandal, M., Collins, J., Lee, M., Roth, A., Sudarsan, N., Jona, I., Wickiser, J. K. and Breaker, R. R. (2004). New RNA motifs suggest an expanded scope for riboswitches in bacterial genetic control. Proc. Natl. Acad. Sci. U.S.A. 101 , 6421–6426. 12. Corbino, K. A., Barrick, J. E., Lim, J., Welz, R., Tucker, B. J., Puskarz, I., Mandal, M.,

Predicting Riboswitch Regulation on a Genomic Scale

13.

14.

15.

16.

17.

18.

Rudnick, N. D. and Breaker, R. R. (2005). Evidence for a second class of S-adenosylmethionine riboswitches and other regulatory RNA motifs in alpha-proteobacteria. Genome Biol. 6, R70. Weinberg, Z., Barrick, J. E., Yao, Z., Roth, A., Kim, J. N., Gore, J., Wang, J. X., Lee, E. R., Block, K. F., Sudarsan, N., Neph, S., Tompa, M., Ruzzo, W. L. and Breaker, R. R. (2007). Identification of 22 candidate structured RNAs in bacteria using the CMfinder comparative genomics pipeline. Nucleic Acids Res. 35, 4809–4819. Yao, Z. Weinberg, Z. and Ruzzo, W. L. (2006). CMfinder – a covariance model based RNA motif finding algorithm. Bioinformatics 22, 445–452. Yao, Z., Barrick, J., Weinberg, Z., Neph, S., Breaker, R., Tompa, M. and Ruzzo, W.L. (2007). A computational pipeline for high-throughput discovery of cis-regulatory noncoding RNA in prokaryotes. PLoS Comput Biol. 3 , e126. Stajich, J. E., Block, D., Boulez, K., Brenner, S. E., Chervitz, S. A., Dagdigian, C., Fuellen, G., Gilbert, J. G., Korf, I., Lapp, H., Lehvaslaiho, H., Matsalla, C., Mungall, C. J., Osborne, B. I., Pocock, M. R., Schattner, P., Senger, M., Stein, L. D., Stupka, E., Wilkinson, M. D. and Birney, E. (2002). The Bioperl toolkit: Perl modules for the life sciences. Genome Res. 12 , 1611–1618. Griffiths-Jones, S. (2005). RALEE–RNA alignment editor in Emacs. Bioinformatics 21 , 257–259. Pruitt, K. D., Tatusova, T. and Maglott, D. R. (2005). NCBI Reference Sequence (RefSeq):

19.

20.

21.

22.

23.

24.

13

a curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res. 33 , D501–504. Marchler-Bauer, A., Anderson, J. B., Cherukuri, P. F., DeWeese-Scott, C., Geer, L. Y., Gwadz, M., He, S., Hurwitz, D. I., Jackson, J. D., Ke, Z., Lanczycki, C. J., Liebert, C. A., Liu, C., Lu, F., Marchler, G. H., Mullokandov, M., Shoemaker, B. A., Simonyan, V., Song, J. S., Thiessen, P. A., Yamashita, R. A., Yin, J. J., Zhang, D. and Bryant, S. H. (2005). CDD: a conserved domain database for protein classification. Nucleic Acids Res. 33, D192–196. Klein, R. J. and Eddy, S. R. (2003). RSEARCH: finding homologs of single structured RNA sequences. BMC Bioinformatics 4, 44. Weinberg, Z. and Ruzzo, W. L. (2006). Sequence-based heuristics for faster annotation of non-coding RNA families. Bioinformatics 22 , 35–39. Weinberg, Z. and Ruzzo, W. L. (2004). Faster genome annotation of non-coding RNA families without loss of accuracy, in Proc. Eighth Annu. Int. Conf. on Comp. Mol. Biol. (RECOMB), ACM Press, New York, pp. 243–251. Weinberg, Z. and Ruzzo, W. L. (2004). Exploiting conserved structure for faster annotation of non-coding RNAs without loss of accuracy. Bioinformatics 20 , i334–i341. Lescoute, A., Leontis, N. B., Massire, C. and Westhof, E. (2005). Recurrent structural RNA motifs, isostericity matrices and sequence alignments. Nucleic Acids Res. 33 , 2395–2409.

Chapter 2 Enzymatic Ligation Strategies for the Preparation of Purine Riboswitches with Site-Specific Chemical Modifications Renate Rieder, Claudia Höbartner, and Ronald Micura Summary One of the most versatile riboswitch classes refers to purine nucleoside metabolism. In the cell, purine riboswitches of the respective mRNAs either act at the transcriptional or translational level and off- or on-regulate genes upon binding to their dedicated ligands. Biophysical studies on ligand-induced folding of these RNA domains in vitro contribute to understanding their regulation mechanisms in vivo. For such studies, in particular, for approaches using fluorescence spectroscopy, the preparation of large RNAs with site-specific chemical modifications is required. Here, we describe a strategy for the preparation of riboswitch aptamers and aptamers adjoined to their expression platforms by chemical synthesis and enzymatic ligation. The modular design enables fast access to a large number of purine riboswitch derivatives with the modification of interest at any strand position. We exemplarily provide a detailed protocol for the preparation of adenosine deaminase (add) A-riboswitch variants with 2-aminopurine (AP) modifications at the 40-nmol scale. Key words: Ligation, Riboswitch, Aminopurine, RNA modification

1. Introduction Purine riboswitches consist of a highly conserved aptamer domain of 60–70 nucleotides and an adjoining expression platform of about the same size. While the aptamer sequences are highly conserved, sequences of the expression platform are more diverse to fulfill the specific requirements that arise from different functional determinants, such as transcriptional vs. translational regulation or gene repression vs. gene activation (1–12). The general construction scheme of purine riboswitches enables a modular strategy for the efficient preparation of riboswitch variants with site-specific Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_2 © Humana Press, a part of Springer Science + Business Media, LLC 2009

15

16

Rieder, Höbartner, and Micura

chemical modifications by ligation of chemically synthesized RNA fragments. The two prime limits that are encountered in this approach are size limitation with respect to chemically synthesized RNA strands (13) and the requirement of high-yielding ligation sites (14). Both issues render such an undertaking methodologically challenging; for the class of purine riboswitches, we offer a satisfying solution described below. In case of purine aptamer domains, two efficient ligation sites have been identified (Fig. 1a) (15). One site resides in loop L2 between nucleotides 35 and 36 (numbering referring to ref. 12;

Fig. 1. Ligation strategies for the assembly of the aptamer domain and the full-length add A-riboswitch. The aptamer domain (a) can be prepared via single-stranded ligation by T4 RNA ligase (top) or via splinted ligation by T4 DNA ligase (bottom). For the full-length domain (b) two RNA fragments can be ligated by T4 RNA ligase (top). Alternatively, the use of T4 DNA ligase and T4 RNA ligase allows the simultaneous ligation of three fragments to the full-length domain (bottom). In the ligation reaction, the free 3¢ hydroxyl group of the acceptor strand (light gray) is joined to the 5¢ monophosphate of the donor strand (black). When three RNA fragments are ligated, the middle fragment acts as a donor at its 5¢ monophosphate (black) and as an acceptor at its 3¢ hydroxyl (light gray).

Enzymatic Ligation Strategies for the Preparation of Purine Riboswitches

17

see also Fig. 2a) and enables ligation of a ~50 nt donor strand and a ~20 nt acceptor strand by employment of T4 RNA ligase (Fig. 1a, top). The alternative ligation site resides in junction J2–3 between nucleotides 52 and 53 (Fig. 1a, bottom) and requires a ~40 nt acceptor strand that is ligated to a ~30 nt donor strand by T4 DNA ligase via a 23 nt DNA splint. Especially when RNA fragments are obtained from commercial suppliers, the latter ligation site is more convenient with regard to readily available RNA lengths; custom synthesis limits are often less than 40 nt at 1 μmol or larger scales. For full-length purine riboswitch domains, two pathways have been explored. One possibility is ligation of two strands, a ~50 nt acceptor and a ~60 nt donor strand. The site of ligation is positioned in loop L3 between nucleotides 64 and 65, and thus enables use of T4 RNA ligase (Fig. 1b, top) for the enzymatic reaction. The more flexible concept, however, is to join three shorter strands (12). Thereby, one ligation site refers to the site optimized for aptamer preparation alone and requires T4 DNA ligase together with a DNA splint. The second site is located in the loop of the repressor element and fulfils all criteria for efficient T4 RNA ligation (Fig. 1b, bottom). The enzymatic preparations described here are typically performed with 40–60 nmol of each ligation fragment in equimolar ratio. For instance, the aptamer domain of add A-riboswitch is ligated at 60-nmol scale by either T4 RNA ligase or T4 DNA ligase (Fig. 1a) and isolated with yields higher than 50% after purification by anion-exchange chromatography (36 nmol; ~820 μg; ~30 OD260) (12, 15). Yields of the corresponding fulllength riboswitch domains, following the three-strand ligation approach at 40-nmol scale with simultaneous usage of T4 RNA ligase and T4 DNA ligase (Fig. 1b, bottom), are 40% after purification by anion-exchange chromatography (15 nmol; ~580 μg; ~20 OD260; see protocol below). Yields of the full-length domain of add A-riboswitch following the two-strand ligation are slightly lower (Fig. 1b , top). The following protocol provides all necessary details for efficient ligation of full-length purine riboswitch domains based on three individual RNA strands, exemplified for the add A-riboswitch (Fig. 2a). Analysis of the ligation reaction and purification of the ligation products are performed by anion-exchange chromatography (Fig. 2b). For characterization of the HPLC-purified riboswitch domains, liquid chromatography electrospray ionization mass spectrometry (LC-ESI-MS) is applied (Fig. 2c). The flexibility of the modular concept is illustrated in Fig. 3. Variable combinations of the three-strand blocks rapidly enable access to a large number of 2-aminopurine modified full-length riboswitch domains with different 5¢ and 3¢ flanking sequences to assess their influence on the ligand-induced folding process.

Fig. 2. (a) Preparation of the 2-aminopurine (AP)-modified full-length add A-riboswitch by the three-strand ligation strategy using T4 DNA ligase and T4 RNA ligase (compare Fig. 1b, bottom). (b) Anion-exchange HPLC analysis of the one-pot ligation reaction. The individual traces show the sample composition before addition of ligase (start), after addition of T4 DNA ligase (2 h reaction time; accumulation of 91 nt intermediate), and after addition of T4 RNA ligase (4.5 h reaction time; formation of 120 nt ligation product). The bottom trace shows the HPLC-purified ligation product that was further characterized by LC-ESI-MS (c).

Fig. 3. Modular concept for the construction of 2-aminopurine (AP)-modified full-length riboswitch domains with varying 5¢ and 3¢ ends. The two-step ligation of three fragments allows a high flexibility in arranging a small number of different RNA fragments (modules) to a large number of riboswitch variants. For instance, a single central module is used for ligation to one out of four different 5¢-modules and one out of three different 3¢-modules. Out of 12 possible combinations, the synthesis of six riboswitch domains ranging from 110 to 120 nt has been experimentally verified.

Enzymatic Ligation Strategies for the Preparation of Purine Riboswitches

19

2. Materials 2.1. Ligation Reaction

1. RNA strands, DNA splint oligonucleotide. We prepared the modified and unmodified RNA and the splint by solid-phase synthesis, but they can also be purchased from commercial suppliers. We recommend HPLC purification of the oligo(ribo) nucleotides. 2. 10 × Ligation buffer (Fermentas, Hanover, MD): 400 mM Tris–HCl, 100 mM MgCl2, 100 mM dithiothreitol (DTT), 5 mM ATP, pH 7.8 at 25°C. 3. 50% (w/v) Polyethylene glycol (PEG) 4,000 solution (Fermentas). 4. T4 DNA ligase, 5 U/mL (Fermentas), in storage solution: 20 mM Tris–HCl, pH 7.5, 1 mM DTT, 50 mM KCl, 0.1 mM EDTA, 50% (v/v) glycerol. 5. T4 RNA ligase, 10 U/mL (Fermentas), in storage solution: 10 mM Tris–HCl, pH 7.5, 1 mM DTT, 50 mM KCl, 0.1 mM EDTA, 50% (v/v) glycerol.

2.2. Purification

1. A phenol/chloroform/isoamyl alcohol (25/24/1, v/v/v) solution is extracted three times with water and stored under a water layer at 4°C (see Notes 1 and 2). 2. Chloroform/isoamyl alcohol (24/1, v/v) solution. 3. HPLC system. 4. Eluant A: 25 mM Tris–HCl, pH 8.0, 6 M urea. Eluant B: 25 mM Tris–HCl, pH 8.0, 0.5 M NaClO4, 6 M urea. Prepare the eluants from a stock solution of 250 mM Tris–HCl, pH 8.0. Filtration of the eluants through a cellulose acetate filter, 0.2 μm pore size (Sartorius, Goettingen, Germany), is strongly recommended. 5. Anion exchange column DNAPac PA-100 or 200, 4 × 250 mm, (Dionex, Sunnyvale, CA). 6. Anion exchange column DNAPac PA-100, 9 × 250 mm (Dionex). 7. Sep-Pak Plus C18 Environmental cartridges (Waters, Milford, MA). 8. 0.15 M triethylamine bicarbonate buffer (Et3NH)HCO3: Prepare 1 M Et3N in water and pass CO2 into the solution until the pH reaches about 8. Store at 4°C. Dilute to 0.15 M just before use. 9. HPLC-grade acetonitrile (CH3CN) (Acros Organics, Geel, Belgium), 50% CH3CN in H2O.

20

Rieder, Höbartner, and Micura

2.3. Mass Spectrometry

1. 20 mM EDTA solution. Dissolve ethylenediaminetetraacetic acid disodium salt dihydrate in water; the pH does not need to be adjusted. 2. LC-ESI-MS system. We recommend the use of a Finnigan LCQ Advantage ion trap instrumentation (Thermo Fisher Scientific, Waltham, MA) connected to an Ettan micro HPLC system (GE Healthcare, Buckinghamshire, UK) with −4 kV applied to the spray needle. 3. XTerra MS C18, particle size 2.5 μm, 1 × 50 mm (Waters). 4. Eluant A: 8.6 mM Et3N, 100 mM 1, 1, 1, 3, 3, 3-hexafluoroisopropanol, pH 8.3. Prior to the addition of 1, 1, 1, 3, 3, 3-hexafluoroisopropanol, dissolve triethylamine in water. Store at 4°C. Eluant B: methanol, HPLC grade (Acros Organics).

3. Methods 3.1. General Aspects on the Design of Ligation Sites for Purine Riboswitch Domains 3.1.1. Requirements for the 5¢ and 3¢ Ends of the RNA Fragments

T4 ligases catalyze the formation of a phosphodiester linkage between the 5¢ monophosphate of a donor oligonucleotide and the free 3¢ hydroxyl of an acceptor oligonucleotide with the assistance of exogenous ATP. To avoid undesired by-products during ligation, it is advisable to use a donor that is blocked at the 3¢ end (e.g., phosphorylated) and an acceptor that has a free 5¢ hydroxyl group in addition to the 3¢-OH. If three RNA fragments are ligated consecutively (see Fig. 2a), the middle RNA fragment acts as donor and simultaneously as acceptor and therefore requires a monophosphate at the 5¢ end and a free hydroxyl at the 3¢ end (see also Fig. 3). Such a fragment is prone to circularization and to competitive ligation to any unblocked 3¢ termini. It should be embedded in a preligation complex that adapts a defined and stable secondary structure that resembles the structure of the intended ligation product. To achieve such a design, the ligation sites must be selected according to the individual substrate characteristics and preferences outlined under Subheading 3.1.3.

3.1.2. Length of RNA Strands

To meet no restriction for positioning of chemical modifications within the riboswitch target, we choose a ligation design that relies on all RNA fragments synthesized by solid-phase synthesis. This means that the individual strands are preferably shorter than 45 nt and none is larger than 60–65 nt. The respective 3¢ and 5¢ phosphate groups are directly introduced by solid-phase synthesis (see Fig. 2a).

3.1.3. T4 RNA versus T4 DNA Ligation Site

T4 RNA ligase and T4 DNA ligase differ in their substrate specificity. T4 RNA ligase prefers single-stranded RNA; in particular,

Enzymatic Ligation Strategies for the Preparation of Purine Riboswitches

21

the 5¢ phosphate should be accessible for the enzyme without restrictions (16). In contrast, T4 DNA ligase recognizes a nicked, double-stranded substrate, which is generated by the addition of a splint oligonucleotide to the two RNA molecules. Importantly, only perfectly aligned junctions get ligated (17). The efficiency of T4 RNA ligation is influenced by the ribonucleoside at the 3¢ terminus of the acceptor (A > C ³ G > U) and the ribonucleoside at the 5¢ terminus of the donor (pC > pU » pA > pG). The yield of ligation is even more dependent on structural premises that are given by the RNA secondary structure. T4 RNA ligation is most advantageous if the ligation site mimics its natural substrate, the 7 nt tRNA anticodon loop. In this sense, and with respect to purine riboswitches investigated here, nicks in loops L2, L3, and in the loop of the repressor or terminator element are highly appropriate for usage of T4 RNA ligase (Fig. 1). Different to T4 RNA ligase, T4 DNA ligase is rather unaffected by the nature of the terminal ribonucleoside. As long as the junction is without gaps or mispaired nucleotides, no further structural requirements are needed. Significantly, the size and the nature of the splint are crucial. The interaction of the splint and RNA has to be potent enough to break up intramolecular base pairing within the RNA substrate that interferes with intermolecular base pairing to form the preligation complex. In our experience, T4 RNA ligase is preferable over T4 DNA ligase whenever possible because of lower amounts of enzyme needed, easier product purification, and therefore higher isolated yields. 3.1.4. Ligation Conditions

Different conditions are needed for T4 RNA ligation compared to T4 DNA ligation. For riboswitch aptamer sequences, we observed that high RNA concentrations (40 mM each strand) and low temperature (21°C) are optimal for T4 RNA ligation, while lower concentration (10 μM each strand) and higher temperatures (37°C) are recommended for T4 DNA ligation. When both ligases are used in a one-pot ligation of three RNA fragments (Fig. 2), it is important to carefully optimize the ligation conditions with regard to RNA and ligase concentrations, buffer composition, and reaction temperature (see Note 3).

3.2. Ligation of Three Purine Riboswitch Fragments Using T4 DNA Ligase and T4 RNA Ligase

1. Mix 40 nmol containing aliquots from aqueous stock solutions of the three RNA strands and the splint oligonucleotide. Adjust the volume to 2,940 μL by the addition of water. 2. Heat to 90°C for 3 min. Then cool to room temperature for 20 min. 3. Add 400 μL of 10× ligation buffer and 400 μL of PEG 4,000 solution. Add 200 μL of T4 DNA ligase in storage solution.

22

Rieder, Höbartner, and Micura

4. Incubate the reaction solution at 33°C for 2 h. 5. Add 60 μL of T4 RNA ligase in storage solution and incubate for another 2.5 h at 33°C. 3.3. Purification of Ligation Product by Anion-Exchange HPLC

1. Extract the reaction mixture twice with 4 mL phenol/chloroform/isoamyl alcohol, and once with 4 mL chloroform/ isoamyl alcohol. Keep the organic layers separated. 2. Wash the organic layers stepwise with 4 mL water (twice) and combine with the extracted reaction mixture (see Note 4). 3. Concentrate to a volume of ~1 mL to make the stock solution of crude oligonucleotide (see Note 5). 4. Analyze the crude product by anion-exchange chromatography. Use 10–20 μL of the stock solution. We recommend the use of Dionex DNAPac PA-100 or 200 columns (4 × 250 mm) at 80°C. Flow rate: 1 mL/min; gradient: 0–60% eluant B in eluant A within 45 min; UV detection at 260 nm. 5. Purify the ligation product by anion-exchange chromatography. Use 100–150 μL portions of stock solution to avoid overloading of the column. Collect and combine productcontaining fractions (see Note 6). We recommend the use of Dionex DNAPac PA-100 columns (9 × 250 mm) at 80°C. Flow rate: 2 mL/min; gradient: approximately 30–50% eluant B in eluant A within 20 min; UV detection at 275 nm. 6. Dilute product-containing fractions 1 to 1 with 0.15 M (Et3NH)HCO3 buffer. 7. Pretreat the C18 cartridge consecutively with 20 mL CH3CN, 30 mL CH3CN/H2O, 30 mL H2O, and 30 mL 0.15 M (Et3NH)HCO3. 8. Load the product on the cartridge. 9. Wash with 10 mL 0.15 M (Et3NH)HCO3 buffer, and then twice with 10 mL H2O. 10. Elute product with 40 mL H2O/CH3CN (1/1) and evaporate to dryness. 11. Dissolve RNA in 1 mL H2O (see Note 7). 12. Quantify isolated product by UV spectroscopy (A260).

3.4. Characterization of Ligation Product by Mass Spectrometry

1. Dry an aliquot of the RNA solution containing 250 pmol of RNA and redissolve it in 20 μL of 20 mM EDTA solution. 2. Analyze the RNA by LC-ESI-MS in the negative ion mode. Inject 10–20 μL. In our setup, the ion trap instrumentation is connected to a micro-HPLC system equipped with XTerra MS C18 column.

Enzymatic Ligation Strategies for the Preparation of Purine Riboswitches

23

The samples are analyzed at 21°C; flow rate: 30 μL/min; gradient: 0–100% eluant B in eluant A within 30 min; UV detection at 254 nm.

4. Notes 1. Unless denoted otherwise, chemicals are available at SigmaAldrich, St. Louis, MO. 2. Only HPLC-grade water should be used in the protocols. 3. For optimization of ligation conditions, we recommend ligation on a small scale (using 400 pmol of each RNA fragment) and analysis by HPLC. 4. Stepwise back-extraction of the organic layers is required to prevent phenol accumulation in the product-containing aqueous layer. 5. Do not evaporate to dryness. Difficulties in redissolving the reaction mixture in 1 mL may occur. Mostly, pellets can be dissolved in additional water. 6. Store fractions at 4°C until they are desalted all together. 7. Store at −20°C. For prolonged storage, lyophilize the stock solution.

Acknowledgments We thank the Austrian Science Fund FWF (P17864) and the BMWF (Gen-AU program; projects “Non-coding RNAs” No. P7260-012-011 and No. P7260-012-012) for funding. References 1. Mandal, M., Boese, B., Barrick, J. E., Winkler, W. C., and Breaker, R. R. (2003). Riboswitches control fundamental biochemical pathways in Bacillus subtilis and other bacteria. Cell 113, 577–586. 2. Mandal, M. and Breaker, R. R. (2004). Adenine riboswitches and gene activation by disruption of a transcription terminator. Nat. Struct. Mol. Biol. 11, 29–35. 3. Batey, R. T. , Gilbert, S. D. , and Montange, R. K. (2004). Structure of a natural guanine-responsive riboswitch complexed with

the metabolite hypoxanthine . Nature 432, 411 – 415 . 4. Serganov, A., Yuan, Y. R., Pikovskaya, O., Polonskaia, A., Malinina, L., Phan, A. T., Höbartner, C., Micura, R., Breaker, R. R., and Patel, D. J. (2004). Structural basis for discriminative regulation of gene expression by adenine- and guanine-sensing mRNAs. Chem. Biol. 11, 1729–1741. 5. Gilbert, S. D., Stoddard, C. D., Wise, S. J., and Batey, R. T. (2006). Thermodynamic and kinetic characterization of ligand binding to

24

6.

7.

8.

9.

10.

11.

12.

Rieder, Höbartner, and Micura the purine riboswitch aptamer domain. J. Mol. Biol. 359, 754–768. Lemay, J.-F., Penedo, J. C., Tremblay, R., Lilley, D. M. J., and Lafontaine, D. A. (2006). Folding of the adenine riboswitch. Chem. Biol. 13, 857–868. Wickiser, J. K., Cheah, M. T., Breaker, R. R., and Crothers, D. M. (2005). The Kinetics of ligand binding by an adenine-sensing riboswitch. Biochemistry 44, 13404–13414. Lemay, J. F. and Lafontaine, D. A. (2007). Core requirements of the adenine riboswitch aptamer for ligand binding. RNA 13, 339–350. Mulhbacher, J. and Lafontaine, D. A. (2007). Ligand recognition determinants of guanine riboswitches. Nucl. Acids Res. 35, 5568–5580. Noeske, J. Buck, J. Fürtig, B. Nasiri, H. R., Schwalbe, H., and Wöhnert, J. (2007). Interplay of ‘induced fit’ and preorganization in the ligand induced folding of the aptamer domain of the guanine binding riboswitch. Nucl. Acids Res 35, 572–583. Noeske, J., Schwalbe, H., and Wöhnert, J. (2007). Metal-ion binding and metal-ion induced folding of the adenine-sensing riboswitch aptamer domain. Nucl. Acids Res. 35, 5262–5273. Rieder, R., Lang, K., Graber, D., and Micura, R. (2007). Ligand-induced folding of the

13.

14.

15.

16.

17.

adenosine deaminase A-riboswitch and implications on riboswitch translational control. ChemBioChem 8, 896–902. Micura, R. (2002). Small interfering RNAs and their chemical synthesis. Angew. Chem. Int. Ed. 41, 2265–2269. Höbartner, C. and Micura, R. (2004). Chemical synthesis of selenium-modified oligoribonucleotides and their enzymatic ligation leading to an U6 snRNA stem-loop segment. J. Am. Chem. Soc. 126, 1141–1149. Höbartner, C., Rieder, R., Puffer, B., Lang, K., Polonskaia, A., Serganov, A., and Micura, R. (2005). Syntheses of RNAs with up to 100 nucleotides containing site-specific 2¢-methylseleno labels for use in X-ray crystallography. J. Am. Chem. Soc. 127, 12035–45. Persson, T., Willkomm, D. K., and Hartmann, R. K. (2005). T4 RNA ligase, in Handbook of RNA Biochemistry (Hartmann, R. K., Bindereif, A., Schön, A., and Westhof, E., eds.), Wiley-VCH, Weinheim, Germany, pp. 53–74. Frilander, J. M. and Turunen, J. J. (2005). RNA ligation using T4 DNA ligase, in Handbook of RNA Biochemistry (Hartmann, R. K., Bindereif, A., Schön, A., and Westhof, E., eds.), Wiley-VCH, Weinheim, Germany, pp. 36–52.

Chapter 3 Application of Fluorescent Measurements for Characterization of Riboswitch–Ligand Interactions Benoit Heppell, Jérôme Mulhbacher, J. Carlos Penedo, and Daniel A. Lafontaine Summary Riboswitches are recently discovered messenger RNA motifs involved in gene regulation. They modulate gene expression at various levels, such as transcription, translation, splicing, and mRNA degradation. Because riboswitches exhibit relatively complex structures, they are able to form highly complex ligandbinding sites, which enable the specific recognition of target metabolites in a complex cellular environment. Practically in all studied cases, riboswitches use ligand-induced conformational changes to control gene expression. To monitor the structural reorganization of riboswitches, we use the local fluorescent reporter 2-aminopurine (2AP), which is a structural analog of adenine. The 2AP fluorescence is strongly quenched when the fluorophore is involved in stacking interactions with surrounding bases, and can, therefore, be used to monitor local structural rearrangements. Here, we show specific examples in which 2AP fluorescence can be used to monitor structural changes in the aptamer domain of the S-adenosyl methionine (SAM) riboswitch and where it can be used as a ligand for the guanine riboswitch. Key words: Riboswitch, 2-Aminopurine, SAM, Guanine, Fluorescence, Competition assays

1. Introduction Riboswitches are typically found in the 5¢-untranslated regions of mRNA and are intimately involved in gene regulation by binding specific cellular metabolites (1–4). Riboswitches do not require proteins that act as metabolite sensors (1). The formation of the complex between the metabolite-sensing (aptamer) domain of the riboswitch and the metabolite promotes a structural reorganization of the mRNA, which modulates gene expression via a classical retro-inhibition feedback mechanism (3). The structural Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_3 © Humana Press, a part of Springer Science + Business Media, LLC 2009

25

26

Heppell et al.

characterization of riboswitches is crucial for grasping how target metabolites can be directly used to perform gene regulation. Using the local fluorescence reporter 2-aminopurine (2AP), we have characterized the adenine (5, 6), guanine (7) and lysine riboswitches (8). For riboswitch studies, 2AP can be utilized in several experimental variations, each characterizing a different aspect of riboswitch activity. For instance, 2AP can be incorporated internally into the aptamer domain of a riboswitch in both base paired and singlestranded regions (8–11), or it can serve as a ligand in the case of purine riboswitches (5–7, 9, 12). Adenine and 2AP can form isosteric base pairs with uracil and, therefore, 2AP can be used as a replacement for adenine in nucleic acid helices without introducing local distortions (Fig. 1). 2AP can be selectively excited in the presence of DNA, RNA, and proteins, given that the absorption of 2AP occurs at longer wavelengths than those of nucleobases and aromatic amino acids. Because the length of riboswitch sequences often exceeds the practical limit of chemical RNA synthesis, it is sometimes necessary to employ enzymatic strategies to join several RNA strands into a complete RNA molecule.

Fig. 1. Adenine and 2-aminopurine form isosteric base pairs with uracil. Hydrogenbonding interactions are shown in dashed lines.

Application of Fluorescent Measurements

27

2. Materials 2.1. In Vitro Transcription

1. DNA oligonucleotides: synthetic oligonucleotides are purchased from Sigma Genosys (Oakville, Canada). All ordered oligonucleotides are purified on reverse-phase cartridges, typically ensuring the presence of 80–90% of full-length product. The basis of this purification is the presence of the hydrophobic 5¢-dimethoxy-trityl (DMT) group, carried by the last nucleotide added during chemical synthesis. This group increases the overall hydrophobicity of the full-length DNA oligonucleotides and enhances their binding to the reverse-phase column, ensuring their separation from incompletely synthesized oligonucleotides. However, the difference in hydrophobicity between the full-length DMT products and the non-DMT failure sequences diminishes as oligonucleotide length increases. Thus, longer molecules require polyacrylamide gel electrophoresis (PAGE) purification to ensure adequate purity for further steps. 2. Templates for transcription: DNA templates are generated in two ways. For templates shorter than 100 nucleotides, deoxyoligonucleotides that contain the sequence to be transcribed together with the T7 RNA polymerase promoter (GGTAATACGACTCACTATA) are paired during a slow cooling procedure (heat at 70°C and cool slowly to 35°C) in 10 mM Tris–HCl, pH 7.9, 10 mM MgCl2, and 50 mM KCl. However, for longer templates, a recursive PCR amplification procedure is employed using the appropriate oligonucleotides. When possible, a GCG sequence is used at the beginning of transcribed sequences to reduce the incorporation of incorrect nucleotides during transcription (13). 3. Transcription buffer (10×): 400 mM Tris–HCl, pH 7.9, 0.1% (v/v) Triton X-100, 200 mM MgCl2, 20 mM spermidine. 4. Transcription reaction: 1× transcription buffer, 10 mM dithiothreitol (DTT), 2.5 mM rNTPs, 250 pmol of template, and T7 polymerase. The amount of T7 enzyme added mostly depends on the scale of the transcription and, therefore, needs to be optimized for most reactions. Moreover, to improve the yield of transcription, pyrophosphatase may be added at the beginning of the reaction to reduce the amount of pyrophosphate produced during transcription.

2.2. Synthetic RNA

1. RNA that contain 2AP are purchased from Dharmacon (Boulder, CO). The deprotection buffer and the protocol are provided by the manufacturer. If the RNA molecule requires a 5¢-PO4 extremity to be used as a substrate for the T4 RNA ligase enzyme, it is appropriate to ask the company to add it during chemical synthesis (see Note 1).

28

Heppell et al.

2.3. Polyacrylamide Gel Electrophoresis

1. 5× TBE: 450 mM Tris base, 450 mM borate, and 10 mM ethylenediamine tetraacetic acid disodium salt (EDTA). 2. Dilution buffer: 7 M urea dissolved in 1× TBE. Store at room temperature. 3. 20% Acrylamide/bis-acrylamide solution (19:1) with 7 M urea in 1× TBE. This solution is neurotoxic when not polymerized. Keep the solution in the dark at room temperature. 4. Ammonium persulfate (APS): prepare 10% (w/v) solution in 10 mL water. Store solution in the dark at 4°C for 2–3 weeks. 5. TEMED (N,N,N,N¢-Tetramethyl-ethylenediamine). 6. Molecular weight markers: xylene cyanol FF 0.02% and bromophenol blue 0.02% dissolved in 100% formamide. 7. Glass plates (20 cm× 40 cm), 1.5-mm thick spacers, and comb for gel preparation. 8. Thin layer chromatography (TLC) plates (Voigt Global Distribution Inc, Lawrence, KS).

2.4. RNA Elution

1. Electroeluter from Harvard lab shops (http://www.mcb. harvard.edu/bioshop). 2. Ammonium acetate (NH4OAc), 8 M in water. Keep the solution at room temperature. 3. RNA precipitation solutions: 3 M sodium acetate (NaOAc), pH 5.2, 100% ethanol. Keep these solutions at room temperature.

2.5. RNA Ligation

1. T4 RNA ligase (New England Biolabs, Ipswich, MA) supplied with 10× buffer.

2.6. Fluorescence Spectroscopy

1. Quanta-Master fluorometer (Photon Technology International, Birmingham, NJ) or similar fluorometers that allow the collection of emission spectra from 330 to 420 nm using an excitation wavelength of 300 nm (see Note 2).

3. Methods The 2AP reporter can be specifically incorporated into an aptamer molecule or used as an external ligand. In either case, special care should be taken not to perturb the native molecular interactions. To assess whether the presence of 2AP causes unwanted effects in the studied riboswitch, a variety of controls can be performed, such as in-line probing, KD measurements, and competition assays.

Application of Fluorescent Measurements

29

As in most nucleotide replacements, the engineered molecule is rarely as active as the wild-type molecule. Thus, it is important to have an assay that can directly monitor the influence of 2AP substitution on activity. The formation of the riboswitch–ligand complex is typically accompanied by structural changes in the riboswitch, which are essential for gene regulation control (1–4). The probing of these changes is, therefore, very important for understanding how these accommodations are harnessed to drive riboswitch function. The introduction of 2AP into strategic locations allows for the observation of specific ligand-induced structural changes, which, in turn, can hint at large conformational transitions associated with the control of gene expression. In this section, we describe techniques that can be utilized to successfully introduce an internal 2AP into an RNA molecule and to apply the 2AP-modified RNA molecules to monitor RNA rearrangements. In addition, we present methods to study the conformational changes of riboswitches by using external 2AP. 3.1. In Vitro Transcription

1. For each transcription sample (100 µL), combine 2.5 mM rNTPs, 1× transcription buffer, 10 mM DTT, 250 pmol DNA template, and 50 U of T7 RNA polymerase. 2. Gently mix the sample and centrifuge shortly. 3. Incubate 3 h at 37°C. Gently shake the tubes every 30 min. 4. Precipitate the sample by adding 3 volumes of 100% ethanol and 0.1 volume of 3 M sodium acetate. Place the sample at −20°C overnight or at −80°C for 30 min. 5. Centrifuge tubes at 13,000 × g for 30 min at 4°C. 6. Discard the supernatant and perform a quick spin to remove residual ethanol. Air dry pellets for 5 min. Resuspend pellets in 75% (v/v) formamide for PAGE purification.

3.2. Synthetic RNA

1. RNA molecules that contain 2AP need to be chemically synthesized. We purchase RNA molecules from Dharmacon (Boulder, Co), which employs 2¢-ACE chemistry for RNA synthesis. However, synthetic RNA can be purchased from any other RNA suppliers, which use different RNA chemistry (e.g., t-butyldimethylsilyl (TBDMS), 2¢-o-triisopropylsilyloxymethyl (ToM), etc.). 2. 2¢-ACE synthetic RNA are deprotected (2¢-OH protecting group removal), as recommended by Dharmacon. The deprotection buffer is removed using the Speedvac and pellets are resuspended in 75% (v/v) formamide for PAGE purification.

3.3. Polyacrylamide Gel Electrophoresis

1. These instructions have been prepared for standard gel running systems (40 cm × 20 cm) with a 1.5-mm thick comb and spacers.

30

Heppell et al.

The following protocol describes the preparation of a 10% acrylamide gel, which is most frequently used in our experiments. 2. Prepare a 10% gel by mixing 30 mL of 20% (w/v) acrylamide/7 M urea solution with 30 mL dilution buffer. Add 450 µL ammonium persulfate solution and 45 µL TEMED consecutively. Shake well after each addition. Insert the comb and allow the gel to polymerize for at least 1 h. 3. After polymerization, remove the comb and carefully wash the wells using a syringe filled with 1× TBE. Prerun the gel at 18 W for at least 30 min using 1× TBE as the running buffer. 4. Just before loading the samples, rewash the wells with a syringe. Load the samples, and run the gel at 18 W for 3 h. On a 10% gel, xylene cyanol and bromophenol dyes migrate at approximately the same rate as oligonucleotides of 55 and 10 nucleotides, respectively, do (see Note 3). 5. Once the migration is complete, remove the glass plates and carefully put the acrylamide gel on a TLC plate to visualize RNA bands by UV shadowing (254 nm). 6. Cut the bands of interest using clean blades. Samples can either be stored in a freezer or used immediately for electroelution (see Note 4). 7. Cut a band containing xylene cyanol or bromophenol blue as a control for electroelution. 3.4. RNA Electroelution

1. Various techniques are used to extract RNA from the acrylamide. For the large-scale purification of RNA, we use electroelution since we have typically attained good recovery yields using this technique. 2. Fill the electroeluter apparatus with 0.25× TBE and run it at 200 V for 30 min to clean the chamber. Discard the running buffer and fill it with freshly made 0.25× TBE buffer. 3. Add 200 µL of 8 M NH4OAc to the small “trapping” wells, where RNA will be collected during electroelution. 4. Carefully place acrylamide slices in the large wells. As a control for the electroelution process, place the acrylamide pieces containing dye in a separate well. Add enough of the running buffer to cover the gel. RNA recovery is more efficient when the acrylamide slices are cut into small pieces. 5. Run the electroelution at 120 V for 1 h or until the dye is completely removed from the acrylamide. 6. Cut ~5 mm off of P200 tips and insert them into the trapping wells. 7. Remove running buffer from the electroeluter using a syringe.

Application of Fluorescent Measurements

31

8. Collect the RNA samples from each trapping well and transfer them to separate Eppendorf tubes. Rinse each trapping well with 1× TBE to collect residual RNA. 9. Fill up each tube with 100% ethanol for precipitation. 10. Recover RNA as described in Subheading 3.1. Resuspend each pellet in water. Determine RNA concentration by using UV spectroscopy. An optical density (O.D.) of 1.0 corresponds to an RNA concentration ~40 µg/mL. 3.5. RNA Ligation

1. Depending on the size of the molecule studied (i.e., >60 nucleotides), the chemical synthesis of an entire RNA strand is not always possible. Therefore, alternative methods, such as the ligation of two RNA pieces, are required. In the present case, the SAM aptamer (Fig. 2), which contains 140 nucleotides, clearly necessitates the use of ligation to join an RNA transcript to a 2AP-containing synthetic RNA. The ligation reaction requires 3¢ OH and 5¢ PO4 termini, which are carried by the transcript and the synthetic RNA, respectively.

Fig. 2. The sequence and secondary structure of a class I SAM riboswitch aptamer. The SAM aptamer is organized around a four-way junction formed by stems P1, P2, P3, and P4. A star denotes the position where 2AP is introduced. The ligation site is indicated with an arrow.

32

Heppell et al.

The PO4 group may be added during chemical synthesis or postsynthetically using a kinase reaction. 2. For the ligation reaction to be efficient, the RNA strands should be held together. In the SAM aptamer, the chemically synthesized RNA fragment can be successfully paired with the in vitro transcribed fragment. The efficiency of pairing can be improved by adding extra nucleotides to both RNA to extend the P1 and the P4 stems of the aptamer (Fig. 2). In addition, the P4 stem has been increased by two base pairs, and the P4 loop sequence has been changed to GCUUAAA to increase the ligation efficiency. 3. Typically, the ligation reaction is performed using 1 and 1.5 nmol of transcribed and synthetic RNA, respectively. Mix transcribed and synthetic RNA molecules in T4 RNA ligase buffer (~30 µL total). Proceed with a slow cooling procedure to promote hybridization between both RNA strands. 4. After slow cooling, add 4 µL of T4 RNA ligase and incubate at 37°C for 3 h. To increase the efficiency of ligation, add an aliquot of T4 RNA ligase after 90 min. The typical ligation efficiency is ~75%. 5. Upon completion of the reaction, 1 volume of 100% formamide is added to the mixture and samples are loaded onto a polyacrylamide gel. Two major bands, corresponding to ligated and nonligated molecules, should be observed on gel. Cut the band corresponding to the ligated product. A transcript that is similar in size to the ligated product can be loaded on the same gel and used as a molecular marker. Purify the ligated product as described in Subheading 3.4. 3.6. Fluorescence Spectroscopy Using Internal 2AP

1. In the following example, 2AP was introduced into the singlestranded region between stems P3 and P4 (Fig. 3a). In the crystal, this region is involved in coaxial stacking between helices P1 and P4 (14). By using a 2AP-containing RNA, it is possible to monitor structural changes in the P1–P4 region induced by salts or ligand binding. Here, we show a representative titration using SAM, which decreases the 2AP fluorescence intensity of the RNA upon binding to it (Fig. 3b). During titrations, it is preferable to use concentrated ligand solutions to avoid diluting the sample. After each addition of the ligand, the sample should be incubated for 3–5 min to complete complex formation and improve reproducibility. A broad range of concentrations and at least 15–20 data points are necessary for the accurate determination of binding constants (Fig. 3c). 2. Prepare 100 µL of fluorescent mixture by combining 1.5 µM RNA, 25 mM NaCl, and 50 mM Tris–HCl, pH 8.3. In some

Application of Fluorescent Measurements

33

Fig. 3. SAM binding to the aptamer domain of the riboswitch results in 2AP fluorescence quenching. (a) Schematic showing that the fluorescence of the engineered 2AP-labeled SAM aptamer is quenched upon ligand binding. 2AP is represented by a black oval. The star shape around 2AP indicates its fluorescence when it is not bound to the aptamer. The ligand is labeled with the letter S. (b) Fluorescence emission spectra of 2AP from 330 to 420 nm for various SAM concentrations. The bold line represents the fluorescence spectrum in the absence of ligand, and the arrow indicates the decrease in fluorescence as ligand concentration is increased. (c) Normalized 2AP fluorescence quenching plotted as a function of SAM concentration. The line represents the best fit to a binding model.

cases, it may be appropriate to increase the RNA concentration to obtain a better signal. 3. Perform a slow cooling procedure to reduce misfolding within the RNA. 4. Add the ligand, incubate the reaction mixture for 3–5 min, and collect data. 3.7. Analysis of 2AP Fluorescence Assay

In our experiments, the addition of ligand quenches 2AP fluorescence. To quantify the fluorescence changes, it is useful to normalize data to the maximum fluorescence measured in the absence of ligand (i.e., in the absence of ligand-induced fluorescence quenching). This value is used as a reference for variations observed at all concentrations. The binding can thus be described by the binding model:

34

Heppell et al.

a = KA[SAM]/(1 + KA[SAM])

(1)

where a is the change in fluorescence intensity and KA is the association constant. An apparent dissociation constant, the concentration of the ligand at which the fluorescence transition is half-completed, is related to KA by the equation KDapp = (1/KA). Substitution of KA by KDapp in Eq. 1 allows for the calculation of a dissociation constant of 5.3 ± 0.8 nM for the interaction between SAM and the riboswitch. This value is in excellent agreement with data from previous studies (15). 3.8. Fluorescence Spectroscopy Using External 2AP

For adenine and guanine riboswitches, 2AP can be used to form an aptamer–ligand complex (5–7). Given that 2AP is used as a ligand, the RNA strand can be entirely produced by T7 RNA polymerase. In general, comments in Subheading 3.6 are applicable to fluorescence studies using external 2AP. However, the titration is performed by increasing RNA concentration. In addition, it is possible to perform a variety of titration experiments. For instance, in the presence of excess 2AP, riboswitch folding can be monitored at various concentrations of magnesium ion.

3.9. Competition Experiments Using External 2AP

1. Purine riboswitches are known to bind various ligands (1–4). To determine their binding affinities for diverse ligands (Fig. 4a), it is often advantageous to use a competition assay, where a competing ligand displaces 2AP upon binding (Fig. 4b). One tube is prepared for each competitor concentration used. 2. Prepare 100 µL of solution for fluorescent measurements with final concentrations of 100 mM KCl, 10 mM MgCl2, 50 mM Tris–HCl, pH 8.3, and 50 nM 2AP. Add the appropriate amount of aptamer to get >80% of fluorescence quenching (~ 1 µM). It is important to observe significant quenching so that the restoration of the fluorescence upon the binding of the competing ligand can be easily detected. 3. Add the competing ligand in the range 1–100 µM. For initial experiments, it is appropriate to use broad concentrations of the competitor that vary by a factor of 10. 4. Perform slow cooling (70–35°C) before collecting data. 5. Collect the data. We typically perform experiments in triplicate. 6. The affinity of each competitor is normalized to the 2AP quenching value in the absence of the competing ligand. The results are presented as histograms (Fig. 4c).

Application of Fluorescent Measurements

35

Fig. 4. 2AP can be used as a ligand for a purine riboswitch aptamer in competition assays. (a) Schematic showing the residue C74 of the guanine riboswitch ligand-binding site bound to guanine (left ). Possible hydrogen-bonding schematics of various ligands with the residue C74. A gray oval indicates an additional imino proton present at the N3 position of xanthine. (b) Schematic showing the competition between 2AP and guanine (G) for the purine-binding site in the riboswitch. The star shape around 2AP indicates its fluorescence when it is not bound to the aptamer. (c) 2AP fluorescence competition assays performed in the absence of competing ligand (−) or in presence of guanine (G), adenine (A), xanthine (X), or hypoxanthine (H). Each ligand was tested at 1 and 10 nM concentrations.

4. Notes 1. 2AP location should not affect the structure and function of the studied aptamer. A control must be done to eliminate this possibility. In-line probing is most appropriate for ensuring

36

Heppell et al.

that ligand-induced RNA structural changes are not affected (16). For other systems, such as ribozymes, functional assays can be used to confirm the structural integrity of the RNA construct (17). In addition, gel shift assays or enzymatic reactions can be used for testing the formation of protein–RNA complexes. 2. 2AP can be selectively excited at wavelengths between 300 and 330 nm. However, to obtain a good separation between the Raman and the fluorescence peaks, it is preferable to perform the excitation at 300 nm. 3. Both xylene cyanol FF and bromophenol blue dyes show UV absorption when visualized on TLC plates. If RNA is expected to comigrate with the dyes, it is preferable to load the dyes in a separate well. 4. It is important not to overload the gel during PAGE purification since this will greatly compromise the resolution of the gel. When optimizing transcription conditions, it is preferable to test different volumes of the transcription mixture to be loaded on the gel by, for instance, consecutively decreasing the loading volume twofold several times. This simple technique enables the correct volume of the transcription reaction to be quickly chosen and allows for the detection of other, if any, major RNA molecules, produced during transcription.

Acknowledgments This work was supported by a graduate scholarship (BH) and a postdoctoral fellowship (JM) from the National Sciences and Engineering Research Council of Canada (NSERC) and by operating grants from the Canadian Institutes of Health Research (DAL) and the Scottish Universities Physics Alliance (JCP). DAL is a CIHR New Investigator Scholar. References 1. Mandal, M. and Breaker, R. R. (2004). Gene regulation by riboswitches. Nat. Rev. Mol. Cell. Biol. 5, 451–463. 2. Winkler, W. C. (2005). Riboswitches and the role of noncoding RNAs in bacterial metabolic control. Curr. Opin. Chem. Biol. 9, 594–602. 3. Winkler, W. C. and Breaker, R. R. (2005). Regulation of bacterial gene expression by riboswitches. Annu. Rev. Microbiol. 59, 487–517. 4. Soukup, J. K. and Soukup, G. A. (2004). Riboswitches exert genetic control through

metabolite-induced conformational change. Curr. Opin. Struct. Biol. 14, 344–349. 5. Lemay, J. F., Penedo, J. C., Tremblay, R., Lilley, D. M. and Lafontaine, D. A. (2006). Folding of the adenine riboswitch. Chem. Biol. 13, 857–868. 6. Lemay, J. F. and Lafontaine, D. A. (2007). Core requirements of the adenine riboswitch aptamer for ligand binding. RNA 13, 339–350. 7. Mulhbacher, J. and Lafontaine, D. A. (2007). Ligand recognition determinants of guanine riboswitches. Nucleic Acids Res. 35, 5568–5580.

Application of Fluorescent Measurements 8. Blouin, S. and Lafontaine, D. A. (2007). A loop–loop interaction and a K-turn motif located in the lysine aptamer domain are important for the riboswitch gene regulation control. RNA 13, 1256–1267. 9. Gilbert, S. D., Stoddard, C. D., Wise, S. J. and Batey, R. T. (2006). Thermodynamic and kinetic characterization of ligand binding to the purine riboswitch aptamer domain. J. Mol. Biol. 359, 754–768. 10. Rieder, R., Lang, K., Graber, D. and Micura, R. (2007). Ligand-induced folding of the adenosine deaminase A-riboswitch and implications on riboswitch translational control. Chembiochem 8, 896–902. 11. Lang, K., Rieder, R. and Micura, R. (2007). Ligand-induced folding of the thiM TPP riboswitch investigated by a structure-based fluorescence spectroscopic approach. Nucleic Acids Res. 35, 5370–5378. 12. Wickiser, J. K., Cheah, M. T., Breaker, R. R. and Crothers, D. M. (2005). The kinetics of

13.

14.

15.

16.

17.

37

ligand binding by an adenine-sensing riboswitch. Biochemistry 44, 13404–13414. Pleiss, J. A., Derrick, M. L. and Uhlenbeck, O. C. (1998). T7 RNA polymerase produces 5¢end heterogeneity during in vitro transcription from certain templates. RNA 4, 1313–1317. Montange, R. K. and Batey, R. T. (2006). Structure of the S-adenosylmethionine riboswitch regulatory mRNA element. Nature 441, 1172–1175. Winkler, W. C., Nahvi, A., Sudarsan, N., Barrick, J. E. and Breaker, R. R. (2003). An mRNA structure that controls gene expression by binding S-adenosylmethionine. Nat. Struct. Biol. 10, 701–707. Soukup, G. A. and Breaker, R. R. (1999). Relationship between internucleotide linkage geometry and the stability of RNA. RNA 5, 1308–1325. Lafontaine, D. A., Norman, D. G. and Lilley, D. M. (2002). The global structure of the VS ribozyme. EMBO J. 21, 2461–2471.

Chapter 4 Transcriptional Approaches to Riboswitch Studies Alexander Mironov, Vitaly Epshtein, and Evgeny Nudler Summary Natural RNA sensors of small molecules (a.k.a. riboswitches) regulate numerous metabolic genes. In bacteria, these RNA elements control transcription termination and translation initiation by changing the folding pathway of nascent RNA upon direct binding of a metabolite. To identify and study riboswitches we used in vitro reconstituted solid-phase transcription elongation/termination system. This approach allows for direct monitoring of ligand binding and riboswitch functioning, establishing the working concentration of a ligand as a function of RNA polymerase speed, and also probing RNA structure of the riboswitch. Using this system we have been able to identify and characterize first several riboswitches including those involved in vitamin biosynthesis and sulfur metabolism. The system can be utilized to facilitate biochemical studies of riboswitches in general, i.e., to simplify analysis of riboswitches that are not necessarily involved in transcriptional control. Key words: Riboswitch, RNA polymerase, Transcription elongation, Termination, Attenuation, Metabolites

1. Introduction In the last several years, riboswitches have become a paradigm of the RNA-driven regulation of gene expression (1–5). These noncoding stretches of RNA adopt an alternative conformation upon direct binding of a small molecular ligand to activate or repress cognate genes. In bacteria, riboswitches are ubiquitous and control over 3% of all metabolic genes. They usually reside in the 5¢ untranslated regions (UTR) and consist of two coupled elements: evolutionary conserved sensor domain, which directly binds the ligand; and a variable expression platform that transmits the signal to the gene expression machinery such as RNA polymerase or a ribosome (1, 2). Depending on its design, a Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_4 © Humana Press, a part of Springer Science + Business Media, LLC 2009

39

40

Mironov, Epshtein, and Nudler

riboswitch either prevents or permits the formation of an RNA attenuator that in turn dictates whether a gene is expressed or not. The attenuator functions as sequester of the ribosomebinding site or intrinsic terminator positioned in front of an operon or both. A few riboswitches that have been found in fungi and plants work at the level of alternative RNA splicing (6–8). Since the sensor domain is highly conserved within each family of riboswitches, it can be placed within a context of a standard transcription termination expression platform to benefit from a simple and quantitative in vitro readout system. Such a strategy also takes full advantage of the solid-phase reconstituted transcription system for thorough biochemical analysis of riboswitch functioning. The system utilizes “walking” technology to obtain individual homogenous elongation complexes (EC) stalled at any desired position within the transcription unit (9). This allows for various types of RNA structural analysis (e.g., chemical and enzymatic probing, crosslinking, and real-time spectrometry) to be performed during transcription thus reflecting a natural situation of cotranscriptional RNA folding. The system also takes into account the kinetics of transcription elongation.

2. Materials 2.1. Protein Purification

1. Escherichia coli and Bacillus subtilis strains recommended for this procedure should express a chromosome-encoded β¢ subunit of RNA polymerase with a His6-tag at the C terminus (10). We used the E. coli strain RL 721 (F-l-thr-1 leu-6 pro-A2 his-4 thi-1 argE3 lacY1 galK2 ara-14 syl-5 mtl-1 tsx-33 rpsL31 supE37 recB21 recC22 sbcB15 sbcC201 rpoC3531(His6) zja::kan). 2. The E. coli strains M5248 (l bio275 c1857 DH1) and JC4588 (F−endoI−recA56 gal his322 thi) and the plasmid pEcoRQ111 (11), which is a derivative of the pSCC2 (ApR), were kindly provided by Dr. Paul Modrich (Duke University). The pEcoRQ111 plasmid was purified from the JC4588 strain and was introduced into the M5248 expression strain just prior to preparation of the enzyme. 3. E. coli or Bacillus culture medium LBD (Gibco/BRL, Bethesda, MD): 0.5% (w/v) Bactotryptone, 1.0% (w/v) Bacto-yeast extract, 0.5% (w/v) NaCl, and 1% (w/v) dextrose. 4. Lysis buffer: 300 mM NaCl, 50 mM Tris–HCl, pH 7.9, 5% (v/v) glycerol, 10 mM ethylenediaminetetraacetic acid (EDTA), 0.1% (w/v) phenylmethanesulfonylfluoride (PMSF), 10 mM mercaptoethanol, 133 μg/mL lysozyme (Sigma).

Transcriptional Approaches to Riboswitch Studies

41

5. 10% (w/v) sodium deoxycholate (Sigma). 6. 10% (v/v) polymine P solution (Sigma). 7. Tris–glycerol–EDTA–DTT (TGED) buffer: 50 mM Tris– HCl, pH 7.9, 5% (v/v) glycerol, 0.1 mM EDTA, and 0.1 mM DTT. 8. TGED50, TGED200, TGED300, TGED500, TGED600, TGED1000 buffer: TGED buffer supplemented with 50, 200, 300, 500, 600, and 1,000 mM NaCl, respectively. 9. Sonic Dismembrator 60 (Fisher Scientific). 10. Solid NH2SO4. 11. Centricon-50 filter cartridges (Amicon/Millipore, Billerica, MA). 12. Chromatography columns: Superose 6 HR 10/30; Mono Q HR 5/5; 10 mL Heparin-Sepharose (GE Healthcare, Piscataway, NJ). 13. Template-dependent forward and reverse oligonucleotides: 50 pmol/μL stock. As an example: 5¢-CCAGATCCCGAAAATTTATC (forward) and 5¢-CACTGACCCTTTTGGGACCGC (reverse). 2.2. DNA Templates and NTP Substrates Preparation

1. Peristaltic pump, gradient mixer, and UV detector (GE Healthcare). 2. dNTP mixture (1 mM each nucleotide) (Promega, Madison, WI). Store at −20°C. 3. Tris–EDTA (TE) buffer: 10 mM Tris–HCl, pH 7.9, and 1 mM EDTA. 4. SeaKem low-melting agarose (Cambrex, East Rutherford, NJ). 5. Phusion DNA polymerase and 10× Phusion reaction buffer (Fermentas, Glen Burnie, MD) as provided by the manufacturer.

2.3. Solid-Phase In Vitro Transcription

1. Ultrapure ribonucleotide triphosphates (GE Healthcare), prepared as 10 mM individual solutions and 1 mM (each) mixture (10× NTPs solution). Store at −20°C. 2. Transcription buffer (TB): 10 mM Tris–HCl, pH 7.9, 10 mM MgCl2, and 100 mM KCl. 3. “High salt” TB buffer: TB supplemented with 1 M KCl. 4. ApUpC RNA primer (Oligos Etc., Wilsonville, OR). 5. 10× starting mix: 100 μM ApUpC RNA primer, 250 μM ATP and GTP. 6. [α-32P] CTP, 3,000 Ci/mmol, 0.5 μM (NEN/PerkinElmer, Waltham, MA)

42

Mironov, Epshtein, and Nudler

7. Loading mix: 12 M urea, 20 mM EDTA, 10 mM Tris–HCl, pH 7.5, 0.0125% (w/v) each of bromophenol blue (BPB) and xylene cyanol (XC). 8. Resins: Ni2+-NTA (Qiagen, Valencia, CA); TALON (Clontech, Mountain View, CA); NutrAvidin (Pierce, Biotechnology, Rockford, IL). 9. 1 mg/mL rifampicin (Sigma) dissolved in water (10× solution). 10. Restriction enzyme EcoRQ111, isolated as described in (15) and stored at ~1 mg/mL at −20°C in a buffer containing 20 mM KPO4, pH 7.4, 0.4 M KCl, 1 mM EDTA, 0.2 mM DTT, 50% (v/v) glycerol. 11. Sequencing gel: 8 M urea, 10% acrylamide:bisacrylamide solution (29:1). 12. RNase H (Roche, Basel, Switzerland). 13. Microcuvettes for fluorescent studies (StarnaCells, Atascadero, CA). 14. Template-dependent antisense oligonucleotides.

3. Methods 3.1. Preparation of the DNA Template

The common structure of the template for riboswitch studies is shown in Fig. 1. It consists of the A1 promoter of bacteriophage T7, adapter sequence, the sensor domain of a riboswitch, and an intrinsic transcription terminator to be coupled with a riboswitch of choice. T7A1 is one of the strongest known promoters to support in vitro transcription by RNA polymerase (RNAP) from Gram-negative or Gram-positive bacteria, including E. coli and B. subtilis. It works within a broad range of temperatures and salt concentrations, and supports multiround transcription in vitro without additional factors. The adapter sequence is a common initial transcribed sequence (ITS), which allows for one-step preparation of the homogenous highly radiolabeled elongation complex stalled at position +20°C (EC20). The terminator

−35 −10 +1 +20 T7A1 promoter ITC Sensor domain of the riboswitch

T terminator

Fig. 1. Schematics of the unified riboswitch template. The bar represents a PCR fragment containing T7A1 promoter (~40 bp), the initial transcribed sequence (ITC; 20 bp), the sensor domain of the riboswitch under study; and the intrinsic terminator (solid arrows indicate inverted repeat of the termination hairpin followed by the U-stretch). The terminator is designed so that the left shoulder of the hairpin stem is complementary to potential antiterminator sequence of the sensor domain (open arrows).

Transcriptional Approaches to Riboswitch Studies

43

hairpin sequence should be designed to match the structure of the sensor domain of a riboswitch. The 3¢ proximal part of the terminator hairpin stem is variable and should be complementary to the antiterminator part of the expression platform. It is followed by a stretch of 8–9 thymidines (T-stretch) and GC-rich “front element” of the terminator. The latter part improves the termination efficiency (12). The following procedure exemplifies the walking technique for the T7 A1 promoter template with the initial transcribed sequence: ATCGAGAGGG10 ACACGGCGAA20 TAGCCATCCC30 AAT33. T7A1 promoter containing linear DNA templates can be easily produced by PCR and purified in a relatively large concentration (e.g., 1 pmol/μL). At least twofold molar excess of DNA over RNAP is recommended to achieve the maximum yield of the initial EC. 1. Prepare an amplification reaction (50 μL) that includes 41.5 μL water, 5 μL 10× Phusion reaction buffer, 1 μL dNTPs mixture, 1 μL of each forward and reverse oligonucleotides, and 0.5 μL Phusion DNA polymerase. 2. Run the PCR reaction as follows: 95°C 45 s, 55°C 1 min, 72°C 45 s (29 cycles). Master mixture is usually prepared for twenty PCR reactions and then distributed in 50-μL aliquots into 250 μL thin-wall plastic tubes. 3. Resolve the amplification products in a 1.2% (w/v) agarose gel, cut the desired fragment, extract it by electroelution, followed by phenol–chloroform treatment and precipitation with ethanol (see Note 1). 4. Dilute the DNA pellet in 50 μL TE. The final concentration of DNA is expected to be ~2 pmol/μL as determined by UV spectrometry. 3.2. Preparation of the RNA Polymerase

The following procedure is optimized for E. coli RL 721 strain (10), a KmR derivative of JC7623 strain (13). This bacterial strain carries a chromosomal copy of the rpoC gene encoding a β¢ sub unit of RNA polymerase with the His-tag at its COOH terminus. The use of this strain guarantees that all purified RNAP would be His-tagged, although essentially the same procedure could be applied to conventional E. coli and B. subtilis strains. The purification method described here yields ~1 mg of RNA polymerase σ70 (RNAPσ70) holoenzyme per 10 g of cells. The enzymatic activity of the holoenzyme is ~2- to 3-fold higher than that of a commercially available preparation of E. coli RNAP. 1. Dilute an overnight culture (50 mL) of RL 721 strain in 4 L of LBD broth containing 50 μg/mL kanamycine. 2. Grow cells with aeration to OD600~1.0 at 37°C (usually 3–4 h). 3. Harvest cells by centrifugation at 3,000 × g for 7 min. The cell pellet can be stored at −70°C.

44

Mironov, Epshtein, and Nudler

4. Resuspend the cell pellet in 100 mL of lysis buffer and incubate on ice for 20 min. Add sodium deoxycholate to the final concentration of ~0.2% (v/v) constantly mixing the cell suspension. 5. Sonicate 20-mL aliquots of cells at the maximum output power by 30 s pulses interrupted by 30 s pauses (5–7 rounds). Completion of cell lysis can be monitored by reduction of viscosity in the mixture. 6. Clarify the lysate by centrifugation at 10,000 × g for 30 min. 7. Slowly add polymine P to the supernatant to the final concentration of ~0.35% (v/v) constantly stirring the mixture. Continue stirring for another 10 min. Centrifuge the suspension at 8,000 × g for 5 min. 8. Resuspend the pellet thoroughly by a glass homogenizer in 100 mL of TGED300 buffer. Centrifuge the suspension at 8,000 × g for 5 min. Discard the supernatant. 9. Resuspend the pellet in 100 mL of TGED500 buffer, homogenize, and centrifuge as described earlier. Discard the supernatant. 10. To extract RNAP from polymine P, resuspend the pellet in 100 mL of TGED1000 buffer for 1 h. Clarify the suspension at 8,000 × g for 5 min. 11. Slowly add solid NH2SO4 to the supernatant to 65% (39.8 g /100 mL) and stir for 30 min. The precipitate can be left overnight without loss of RNAP activity. Centrifuge the suspension at 10,000 × g for 45 min. Dissolve the pellet in 50 mL of TGED buffer and centrifuge again to remove the insoluble material. The final protein concentration should not exceed 5 mg/mL as determined by UV spectrometry using an extinction coefficient of 6.2 mL/cm/mg at 280 nm. If protein concentration is too high, sedimentation may be compromised. In this case, dilute the solution with water until it becomes colorless. 12. Load the protein solution onto 10 mL Heparin-Sepharose column pre-equilibrated with TGED50 buffer. Loading proceeds in 15–20 mL portions using 50 mL superloop at 0.5 mL/min. 13. Wash the column with 20 mL of the same buffer. Change the flow rate to 1 mL/min and wash the column with 20 mL TGED300 buffer. Elute RNAP from the column with 20 mL TGED600 buffer. Fractionate the protein into 1-mL aliquots. 14. Concentrate the eluate from several protein-containing fractions using Centricon-50 filter cartridge at 1,000 × g up to 0.5–1 mL so that the protein concentration would not exceed 10 mg/mL.

Transcriptional Approaches to Riboswitch Studies

45

15. Load concentrated RNAP sample onto a Superose-6 10/30 HR column in 0.5 mL portions (usually twice) at 0.3 mL/ min in TGED200 buffer. 16. Carry out elution at the same rate in 1 mL fractions for 90 min. RNAP is eluted as a separate peak after ~35 min. Combine protein-containing fractions, concentrate them up to 2 mL by Centricon-50, and dilute to 10 mL by TGED50. 17. Load RNAP sample onto Mono Q 5/5 HR column in TGED50 buffer using the 50 mL superloop at 0.2 mL/min flow rate. After washing the column with 5 mL of the same buffer, apply the 0.05–1 M NaCl gradient as follows: 0–20% – 2 mL, 21–80% – 60 mL, 81–100% – 2 mL, and collect 1 mL fractions. The RNAP core enzyme and the holoenzyme are eluted as separate peaks between 30 and 45% of salt. 18. Concentrate the fractions containing RNAP holoenzyme with Centricon-50, mix with equal volume of cold glycerol, and store the protein at −20°C. 3.3. Transcription in Solid Phase (Walking)

3.3.1. Preparation of the Initial EC and Walking

The principle of the walking reaction is that the initial EC immobilized onto a solid support undergoes rounds of washing to remove the unincorporated NTP substrates, followed by addition of the incomplete set of NTPs that allow transcription to proceed to the next DNA position corresponding to the first missing NTP. Each component of the EC (RNAP, DNA, or RNA) can be tagged for walking. Biotin tag can be put at the 5¢ terminus of the DNA strand (see later). For the His-tag, two types of metalchelating beads are available: Ni2+-NTA-agarose and TALON Co2+-Sepharose beads. The latter resin has less nonspecific binding and requires less time and concentration of imidazole to elute the EC off the beads, if necessary. For the biotin tag, the best resin for walking is NeutrAvidine UltraLink from Pierce. A representative walking protocol, which is described here for the His-tagged β¢ subunit of RNA polymerase, can be utilized with minimal changes for other tags (see Note 2). 1. To prepare a standard reaction (for 10–15 samples) incubate 1 pmol (~0.5 μL) of His-tagged RNAP and 2 pmol (~1 μL) of T7A1 promoter DNA fragment (~200 bp) in 8 μL of TB at 37°C for 5 min. Add 1.5 μL of the starting mix and 5 μL of TB-pre-equilibrated Ni2+-NTA or TALON resin. Incubate the mixture at 37°C with slight agitation for 8 min. 2. Wash the beads four times with 1.5 mL TB and resuspend in 10 μL TB containing 0.3 μM [α-32P] CTP with or without ATP and GTP (25 μM). Incubate the reaction at room temperature for 8 min and perform two rounds of washing with “high salt” TB and two rounds of standard

46

Mironov, Epshtein, and Nudler

TB washing. Without ATP and GTP, the initial [α-32P] EC is stalled at position +11 (EC11); with ATP and GTP, the initial EC is stalled at position +20 (EC20). It is expected that the washed EC11 and EC20 count 1,000–2,000 and 3,000–5,000 cpm, respectively, if the bottom of the tube is placed next to the membrane of the Geiger counter. 3. Add the appropriate composition of NTP (1 μL) to the washed initial EC (~20 μL) and incubate for 3–5 min. Wash with TB as described earlier. Repeat this procedure as many times as needed to move the EC to a desired position (Fig. 2). All walking steps are usually done at room temperature. The concentration of NTPs for each step has to be selected empirically to minimize the read-through as well as incomplete chase (see Note 3). If purified NTPs are used, 20 μM of each NTP is generally recommended. If nonpurified NTPs are used, 3–5 μM NTP and longer elongation time (~5–8 min) are more appropriate. At the last walking step, TB can be changed for any other buffer that suits the purpose of the experiment. 3.3.2. Roadblocking

Walking the EC far from the promoter (e.g., over 200 bp) may be laborious and time consuming. The yield of the final EC in this case can be decreased substantially due to partial loss of material (beads) during multiple washing steps and also due to partial loss of activity at certain (arrest) positions. To avoid these complications, a transcriptional roadblock can be used. The roadblock is a site-specific DNA-binding protein that is able to stop the EC completely without termination at any distance from the

EC

20 23 34 46 53 56 59

68 T

Fig. 2. RNAP walking. The autoradiogram shows [32P]-labeled RNA transcripts isolated from the indicated ECs. The panel shows walking using His6-tagged RNAP. “T” corresponds to the termination point (position +68) of the λ tR2 terminator. The two lanes of EC68 show RNA recovery before and after washing the beads with TB.

Transcriptional Approaches to Riboswitch Studies

47

promoter. The roadblock, which worked well in our hands, is the mutant form of EcoRI restriction endonuclease, EcoRQ111 (14). EcoRQ111 can be removed from DNA without interfering with the EC by the high salt buffers (e.g., 500 mM KCl). The use of EcoRQ111 has an advantage over other site-specific DNAbinding proteins in that it requires small changes in DNA for generating its binding site. 1. Engineer EcoRI site by PCR a few base pairs upstream of the site of interest. 2. Dilute an aliquot of EcoRQ111 in TB to prepare 10× stock (~0.02 mg/mL), which can be stored at +4°C for a few days. Add 1 μL of EcoRQ111 from 10× stock to the initial EC for 2 min, followed by addition of all four NTPs (0.5 mM). The working concentration of EcoRQ111 is about 8 pmol per 1 pmol of the EcoRI site. The chase reaction proceeds for 1–3 min. More than 90% of the EC is expected to be blocked by EcoRQ111 under these conditions. 3.4. Applications for Riboswitch Studies 3.4.1. Single-Round (Anti) Termination Assay

The following assay is designed for a quick assessment of the intrinsic termination efficiency as a function of riboswitch activity (Fig. 3, last two lanes). 1. Prepare the initial [α-32P]-labeled EC20 as described in Subheading 3.3.1, steps 1–2. All following reactions are performed at room temperature. 2. Add rifampicin to the final concentration 10 μM to prevent reinitiation by RNAP. 3. Distribute EC20 in 10-μL aliquots in ten Eppendorf tubes. Aliquots can be stored at +4°C for 12 h. 4. Add small molecular ligands (usually 1 μL) followed by the chase reaction induced by adding 1 μL of 10× NTP mix for 5 min. 5. Stop the reaction by 2 volumes of the hot loading mix. Heat the samples at 100°C for 30 s and immediately load on hot 10% denaturing sequencing gel that had been pre-run for at least 10 min. Freeze the rest of the samples and store at −70°C. 6. To calculate the efficiency of termination (%T), divide the amount of radioactivity in a termination band by the total radioactivity present in the termination and all read-through bands. %T depends on monovalent and divalent salt concentrations, NTP concentration, and temperature. Variation of the ligand concentration and amount of NTPs added to the washed EC20 allow estimation of the riboswitch termination efficiency as a function of the elongation rate. To study the process in more detail, the assay can be modified to include walking and roadblocking steps. Once the initial immobilized

48

Mironov, Epshtein, and Nudler

Fig. 3. A solid-phase transcription assay to probe riboswitch functioning. A representative radiogram shows RNA products after a single-round transcription reaction performed in solid phase using T7A1-thi-box riboswitch template (last two lanes). The initial EC20 carrying [32P]-labeled transcript was chased through the terminator in the presence or absence of 10 μM TPP. The first four lanes demonstrate walking reaction using the same template. EC20 was walked four steps to reach the thi-box domain.

EC is prepared as described earlier, it can be walked to any desired position along the riboswitch sequence. For example, to reach the sensor domain of the TPP-sensing riboswitch (thi-box) (16) at least four walking steps are needed (Fig. 3). Alternatively, the EcoRI site can be engineered to halt the EC at any distance within the riboswitch sequence. Using such walking techniques one can study the effect of cotranscriptional RNA folding on riboswitch functioning and also probe the RNA structure directly during elongation. 3.4.2. Riboswitch Structure Probing

Many types of chemical and enzymatic RNA probing can be performed without releasing the EC from the solid support. The following methods of RNA probing have been used to monitor riboswitch conformational changes during transcription in response to ligand binding. Major conformational changes in RNA can be readily detected by ribonuclease H (RNase H) following the annealing of short (8–10 nt) antisense DNA oligos to the different parts of the

Transcriptional Approaches to Riboswitch Studies

49

riboswitch nascent transcript (16, 17). RNase H specifically recognizes RNA:DNA heteroduplexes and works in a standard TB. 1. Add antisense oligonucleotide at 50-fold molar excess over the DNA template. 2. Add 0.5 U of RNase H per 10 μL reaction. Incubate for 5 min at room temperature. Stop the reaction by the hot loading mix and analyze the products on the sequencing gel. Walking can be used for generating specifically modified RNA transcripts for structural analysis (12). Many different NTP analogs can be utilized by E. coli RNAP and incorporated into the nascent RNA at specific positions during walking, for example, within the sensor domain of the riboswitch. The crosslinkable NTP analogs (such as 4-thio-UTP and 6-thio-GTP) and fluorescent analogs (e.g., 2-aminopurinetriphosphate (2AP); 5-(fur-2-yl) UTP; or pyrollo-C) are particularly useful to probe the local RNA conformational changes and interactions directly in the elongation complexes. For example, stacking interactions with the neighboring bases quench the 2AP fluorescence. Many NTP analogs and conditions for their incorporation are available from our laboratory upon request. 3.4.3. Spectroscopic Detection of the RNA– Ligand Interaction

Direct RNA binding of a small molecule, which is intrinsically fluorescent (e.g., FMN) or carries a fluorescent tag, can be detected in real time during transcription by scanning spectrofluorometry (16). The change of fluorescence during steady-state transcription is monitored on a Perkin-Elmer LS50B scanning fluorometer equipped with the quartz “submicrocuvette” with the 10-mm pathlength (see Note 4). For example, the formation of the riboswitch– FMN complex quenches FMN fluorescence due to photoinduced electron transfer from FMN to the aromatic rings of RNA bases. 1. To measure FMN binding, prepare the start-up transcription complex in 25 μL as described in Subheading 3.3.1, steps 1–2, except that 20 μM CTP is used instead of [32P] CTP. Add FMN to 4 μM. 2. Transfer the mixture to the submicrocuvette and take the initial control spectrum (λ = 519em/450ex nm, 5-nm slit widths). Record spectra at various time intervals after addition of NTPs to 100 μM.

4. Notes 1. Phenol and high temperature (>70°C) treatments affect the quality of DNA templates. Avoid phenol extraction or keep phenol treatment as brief as possible. To obtain DNA for

50

Mironov, Epshtein, and Nudler

biotin-tag-based walking, use for PCR the PAGE-purified forward DNA oligonucleotide carrying the 5¢-biotin. 2. Perform all reactions in siliconized Eppendorf plastic tubes. The standard round of washing, i.e., pelleting and resuspending the beads, includes 3–5 s centrifugation in a table-top microcentrifuge, removal of supernatant leaving ~50 μL above the pellet, and resuspension in 1.5 mL of the appropriate transcription buffer. 3. Commercially available NTPs are not pure enough to support walking on every DNA sequences. A read-through of certain positions due to small contamination in the NTP stocks is common. To avoid this problem, it is strongly recommended to purify original NTP stocks. The purification procedure has been described in detail (9). 4. The method detects the effect of various conditions (e.g., the rate of elongation, salt concentration, etc.) on ligand binding during transcription. Spectroscopic detection can be coupled with the fast kinetic and walking techniques.

Acknowledgment This work was supported by the NIH grants R01 GM58750 and GM72814 (E.N.) References 1. Nudler, E., and Mironov, A.S. (2004). The riboswitch control of bacterial metabolism. Trends Biochem. Sci. 29, 11–17. 2. Tucker, B.J., and Breaker, R.R. (2005). Riboswitches as versatile gene control elements. Curr. Opin. Struct. Biol. 15, 342–348. 3. Grundy, F.J., and Henkin, T.M. (2006). From ribosome to riboswitch: control of gene expression in bacteria by RNA structural rearrangements. Crit. Rev. Biochem. Mol. Biol. 41, 329–338. 4. Coppins, R.L., Hall, K.B., and Groisman, E.A. (2007). The intricate world of riboswitches. Curr. Opin. Microbiol. 10, 176–181. 5. Serganov, A., and Patel, D.J. (2007). Ribozymes, riboswitches and beyond: regulation of gene expression without proteins. Nat. Rev. Genet. 8, 776–790. 6. Kubodera, T., Watanabe, M., Yoshiuchi, K., Yamashita, N., Nishimura, A., Nakai, S., Gomi, K., and Hanamoto, H. (2003). Thiamineregulated gene expression of Aspergillus oryzae thiA requires splicing of the intron containing

7.

8.

9.

10.

11.

a riboswitch-like domain in the 5¢-UTR. FEBS Lett 555, 516–520. Cheah, M.T., Wachter, A., Sudarsan, N., and Breaker, R.R. (2007). Control of alternative RNA splicing and gene expression by eukaryotic riboswitches. Nature 447, 497–500. Bocobza, S., Adato, A., Mandel, T., Shapira, M., Nudler, E., and Aharoni, A. (2007). Riboswitch-dependent gene regulation and its evolution in the plant kingdom. Genes Dev. 21, 2874–2879. Nudler, E. , Gusarov, I. , and Bar-Nahum , G. (2003). Methods of walking with the RNA polymerase . Methods Enzymol. 371, 160 – 169 . Anthony, L.C., Artsimovitch, I., Svetlov, V., Landick, R., and Burgess, R.R. (2000). Rapid purification of His(6)-tagged Bacillus subtilis core RNA polymerase. Protein Expr. Purif. 19, 350–354. Wright, D.J., King, K., and Modrich, P. (1989). The negative charge of Glu-111 is required to

Transcriptional Approaches to Riboswitch Studies activate the cleavage center of EcoRI endonuclease. J. Biol. Chem. 264, 11816–11821. 12. Epshtein, V., Cardinale, C., Ruckenstein, A.E., Borukhov, S., and Nudler, E. (2007). Allosteric path to transcription termination. Mol. Cell. 28, 991–1001. 13. Kushner, S.R., Nagaishi, H., Templin, A., and Clark, A.J. (1971). Genetic recombination in Escherichia coli: the role of exonuclease I. Proc. Natl Acad. Sci. U. S. A. 68, 824–827. 14. Epshtein, V., Toulmé, F., Rahmouni, A.R., Borukhov, S., and Nudler, E. (2003). Transcription through the roadblocks: the role of RNA polymerase cooperation. EMBO J. 22, 4719–4727.

51

15. Cheng, S.C., Kim, R., King, K., Kim, S.H., and Modrich, P. (1984). Isolation of gram quantities of EcoRI restriction and modification enzymes from an overproducing strain. J. Biol. Chem. 259, 11571–11575. 16. Mironov, A.S., Gusarov, I., Rafikov, R., Errais-Lopez, L., Shatalin, K., Kreneva, R.A., Perumov, D.A., and Nudler, E. (2002). Sensing small molecules by nascent RNA: a mechanism to control transcription in bacteria. Cell 111, 747–756. 17. Epshtein, V., Mironov, A.S., and Nudler, E. (2003). The riboswitch-mediated control of sulfur metabolism in bacteria. Proc. Natl Acad. Sci. U. S. A. 100, 5052–5026.

Chapter 5 Kinetics of Riboswitch Regulation Studied By In Vitro Transcription J. Kenneth Wickiser Summary Riboswitches encompass messenger ribonucleic acid transcripts that sense the concentration of small molecule metabolites through binding the target compound and then control the expression of metaboliterelated genes in response to the metabolite concentration. While much of the riboswitch-related research has focused on the remarkable ability of different aptamer domains to adopt the intricate structures required to bind a spectrum of biological metabolites with high affinity and specificity, less attention has been paid to the mechanism of riboswitch action. Specifically, the genetic control element of the riboswitch, known as the expression platform, must function cotranscriptionally in the case of transcription termination-controlled riboswitches. By correlating the transcriptional kinetics of the entire switch and the kinetics and thermodynamics of metabolite binding of the aptamer domain, it was found that the FMN-binding riboswitch in the 5¢ UTR of the Bacillus subtilis ribGBAHT operon functions as a kinetically controlled genetic switch chiefly dependent upon transcriptional pausing and the concentration of the target metabolite. This study has emphasized the importance of studying the switch in its entirety and in the context of an actively transcribing RNA polymerase. Herein I will describe the study of the kinetics of riboswitch transcription and the proposed mechanism for the transcription termination-associated riboswitch control of riboflavin-related genes. Key words: Riboswitch, Transcription, Kinetics, Thermodynamics, RNA Polymerase, Genetic control, Transcriptional pausing, Electrophoresis

1. Introduction Riboswitches have been found in all three kingdoms of life. Riboswitches respond to a variety of small molecule metabolites including adenosylcobalamin (Ado-Cbl), flavin mononucleotide (FMN), thiamine pyrophosphate (TPP), L-lysine, adenine, guanine, S-adenosylmethionine (SAM), glucosamine-6-phosphate, Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_5 © Humana Press, a part of Springer Science + Business Media, LLC 2009

53

54

Wickiser

pre-queuosine, and glycine (1–3). The mechanisms of genetic control involve transcription termination, the inhibition of translational initiation, transcript stability, and alternative splicing (4, 5). Several excellent studies have elucidated the structure

Fig. 1. Progression of RNAP along DNA template and kinetics of the transcription elongation. (a) Halted Complex where the RNAP is starved for the fourth nucleotide. (b) During elongation phase, the RNAP progresses to allow the aptamer domain to exit the polymerase channel. This marks the first instance when the riboswitch is competent for ligand binding. The gray oval indicates the upstream half of the antiterminator helix. (c) A representative gel showing transcription elongation in the presence of FMN. PA, PB, T, and FL stand for the Pause A, Pause B, Termination, and Full-Length bands, respectively. (d) A representative plot showing kinetics of Pause A from 4 to 120 s. The solid line represents a Gaussian fit of the discrete data. The values on the right refer to the numerical fit of the data. The width value is taken as the lifetime of the RNAP stalled at the Pause Site A.

Kinetics of Riboswitch Regulation Studied By In Vitro Transcription

55

and biochemical properties of the aptamer domain of the known riboswitches (6–9), but missing was a study of the dynamic interplay between the aptamer and expression platform domains and the dependence of structure formation on the kinetics of transcription itself. To elucidate the mechanism of a riboswitch in its natural context, we chose the FMN-sensing riboswitch in the 5¢ UTR of the ribG gene in Bacillus subtilis that controls gene expression by transcription termination (10). The kinetics of transcription termination in the presence and absence of FMN was assessed using the clear distinction between a fulllength transcript and a terminated product provided by gel electrophoresis. It was reasoned that the aptamer domain, upstream of the expression platform (Fig. 1b), would have to be free and clear of the polymerase exit channel so that the decision whether to terminate transcription could be made after enough sequence was transcribed to allow the formation of the aptamer, but not so much as to have passed the termination site. Given the relatively short sequence between the 3¢ end of the aptamer domain and the site of termination (11), it became clear that time was a major contributor to the riboswitch mechanism. The kinetics of transcription in prokaryotes has been the subject of many studies over the last 25 years. The focus of these studies has been to identify the proteinaceous transacting components of the transcriptional machinery as well as to investigate signals in the DNA or RNA that communicate to the polymerase the need to initiate, pause, or terminate the transcription reaction (12–15). The early examples of transcriptional pausing investigated by Chamberlin, Yanofsky, Platt, and others studied the kinetics of transcription of the bacteriophage T7 and the interplay between the ribosome, amino acid concentration, and transcriptional kinetics in the trp operon. Their main tool was pulse-labeled, synchronized RNA transcription using radionucleotides to monitor the progress of the transcription (16). Additionally, the Crother laboratory and others used the kinetics of transcription to investigate DNA–drug interactions in vitro (17). We chose to investigate pausing during transcription of the riboswitch RNA to identify the temporal context required for the aptamer domain to bind the target metabolite.

2. Materials 2.1. Polyacrylamide Gel Electrophoresis

All solutions should be prepared with high-quality water, filtered and autoclaved, if possible.

56

Wickiser

1. N,N,N¢,N¢-Tetramethylethylenediamine (TEMED) (Sigma). 2. Ammonium persulfate (APS): 10% g/mL, protect from light and store at 4°C for up to several months. 3. 10 × Tris–Borate–EDTA (TBE) buffer: 108 g Tris base, 55 g boric acid, and 3.725 g ethylenediaminetetraacetic acid (EDTA) in 1 L water. 4. Urea (Sigma). 5. Acrylamide gel solution. 6% acrylamide:bis-acrylamide 19:1 (w/w) mix, 8 M urea, 1× TBE; store at room temperature and protect from light. 2.2. Synchronized Transcription Elongation

The purpose of the Synchronized Transcription assay is to ensure a uniform start of transcript elongation for the kinetic analysis and to ensure uniform labeling of the transcripts regardless of the RNA length. 1. E. coli RNA Polymerase Holoenzyme (EcRNAP) (Epicentre Biotechnologies, Madison, WI). 2. 50 mM individual ribonucleotide triphosphates (NTPs) (Sigma): brought to neutral pH using NaOH (Sigma) and stored at −20°C. 3. Transcription buffer: 20 mM Tris–HCl, pH 8.0, 20 mM NaCl, 14 mM MgCl2, 14 mM ß-mercaptoethanol, and 0.1 mM EDTA. 4. The initiation buffer: transcription buffer supplemented with 80 µg/mL bovine serum albumin (BSA) (Sigma), 170 µM dinucleotide ApA (ApA) (BioLog Life Sciences Institute, Germany), 3.75% glycerol (Sigma), 8.8 pmol of EcRNAP, 2.5 µM each of GTP and UTP, 1 µM ATP, [α-32P]- ATP, or UTP (ATP* or UTP*) (Amersham). ApA is not stable and should be kept in small aliquots at −20°C and should remain on ice prior to use. Low concentration of standard nucleotides in this buffer increases radioactive signal of the leader region of the nascent transcript. 5. The elongation buffer: the initiation buffer omitting radionucleotides (ATP* or UTP*) and ApA, and supplemented with heparin (Sigma) at 1 mg/mL and NTPs at a concentration (i.e. 125 µM each nucleotide) specific for the experiment (see Note 1). Store the buffer at −20°C. 6. Sequencing buffer: the elongation buffer with three of the four NTPs at 2–10 mM (final concentration needs optimization for each template), whereas the fourth NTP is a mixture of 3¢-O-Me-NTP and NTP in a 4:1 ratio at lower concentration. For a U-ladder we typically used the buffer containing 2 mM ATP, 2 mM GTP, 2 mM CTP, 400 µM 3¢-O-Methyl(Me)-UTP, and 100 µM UTP. For templates shorter than 150 nucleotides

Kinetics of Riboswitch Regulation Studied By In Vitro Transcription

57

a 3¢-O-Me-NTP:NTP ratio can be as high as 9:1, whereas longer templates require a lower ratio to ensure longer elongation prior reaction termination (see Note 2). 7. Stop buffer (2×): 7 M urea, 80 mM EDTA, pH 8.0, and trace amounts of bromophenol blue (BPB) (Sigma) and xylene cyanol (XC) (Sigma). Preheat stop buffer at 37°C prior to use. 8. Flavin mononucleotide (FMN) (Sigma): prepared at 50 µM, stored at −20°C in small aliquots, and protected from light using aluminum foil. It is highly recommended to check the concentration of FMN by UV absorbance using an extinction coefficient of 12,200 M−1 cm−1 at 450 nm or as designated by the supplier’s certificate of analysis. 9. Template DNA. The wild type ribG sequence is as follows: 5 ¢ -AAGGAcAAATGAATAAAGATTGTATCTTCGGGGCAGGGTGGAAATCCCGACCGGCGGTAGTAAAGCAC ATTTGCTTTAGAGCCCGTGACCCGTGTGCATAAGC ACGCGGTGGATTCAGTTTAAGCTGAAGCCGACAGTGAAAGTCTGGATGGGAGAAGGATGATGAGCCGCTATGCAAAATGTTTAAAAATGCATAGTGTTATTTCCTATTGCGTAAAATACCTAAAGCCCCGAATTTTTTATAAATTCGGGGCTTTTTT-3¢ To generate the synchronized transcription template, mutate the sixth base from a cytosine (lowercase to assist identification) to a thymidine (C6T) using an appropriate PCR primer so that the RNA polymerase starved for CTP would halt at position 26 (C26). 2.3. Transcript Visualization

1. Phosphorimager. 2. Phosphor screen and plate (Kodak) (see Note 3). 3. ImagQuant software (Molecular Dynamics).

3. Methods 3.1. Polyacrylamide Gel Preparation

1. Clean two glass plates with methanol or ethanol, place 0.75-mm-thick spacers between plates and fasten the plates together with large binding clips. The clipped plates are then laid on a small box (e.g., an empty pipet tip box) parallel to the bench top. 2. Pour the appropriate amount of premixed acrylamide solution into a small beaker (see Note 4). 3. Add 40 µL TEMED for each 100 mL of acrylamide solution. Mix with pipet tip.

58

Wickiser

Fig. 2. A representative polyacrylamide gel showing separated transcription products. A time course experiment was conducted using a DNA template corresponding to a full-length 304-nt riboswitch. G, A, U, and C lanes are 20-min sequencing reactions. To validate the location of the pause and termination sites, templates corresponding to runoff transcripts of length 210, 215, 230, and 244 nucleotides were transcribed as markers.

4. Add 400 µL APS for each 100 mL of acrylamide solution. Mix with pipet. 5. Using either the beaker or a 50-mL plastic pipet, pour the solution slowly, but consistently (to avoid breaks in the flow and bubbles), into the space between the two plates. Gentle knocking on the glass plates with the knuckles of the other hand will help to distribute the solution and avoid bubbles.

Kinetics of Riboswitch Regulation Studied By In Vitro Transcription

59

6. Gently insert the comb. To avoid bubbles at the tips of the comb, ensure that the tips are submerged in acrylamide before inserting the comb into the notch. Should any bubbles form, gently pull the comb out and pour additional acrylamide onto the notch if needed. 7. Allow the gel to polymerize for ~30 min. Polymerization time depends on the temperature and may be extended, if needed. 8. After polymerization is complete, remove the comb and flush the wells with deionized water. 9. Insert the plates into the gel rig, add the running buffer (1× TBE), and preheat the gel for ~15 min. 3.2. Synchronized Transcription Elongation

1. Place all components of the transcription initiation and elongation reactions on ice. Prepare initiation reaction mixture (8 µL per reaction) with template DNA (14 pmol per reaction). Mix solution with a pipet or by gently flicking the test tube. Briefly spin the tube using a small benchtop microfuge. Incubate the reaction in the heating block at 37°C for 10 min. 2. Place the initiation reaction on ice. This reaction solution is now the Halted Complex Mixture (Fig. 1a and Fig. 2, lane 0 ). The Halted Complex can be stored on ice for at least 1 h. 3. Draw 8 µL of the Halted Complex Mixture and add it into an empty test tube preheated in the heating block at 37°C. Keep the tube cap open. 4. Add 2 µL of the Elongation Mixture into an empty tube preheated in the heating block at 37°C. Keep the tube cap open. To avoid evaporation, do not leave the tube open longer than 1 min. Keep the same heating time throughout the experiment. 5. Draw the warm Halted Complex Mixture and add to the warm Elongation Mixture. Mix by pipetting up and down 2–3 times. Start the timer. 6. If incubation time allows, close the lid of the test tube and leave it on the heating block. Avoid splashing the reaction mixture inside the tube. 7. Prior to the time point, open the test tube cap and at the appropriate time, add 10 µL of 2× Stop Buffer (see Note 5). Pipet up and down several times. Recap the tube and place the tube on ice. If time permits, vortex and centrifuge the reaction tube. This reaction state is presented in Fig. 1b, Fig. 2 (lanes 0″–5¢), and Fig. 3a–d. 8. RNA Sequencing was performed using the same synchronized, single turnover transcription described in steps 1–7 of this section. However, a specific chain-terminating 3¢-O-Meribonucleotide complemented the four natural ribonucleotides and a radiolabeled nucleotide (ATP* or UTP*) were

60

Wickiser

Fig. 3. A model of the FMN riboswitch transcription termination. (a) The RNAP has progressed into the expression platform and has paused at the site A. (b) The RNAP continues and then pauses at the site B. The exit channel likely obscures the downstream half of the antiterminator helix during the pause B. (c) With the FMN-bound aptamer, the terminator structure is formed thus signaling to the RNAP to halt and dissociate from the template. (d) If FMN is not bound to the aptamer domain while the RNAP is at the termination site, the upstream and downstream halves of the antiterminator helix pair and prevent formation of the transcription terminator thereby leading to transcription of entire riboswitch.

Kinetics of Riboswitch Regulation Studied By In Vitro Transcription

61

used as described in Subheading 2.2, steps 1–3, 5, and 6. Further, the elongation reaction was quenched after proceeding for 20 min. 3.3. Polyacrylamide Gel Electrophoresis and Transcription Analysis

1. Load ~10 µL of the stopped reaction mixture onto preheated gel with a long gel-loading tip. During loading, the tip must be as close to the bottom of the well as possible, and remain straight down to eliminate mixing of the sample with the running buffer. 2. Run the gel until the desired products are separated (~1.5–2 h depending on power settings). 3. Turn off the power to the gel rig, dismantle the rig, and put the glass plates on top of a box (e.g., empty pipet box). 4. Remove the spacers from the gel plates with a spatula. Split the plates apart using a spatula or a thin spacer. 5. Ensure that the plate to which the gel is sticking is on the bottom and gently lift the other plate up taking care not to rip the gel. 6. Place a piece of Whatman paper, slightly larger than the gel itself, on top of the gel starting at one corner and rolling to an opposite corner. Avoid air pockets between the gel and the paper. 7. Turn the plate over so that the glass is positioned up. 8. Slowly drag the plate toward off the edge of the bench letting the paper and the gel to hang down. 9. Continue separation of the gel from the glass by moving the paper/gel/glass sandwich further from the edge of the bench. 10. Just prior to the end of separation, shift the sandwich back to the bench top so that when the separation is complete, the paper-gel portion rests on the flat surface. 11. Cover the gel with plastic wrap (e.g., Saran Wrap). 12. Trim the plastic wrap and paper so that it is only slightly larger than the gel itself. Trim one corner to mark the orientation of the gel. 13. Place the plastic wrap-gel-paper sandwich on an additional and slightly larger piece of Whatman paper. 14. Place the sandwich in a gel dryer with the plastic wrap upward. Dry the gel. 15. Expose the gel to a phosphorimager screen for ~3–12 h depending on the activity of the radiolabel. 16. Scan the gel using the phosphorimager (see Figs. 1c and 2).

62

Wickiser

17. Quantify the bands using ImageQuant software. Draw boxes around the bands of interest. Ensure that the same area is used for each band. Subtract the background signal using a box of the same size. If the background signal changes from lane to lane, then lane-specific background correction should be done. 18. Plot the reaction kinetics for a particular band corresponding to a transcriptional pause, a terminated product, or the fulllength product as in Fig. 1d. 19. To estimate the lifetime of a pause, measure the time window corresponding to the half maximum value of the maximal occupancy of that site. This value is termed the full width at half maximum (FWHM) value (Fig. 1d) (see Note 6). 20. Once the lifetimes of the transcriptional pauses are known, they should be correlated with the kinetics of metabolite binding and the kinetics of structure formation to devise a working mechanistic model of the riboswitch action.

4. Notes 1. If proteins (such as NusA) in the glycerol-containing storage buffer are added to the reaction, make sure that the glycerol concentration in both protein-containing and protein-omitting reactions is the same. The glycerol concentration should never exceed 10%. 2. In order to obtain the bands of similar intensities in the sequencing and runoff reactions, the amount of radiolabeled NTP should be 3–4 times higher in the sequencing reaction than in the runoff reactions. 3. It is recommended to erase the screen prior to each use. In case of radioactive contamination, clean the screen using Kodak phosphor-screen cleaning solution. Do not use organic solvents such as ethanol for cleaning. 4. Caution: unpolymerized polyacrylamide is neurotoxic. 5. For short incubations, use additional pipet with a stop buffer in the tip. 6. The estimation of the pause lifetime using the FWHM value is crude because the actual distribution is nonGaussian. Yet this simple calculation yields a first-order approximation of the pause lifetime and is expected to be fairly accurate.

Kinetics of Riboswitch Regulation Studied By In Vitro Transcription

63

Acknowledgments The author would like to thank Profs. Donald M. Crothers and Ronald R. Breaker for their mentorship and support during the experiment phase of this work. In addition, the author would like to thank Dr. Ali Nahvi for his insightful comments during manuscript preparation. References 1. Roth, A., Winkler, W. C., Regulski, E. E., Lee, B. W., Lim, J., Jona, I., Barrick, J. E., Ritwik, A., Kim, J. N., Welz, R., Iwata-Reuyl, D., and Breaker, R. R. (2007). A riboswitch selective for the queuosine precursor preQ1 contains an unusually small aptamer domain. Nat. Struct. Mol. Biol. 14, 308–317. 2. Dann, C. E., III, Wakeman, C. A., Sieling, C. L., Baker, S. C., Irnov, I., and Winkler, W. C. (2007). Structure and mechanism of a metal-sensing regulatory RNA. Cell 130, 878–892. 3. Winkler, W. C., and Breaker, R. R. (2005). Regulation of bacterial gene expression by riboswitches. Annu. Rev. Microbiol. 59, 487–517. 4. Cheah, M. T., Wachter, A., Sudarsan, N., and Breaker, R. R. (2007). Control of alternative RNA splicing and gene expression by eukaryotic riboswitches. Nature 447, 497–500. 5. Winkler, W. C. (2005). Riboswitches and the role of noncoding RNAs in bacterial metabolic control. Curr. Opin. Chem. Biol. 9, 596–602. 6. Serganov, A., Polonskaia, A., Phan, A. T., Breaker, R. R., and Patel, D. J. (2006). Structural basis for gene regulation by a thiamine pyrophosphate-sensing riboswitch. Nature 441, 1167–1171. 7. Montange, R. K., and Batey, R. T. (2006). Structure of the S-adenosylmethionine riboswitch regulatory mRNA element. Nature 441, 1172–1175 8. Serganov, A., Yuan, Y. R., Pikovskaya, O., Polonskaia, A., Malinina, L., Phan, A. T., Hobartner, C., Micura, R., Breaker, R. R., and Patel, D. J. (2004). Structural basis for discriminative regulation of gene expression by adenine- and guanine-sensing mRNAs. Chem. Biol. 11, 1729–1741.

9. Batey, R. T., Gilbert, S. D., and Montange, R. K. (2004). Structure of a natural guanineresponsive riboswitch complexed with the metabolite hypoxanthine. Nature 432, 411–415. 10. Wickiser, J. K., Winkler, W. C., Breaker, R. R., and Crothers, D. M. (2005). The speed of RNA transcription and metabolite binding kinetics operate an FMN riboswitch. Mol. Cell 18, 49–60. 11. Winkler, W. C., Cohen-Chalamish, S., and Breaker, R. R. (2002). An mRNA structure that controls gene expression by binding FMN. Proc. Natl. Acad. Sci. U S A 99, 15908–15913. 12. Lau, L. F., Roberts, J. W., and Wu, R. (1983). RNA polymerase pausing and transcript release at the lambda tR1 terminator in vitro. J. Biol. Chem. 258, 9391–9397. 13. Farnham, P. J., Greenblatt, J., and Platt, T. (1982). Effects of NusA protein on transcription termination in the tryptophan operon of Escherichia coli. Cell 29, 945–951. 14. Winkler, M. E., and Yanofsky, C. (1981). Pausing of RNA polymerase during in vitro transcription of the tryptophan operon leader region. Biochemistry 20, 3738–3744. 15. Kassavetis, G. A., and Chamberlin, M. J. (1981). Pausing and termination of transcription within the early region of bacteriophage T7 DNA in vitro. J. Biol. Chem. 256, 2777–2786. 16. Landick, R., Wang, D., and Chan, C. L. (1996). Quantitative analysis of transcriptional pausing by Escherichia coli RNA polymerase: his leader pause site as paradigm. Methods Enzymol. 274, 334–353 17. Phillips, D. R., and Crothers, D. M. (1986). Kinetics and sequence specificity of drug– DNA interactions: an in vitro transcription assay. Biochemistry 25, 7355–7362.

Chapter 6 Molecular Basis of RNA-Mediated Gene Regulation on the Adenine Riboswitch by Single-Molecule Approaches Jean-François Lemay, J. Carlos Penedo, Jérôme Mulhbacher, and Daniel A. Lafontaine Summary The adenine-specific pbuE riboswitch undergoes metal ion-dependent folding that involves a long-range tertiary loop–loop interaction between two stem loops. Fluorescence resonance energy transfer (FRET) and single-molecule FRET studies demonstrate the ability of the loops to interact in the absence of the ligand. Although the riboswitch can fold in the absence of adenine, ligand binding stabilizes this folded conformation by increasing the folding and decreasing the unfolding rates of the riboswitch. The presence of the ligand also decreases the magnesium ion concentration required to promote the loop–loop interaction. Single-molecule FRET studies demonstrate that individual aptamer molecules exhibit great heterogeneity in the rates of folding and unfolding, which is reduced in the presence of adenine. Moreover, single-molecule FRET proposes that riboswitch folding proceeds through a complex landscape that involves a discrete intermediate. Key words: Riboswitch, Adenine, Fluorescence, RNA regulation, FRET, Single molecule

1. Introduction Riboswitches are genetic control elements found in the untranslated regions (UTRs) of certain messenger RNA in both prokaryotes and eukaryotes (1, 2). Riboswitches are composed of two modular domains: a ligand-binding aptamer domain and an expression platform. The most conserved region of riboswitches is the aptamer domain, a receptor that specifically binds a given metabolite (3). Expression platforms vary widely in sequence and structure, and they typically affect gene regulation by modulating

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_6 © Humana Press, a part of Springer Science + Business Media, LLC 2009

65

66

Lemay et al.

the formation of Rho-independent transcriptional terminators or the sequestration of the Shine-Dalgarno sequences required for translation initiation (4). The adenine riboswitch aptamer domain folds in a three-way junction (Fig. 1) in which the core region is reorganized upon ligand binding, the latter being important for the control of gene expression (5). The three-dimensional fold of the aptamer domain is characterized by a long-range tertiary loop–loop interaction that occurs between stem loops P2 and P3 (6, 7). The crystal structures of both the guanine- and adenine-specific aptamers are very similar, and the folds of the RNAs are very compact and are characterized by intricate tertiary interactions (6, 7). The ligands are buried deep within the structure where they make specific hydrogen bonds that provide a structural ground for the high specificity that each riboswitch exhibits for its cognate ligand (5, 8). Because riboswitches heavily rely on structural changes to regulate gene expression, it is important to seek alternative techniques that provide vital information about folding and dynamics of RNA molecules. Fluorescence Resonance Energy Transfer (FRET) is unique in its ability to provide accurate long-range distance information (20–80 Å) in solution under physiological conditions (9). Recently, FRET measurements at the single-molecule level (sm-FRET) have become possible, thus providing information on nucleic acid dynamics that was

Fig. 1. Secondary structure of the adenine riboswitch-binding domain with donor (fluorescein) and acceptor (Cy3) fluorophores attached to the loops at shaded positions. The RNA sequence is based on the VV1 aptamer variant (6). The ligation site is shown by an arrow.

Molecular Basis of RNA-Mediated Gene Regulation on the Adenine Riboswitch

67

previously hidden when using bulk assays (10). Using smFRET, we have recently showed that the loops of the adenine riboswitch aptamer interact in the presence of low-micromolar concentrations of magnesium ions (11). We have also uncovered the existence of an intermediate state in riboswitch folding. Interestingly, the addition of adenine promotes folding of the intermediate structure, suggesting that the ligand actively participates in the formation of the loop–loop interaction and in the folding process of the core of the riboswitch aptamer domain. Sm-FRET can be applied to any nucleic acid molecule and may prove useful in the identification of subpopulations in a heterogeneous mixture. Such detection is desirable for establishing the folding pathway and dynamics of the studied molecule.

2. Materials 2.1. Synthetic RNA

1. The aptamer RNA was obtained by the ligation of the following oligonucleotides: 5¢ strand, GCGCGAGCGUUGUAUAAUCCUAAUGAUAUGGUUU-GGGAGU; 3¢ strand, PO4-UUCUACCAAGAGCCUUAAACUCUUGAUUACAACGCUCCGC. RNA molecules were purchased from Dharmacon (Boulder, CO) with 5¢-amino-allyl uridines at defined positions (underlined nucleotides). 2. Biotin was incorporated at the 5¢ end of the 5¢ RNA strand for single-molecule experiments.

2.2. Fluorophores

1. Fluorescein succinimidyl ester derivative (Invitrogen, Carlsbad, CA): Oregon Green 488 carboxylic acid predissolved in DMSO at 18 µg/µL. 2. Indocarbocyanine-3 (Cy3) (Amersham Biosciences). 3. For single-molecule experiments, fluorescein is replaced by indocarbocyanine-5 (Cy5) (Amersham Biosciences) as an acceptor for Cy3 because of its higher photostability.

2.3. Preparation of Labeled-RNA

1. 2’-Deprotection buffer: 100 mM acetic acid, pH is adjusted to 3.8 with N,N,N,N¢-tetramethyl-ethylenediamine (TEMED) according to the Dharmacon protocol. 2. Labeling buffer: 70 mM tetraborate sodium (Sigma), pH 8.5, 14% (v/v) dimethyl sulfoxide (DMSO) (Sigma). 3. Annealing buffer: 10 mM HEPES–Na, pH 7.5, 50 mM NaCl.

68

Lemay et al.

4. T4 RNA ligase (New Englands Biolabs, Ipswich, MA). 5. Precipitation solutions: 3 M sodium acetate (NaOAc), pH 5.2, 100% ethanol. Keep these solutions at room temperature. Glycogen: 1 mg/mL. Store at −20°C. 2.4. Polyacrylamide Gel Electrophoresis

1. 5× TBE: 450 mM Tris base, 450 mM borate and 10 mM ethylenediamine tetraacetic acid disodium salt (EDTA). 2. 1× Dilution buffer: 7 M urea dissolved in 1× TBE (store at room temperature). 3. 20% acrylamide/bis-acrylamide solution (19:1) with 7 M urea in 1× TBE. This solution is neurotoxic when not polymerized. Keep this solution in the dark at room temperature. 4. Ammonium persulfate (APS): prepare 10% (w/v) solution in water. Keep this solution in the dark at 4°C (see Note 1). 5. Molecular weight markers: xylene cyanol 0.02% and bromophenol blue 0.02% dissolved in pure formamide. 6. Thin layer chromatography (TLC) plates, 20 × 20 cm (Selecto Scientific, Suwanee, GA).

2.5. RNA Elution

1. Electroeluter from Harvard lab shops (http://www.mcb. harvard.edu/bioshop). 2. 8 M ammonium acetate (NH4OAc). Keep the solution at room temperature.

2.6. Steady-State Fluorescence Measurement

1. Fluorescence buffer: 1× TBE. To remove trace amounts of metal ions, pour a Chelex 100 resin bed (Sigma) in a ~1 × 15 cm plastic column (Bio-Rad). Wash the resin with 500 mL of nanopure water. 2. Ligand: 1 mM adenine dissolved in 1 M HCl.

2.7. Single-Molecule Fluorescence

In this section, we assume that total-internal reflection (TIR) equipment for sm-FRET is already available and the reader is familiar with the principles of sm-TIR FRET. For an excellent review about building the single-molecule equipment, refer to (12). 1. Riboswitch buffer: 50 mM Tris–HCl, pH 8.1, 25 mM NaCl. 2. Biotinylated and nonbiotinylated bovine serum albumin (Sigma): prepare 10 mg/mL stock solutions in the riboswitch buffer. Keep at 4°C after preparation and use (see Note 2). 3. Streptavidin (Invitrogen, Carlsbad, CA): 5 mg/mL in the riboswitch buffer. The solution should be stored at 4°C immediately after preparation and use. 4. Base buffer used for imaging: 50 mM Tris–HCl, pH 8.1, 6% (w/w) glucose, 1% (v/v) 2-mercaptoethanol, 0.1 mg/mL glucose oxidase type II-S from Aspergillus niger (Sigma), 0.02 mg/ mL glucose catalase (Roche Diagnostics) (see Notes 1 and 4).

Molecular Basis of RNA-Mediated Gene Regulation on the Adenine Riboswitch

69

5. Quartz microscope slides of 2.5 cm × 7.5 cm × 0.1 cm (Finkenbeiner, Waltham) for prism-type sm-TIR. 6. Fluorescence beads (FluoSpheres carboxylated, 0.2 µm) for instrument alignment and for the generation of the mapping algorithm to correlate individual donor spots with individual acceptor spots (Invitrogen, Carlsbad, CA). Commercial stock solution should be diluted 1,000-fold in 5 mM HCl (see Note 5).

3. Methods To prevent RNA degradation, the preparation of the labeled RNA should be done in an RNase-free environment using RNase-free reagents and water. The ratio between the donor and the acceptor must be near 1 to obtain reliable FRET values. The buffer used for the steady-state FRET is devoid of ions in order to detect the true influence of the ligand and ions on the aptamer folding. 3.1. Preparation of Labeled Fluorescent Molecules

1. Dissolve the RNA sample in 100 µL water and precipitate RNA overnight at −20°C with 0.1 volume of 3 M NaOAc, 3 volumes of 100% ethanol, and 2 µL glycogen. Centrifuge the RNA solution at 13,000 × g for 30 min at 4°C. 2. Discard the supernatant and perform a quick spin to remove residual ethanol. Air dry the pellets for 5 min and resuspend them in water. Determine RNA concentration by UV spectroscopy. An optical density (OD) of 1.0 at 260 nm corresponds to an RNA concentration of ~40 µg/mL. 3. Mix 125 µg RNA with 250 µg succinimidyl ester fluorescent label. Bring the total reaction volume to 100 µL by the labeling buffer. Tumble gently on a shaker overnight at room temperature and precipitate with ethanol. 4. Resuspend RNA in 400 µL 2¢-deprotection buffer and heat for 30 min at 60°C for fluorescein labeling or 2 h for Cy3, Cy5 and biotinylated RNA. 5. Divide the reaction into two tubes and dry them in the SpeedVac. Resuspend the deprotected RNA in 50% formamide and purify as described in Subheadings 3.2 and 3.3.

3.2. Polyacrylamide Gel Electrophoresis

1. These instructions are prepared for standard gel running systems (40 cm × 20 cm) and 10% acrylamide gels, which are most often used for the purification of the RNA molecules of ~50 nucleotides.

70

Lemay et al.

2. Prepare a 1.5-mm-thick gel by mixing 30 mL 20% acrylamide/7 M urea solution with 30 mL 1× dilution buffer. Add 450 µL of 10% APS, followed by 45 µL of TEMED while constantly mixing the gel solution. Pour the gel, insert the comb, and allow the gel to polymerize for at least 1 h. 3. After polymerization, remove the comb, carefully wash the wells with TBE, and prerun the gel at 18 W for at least 30 min using 1× TBE as a running buffer. 4. Just before loading samples, rewash the loading wells with a syringe. Load the samples and run Polyacrylamide Gel Electrophoresis (PAGE) at 18 W. Xylene cyanol FF and bromophenol blue dyes, which migrate similarly to RNAs of 55 and 10 nucleotides, respectively, on 10% gels, can be used as size markers (see Note 6). 5. After migration, remove glass plates and visualize the RNA bands using the “UV shadow” technique on TLC plates (see Pikovskaya et al., this issue). Cut the bands using clean blades. Gel slices can either be stored in the freezer or used immediately for electroelution (see Note 7). Cut a band containing one of the dyes as a control for electroelution. 3.3. RNA Electroelution

1. Among two RNA extraction techniques, crush-and-soak and electroelution, the latter technique provides better recovery yields for preparative scale RNA extraction. 2. Fill the electroeluter apparatus with 0.25× TBE and run at 200 V for 30 min to clean it. Discard the running buffer and fill the apparatus with fresh 0.25× TBE buffer. Add 200 µL of 8 M ammonium acetate into the “trapping” wells, where RNA will be concentrated. 3. Carefully place the acrylamide slices into the large wells. As a control for the electroelution process, place the acrylamide slice containing dye in a separate well. The efficiency of recovery can be increased by cutting acrylamide slices in small pieces. Run the electroelution at 120 V for 1 h or until the dye is completely removed from the acrylamide. 4. Cut ~5 mm off of P200 tips and insert the tips into the trapping wells. Remove running buffer from the electroeluter using a syringe. Collect samples from each trapping well and transfer them into Eppendorf tubes. Rinse each well with 1×TBE to collect residual RNA molecules. Precipitate RNA with ethanol as described in Subheading 3.1, steps 1 and 2. 5. Resuspend each pellet in 50 mM Tris–HCl, pH 8.0. Calculate the labeling efficiency as described in Subheading 3.5.

3.4. RNA Ligation

1. Anneal purified RNA strands by heating a mixture (1:1 ratio) in annealing buffer at 70°C and slowly cooling it to room

Molecular Basis of RNA-Mediated Gene Regulation on the Adenine Riboswitch

71

temperature. Add 80 U of T4 RNA ligase and incubate at 37°C for 4 h. 2. Add one volume of 100% formamide to the mixture and load the samples on a polyacrylamide gel. A single band corresponding to ligated molecules should be observed on the gel. A transcript corresponding to the ligated product should be loaded on the same gel and used as a molecular marker. Cut the band corresponding to the ligated product, purify it as described in Subheadings 3.2 and 3.3, and calculate the labeling efficiency as described in Subheading 3.5. 3.5. Labeling Efficiency

1. To ensure that the ratio between the two fluorophores is near 1, it is necessary to record a spectrum of the labeled RNAs from 220 to 700 nm. 2. Take 10 pmol of the aptamer molecule in 490 µL of Fluorescence buffer. Record the spectra at 4°C and wait 3 min between each spectrum. 3. The molar extinction coefficient of the RNA strand can be calculated using the molecular extinction coefficient of one nucleotide (9,500 L/mole/cm). The molar extinction coefficients of the dyes are: fluorescein, 68,000 L/mole/cm; Cy3, 150,000 L/mole/cm; and Cy5, 250,000 L/mole/cm. The labeling efficiency of the RNA is determined by comparing the ratios between theoretical extinction coefficients of a dye and RNA with the measured absorbance of a dye and RNA.

3.6. Steady-State FRET Measurement

By monitoring the change in the FRET signal, steady-state FRET experiments can provide information about RNA conformations that result from a change in solution conditions (9). Typically, a nucleic acid is labeled with two fluorophores, a donor (fluorescein) and an acceptor (indocarbocyanine-3, Cy3), that are covalently attached at different locations (Fig. 1). The absorption of the donor occurs at a higher frequency than that of the acceptor, which leads to a transfer of excitation energy. FRET can be observed in a variety of ways, including an increased fluorescent emission from the acceptor, which is determined using the acceptor normalization method (9). Because the fluorescence emission at any given wavelength of a double-labeled sample excited at the donor wavelength contains emission from (1) the donor, (2) the directly excited acceptor, and (3) the acceptor excited by energy transfer from the donor, it is important to extract emissions which are not produced through FRET. Thus, the first part of the analysis requires the subtraction of the spectrum of the RNA labeled only with donor, leaving just the acceptor fluorescence components. Then, the derived acceptor spectrum is normalized to one from the sample excited at a wavelength at which only the acceptor is excited, which yields the normalized acceptor ratio.

72

Lemay et al.

This ratio can be directly converted to FRET by using the molar fraction of labeled molecules (9). 1. Using a double-labeled molecule, the attached fluorescein is excited at 490 nm, and an emission spectrum is collected from 500 to 650 nm at 1 s/nm. In addition, collect a donor-only fluorescence spectrum at the same settings using a donor-only labeled RNA molecule instead of a double-labeled RNA. By subtracting the donor-only fluorescence spectrum from the doublelabeled one, we obtain a spectrum of the acceptor components. 2. Using the double-labeled molecule, record an acceptor-only spectrum by using an excitation wavelength of 547 nm and by collecting an emission spectrum from 550 to 650 nm at 1 s/nm. 3. Values of EFRET are calculated by normalizing the spectrum containing the acceptor components to that of the acceptoronly (9). Using this methodology, it is possible to monitor the formation of the P2-P3 loop–loop interaction as a function of magnesium ions (11). The experimental data clearly show a conformational transition, which is revealed as an enhanced energy transfer upon increasing Mg2+ concentration (Fig. 2). Fitting the data to a simple two-state binding model allows for the calculations of the values of the Hill coefficient n and the magnesium ion concentration at which the transition is 50% complete ([Mg2+]1/2). In this experiment (Fig. 2), the values are: [Mg2+]1/2 =22 µM and n = 1.1 (11). These values suggest

Fig. 2. Plot of the efficiency of FRET (EFRET) as a function of magnesium concentration in the absence of adenine. The data were fitted (line) by regression to a two-state binding model, in which the binding of magnesium ions to the aptamer induces the formation of the loop–loop interaction.

Molecular Basis of RNA-Mediated Gene Regulation on the Adenine Riboswitch

73

that the transition occurs in the low micromolar range with an apparent lack of cooperativity. 3.7. Single-Molecule Analysis

Sm-FRET techniques provide a new approach for studying the structure-dynamics-function relationship in nucleic acid molecules, a method useful for the identification of subpopulations in a heterogeneous population, the analysis of discrete lifetimes, and the recovery of FRET efficiency distributions.

3.7.1. Quartz Slides Preparation

Cleaning the quartz slides and cover slips is a crucial step for single-molecule applications. In this section, we provide the protocol that is commonly used in the laboratory. 1. Sonicate quartz slides and cover slips sequentially in the following reagents: 20% detergent solution for 15 min, milliQ water for 5 min, acetone for 15 min, milliQ water for 5 min, 1 M KOH for 15 min, methanol for 15 min, 1 M KOH for 15 min, water 5 min. Dry the slides and slips using compressed air and pass them through a torch flame to remove impurities and moisture. 2. Form a channel by, first, attaching the ends of two pieces of tape such that there is a ~5-mm gap between them; then sandwich the tape pieces between the slide and the cover slip, as a means to create a channel for injecting reagents. 3. Add water to the channel and look in the microscope to see if the setup is clean before adding the immobilization reagents (see Note 8).

3.7.2. Immobilization of Single RNA Molecules

1. Dilute the initial biotinylated BSA stock tenfold in the riboswitch buffer and add 50 μL of this solution to the slide channel. 2. Wait 10 min to allow the biotinylated BSA to bind to the glass surface. Wash away unbound material with 60 μL of the riboswitch buffer. 3. Dilute the initial Streptavidin stock solution 25-fold in the riboswitch buffer and add 50 μL to the slide channel. 4. Allow 10 min for Streptavidin binding. Wash away unbound Streptavidin with 60 μL of the riboswitch buffer. 5. Add 60 μL of 50–100 pM solution of the ligated RNA to the riboswitch buffer and allow 5 min for the RNA to bind to the Streptavidin-coated surface (see Note 2). 6. Estimate the density and quality of labeling in the sm-TIR microscope; if these are found to be adequate, proceed with the addition of the imaging buffer. If the amount of immobilized molecules is not sufficient, repeat the RNAbinding step.

3.7.3. Data Collection and Analysis

1. Collection of single-molecule FRET data requires a mapping algorithm that correlates donor spots (left half of the EMCCD

74

Lemay et al.

Fig. 3. FRET trace for a single molecule (50 ms integration time) in the presence of 20 µM magnesium. The three different states corresponding to an unfolded (U), an intermediate (I), and a folded (F) state are shown on the right.

camera) with their acceptor counterparts (right half). In our laboratory, data collection is performed using 200-nm fluorescence beads and a program written in IDL v. 6 software (ITT Visual Information Systems, USA). 2. Fluorescence data at donor and acceptor wavelengths are acquired at room temperature from single molecules using total-internal reflection fluorescence microscopy with 532 nm laser excitation and a laboratory-written Visual C + + v. 6 program. Integration times range from 16 to 100 ms, depending on the sample dynamics. A detailed analysis of folding kinetics observed at Mg2+ concentration below 50 µM revealed that the aptamer domain has at least three different conformational states (11) (Fig. 3), which correspond to an unfolded (U), an intermediate (I), and a folded (F) state. 3. Single-molecule FRET efficiency after background correction is estimated by the equation (IA/[IA + ID]), where IA and ID are the fluorescence intensities of the acceptor and the donor, respectively. Because the quantum yields and detection efficiencies of Cy3 and Cy5 are very close, Eapp (the apparent FRET value observed) closely matches the true efficiency of energy transfer. Data analysis is performed using the laboratorywritten analysis routines developed in MATLAB 7 (The MathWorks Inc., Natick, MA). 4. Single-molecule FRET histograms are obtained by averaging the first ten frames of each FRET trace for every individual molecule after manually filtering the photobleaching and blinking effects (11). The conformational states are identified from the Eapp histograms, and dwell times are analyzed only if the time resolution allows for the clear observation of transitions (more than five data points per dwell time). Rapidly fluctuating molecules undergo more transitions than slowly fluctuating ones and, thus, in order to avoid bias toward fast rates, dwell time

Molecular Basis of RNA-Mediated Gene Regulation on the Adenine Riboswitch

75

histograms are obtained by using a weighting factor inversely proportional to the number of transitions observed for each molecule. These dwell time histograms are then fitted to a single-exponential function to obtain the lifetimes of each state, which are the inverses of the rates of conversion. For the heterogeneity analysis, the average of the dwell times is calculated for every state for each individual molecule (see Note 9).

4. Notes 1. APS is generally prepared for a volume of 10 mL. This solution is kept at 4°C in the dark for 2–3 weeks to yield efficient polymerization. 2. Addition of nonbiotinylated BSA precludes nonspecific binding of RNA molecules to the plastic containers, which can be significant at low (50–250 pM) RNA concentrations required for single-molecule FRET. 3. Recently, TROLOX (6-hydroxy-2,7,5,8-tetramethylchroman2-carboxylic-acid, a water-soluble derivative of Vitamin E, Sigma) has become an interesting alternative to 2-mercaptoethanol as a triplet-state quencher that greatly reduces undesired Cy5 blinking events. 4. Care should be taken when using higher glucose concentrations since an increase in solution viscosity could alter the intrinsic RNA dynamics. 5. Quartz slides coated with fluorescence beads for alignment and mapping can be stored and used for long periods of time if they are sealed with colorless epoxy resin, which reduces solvent evaporation and oxygen exchange. 6. Since both xylene cyanol FF and bromophenol blue absorb UV light when visualized on TLC plates, it is thus preferable not to load these dyes in the same wells with RNA samples if RNA is expected to comigrate with the dyes. 7. It is important not to overload the gel during PAGE purification since this could greatly compromise the resolution of the gel. 8. The use of ultrapure water for single-molecule applications is a crucial requirement since even the smallest fluorescence contaminants, undetectable in bulk solution experiments, could have a drastic effect at the single-molecular level. Thus, maximum care should be taken to ensure the highest water quality. Water that has been stored for long periods of time should not be used in single-molecule experiments. If slides are not clean enough after the cleaning protocol, use commercial bottled water (Oakville, Canada).

76

Lemay et al.

9. Single-molecule FRET histograms usually show a “zero peak” gaussian contribution caused mainly by incomplete labeling (lack of acceptor) or fast photobleaching during the first few frames of the experiment. Thus, care should be taken to ensure that the lowest value of FRET that could be obtained from the RNA molecule far from the “zero peak.” Usually, low FRET values over 0.15 can be safely assumed to have arisen from the active FRET state of the RNA molecule.

Acknowledgments This work was supported by a graduate scholarship (JFL) and a postdoctoral fellowship (JM) from the National Sciences and Engineering Research Council of Canada (NSERC) and by operating grants from the Canadian Institutes of Health Research (DAL) and the Scottish Universities Physics Alliance (JCP). DAL is a CIHR New Investigator Scholar.

References 1. Sudarsan, N., Barrick, J.E., and Breaker, R.R. (2003). Metabolite-binding RNA domains are present in the genes of eukaryotes. RNA 9, 644–647 2. Kubodera, T., Watanabe, M., Yoshiuchi, K., Yamashita, N., Nishimura, A., Nakai, S., Gomi, K., and Hanamoto, H. (2003). Thiamine-regulated gene expression of Aspergillus oryzae thiA requires splicing of the intron containing a riboswitch-like domain in the 5¢-UTR. FEBS Lett. 555, 516–520 3. Serganov, A. and Patel, D.J. (2007). Ribozymes, riboswitches and beyond: regulation of gene expression without proteins. Nat. Rev. Gen. 8, 776–790 4. Edwards, T.E. and Ferré-D’Amaré, A.R. (2007). Riboswitches: small-molecule recognition by gene regulatory RNAs. Curr. Opin. Struct. Biol. 17, 273–279 5. Mandal, M. and Breaker, R.R. (2004). Adenine riboswitches and gene activation by disruption of a transcription terminator. Nat. Struct. Mol. Biol. 11, 29–35 6. Serganov, A., Yuan, Y.R., Pikovskaya, O., Polonskaia, A., Malinina, L., Phan, A.T., Hobartner, C., Micura, R., Breaker, R.R.,

7.

8.

9.

10.

11.

12.

and Patel, D.J. (2004). Structural basis for discriminative regulation of gene expression by adenine- and guanine-sensing mRNAs. Chem. Biol. 11, 1729–1741 Batey, R.T., Gilbert, S.D., and Montange, R.K. (2004). Structure of a natural guanine-responsive riboswitch complexed with the metabolite hypoxanthine. Nature 432, 411–415 Mandal, M., Boese, B., Barrick, J.E., Winkler, W.C., and Breaker, R.R. (2003). Riboswitches control fundamental biochemical pathways in Bacillus subtilis and other bacteria. Cell 113, 577–586 Clegg, R.M. (1992). Fluorescence resonance energy transfer and nucleic acids. Methods Enzymol. 211, 353–388 Cornish, P.V. and Ha, T. (2007). A survey of single-molecule techniques in chemical biology. ACS Chem. Biol. 2, 53–61 Lemay, J.F., Penedo J.C., Tremblay, R., Lilley D.M. and Lafontaine, D.A. (2006). Folding of the adenine riboswitch. Chem. Biol. 13, 857–858 Ha, T. (2001). Single-molecule fluorescence resonance energy transfer. Methods 25, 78–86

Chapter 7 Methods for Analysis of Ligand-Induced RNA Conformational Changes Chad A. Brautigam, Catherine A. Wakeman, and Wade C. Winkler Summary Encoded within many RNA sequences is the requisite information for folding of intricate three-dimensional structures. Moreover, many noncoding RNAs can adopt structurally distinct and functionally specialized conformations in response to specific cellular signals. These conformational transitions are oftentimes accompanied by changes in hydrodynamic radii. Therefore, experimental methods that measure changes in hydrodynamic radius can be employed for study of signal-induced RNA conformational changes. Several hydrodynamic methods, including analytical ultracentrifugation, size-exclusion chromatography, and nondenaturing gel electrophoresis, are briefly discussed herein. Key words: RNA, RNA folding, Analytical ultracentrifugation, Riboswitch, RNA conformational changes

1. Introduction Hydrodynamic methods have been at the forefront of RNA research for more than 50 years (e.g., see ref. 1). A very sensitive hydrodynamic method, analytical ultracentrifugation (AUC), has been used to characterize ribosomes, ribosomal RNA, tRNA, and catalytic RNA (1–4). Also, centrifugation and size-exclusion chromatography (SEC) were essential to early efforts to purify RNAs (5, 6). Another hydrodynamic technique, denaturing polyacrylamide gel electrophoresis (PAGE), has long been used to separate RNAs (7), and native PAGE has been used to probe RNA structure more recently (e.g., see refs. 8–10).

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_7 © Humana Press, a part of Springer Science + Business Media, LLC 2009

77

78

Brautigam, Wakeman, and Winkler

Riboswitches are cis-acting RNA elements that are capable of regulating the expression of genes in response to smallmolecule metabolites (11, 12). They are thought to undergo significant structural rearrangements upon binding their cognate ligands (13, 14). These conformational changes are likely to be accompanied by changes in the hydrodynamic radius of the isolated riboswitch. Because AUC, SEC, and native PAGE are all capable of sensing such changes, these methods should be useful in the characterization of ligand binding by riboswitches. In this chapter, we present protocols for analyzing ligandinduced changes in the hydrodynamic radii of riboswitches. Because AUC is the most sensitive technique, it is presented in most detail; recent advances in computer programs that can analyze AUC data have brought about a renaissance of the methodology (15, 16). However, orthogonal methods for studying a problem are always helpful, and thus brief protocols for the SEC and native PAGE analyses of these RNAs are also included.

2. Materials 2.1. Analytical Ultracentrifugation

The following method assumes that the experimenter is using a Beckman analytical ultracentrifuge model XL-A or XL-I. Because the XL-A is not capable of laser interferometry, the detection method will take advantage of the spectrophotometer built into both models of ultracentrifuge. 1. Materials for the assembly of Beckman ultracentrifugation cells (at least 3). To familiarize the inexperienced reader with the apparatus and its accompanying terminology, a brief description of the cells follows. The experiment is carried out in a centerpiece that has two sector-shaped chambers. One chamber is for the reference buffer, and the other is for the sample. A central rib separates the two chambers. At the narrow ends of the sectors are two small fill ports that are open to the outside of the centerpiece; they will be used to introduce the sample and reference buffers into the sectors after cell assembly. The centerpiece is to be sandwiched between two transparent windows such that the windows form the sides of the sectors and prevent liquid from escaping. The on-board spectrophotometer in the Beckman instrument compares the intensity of light coming through the reference sector with that coming through the sample sector, and the absorbance is calculated from this difference.

Methods for Analysis of Ligand-Induced RNA Conformational Changes

79

The window/centerpiece sandwich is placed in a centerpiece housing, which is simply a hollow aluminum cylinder. A screw ring will be screwed into the housing on top of the sandwich to hold it in place, and it will be tightened using a torque wrench to provide a strong seal between the centerpiece and the windows. The fill ports will align with two holes in the centerpiece housing. After filling, the ports will be sealed with two housing plugs that screw into the holes and rest against the centerpiece. The assembled, filled cell is placed in the centrifugation rotor such that the windows are “up” and “down” and the narrow portions of the sectors are closest to the center of the centrifuge. It should be noted that the thickness of the centerpiece is 1.2 cm; therefore, a solution with an absorbance of 1.0 AU in a 1-cm cuvette will have an absorbance of approximately 1.2 AU in the centerpiece. 2. One Beckman rotor, either model An60-Ti (with four holes) or An50-Ti (with eight holes). 3. Round gel-loading pipettor tips, USA Scientific (Ocala, FL), catalog number 1022–0600. 4. Standard Cleaning Solution: Beckman 555 detergent, diluted 1:10 with deionized H2O, total volume of 200 mL. 5. Rigorous Cleaning Solution: 0.5% (w/v) sodium dodecyl sulfate, 0.5 mM EDTA, pH 8.0, total volume of 250 mL. 6. Small plastic beakers. 7. Bath sonicator. 2.2. RNA Preparation by In Vitro Transcription

1. T7 RNA polymerase (~50 μg/mL reaction); store at −20°C. 2. DNA template (~1–2 μg/reaction). This template can be generated through PCR-amplification using a forward primer that incorporates the T7 promoter sequence (TAATACGACTCACTATAG). 3. 10× transcription buffer: 300 mM Tris–HCl, pH 8.0, 100 mM DTT, 1% Triton X-100, 1 mM spermidine, 400 mM MgCl2; store at −20°C. 4. 25 mM NTP mix: 25 mM rATP, rCTP, rGTP, and rUTP (Roche); store at −20°C. 5. Optional: yeast inorganic pyrophosphatase (Sigma) (see Note 1); store at −20°C. 6. Phenol:Chloroform:Isoamyl (Roche); store at 4°C.

alcohol

25:24:1

7. Chloroform (EMD, Gibbstown, NJ). 8. Glycogen (20 mg/mL) (Roche); store at −20°C. 9. 3 M sodium acetate, pH 5.2.

(v/v/v)

80

Brautigam, Wakeman, and Winkler

10. A SEC column and a liquid chromatography system. We use a Superdex 200 10/300 GL or a Superdex 75 10/300 GL (GE Healthcare) column and an AKTA FPLC system (Amersham). 11. Uridine 5¢-triphosphate, [α-32P] (α-UTP, Amersham); store at 4°C. This is only required to produce body-labeled radioactive RNA for the native PAGE method. 2.3. Size-Exclusion Chromatography

1. SEC running buffer: 10 mM Tris–HCl, pH 7.5, 100 mM KCl, 10 mM MgCl2. Other buffers with similar ionic strengths should produce comparable results. 2. ~100 pmols RNA, store at −20°C.

2.4. Native Polyacrylamide Gel Electrophoresis

1. ~10–50 kcpm radiolabeled RNA, store at −20°C. 2. 10× TB: 108 g Tris base, 55 g boric acid, adjust with water to 1 L. 3. Running buffer (1× TBM): 10× TB diluted 1:10 in H2O supplemented with 1 mM MgCl2. 4. 10% (w/v) ammonium persulfate (APS), prepare fresh. 5. TEMED; store at 4°C. 6. 30% (w/v) acrylamide/bis-acrylamide [29:1], store at 4°C. 7. 1 M MgCl2. 8. 2× native gel buffer: 400 mM KCl, 100 mM Tris–HCl, pH 7.4, 20 mM MgCl2. 9. Glycerol. 10. Glass plates (sized approximately 28 cm × 16.5 cm with 0.75-mm spacers, 8 well combs) and appropriately sized gel rigs. 11. Gel-drying equipment, Whatman paper, and plastic wrap. 12. 20 cm × 25 cm phosphor screens (Amersham). 13. Typhoon 9,200 Variable Mode Imager (Molecular Dynamics, Sunnyvale, CA). 14. ImageQuant version 5.2 software.

3. Methods Although many different kinds of experiments may be carried out in the analytical ultracentrifuge, two methodologies are currently dominant among researchers studying macromolecules: sedimentation equilibrium (SE) and sedimentation velocity (SV). In an SE experiment, a solution containing a macromolecule (or mixture of macromolecules) is centrifuged for a long period

Methods for Analysis of Ligand-Induced RNA Conformational Changes

81

of time (hours to days) at an angular velocity that allows the centrifugal, diffusional, and buoyant forces acting on the macromolecules to be in balance at all radial positions in the sample cell. The appearance of the resultant concentration gradient is dependent on the molecular weight of the macromolecules, not their shape. SE is therefore inappropriate to study conformational changes of macromolecules. Instead, SV should be used. Solutions of macromolecules subjected to SV are centrifuged at high angular velocity; in most cases, the centrifugal force overwhelms the diffusional and buoyant forces, causing the macromolecules to accumulate near to the bottom of the sample cell. By measuring the concentration gradient across the sample cell as the macromolecule sediments toward the bottom of the sample cell at regular time intervals (Fig. 1a), the sedimentation coefficient of the macromolecule can be calculated. This “s-value” is dependent on the shape of the macromolecule; a spherical species sediments faster than an elongated species of identical mass. The shape dependence of the s-value makes SV well suited to studying conformational changes in macromolecules. Indeed, it has been used to study ligand-induced shape changes in proteins (17) and metal-ion-induced compaction of RNAs (18, 19). Size-exclusion chromatography can also be a useful tool to assess changes in the size and shape of an RNA molecule during folding or in response to ligand association (e.g., 20, 21). Changes in the RNA elution profile under different conditions can be employed as a qualitative assay for detecting alterations in hydrodynamic radius and also for preparative purposes. SEC can also be adapted for analytical analyses by employing data obtained from a comparison of the elution profiles for proteins of known Stokes radii with the target RNA(s) for calculation of the apparent Stokes radii for the target RNA(s) (20). Nondenaturing electrophoresis can also be a rapid and sensitive method for accessing the global architecture of structured RNAs. Indeed, comparative electrophoretic analyses have been extensively employed in recent decades as a means for evaluating RNA folding pathways (e.g., 8–10). 3.1. Analytical Ultracentrifugation 3.1.1. Preparation of the Samples for AUC

Before the samples are prepared, some thought must be given to the composition of the sample buffer (see Note 2). Foremost, buffer components that absorb light strongly at the monitoring wavelength should be minimized or eliminated. For example, Tris buffer absorbs light in the far-UV region of the spectrum. However, we have successfully used Tris in solutions that were monitored at 260 nm if the concentration of the buffer is low (10– 20 mM). Additionally, the ionic strength of the solution should be considered carefully. Because RNA is a highly charged polymer, sufficient supporting electrolyte is required to avoid molecular repulsion that could lead to nonideal sedimentation. At least 100 mM of monovalent cation is recommended. At this point, the

82

Brautigam, Wakeman, and Winkler

Fig. 1. The results of a c(s) fitting session in SEDFIT. (a) The raw sedimentation velocity data with the fit. In the upper panel, the raw absorbance data are represented as open circles. For clarity, only every sixth scan is shown, and only every third data point from each scan is shown. In this figure, the M-box RNA was sedimented in the presence of 50 μM MgCl2. The solid lines represent the fit to the data that was obtained by optimization in SEDFIT. The lower panel shows the residuals between the data and the fit. (b) The c(s) distribution. The solid line shows the c(s) distribution that was obtained from the fit in (a). The dotted line is the c(s) distribution for a separate experiment that had 7.5 mM MgCl2 in solution. The raw data for this latter experiment are not shown. Both distributions have significant evidence of aggregation at s-values greater than 8 S.

Methods for Analysis of Ligand-Induced RNA Conformational Changes

83

buffer density (r), the buffer viscosity (h), and the partial specific volume of the RNA (v) should be determined, experimentally if possible. However, not all researchers have access to the densimeters and viscometers that are necessary for such an experimental determination. In such cases, r and h may be calculated using the freeware program SEDNTERP (22). A standard value for the v of RNA of 0.53 cm3/g can be used (19). If the conformation of the RNA under study is dependent on the concentration of ligand, either three (in the four-hole rotor) or seven (in the eight-hole rotor) different concentrations of ligand can be studied in one experiment (see Note 3). Ligand concentrations should be chosen such that they span the EC50 of the interaction. Obviously, with the limited number of data points available, it will be necessary to run the centrifuge several times for a complete characterization of the conformational change. 3.1.2. Quantities of Samples to Be Used

It behooves the experimenter to use as much volume as possible in the SV experiment. The sectors in the centerpieces are capable of holding 450 μL each. We typically use 400 μL of material per experiment. Thus, if three samples are to be analyzed simultaneously, at least 1.2 mL of sample is required.

3.1.3. Wavelength, Window, and Centerpiece Choices

RNA obviously absorbs UV light maximally at about 260 nm, making this wavelength a natural choice to monitor the sedimentation. One must be aware that the usable dynamic range of the on-board spectrophotometer is from about 0.1 to 1.2 AU at this wavelength. If the experiment requires an RNA concentration whose absorbance exceeds 1.2 AU, the spectrophotometer can be tuned to an off-maximal wavelength (e.g., 270 nm) to put the reading within the dynamic range of the instrument. Either quartz or sapphire windows may be used. Because of their superior durability, we prefer sapphire windows. However, the experimenter is cautioned that such windows may have significant absorbance in the UV region of the spectrum. A spectrum of the sapphire window should be taken to ensure that the absorbance at the wavelength of interest is not too great. Quartz windows must be used for work in the far-UV region of the spectrum because of the strong absorbance of the sapphire windows there. Charcoal-filled epon centerpieces are appropriate for RNA. Beckman also manufactures aluminum centerpieces, but they may be incompatible with the solution conditions needed for RNA SV experiments.

3.1.4. Cell Assembly

Instructions for cell assembly are provided by Beckman. They should be carefully followed. At least 120 in-lb of torque should be applied to the screw ring that holds the window/centerpiece sandwich in place because leakage of the centerpiece can occur

84

Brautigam, Wakeman, and Winkler

at lesser values. Caution should be taken not to apply too much torque; damage to the expensive windows can occur. 3.1.5. Cell Loading

Again, the instructions supplied by Beckman should be followed. In addition, we offer the following hints. The solution that is loaded into the reference sector should be the same as that loaded into the sample sector, except, of course, that the RNA should be absent. The gel-loading pipettor tips fit on pipettes that are capable of dispensing only 200 μL. Therefore, two applications to each sector will be necessary. We have noticed that reflux can occur during the second application. To avoid this problem, try not to wet the fill port during the first dispensation. Also, during the second application, it is helpful to place the end of the tip below the level of the liquid that has already been dispensed into the sector. Some researchers routinely fill the sample sector with 5–20 μL less volume than the reference sector. This practice allows them to readily identify the sample meniscus during the data analysis step (see later). To achieve a good seal over the fill ports, put two red plug gaskets over the fill port before sealing the fill ports with the housing plugs. Obtain a mass of all filled, assembled cells and ensure that cells that are to be centrifuged in opposite rotor holes are the same mass (within 0.5 g). One cell will be balanced against the counterbalance, the mass of which can be adjusted.

3.1.6. Cell Insertion into the Rotor

The cells should be inserted into the rotor such that the fill ports are most proximal to the rotor’s center, and the screw rings should be up. The centerpiece housings and the rotor holes have registration marks to allow the proper alignment of the cells (see the Beckman manual). It is important to be very precise with this alignment; errors that this stage can cause turbulence in the sample, which can deleteriously affect the data analysis.

3.1.7. Rotor Placement; Temperature Equilibration

The rotor and optical assembly should be placed in the centrifuge as directed by Beckman. Once the door is closed, the vacuum can be activated. Set the temperature of the centrifuge to the desired experimental temperature. We routinely use a temperature of 20°C. Allow the rotor to rest for 1–2 h after the desired temperature has been reached. This step ensures that the cell contents are at the same temperature as the rotor, which will minimize the possibility of convection caused by temperature gradients in the sample.

3.1.8. Starting the Experiment

The Beckman data acquisition software is described in a manual printed by Beckman; its contents will not be reproduced here. However, there are several choices the experimenter must make at this stage. One choice is rotor speed. Most riboswitches fall in a mass range of 20,000–60,000 Da. For such masses, choose the

Methods for Analysis of Ligand-Induced RNA Conformational Changes

85

fastest possible speed up to 50,000 rpm. The Beckman charcoalfilled epon centerpieces are rated to 42,000 rpm. Third-party suppliers make similar centerpieces that can be used at speeds up to 60,000 rpm. Another choice is how often to acquire the data. SV data analyses generally benefit from large sets of scans acquired close together in time; therefore, data should be acquired as often as possible. In the Beckman data acquisition software, the user sets the time interval between scans. We set this parameter to 1 min. Because it takes about 1.5 min to scan a single cell in absorbance mode, this setting ensures that data are continually acquired. A final consideration is how long to acquire data. It is best to allow the experiment to reach its conclusion. Therefore, one should continue the experiment until all the RNA is localized at the bottom of the sample sector. One accomplishes this by inputting the maximum number of scans (999) into the software. In reality, the experimenter will rarely collect all these scans; the RNA will have fully sedimented long before so many scans can be collected, at which time the experiment is terminated (see later). Once all these choices have been made, and temperature equilibrium has been reached, the centrifuge may be started. Only after the rotor has achieved its final speed should data acquisition be initiated. 3.1.9. Stopping the Experiment

After all the RNA has sedimented to the bottom of the cell, it is time to stop the experiment. The appearance of the absorbance scans will demonstrate that the endpoint of the experiment has been reached. The scans should read zero absorbance throughout the cell except in the meniscus region and near the bottom of the cell. To end data acquisition, choose “Stop Scans” in the Scan menu. After the scans have ceased, stop the centrifuge. When the display on the centrifuge instructs the user to do so, release the vacuum, open the door, remove the optical assembly, and recover the rotor.

3.1.10. Sample Recovery and Cell Disassembly

Remove the cells from the rotor. Remove the housing plugs and the red plug gaskets. The reference buffer and sample should be removed before disassembly. If prolonged exposure to the experimental temperature and UV irradiation do not harm your RNA, the sample may be recovered and used for further characterization. At this point, some researchers simply clean the centerpieces with water and leave the cells assembled. Because our facility accommodates multiple users and RNA is notoriously labile, we feel that the cells should be disassembled and cleaned after every use. The cells should be disassembled according to Beckman’s instructions.

3.1.11. Cleaning Cell Components

At the least, the windows and centerpieces should be rinsed with distilled water, soaked for ~5 min in the Standard Cleaning Solution, then rinsed very thoroughly with distilled water. We occasionally clean all cell components. Once every few weeks, we place the

86

Brautigam, Wakeman, and Winkler

windows and centerpieces in the Rigorous Cleaning Solution (in individual plastic beakers) and sonicate them in a bath sonicator for ~5 min. This treatment is followed by copious rinses with distilled water, 95% ethanol, and distilled water again. 3.1.12. Data Analysis

Several different approaches are available to the modern SV practitioner who wishes to analyze data. Our preferred method is to perform a c(s) analysis (see Note 4). The c(s) approach, introduced by Peter Schuck and implemented in his program SEDFIT, uses superpositions of Lamm equation solutions to provide a fit to the SV data (23–25) (see Fig. 1a). The result is a continuous, diffusion-deconvoluted differential distribution, called c(s), that shows the relative amounts of species with a given sedimentation coefficient s. Although the analogy is not precise, appearance of the c(s) distribution is similar to that of a signal trace from a gel filtration column, but with smaller species appearing earlier in the distribution. SEDFIT is freely available from the web site www. analyticalultracentrifugation.com. Very instructive tutorials and extensive documentation for the program are available on that web site as well. For this reason, we will only cover the basics of how to analyze SV data using SEDFIT. 1. Load data by choosing “Load New Files” under the “Data” menu. Choose files that will cover the entire time course of the experiment. The program will prompt the user to load every nth file. This feature is useful if many files were selected; choose a value of n such that about 40–50 files are actually loaded. For example, if 150 files were selected, choose an n of 3. 2. Choose the sample meniscus and cell bottom positions. This action is accomplished by placing the pointer at a radial position, holding down the “ctrl” key, and double clicking. In absorbance data, the sample meniscus usually has the appearance of a sharp, positive peak. If the user has loaded less volume into the sample sector (see Subheading 3.5), then the sample meniscus should appear to the right of the reference cell meniscus. The cell bottom can be found by locating the position where the absorbance profiles become flat. This position is usually very close to 7.2 cm, using the Beckman charcoal-filled epon centerpieces. 3. Choose the fitting limits. These limits are chosen by simply double clicking after placing the cursor at the desired position. The limits should be chosen to exclude optical artifacts near the meniscus. Also, because of the very steep concentration gradients near the bottom of the cell, this region should be excluded from the analysis. 4. Choose the model. Under the “Model” menu, choose “Continuous c(s) Distribution.” 5. Set the parameters. Choose the “Parameters” option. A dialog box will open. For the beginning of the analysis, the resolution

Methods for Analysis of Ligand-Induced RNA Conformational Changes

87

(the number of s-values that SEDFIT considers) can be set to 50. The values of “s min” and “s max” can be set to a range where it is likely that the RNA will sediment. For most riboswitches, a range of 1–10 S is sufficient. For “frictional ratio,” the default value of 1.2 can be taken. The box to the left of “frictional ratio” indicates that this parameter can be fit if the user checks the box. It is appropriate to fit the frictional ratio. The values for “partial specific volume,” “buffer density,” and “buffer viscosity (Poise)” should be entered. The fit boxes for “Baseline” and “Fit Time Independent Noise” should be checked. It is usually not appropriate to check the box for “Fit RI Noise” for absorbance data. Because the actual position of the sample meniscus is usually uncertain, allow the program to fit it. Because the region near to the bottom of the sector has been excluded from the analysis, the “Bottom” position does not need to be optimized. The current default for “confidence level (F-ratio)” is 0.95. This value controls the regularization of the c(s) distribution. The default value will produce a smoother, broader distribution than lower values. We usually choose to retain this default value. The other parameters are for advanced users; the reader is encouraged to explore their meanings in the online SEDFIT documentation. When all parameters have been input, press OK. 6. Do an initial run by choosing “Run.” During a run, the linear parameters (baseline and time-invariant noise, in this case) are optimized. A c(s) distribution (Fig. 1b) will appear at the bottom of the screen, and a grayscale bitmap will appear in the middle of the screen. Without going into details, strong diagonal features in the bitmap indicate systematic deviations of the fit from the data. If such a feature is seen, it has been our experience that the most likely incorrect parameter is the frictional ratio. It should be adjusted (followed by subsequent Runs) until the diagonal feature is minimized. The root-mean-square deviation (r.m.s.d.) of the fit to the data is displayed in the upper left of the display. For a Beckman centrifuge in good working order one should generally expect r.m.s.d. values of less than 0.007 AU. 7. Once the parameters have been optimized, choose “Fit.” Now all parameters checked in the parameter dialog will be fit using a nonlinear simplex procedure. This process may take several minutes. The result is an optimized fit. The user may wish to repeat the fit to ensure that the global minimum in the error surface has been reached. 8. Finalize the analysis. At this point, all data (not just every nth scan) may be analyzed. Choose “Add More Files” under the “Data” menu, and choose all appropriate data. The resolution of the analysis can also be increased; we often use a resolution of 100–150. Do a Run to finalize the analysis.

88

Brautigam, Wakeman, and Winkler

9. Analyze the distribution. SEDFIT has an integration feature that allows the user to find the signal-average s-value under a selected peak (see Note 5). It is possible that more than one peak will be evident. For example, two peaks could be present: one for an elongated, relatively unfolded structure, the other for a compactly folded structure. In such a case, it may be most useful to obtain the signal-averaged sedimentation coefficient by integrating the entire range in which the two peaks are encountered. 10. If it is desirable to know the value of the hydrodynamic radius of species that have been characterized by SEDFIT, the program has a built-in calculator for determining it. Under “Options,” choose “Calculator,” then “calculate axial and frictional coefficient ratios.” The user must know the molar mass, s-value, v, hydration (we have used 0.59 g/g (22)), temperature, r, and h. A dialog box will appear with the hydrodynamic radius. 11. Document the result. SEDFIT does not have advanced graphic or documentation capabilities. In the “Copy” menu, however, the c(s) distribution, raw data, fit data, etc. may be copied to the operating system’s clipboard and pasted into a spreadsheet program. For the widest applicability, the s-values in the distribution may be converted to standard conditions, i.e., the s-value at 20°C in water (s20,w). This is accomplished by multiplying the s-values by: ⎛ hT,b ⎞ ⎛ 1 − v r 20, w ⎞ ⎟, ⎜h ⎟⎜ ⎝ 20, w ⎠ ⎝ 1 − v r T,b ⎠ where hT,b is the viscosity of the buffer at the experimental temperature, h20,w is the viscosity of water at 20°C, r20,w is the density of water at 20°C, and rT,b is the density of the buffer at the experimental temperature. This formula assumes that the partial specific volume of the RNA does not vary much with temperature or buffer conditions. For any given RNA, this assumption may or may not be well founded. 3.1.13. Overall Analysis

In our hands, the data could be analyzed by plotting log[Mg2+] vs. the observed sedimentation coefficient of the riboswitch (21). Such a plot could be fitted to the Hill-type sigmoidal doseresponse equation: s = s min +

s max − s min (log EC50 − log[Mg 2+ ])(n )

1 + 10

,

Methods for Analysis of Ligand-Induced RNA Conformational Changes

89

where smin is the minimum s-value, smax is the maximum s-value, EC50 is the concentration of ligand at which the average of smin and smax is obtained, and n is the Hill coefficient. Nonlinear minimization algorithms may be used to optimize the values of smax, smin, EC50, and n. We used the commercial program SigmaPlot version 10.0 to perform this analysis. Such an analysis might also be used if analyzing signal-averaged s-values. 3.2. RNA Preparation by In Vitro Transcription

1. For a yield of ~400–800 pmols of RNA, combine ~4–8 μg DNA template, 10 μL 10× transcription buffer, 10 μL 25 mM NTP mix, ~50 μg/mL T7 RNA polymerase, and 0.01 U inorganic pyrophosphatase (optional) in a final volume of 100 μL. If more RNA is desired, this reaction can easily be scaled up. For producing internally radiolabeled RNA for native PAGE experiments, supplement the reaction with 2–6 μL α-UTP (3,000 Ci/mmol). At a relatively low frequency, radiolabeled UTP will be incorporated into the RNA molecules in place of normal UTP. The reactions should be incubated for 2–3 h at 37°C (see Note 6). At this point, the reactions are terminated with the addition of 100 μL H2O and 200 μL phenol:chloroform:isoamyl alcohol. Mix thoroughly. 2. Centrifuge at ~20,000× g for 5 min to separate into 2 phases. Retain the top, aqueous phase, which contains the RNA, and discard the bottom phase. Repeat the extraction and centrifugation with 200 μL chloroform to remove traces of phenol. 3. At this point, the RNA should be concentrated through the use of ethanol precipitation. Add 20 μL (1/10th volume) 3 M sodium acetate and 1 μL glycogen to the RNA (see Note 7). Then add 500 μL (2.5× volume) 100% ethanol, mix by inversion, and incubate at −20°C for 30 min. Pellet the RNA by centrifuging at ~20,000× g for 15 min at 4°C. Wash the pellet with the addition of 200 μL 70% ethanol and centrifuge at ~20,000× g for 5 min at 4°C. Discard supernatant and air dry the pellet for 2–4 min. Resuspend the RNA in 100 μL H2O. Load onto a gel filtration column as detailed in Subheading 3.3 and collect the fractions with high absorbance at 254 nm. In this case, no ligand should be included in the RNA sample or in the gel-filtration buffer. 4. The RNA should be concentrated through the use of ethanol precipitation as performed in the previous step (using proportional volumes). Resuspend the RNA in 50 μL H2O for use in native PAGE and AUC, and 125 μL SEC running buffer (see Subheading 2.2) for use in SEC experiments.

3.3. Analytical Size-Exclusion Chromatography

1. This section assumes the use of a Superdex 200 10/300 GL column and an AKTA FPLC, which has been demonstrated to be useful for RNA constructs ranging from 150 to 250

90

Brautigam, Wakeman, and Winkler

nucleotides (~45–75 kDa assuming ~300 Da/nucleotide) and should be also be useful for constructs ranging from ~40– 200 nucleotides although a Superdex 75 10/300 GL column might be better suited for constructs smaller than 150 nucleotides. 2. Equilibrate the column with SEC running buffer at 4°C. Pump at least 1.5 column volumes of running buffer over the column prior to the experiment. In experiments where a ligand will be added to assess the effect of the molecule on global architecture, the running buffer should also contain the desired concentration (at least tenfold above the dissociation constant) of the ligand to be tested. 3. Set the UV-absorbance system to measure at 254 nm (the AKTA allows two detection wavelengths: either 254 nm or 280 nm). 4. In runs where the ligand is to be added to the RNA, add the desired concentration of ligand to the 125 μL volume of RNA. Heat the sample to 55°C for 2 min and allow it to slowly cool to room temperature for 15 min on the benchtop. This step allows the RNA to refold in the presence of the potential ligand. 5. Load 125 μL RNA into a 100-μL loop with a syringe for injection onto the column. 25 μL of extra volume is used to prevent air from being trapped in the loop and injected onto the column. 6. Apply the RNA to the column, using a flow rate of 0.5 mL/ min. The absorbance profile at 254 nm can be compared between the ligand bound and unbound samples. A shift in elution time should occur if the ligand induces a significant change in the hydrodynamic radius of the RNA. For example, if the ligand-treated RNA elutes at a later time, a smaller hydrodynamic radius for the RNA–ligand complex is indicated. This technique can be used to distinguish between magnesium-bound and unbound species of a magnesium-responsive B. cereus riboswitch as demonstrated in Fig. 2a due to the significant conformational change that occurs upon magnesium binding (see Fig. 2b). Fig. 2. (continued) magnesium concentrations, the RNA elutes at different retention times. These data suggest that the RNA undergoes significant compaction of the global fold upon magnesium binding. (c) Separation of compact and extended states of the B. subtilis M-box RNA by native gel electrophoresis on an 8% polyacrylamide gel containing 2 mM MgCl2 in the gel and the running buffer. The mutant RNA is the M3 mutant previously demonstrated to have the structural characteristics of unbound RNA (extended conformation) even under magnesium-replete conditions (21). (d) Separation of compact and extended states of a thiamine pyrophosphate (TPP)-binding riboswitch aptamer from the 5¢ UTR of the B. subtilis tenA gene. This separation was performed on a 10% nondenaturing polyacrylamide gel with 1 mM MgCl2 in the gel and the running buffer. Presumably, other riboswitches are also likely to undergo a similar compaction upon ligand binding.

Fig. 2. Separation of ligand-bound and free riboswitch RNAs based on changes in hydrodynamic radius. (a) Schematic demonstrating the genetic mechanism employed by M-box riboswitch RNAs. Association of magnesium with the M-box RNA leads to compaction of a metal-binding domain, as supported by biochemical and biophysical data (21). The numbered boxes highlight the significant base pairing rearrangement that occurs upon the formation of magnesium-induced tertiary contacts. The helical element formed between #3 and #4 sequences promotes termination of transcription and hence lowered expression for downstream genes. (b) Size-exclusion chromatography of the B. cereus M-box magnesium-binding aptamer domain (21). When this RNA element is separated on a Superdex 200 10/300 GL column that has been equilibrated in buffers containing 10 mM Tris pH 7.5, 100 mM KCl, and either high (black line) or low (gray line)

92

Brautigam, Wakeman, and Winkler

3.4. Native Polyacrylamide Gel Electrophoresis

1. To make an 8% native gel solution, combine 4 mL 10× TB, 10.7 mL 30% acrylamide, and 0.08 mL 1 M MgCl2 (or ligand). Bring the final volume to 40 mL with deionized water. While an 8% gel appears to be useful for most constructs, a different percentage gel can be made by adjusting the amount of 30% acrylamide added. 2. Assemble a sandwich of the spacers and the glass plates in preparation for pouring the gel (see Note 8). 3. To initiate polymerization, add 0.4 mL 10% APS and 0.035 mL TEMED to the acrylamide solution. Immediately pour this solution into the space between the glass plates, and slide in the comb. Allow the gel to polymerize for ~30 min. 4. Remove the comb and rinse out the wells with distilled H2O. 5. Attach the plates to the gel rig and fill the upper and lower reservoirs with running buffer. 6. In a 1.5-mL microcentrifuge tube, combine 5 μL 2× native gel buffer, 2 μL glycerol, ~10–50 kcpm RNA, and the desired potential ligand to be tested. A sample with no ligand should also be prepared for the sake of comparison. Bring the volume up to 10 μL with H2O. 7. Heat the samples in a heat block to 55°C for 2 min to partially denature the RNA structure. Remove the block from heat and allow to slowly cool to room temperature for 15 min before loading on the gel. No loading buffer will be required due to the glycerol that was added in the reaction. 8. Load the entire RNA sample into a well of the prepared gel. 9. Apply ~15–20 W to the gel for 3–8 h. These exact experimental conditions will need to be optimized for each particular RNA species. After ~4 h of electrophoresis the buffers should be exchanged. If the glass plates feel warm to the touch, the power should be turned down or the gel rig should be moved to a cold room. 10. After 3–5 h, stop the gel and check the gel rig reservoirs for radioactivity prior to disassembly. 11. Remove one of the glass plates, allowing the gel to remain attached to the other. Press a sheet of Whatman paper on top of the exposed gel. The gel will stick to the Whatman paper and can be peeled away from the remaining glass plate. Cover the exposed side of the gel in plastic wrap. Dry the gel at 80°C for 2 h. 12. Place the dried gel on a phosphorimaging screen such that the plastic wrap is in contact with the screen. Expose for >1 h. In some cases, the gel can be exposed overnight, but this expedient might result in regions that have been saturated.

Methods for Analysis of Ligand-Induced RNA Conformational Changes

93

13. Scan in the image of the gel using the phosphorimager. If the image has areas of saturation, re-expose the gel for a shorter time and rescan the screen. If the ligand binds to the RNA and induces a significant change in hydrodynamic radius, the mobility of the RNA on the gel should be altered relative to that of the free RNA. For example, the magnesium-induced conformational change in the B. subtilis M-box riboswitch can be visualized by native gel electrophoresis when the mobility of the wild-type RNA is compared to that of a mutant incapable of forming the magnesium-mediated compacted state (Fig. 2c). SEC analysis of this mutant resulted in the same elution profile as that of the unbound wild-type RNA (21). The ligand-bound state is significantly more compact than the free state, which results in an RNA species with an increased mobility through the gel. An additional example is shown in Fig. 2d for separation of bound and unbound complexes for a thiamine pyrophosphate (TPP)-binding riboswitch.

4. Notes 1. While not necessary for in vitro transcription to occur, inorganic pyrophosphatase can improve the yield of RNA. During transcription, pyrophosphate will be released into solution and may chelate Mg2+. When too much Mg2+ is sequestered by pyrophosphate, RNA polymerase, a Mg2+-dependent enzyme, will be inhibited. Inorganic pyrophosphatase prevents this inhibition by hydrolyzing the pyrophosphate to inorganic phosphate, which does not sequester Mg2+ as readily. 2. The procedure described in this chapter assumes that the RNA under study is stable for the duration of the SV experiment (hours) and is not particularly labile. If this may not be the case, other precautions can be taken. SV can be performed at lower temperatures. Also, the windows and centerpieces can be treated with a fresh water/DEPC solution or a commercial RNase inhibitory preparation (like RNaseZAP, Ambion). RNase inhibiting proteins may also be included in the RNA solution, though these may have absorbance that must be accounted for in data analysis. 3. The s-values of some RNAs are dependent on concentration (3, 4). If valid comparisons must be drawn between several samples at different conditions, it is important to keep the RNA concentration constant in all experiments.

94

Brautigam, Wakeman, and Winkler

4. The c(s) methodology as implemented in SEDFIT has limitations that the RNA researcher must keep in mind. The program assumes that all species have the same frictional ratio. If two or more species with dissimilar frictional ratios are present, the fit can suffer. SEDFIT also assumes that all species are noninteracting for the c(s) model. Finally, because the distribution is regularized, the meaning of peak widths is not readily interpretable. 5. We found that as Mg2+ was titrated into solutions containing M-box RNA, the riboswitch sedimented as a single species of increasing s-value (21); see Fig. 1b. One possible explanation of these observations is that the riboswitch is interconverting between a highly extended state at lower Mg2+ concentration and a more compact state at higher Mg2+ concentration; in such a case, the two states must be interconverting on a time scale that is fast compared to the time scale of the sedimentation experiment. Simulations and experiments have demonstrated that, in a rapidly re-equilibrating system, ligand binding to a species that causes it to sediment faster can result in the formation of a gradient of ligand in the cell (26). Such a gradient can lead to bifurcated peaks in a sedimentation coefficient distribution. No such peaks are evident in our experiments, indicating that any such gradient in our experiments is probably eliminated by the rapid diffusion of Mg2+. 6. If a white precipitate forms, no further transcription can occur, and the reaction can be stopped prior to the full 2–3 h incubation. The white precipitate indicates Mg2+ chelation by pyrophosphate. The addition of inorganic pyrophosphatase should prevent the formation of this precipitate and increase the RNA yield. 7. Glycogen is optional but recommended in ethanol precipitations. It is inert, and its presence will not affect subsequent reactions. The glycogen precipitates with the RNA and allows for easy visualization of the pellet. 8. Upon purchase, the outside of the glass plates should be permanently marked so that proper orientation of the plates can be maintained in all subsequent runs. References 1. Schachman, H. K., Pardee, A. B., and Stanier, R. Y. (1952). Studies on the macromolecular organization of microbial cells. Arch. Biochem. Biophys. 38, 245–260 2. Deras, M. L., Brenowitz, M., Ralston, C. Y., Chance, M. R., and Woodson, S. A. (2000). Folding mechanism of the Tetrahymena ribozyme P4-P6 domain. Biochemistry 39, 10975–10985

3. Henley, D. D., Lindahl, T., and Fresco, J. R. (1966). Hydrodynamic changes accompanying the thermal denaturation of transfer ribonucleic acid. Proc. Natl. Acad. Sci. U.S.A. 55, 191–198 4. Stanley, W. M. J. and Bock, R. M. (1965). Isolation and physical properties of the ribosomal ribonucleic acid of Escherichia coli. Biochemistry 4, 1302–1311

Methods for Analysis of Ligand-Induced RNA Conformational Changes 5. McConkey, E. H. (1967). The fractionation of RNA’s by sucrose gradient centrifugation. Methods Enzymol. 12, 620–634 6. Moldave, K. and Sutter, R. P. (1967). Purification of aminoacyl-sRNA by molecular sieve chromatography on Sephadex. Methods Enzymol. 12, 598–601 7. Maniatis, T., Jeffrey, A., and van deSande, H. (1975). Chain length determination of small double- and single-stranded DNA molecules by polyacrylamide gel electrophoresis. Biochemistry 14, 3787–3794 8. Emerick, V.L. and Woodson, S.A. (1994). Fingerprinting the folding of a group I precursor RNA. Proc. Natl. Acad. Sci. U.S.A. 91, 9675–9679 9. Pan, J., Thirumalai, D. and Woodson, S.A. (2007). Magnesium-dependent folding of self-splicing RNA: Exploring the link between cooperativity, thermodynamics, and kinetics. Proc. Natl. Acad. Sci. U.S.A. 96, 6149–6154 10. Lilley, D.M. (2004). Analysis of global conformational transitions in ribozymes. Methods Mol. Biol. 252, 77–108 11. Nudler, E. and Mironov, A. S. (2004). The riboswitch control of bacterial metabolism. Trends Biochem. Sci. 29, 11–17 12. Winkler, W. C. (2005). Riboswitches and the role of noncoding RNAs in bacterial metabolic control. Curr. Opin. Chem. Biol. 9, 594–602 13. Mandal, M. and Breaker, R. R. (2004). Gene regulation by riboswitches. Nat. Rev. Mol. Cell. Biol. 5, 451–463 14. Wakeman, C. A., Winkler, W. C., and Dann, C. E., III. (2007). Structural features of metabolite-sensing riboswitches. Trends Biochem. Sci. 32, 415–424 15. Howlett, G. J., Minton, A. P., and Rivas, G. (2006). Analytical ultracentrifugation for the study of protein association and assembly. Curr. Opin. Chem. Biol. 10, 430–436 16. Lebowitz, J., Lewis, M. S., and Schuck, P. (2002). Modern analytical ultracentrifugation in protein science: a tutorial review. Protein Sci. 11, 2067–2079 17. Gerhart, J. C. and Schachman, H. K. (1968). Allosteric interactions in aspartate transcarbamy-

18.

19.

20.

21.

22.

23.

24.

25.

26.

95

lase. II. Evidence for different conformational states of the protein in the presence and absence of specific ligands. Biochemistry 7, 538–552 Constantino, D. and Kieft, J. S. (2005). A preformed compact ribosome-binding domain in the cricket paralysis-like virus IRES RNAs. RNA 11, 332–343 Takamoto, K., He, Q., Morris, S., Chance, M. R., and Brenowitz, M. (2002). Monovalent cations mediate formation of native tertiary structure of the Tetrahymena thermophila ribozyme. Nat. Struct. Mol. Biol. 9, 928–933 Buchmueller,K.L.,Webb,A.E.,Richardson,D.A. and Weeks, K.M. (2000). A collapsed nonnative RNA folding state. Nat. Struct. Biol. 7, 362–366 Dann, C. E., III, Wakeman, C. A., Sieling, C. L., Baker, S. C., Irnov, I., and Winkler, W. C. (2007). Structure and mechanism of a metalsensing regulatory RNA. Cell 130, 878–892 Laue, T. M., Shah, B. D., Ridgeway, R. M., and Pelletier, S. L. (1992). Computer-aided interpretation of analytical sedimentation data for proteins. In: Analytical Ultracentrifugation in Biochemistry and Polymer Science (Harding, S. E., Rowe, A. J., and Horton, J. C., Eds.), The Royal Society of Chemistry, Cambridge, UK Schuck, P. (2000). Size distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and Lamm equation modeling. Biophys. J. 78, 1606–1619 Schuck, P., Perugini, M. A., Gonzales, N. R., Howlett, G. J., and Schubert, D. (2002). Sizedistribution analysis of proteins by analytical ultracentrifugation: strategies and application to model systems. Biophys. J. 82, 1096–1111 Lamm, O. (1929). Die Differentialgleichung der Ultrazentrifugierung, Ark. Mat. Astr. Fys. 21B, 1–4 Cann, J. R. (1994). Computer simulation of the sedimentation of ligand-mediated and kinetically controlled macromolecular interactions. In: Modern Analytical Ultracentrifugation: Acquisition and Interpretation of Data for Biological and Synthetic Polymer Systems (Schuster, T. M. and Laue, T. M., Eds.), Birkhäuser, Boston, MA, pp. 171–188

Chapter 8 Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry Sunny D. Gilbert and Robert T. Batey Summary Isothermal titration calorimetry (ITC) is a biophysical technique that measures the heat evolved or absorbed during a reaction to report the enthalpy, entropy, stoichiometry of binding, and equilibrium association constant. A significant advantage of ITC over other methods is that it can be readily applied to almost any RNA–ligand complex without having to label either molecule and can be performed under a broad range of pH, temperature, and ionic concentrations. During our application of ITC to investigate the thermodynamic details of the interaction of a variety of compounds with the purine riboswitch, we have explored and optimized experimental parameters that yield the most useful and reproducible results for RNAs. In this chapter, we detail this method using the titration of an adenine-binding RNA with 2,6-diaminopurine (DAP) as a practical example. Our insights should be generally applicable to observing the interactions of a broad range of molecules with structured RNAs. Key words: ITC, RNA, Riboswitch, Ligand, Purine, Thermodynamics

1. Introduction The use of isothermal titration calorimetry (ITC) to analyze RNA–ligand interactions has several advantages and challenges. A key advantage is the direct measurement of the heat associated with the binding event, allowing for a determination of enthalpy (DH) and free energy (DG) of binding in a single experiment (1). Consequently, there is no requirement that either molecule be specifically labeled with a fluorescent or radioactive probe; the heat of the reaction is the signal detected by the calorimeter. RNA–ligand interactions are particularly suited for ITC

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_8 © Humana Press, a part of Springer Science + Business Media, LLC 2009

97

98

Gilbert and Batey

analysis because they are often coupled to folding events with large enthalpies (2) that augments the observed signal. Another advantage is that measurements can be performed under a range of temperatures, pHs, and ionic strengths allowing for a detailed exploration of condition space. In addition, ITC easily accommodates diverse ligands and alterations in the RNA sequence within a single study without having to change the basic experimental setup. As a result, the user can obtain a detailed understanding of the nature of the small molecule binding pocket including electrostatic and energetic contributions of direct molecular contacts (3, 4), solvation (5), and, indirectly, induction or restriction of molecular motions (6). Despite these advantages, there are challenges to obtaining reliable data on RNA–ligand interactions using ITC. These issues involve sample preparation, quality and concentration of the RNA, choice of experimental parameters, and data interpretation. Often, the greatest hurdle of this technique is that it requires a large amount of sample; a single experiment typically requires 2.1 mL of 5–10 mM RNA. The concentration of the small molecule (the titrant) in solution is generally 10-fold higher, which can exceed the solubility limit of a hydrophobic molecule that associates with the RNA (the titrate) of interest. For example, one of the natural ligands of the purine riboswitch, guanine, is virtually insoluble in water, presenting difficulties in performing these titrations (7). Working with RNA also requires steps prior to the experiment that ensure the RNA is folded properly. Aggregation, dimerization, or otherwise misfolded RNA will lead to an overestimation of the concentration of the active RNA species, leading to inaccuracies in the data analysis. Finally, all ITC instruments have been contaminated by RNase A at some point, as the titration of 2¢-cytidine monophosphate (2¢CMP) into this protein is the primary standard used by manufacturers and end users to calibrate the instrument. Thus, special cleaning procedures need to be in place to ensure the survival of the RNA during the experiment (see Subheading 3.3.2). In this chapter, we present a detailed step-by-step procedure for implementing ITC as an experimental tool for analyzing RNA–small molecule interactions. As a guide, we present as a standard experiment the titration of 2,6-diaminopurine (DAP) into a 65-nucleotide RNA element derived from the B. subtilis xpt-pbuX guanine riboswitch that bears a point mutation conferring adenine responsiveness (4, 8). For further reading, there are many excellent examples in the primary literature of ways in which RNA has been studied by ITC (reviewed in ref. 9). Among investigations of RNA by ITC are studies of interactions between aminoglycoside antibiotics and ribosomal RNA that revealed a mechanism of translational inhibition (6, 10–12) and a complete characterization of purine recognition by the purine

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry

99

riboswitch (3, 4, 13). ITC has also been successfully applied to studying RNA secondary and tertiary structure stability (14–16), RNA structure ion dependence and binding (17, 18) and ribosomal RNA assembly (19, 20).

2. Materials 2.1. Experimental Considerations

1. Access to MicroCal ITC User’s Manual (http://www. microcal.com).

2.2. Running the Experiment

1. 5.0 U/mL (stock) Taq DNA polymerase (New England Biolabs, Ipswich, MA; Cat. # M0267L).

2.2.1. Making the RNA Sample

2. 10× thermophilic DNA polymerase buffer: 200 mM Tris– HCl of pH 8.8, 100 mM KCl, 100 mM (NH4)2SO4, 20 mM MgSO4, 1% Triton X-100. 3. 100 mM stock concentration of DNA oligonucleotide primers; 5¢-primer sequence: (5¢)GCGCGCGAATTCTAATACGACTCACTATAG; 3¢-primer sequence: (5¢)TGGACATAATCGGACATTTACGGT. 4. pAR7 (stock concentration of 2.5 mg/mL): this plasmid is available from the authors upon request (4). 5. 10 mM dNTP mixture (New England Biolabs, Cat. # N0447L). 6. 10× Transcription buffer: 400 mM Tris–HCl of pH 8.0, 100 mM DTT, 20 mM spermidine, 0.1% Triton X-100. 7. 100 mM rNTP stocks. 8. Inorganic pyrophosphatase (Sigma-Aldrich, St. Louis, MO): 20 U/mL in buffer 20 mM KH2PO4 of pH 7.0, 100 mM NaCl, 50% glycerol (v/v), 10 mM DTT, 0.1 mM EDTA of pH 8.0, 0.2% (w/v) NaN3. 9. T7 RNA polymerase: 0.25 mg/mL working concentration, made in-house or from New England Biolabs. 10. 100% Ethanol. 11. 8 M Urea. 12. 0.5 M Na2EDTA, pH 8.0. 13. 1 M MgCl2. 14. 1 M DTT. 15. 5× TBE buffer: 0.5 M Tris base, 0.42 M boric acid, 5 mM Na2EDTA. 16. Formamide load buffer: 85% formamide, 0.05% (w/v) xylene cyanol, and 0.05% (w/v) bromophenol blue.

100

Gilbert and Batey

17. PAGE solution: 12% acrylamide:bisacrylamide solution (29:1), 8 M urea, 1× TBE. 18. A fluorescent thin layer chromatography (TLC) plate. 19. Electroelution chamber. 20. 10,000 Da molecular weight cutoff concentrators, Centricon YM-10 (Millipore, Billerica, MA). 21. 1× RNA storage buffer: 10 mM K+ HEPES of pH 7.5 and 0.1% NaN3. 2.2.2. Dialysis and Concentration Determination

1. 6–8 kDa Molecular weight cutoff dialysis tubing (Fisher Scientific; Cat. # 08-670B). 2. Magnetic stir bar and plate. 3. ITC buffer: 50 mM K+ HEPES, pH 7.5, 100 mM KCl, 10 mM MgCl2.

2.2.3. Setting Up an Experiment

1. Vacuum manifold (MicroCal, LLC, Northampton, MA). 2. Glass cuvettes, round bottom, 0.7 mL volume (MicroCal, LLC; Cat. # TTB15001C). 3. Plastic cuvettes (MicroCal, LLC; Cat #TTB15002C). 4. Micro magnetic stir bar (VWR Scientific; Cat. #58948-400). 5. Isothermal titration calorimeter (MicroCal, LLC). 6. 10-cc Glass syringe with 8.5-in. 18-gauge needle (for loading cell). 7. 3-cc Plastic syringe with plastic tubing (for loading injector).

2.3. Further Considerations

1. 10 mM Barium chloride (Sigma-Aldrich). 2. 1 mM 18-crown-6 ether (Sigma-Aldrich).

2.3.1. A Useful Standard 2.3.2. Maintaining an RNase-Free ITC

1. ITC cleaning and maintenance attachments as specified in the instruction manual. 2. 10% (v/v) Contrad detergent solution (MicroCal, LLC).

3. Methods 3.1. Experimental Considerations

Successful application of ITC to an investigation of a particular RNA–small molecule interaction depends on several factors. Prior to beginning an ITC study, the issues discussed in this section should be thoroughly considered.

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry 3.1.1. Choosing the Molecules Under Investigation

101

In order to ensure that an accurate thermodynamic analysis can be performed, the solubility and stability of both the RNA and small molecule must be considered. First, the RNA must not be misfolded, aggregated, or multimerized in solution. This is problematic for some RNAs because concentrations greater than 20–100 mM are required to measure low-affinity interactions by ITC. Ideally, the RNA of interest should also be available in sufficient quantities to perform 3–4 successive runs (see Note 1). Another important consideration is that the RNA may have to be modified or engineered to ensure optimal behavior. Our ITC investigations have focused upon the structurally characterized aptamer domain of the B. subtilis xpt-pbuX guanine riboswitch RNA (henceforth this RNA is referred to as “GR”) (7, 8). The crystal structures suggested that purine binding induces a structural rearrangement in GR, a largely exothermic process that can be readily measured by ITC (8). Modifications were made to nonconserved regions of this RNA in order to make it amenable to analysis (see Note 2, Fig. 1). We changed the base-pair identities at the 5¢ end of the P1 helix to engineer an ideal T7 polymerase transcription start site (21). Other avenues for improving the candidacy of an RNA for ITC are choosing a well-behaved homolog (e.g., for a SAM-dependent riboswitch, we typically use a variant from Thermoanaerobacter tengcongensis) (22), engineering nonconserved secondary structures (e.g., mutating a loop region into a stable GAAA tetraloop or changing the base-pairing helix length) (23), and, in the case of large RNAs, isolating well-behaved structured domains (24, 25). A second modification made to GR RNA relates to a fortuitous feature of the purine riboswitch RNA. By changing a single residue, cytidine 74, to a uridine, the RNA’s specificity is switched from guanine to adenine (see Fig. 1) (4, 8, 26). The advantage of this change is that it enables us to use a high affinity and soluble ligand. The natural ligand of GR, guanine, exhibits high affinity (ΔG = −12 kcal/mol; KD = 3 nM) (7, 8) but it is not ideal for study by ITC because the experiment requires ligand concentrations in excess of the guanine solubility at physiological pH and salt concentration (<50 mM); the typical experiment requires the concentration of the ligand be in excess of 100 mM (typically 7–10× the concentration of RNA). The engineered GR(C74U) or GRA (meaning Guanine-Responsive to Adenine switch) variant binds 2,6-diaminopurine (DAP) with an affinity comparable to the GR–guanine interaction (ΔG = −10.7 kcal/mol; KD = 17 nM) (4, 26). DAP is 100-fold more soluble than guanine (~5 mM vs. ~50 mM, respectively) (4). Thus, solubility is an important consideration in choosing both the small molecule and the RNA.

102

Gilbert and Batey

Fig. 1. Structures of the components of the ITC experiment presented in this chapter. The ligand is 2,6-diaminopurine, an analog of adenine (top). The secondary structure of the GRA RNA is shown (bottom), highlighting modifications made to make GRA easier to work with for ITC (sequence changes shown in boxes).

3.1.2. A Useful Calculation to Determine the Amount of Sample Required

There exist several calculations to help estimate the amount of each molecule that is required for an ITC experiment. Primarily, this depends on the affinity of the interaction and the solubility of each sample. Working with a high-affinity interaction (KD = 1–50 nM) lowers the concentrations of the small molecule and the RNA required to obtain a complete titration and an accurate measurement of the enthalpy and equilibrium dissociation constant. Minimizing the amount of RNA required is

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry

103

an important consideration, as each titration requires a large amount of material. 1. Using an estimate of the expected KD (the apparent disassociationbinding constant, which is the inverse of KA, the apparent association constant, that is usually considered by ITC) (see Note 3), the known stoichiometry of the interaction (n, the number of ligands bound to the RNA at saturation), and a proposed concentration of the titrate RNA ([RNA], molar), the quality of data (c) can be calculated (27). c=

n[RNA] . KD

(1)

A c-value between 1 and 1,000 is generally considered to yield the most accurate measurements of the enthalpy and equilibrium dissociationconstant,withthebestresultsobtainedwhencisbetween 10 and 500 (27). The c-value is most readily optimized by adjusting the concentration of the titrate RNA used in the experiment. At the high end of the practical c-value range, situations where binding affinity is subnanomolar, the RNA concentration can be reduced. The need to evolve detectable amounts of heat in the binding interaction establishes the lower limits of the RNA concentration (i.e., sufficient signal to yield interpretable data). This typically means that the highest binding affinities that can be accurately measured have a KD ≥1–10 nM. In situations where KD > 20–100 mM, and increasing the RNA concentration is impractical, experiments with c-values as low as 10−4 can be used with additional optimization (see Note 4) (28). For DAP binding to GRA at an RNA concentration of 5 mM, a KD = 1.7 × 10−8 M, and a stoichiometric equivalent (n) equal to 1, the c-value is calculated to be 294. This will be the condition used for the DAP-GRA RNA titration described in this chapter. 2. The optimal ligand concentration can also be calculated from the c-value and the known volumes of the titrant (ligand) and titrate (RNA), assuming the stoichiometry is equal to 1, using the relationship Rm =

6.4 13 + , c 0.2 c

(2)

where Rm is the ratio of the total titrant to total tirate in the cell after the last injection (mth) (29). In this relationship, the smaller the c-value, the larger the final ratio of titrant to titrate that is required for the reaction to reach equilibrium. Using a c-value of 294, the Rm for the interaction of DAP with GRA is 2.1 (see Note 5). The initial concentration of the small molecule, [X], can then be obtained using the equation [X] =

Rm [RNA]V cell , V injector

(3)

104

Gilbert and Batey

where the corresponding cell and injector volumes (Vcell and Vinjector, respectively) can be found in the dialog window for the VP-ITC. Typically, these values are set at 1.45 and 0.28 mL, respectively, for the MicroCal VP-ITC (see Note 6). Therefore, when [RNA] = 5 mM, the optimal starting DAP concentration is 54 mM or ~10-fold the concentration of the RNA. 3. If the binding affinity cannot be estimated, it is generally suggested that the initial concentration of titrant be between 10- and 30-fold higher than the concentration of the titrate. A good starting experiment would be to use [RNA] = 10 mM and [ligand] = 100 mM. 3.2. Running the Experiment 3.2.1. Making the RNA Sample

RNA in sufficient quantities is made by T7 RNA polymerase in vitro transcription followed by denaturing polyacrylamide gel purification (see Note 7). We include all the reagents and steps necessary for generating GRA as an example of a transcription and purification protocol that yields a sufficient amount of highquality RNA for a series of 10–12 titrations. 1. Construct dsDNA template from oligonucleotide primers by PCR as follows: combine in a 1.5-mL Eppendorf tube 10 mL of each 100 mM 5¢- and 3¢-primers, 3 mL of pAR7, 20 mL of 100 mM dNTP mixture, 100 mL of 10× thermophilic DNA polymerase buffer, 10 mL of Taq DNA polymerase; bring to 1 mL with ddH2O. Aliquot 125 mL of this solution into eight 200-mL thin-walled PCR tubes. 2. Amplify DNA in a thermocycling PCR machine with the program: one cycle of 95°C for 3 min; 30 cycles of 95°C for 30 s, 55°C for 30 s, 72°C for 30 s; and a 3 min final extension at 72°C. Pool the PCR reactions and to verify PCR amplification, run 8 mL of the reaction on a 2% agarose gel containing 10 mg/mL ethidium bromide and image the DNA with shortwave UV light. 3. For in vitro RNA transcription, combine in a polypropylene 50-mL conical tube: 1.25 mL of 10× transcription buffer, 400 mL of 1 M MgCl2, 125 mL of 1 M DTT, 500 mL of each 100 mM NTP, 1 mL PCR reaction, 100 mL inorganic pyrophosphatase, and 0.25 mg/mL T7 RNA polymerase, bringing to a final volume of 12.5 mL with ddH2O. Incubate in a 37°C water bath for 2 h. 4. To precipitate the RNA, add 31.5 mL of 100% ice-cold ethanol and incubate for 1 h at −20°C. Centrifuge at 3,000 × g for 20 min to pellet the RNA. 5. Resuspend the pelleted RNA in 2 mL 8 M urea, 0.75 mL 0.5 M Na2EDTA of pH 8.0, and 0.25 mL formamide load buffer. 6. Purify sample by 12% acrylamide:polyacrylamide (29:1) denaturing gel electrophoresis (run the xylene cyanol 1–2 cm from the bottom of the gel).

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry

105

7. Visualize the RNA in the gel using a shortwave handheld UV lamp and a fluorescent TLC plate. Excise gel fragments corresponding to GRA band (just above the xylene cyanol band) and electroelute in 1× TBE buffer. 8. Collect eluted RNA and exchange three times with 15 mL of RNA storage buffer using Centricon YM-10. Transfer RNA sample to a 1.5-mL Eppendorf tube and store at −20°C. 3.2.2. Dialysis and Concentration Determination

An exact buffer match between the RNA and small molecule samples is critical in an ITC experiment in order to minimize heat associated with dilution of the buffer components. Therefore, the RNA should be exhaustively dialyzed against the sample buffer and the ligand subsequently redissolved in the same buffer immediately prior to the experiment. 1. Each 12.5 mL transcription yields approximately 0.5 mL of 200 mM GRA. Therefore, we simultaneously prepare and purify three 12.5-mL transcriptions for each process. This is enough sample to complete 12 ITC runs performed with 10 mM GRA. 2. To set up sample dialysis apparatus prepare 1 L of ITC buffer in a 1-L glass beaker. 3. Refold RNA by heating to 95°C for 3 min in a heat block. Place sample immediately on ice for 10 min to cool. 4. In dialysis tubing, place enough RNA sample for a single day’s worth of experiments (see Note 8). 5. Place RNase-free magnetic stir bar into buffer solution along with dialysis tubing containing RNA and place at 4°C with moderate stirring overnight. In order to prevent sample degradation, do not dialyze the RNAs for longer than 12–18 h at a time (see Note 9). 6. Following overnight dialysis, carefully extract the RNA from the dialysis tubing. 7. Place two 50-mL aliquots of dialysis buffer in 50-mL conical tubes. One tube will be used to make dilutions of the RNA and small molecule. The other tube will be used to prepare the instrument. 8. Determine the RNA concentration by UV absorbance at A260. The concentration can be calculated from A260 = ebc, where b is the path length of the cell (1 cm in most spectrophotometers), c is the concentration (M) of the RNA (if diluted for the reading, this value includes the dilution factor), and e is the molar extinction coefficient (M−1 cm−1) of the RNA. The e can be obtained by knowing the sequence of the RNA and calculating based on e = #A(15,400 M−1 cm−1) + #C(7,400 M−1 cm−1) + #G(11,500 M−1 cm−1) + #U (8,700 M−1 cm−1) (see Note 10). For GRA, we use a molar extinction coefficient of 657,400 M−1 cm−1.

106

Gilbert and Batey

9. Adjust RNA to desired concentration in 2.1 mL total volume per ITC run. Remeasure the UV absorbance to obtain the exact concentration. 10. Dissolve a sufficient mass of ligand (in solid form) into a volume of dialysis buffer to yield the required titrant concentration. DAP is dissolved initially to a concentration of 0.5 mM and diluted further into ITC buffer to yield 54 mM. In this way, the concentration is assumed to be accurate to the dilution made. Alternatively, if the molar extinction coefficient of the ligand is known, the final concentration can be measured by UV absorbance directly. For DAP at neutral pH the molar extinction coefficient at 282 nm (e282) is 10,000 M−1 cm−1. 3.2.3. Setting Up an Experiment

1. Place ~2.1 mL of titrate RNA sample into the plastic cuvette. Though the actual volume in the sample cell is 1.45 mL, having an excess of sample ensures that no air will be introduced when it is loaded into the sample cell. Add the micro stir bar to the cuvette and place in vacuum manifold. 2. Place ~0.6 mL of titrate small molecule into the glass cuvette and place in the vacuum manifold. 3. Turn magnetic stir speed in vacuum to 300 rpm. Set the temperature of the vacuum to 5°C below the temperature at which the sample will be run in the ITC. The DAP-GRA RNA titration is performed at 30°C; thus, the vacuum is set to 25°C. Apply vacuum to samples for 10 min to ensure complete degassing of the samples. 4. After samples are degassed, carefully load the RNA into the sample cell using the needle/syringe apparatus and the small molecule titrant into the injector (see Note 11). Once the titrant is loaded into the injector, perform two purge refill steps (button in dialog window of the operating software), followed by a 0.05-in. mechanical downward motion of the injector pin (set distance in dialog window) (30).

3.2.4. Starting the ITC

Several parameters must be set by the user prior to each titration, as discussed in the following text.

Volume and Number of Injections

The amount of heat evolved upon each injection is modulated by changing the volume of titrant small molecule that is added into the cell per injection. For a high-affinity interaction, such as DAPGRA, in which the majority of ligand will be bound by the RNA in most injections until saturation is reached, an injection volume of 5–10 mL is used, resulting in ~30–50 injections (see Fig. 2). This value is chosen to increase the number of data points at the inflection point of the binding isotherm. For weaker binding small molecules, injections of 15–25 mL with 10–15 total injections are made to increase the amount of heat evolved per injection.

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry

107

Fig. 2. Model experiment of DAP binding GRA RNA, in which DAP at a concentration of 54 mM is titrated into 5 mM GRA in twenty-nine 10-mL injections. The top graph represents the change in the differential power over time. Each peak below the baseline is integrated to calculate the total heat observed per mole of injectant. This particular experiment did not include either a 2-mL initial injection or a 0.05-in. advancement of the plunger in the injector, yielding the observed the decreased signal of the first injection. This point is removed in the single-site binding fit procedure. Reprinted from (4), with permission from Elsevier.

Injection Time

This is the total time for an individual injection to take place and the system to return to an equilibrium temperature. The reaction steadily returns to equilibrium following the completion of the injection. The default rate of injection is 0.5 mL/s. Therefore, a 25-mL injection requires 50 s to complete. In practice, 3–5 min is allowed for each injection to ensure a complete return to baseline prior to the next injection.

Temperature at Which the Cell Will Be Maintained

It is standard to run experiments above room temperature or 30°C. Prior to loading sample into the cell and syringe, they must be degassed under vacuum. Typically this is done at a temperature 5°C below the titration temperature to avoid long equilibration times on the machine and prevent the sample from overly warming as it is transferred to the instrument. The practical temperature range for an experiment performed at room temperature is 10–50°C. Outside this range, the instrument requires extended periods of

108

Gilbert and Batey

time to equilibrate and may be subject to heat gain or loss to the environment over the course of the experiment. Differential Power

ITC measures the heat evolved or absorbed upon each injection of ligand by recording the power required to maintain the sample cell and a reference cell at the same temperature in an adiabatic system (31). The magnitude of the change in power is reported as a Differential Power (DP), or the difference in the power input between the reference cell and the sample cell to keep them at a certain temperature. The DP can be set anywhere between 1 and 39.5 mcal/s on a standard VP-ITC. A setting of 5 mcal/s is suitable for most reactions, including DAP-GRA binding. For more exothermic reactions, the DP must be increased in order to avoid negative compensation, a situation in which the instrument cannot attenuate the power feedback enough to compensate for the heat of the reaction, leading to a large error in the measurement (1, 31).

3.2.5. Data Analysis

Each data point on the binding isotherm represents a single injection (31). The data are plotted as time (s) versus DP (mcal/s) (see Fig. 2). The exothermic reaction of DAP binding to GRA results in negative peaks on this graph, corresponding to the decrease in power input required in the sample cell to maintain a constant temperature while the ligand is binding RNA. In contrast, an endothermic reaction requires more power input to the sample cell during each injection, resulting in positive peaks on the time versus DP plot. The heat of binding, q, correlates to the integration of each injection peak below a baseline. The baseline is automatically generated in the Origin 5.0 data analysis software when ITC data are read into the program. The data are converted to a binding isotherm by plotting the molar ratio of the ligand and the RNA at each injection point in the reaction versus the heat evolved/ absorbed for that injection (see Fig. 2). Certain parameters in the integration and fitting calculation can be adjusted to reflect the actual measured heat of each injection and lower the error associated with each measurement (see Fig. 3). 1. At certain points in the experiment, erroneous events such as the presence of an air bubble or drift in the baseline result in disruption to single injections or an increase in the noise of each injection over the course of the experiment. Single data points can be eliminated with the “Remove Bad Data Points” function, but this strategy should be used sparingly, generally for no more than 10% of the data points in a given experiment. This function is used most often to remove the first injection that is typically smaller in energy than the following injections if the injector was not slightly advanced prior to the titration (30). This injection is set to 2 mL and thus is much smaller than subsequent injections (see Fig. 3, left).

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry

109

Fig. 3. User-controlled data adjustment parameters in ITC data analysis illustrating how several data analysis parameters in Origin 5.0 can be adjusted to more accurately define the heat of the reaction as well as minimize the heat of dilution in the calculation. Before data adjustment, the error of the KD calculation reflects the less than ideal fit of the single-site binding isotherm (left). By adjusting the baseline and area of integration of each injection to more accurately reflect the actual heat of the reaction apart from the baseline instability, as well as eliminating the first injection, the data points converge on the binding isotherm and the percent error decreases (middle). However, the curve does not accurately portray the data at the far ends of the equilibrium titration. Subtracting the reference data set accounts for the heat of sample dilution that is equal to the heat observed at the end of the titration, and the percent error is further lowered and the binding isotherm more accurately defines the data (right). All graphs are shown for the same titration of DAP into GRA.

2. The boundaries of integration for each peak can be changed to reflect only the injection and binding event by readjusting the baseline and limiting the time range to include only the peak for which the integral is taken. Selecting the “adjust baseline” function in the raw data window brings up a second window in which baseline adjustment is performed manually. This adjustment is particularly useful given that each injection is spaced so that adequate time is allowed for the binding interaction to reach equilibrium. In the default integration, this region is included in the calculation even though it does not correspond to the heat of binding but to the noise of the power input. Similarly, the baseline is generated automatically and often does not fit the peaks correctly, especially if there is drift in the experiment. Exclusion of the regions between each injection after the system returns to equilibrium and before the next injection improves the accuracy of the binding measurement in some cases (see Fig. 3, middle). This should be used sparingly, and never used to cut-off portions of the peak,

110

Gilbert and Batey

as it is possible to overanalyze data points, skewing the values of the obtained thermodynamic parameters. 3. A final data adjustment that must be made is the subtraction of background heat associated with dilution and mixing. Typically, a blank run is performed in which ligand is titrated into buffer alone under the same experimental condition in order to generate a reference data set. Subtracting the reference data from the experimental data generates the final ITC-binding isotherm used to calculate the binding affinity, enthalpy, entropy, and n value of the system (see Fig. 3, right). Several labs have examined the error model for the experimental KA and ΔH measured by calorimetry (DHcal), particularly as they compare to the KA and DH obtained from van’t Hoff analysis (DHvH) of the temperature dependence of the binding affinity measured by other methods. These studies have primarily focused on the enthalpic complexation of Ba2+ and 18-crown-6 ether (1, 32, 33) and address the errors in n, KA, and DH that are calculated in Origin 5.0 in the analysis of a single experiment. In practice, these errors tend to be smaller than the systematic error calculated from statistical sampling. Refinements and improvements in the analysis of ITC data are continuous. However, for the time being a 10% rule of thumb (34) for the error of KA (and thus, KD) and DH in a single experiment has been reliably applied to our data and compares favorably where possible with data obtained using other methods such as in-line probing and spectrofluorescence (26, 35). 3.3. Further Considerations

The extreme sensitivity of ITC to changes in temperature as a result of binding means that this method is also very sensitive to perturbations in the system separate from the binding reaction. Maintaining a regular cleaning and maintenance schedule is of the utmost importance. Measuring RNA binding also requires that the instrument be RNase free. However, many ITCs are not dedicated to analysis of RNA alone. To ensure that our shared instrument does not degrade RNA samples due to crosscontamination from other sources, we employ a series of guidelines to maintain the instrument.

3.3.1. A Useful Standard

The RNase A-2¢CMP reaction is typically used by the manufacturer to calibrate the ITC. Needless to say, this reaction has not ever used as a standard in our instrument since it has been in our possession (see Note 12)! Instead, the BaCl2-18-crown-6 ether binding reaction (36) is used to diagnose electrical malfunction, problems associated with DP feedback loop, and baseline drift. We suggest performing a few of these titrations to familiarize yourself with the instrument as well as establish standard conditions under which to perform this analysis so that problems with the ITC can be determined quickly in the future.

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry 3.3.2. Maintaining an RNase-Free ITC

111

After a user has completed a series of experiments, particularly if they involve protein samples, a 10% (v/v) Contrad-80 solution is added to the sample cell and injector. The temperature is brought to 65°C and maintained for 1 h, followed by extensive flushing with Milli-Q (deionized) water. If an unexpected result is obtained (i.e., no binding or an n value significantly below 1), following the titration experiment, RNA samples should be collected and analyzed on a denaturing PAGE gel to ensure that no degradation has taken place. As an extra precaution against contamination, we keep separate loading syringes and micro stir bars for RNA and protein samples.

4. Notes 1. We were able to reuse GRA samples in successive ITC runs by denaturing the RNA and refolding it in the absence of ligand. This is performed by three 15 mL buffer exchanges into 8 M urea in a 10,000 MWCO centricon, followed by three 15 mL buffer exchanges into RNA storage buffer. RNA is then treated exactly as before in the preparation of RNA for ITC (see Subheading 3.2.2). Alternatively, the GRA–ligand complex could be concentrated and used to set up crystal trays for structure determination. 2. An ideal T7 transcription start site begins with successive guanosine residues (21). The nonconserved 5¢ region of the P1 helix (see Fig. 1, bottom) was thus altered to contain a strong transcription initiation sequence. 3. Methods used to obtain estimates of binding affinity include equilibrium dialysis, gel shift, and in-line probing. 4. Joel Tellinghuisen has analyzed ITC simulations and reactions performed at very low c-values (28, 29, 37). His work suggests optimizing parameters such as injection number (decreasing it down to a lower limit of three or employing the single injection mode available in Origin 7.0) (37) and volume (using variable volumes per injection throughout an experiment to redistribute the heat of binding toward the beginning of the titration) (37), as well as freezing the n value when calculating the binding isotherm in the single-site model (28). 5. The utility of the calculated Rm value in determining sample concentrations is most readily observed in situations where the KD is weaker than 1 mM. In such cases, it is necessary to reach a final equivalence ratio that is greater than two in order to saturate the binding interaction at the end point of the titration (29). For weaker binding interactions it is also useful to maximize the signal per injection by setting the instrument to make ten injections of 25 mL each.

112

Gilbert and Batey

6. These numbers represent approximations and may differ from instrument to instrument. 7. Methods have been developed to purify native RNAs if refolding of the RNA of interest is a concern (38). 8. Depending on the experimental parameters, three to six ITC experiments can be performed in a single day, not including buffer–buffer and sample–buffer control experiments. 9. Certain dialysis steps may take longer due to more complex buffer solutions (for example, glycerol and reducing agents). In this case, multiple buffer changes and longer periods of buffer exchange may be required. 10. Remember that this number reflects the precise molar extinction coefficient of each individual base and does not take into account base stacking, secondary structures, nearest neighbor contributions, and tertiary interactions in the RNA that interfere with the absorbance of each base. The most accurate determination of the RNA concentration is made when the RNA is denatured and subjected to base hydrolysis to completely degrade the RNA into individual nucleotides. In our laboratory we use a “fudge factor” of 1.3 to account for hypochromism, such that if the measured RNA concentration is 1 mM, then the estimated real concentration is 1.3 mM. While not perfect, this allows us to easily obtain concentration estimates closer to the true value. 11. In certain instances, it may be useful for the RNA to be in the injector at high concentration and the small molecule to be in the sample cell at the lower concentration. Because of the insolubility of guanine, our original ITC experiments with the GR involved titrating 50 mM RNA into a 5 mM guanine, yielding the same binding affinity. 12. The previous owners of our instrument, however, routinely used the RNase A-2¢CMP binding reaction as a standard. Despite this, we were able to remove the RNase contamination from the sample cell fairly easily and have not noticed any unusually high degree of sample degradation during our experiments.

Acknowledgments The authors would like to thank Deborah Wuttke and Jonas Fast for useful discussions on optimizing ITC experiments. This work was made possible by a Research Scholar Grant from the American Cancer Society to R.T.B.

Monitoring RNA–Ligand Interactions Using Isothermal Titration Calorimetry

113

References 1. Mizoue, L. S., and Tellinghuisen, J. (2004). Calorimetric vs. van’t Hoff binding enthalpies from isothermal titration calorimetry: Ba2+crown ether complexation. Biophys. Chem. 110, 15–24. 2. Leulliot, N., and Varani, G. (2001). Current topics in RNA-protein recognition: control of specificity and biological function through induced fit and conformational capture. Biochemistry 40, 7947–7956. 3. Gilbert, S. D., Mediatore, S. J., and Batey, R. T. (2006). Modified pyrimidines specifically bind the purine riboswitch. J. Am. Chem. Soc. 128, 14214–14215. 4. Gilbert, S. D., Stoddard, C. D., Wise, S. J., and Batey, R. T. (2006). Thermodynamic and kinetic characterization of ligand binding to the purine riboswitch aptamer domain. J. Mol. Biol. 359, 754–768. 5. Wang, L., Kumar, A., Boykin, D. W., Bailly, C., and Wilson, W. D. (2002). Comparative thermodynamics for monomer and dimer sequence-dependent binding of a heterocyclic dication in the DNA minor groove. J. Mol. Biol. 317, 361–374. 6. Kaul, M., Barbieri, C. M., Srinivasan, A. R., and Pilch, D. S. (2007). Molecular determinants of antibiotic recognition and resistance by aminoglycoside phosphotransferase (3¢)-IIIa: a calorimetric and mutational analysis. J. Mol. Biol. 369, 142–156. 7. Batey, R. T., Gilbert, S. D., and Montange, R. K. (2004). Structure of a natural guanine-responsive riboswitch complexed with the metabolite hypoxanthine. Nature 432, 411–415. 8. Mandal, M., Boese, B., Barrick, J. E., Winkler, W. C., and Breaker, R. R. (2003). Riboswitches control fundamental biochemical pathways in Bacillus subtilis and other bacteria. Cell 113, 577–586. 9. Feig, A. L. (2007). Applications of isothermal titration calorimetry in RNA biochemistry and biophysics. Biopolymers 87, 293–301. 10. Kaul, M., Barbieri, C. M., and Pilch, D. S. (2005). Defining the basis for the specificity of aminoglycoside-rRNA recognition: a comparative study of drug binding to the A sites of Escherichia coli and human rRNA. J. Mol. Biol. 346, 119–134. 11. Kaul, M., and Pilch, D. S. (2002). Thermodynamics of aminoglycoside-rRNA recognition: the binding of neomycin-class aminoglycosides to the A site of 16S rRNA. Biochemistry 41, 7695–7706. 12. Pilch, D. S., Kaul, M., Barbieri, C. M., and Kerrigan, J. E. (2003). Thermodynamics of

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

aminoglycoside-rRNA recognition. Biopolymers 70, 58–79. Gilbert, S. D., Love, C. E., Edwards, A. L., and Batey, R. T. (2007). Mutational analysis of the purine riboswitch aptamer domain. Biochemistry 46, 13297–13309. Bernacchi, S., Freisz, S., Maechling, C., Spiess, B., Marquet, R., Dumas, P., and Ennifar, E. (2007). Aminoglycoside binding to the HIV-1 RNA dimerization initiation site: thermodynamics and effect on the kissing-loop to duplex conversion. Nucleic Acids Res. 35, 7128–7139. Diamond, J. M., Turner, D. H., and Mathews, D. H. (2001). Thermodynamics of three-way multibranch loops in RNA. Biochemistry 40, 6971–6981. Mikulecky, P. J., Takach, J. C., and Feig, A. L. (2004). Entropy-driven folding of an RNA helical junction: an isothermal titration calorimetric analysis of the hammerhead ribozyme. Biochemistry 43, 5870–5881. Hammann, C., Cooper, A., and Lilley, D. M. (2001). Thermodynamics of ion-induced RNA folding in the hammerhead ribozyme: an isothermal titration calorimetric study. Biochemistry 40, 1423–1429. Takach, J. C., Mikulecky, P. J., and Feig, A. L. (2004). Salt-dependent heat capacity changes for RNA duplex formation. J. Am. Chem. Soc. 126, 6530–6531. Recht, M. I., and Williamson, J. R. (2001). Central domain assembly: thermodynamics and kinetics of S6 and S18 binding to an S15–RNA complex. J. Mol. Biol. 313, 35–48. Recht, M. I., and Williamson, J. R. (2004). RNA tertiary structure and cooperative assembly of a large ribonucleoprotein complex. J. Mol. Biol. 344, 395–407. Milligan, J. F., Groebe, D. R., Witherell, G. W., and Uhlenbeck, O. C. (1987). Oligoribonucleotide synthesis using T7 RNA polymerase and synthetic DNA templates. Nucleic Acids Res. 15, 8783–8798. Montange, R. K., and Batey, R. T. (2006). Structure of the S-adenosylmethionine riboswitch regulatory mRNA element. Nature 441, 1172–1175. Gilbert, S. D., Montange, R. K., Stoddard, C. D., and Batey, R. T. (2006). Structural studies of the purine and SAM binding riboswitches. Cold Spring Harb. Symp. Quant. Biol. 71, 259–268. Agalarov, S. C., Sridhar Prasad, G., Funke, P. M., Stout, C. D., and Williamson, J. R. (2000). Structure of the S15,S6,S18-rRNA complex: assembly of the 30S ribosome central domain. Science 288, 107–113.

114

Gilbert and Batey

25. Orr, J. W., Hagerman, P. J., and Williamson, J. R. (1998). Protein and Mg(2+)-induced conformational changes in the S15 binding site of 16S ribosomal RNA. J. Mol. Biol. 275, 453–464. 26. Mandal, M., and Breaker, R. R. (2004). Adenine riboswitches and gene activation by disruption of a transcription terminator. Nat. Struct. Mol. Biol. 11, 29–35. 27. Turnbull, W. B., and Daranas, A. H. (2003). On the value of c: can low affinity systems be studied by isothermal titration calorimetry? J. Am. Chem. Soc. 125, 14859–14866. 28. Tellinghuisen, J. (2008). Isothermal titration calorimetry at very low c. Anal. Biochem. 373, 395–397. 29. Tellinghuisen, J. (2005). Optimizing experimental parameters in isothermal titration calorimetry. J. Phys. Chem. B 109, 20027–20035. 30. Mizoue, L. S., and Tellinghuisen, J. (2004). The role of backlash in the “first injection anomaly” in isothermal titration calorimetry. Anal. Biochem. 326, 125–127. 31. MicroCal, LLC. (2003), Northampton, MA. 32. Briggner, L. E., and Wadso, I. (1991). Test and calibration processes for microcalorimeters, with

33.

34.

35.

36.

37.

38.

special reference to heat conduction instruments used with aqueous systems. J. Biochem. Biophys Methods 22, 101–118. Liu, Y., and Sturtevant, J. M. (1997). Significant discrepancies between van’t Hoff and calorimetric enthalpies. III. Biophys. Chem. 64, 121–126. Tellinghuisen, J. (2005). Statistical error in isothermal titration calorimetry: variance function estimation from generalized least squares. Anal. Biochem. 343, 106–115. Lemay, J. F., and Lafontaine, D. A. (2007). Core requirements of the adenine riboswitch aptamer for ligand binding. RNA 13, 339–350. Wadso, I., and Goldberg, R. N. (2001). Standards in isothermal titration calorimetry. Pure Appl. Chem. 73, 1625–1639. Tellinghuisen, J. (2007). Optimizing experimental parameters in isothermal titration calorimetry: variable volume procedures. J. Phys. Chem. B 111, 11531–11537. Batey, R. T., and Kieft, J. S. (2007). Improved native affinity purification of RNA. RNA 13, 1384–1389.

Chapter 9 Preparation and Crystallization of Riboswitch–Ligand Complexes Olga Pikovskaya, Artem A. Serganov, Ann Polonskaia, Alexander Serganov, and Dinshaw J. Patel Summary Riboswitches are mRNA regions that regulate the expression of genes in response to various cellular metabolites. These RNA sequences, typically situated in the untranslated regions of mRNAs, possess complex structures that dictate highly specific binding to certain ligands, such as nucleobases, coenzymes, amino acids, and sugars, without protein assistance. Depending on the presence of the ligand, metabolite-binding domains of riboswitches can adopt two alternative conformations, which define the conformations of the adjacent sequences involved in the regulation of gene expression. In order to understand in detail the nature of riboswitch–ligand interactions and the molecular basis of riboswitch-based gene expression control, it is necessary to determine the three-dimensional structures of riboswitch–ligand complexes. This chapter outlines the techniques that are employed to prepare riboswitch–ligand complexes for structure determination using X-ray crystallography. The chapter describes the principles of construct design, in vitro transcription, RNA purification, complex formation, and crystallization screening utilized during the successful crystallization of several riboswitches. Key words: Purine riboswitch, preQ1 riboswitch, RNA secondary structure, Crystallization, NMR spectroscopy

1. Introduction Riboswitches are one of the latest additions to the growing field of RNA-based gene expression control systems (1–3). In contrast to many regulatory circuits, feedback gene regulation by riboswitches is primarily RNA dependent; in particular, riboswitches can sense and selectively bind ligands without the assistance of proteins, thereby directing the expression of the genes involved in the synthesis and transport of these ligands and related compounds. Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_9 © Humana Press, a part of Springer Science + Business Media, LLC 2009

115

116

Pikovskaya et al.

Riboswitches are typically situated in the 5¢ untranslated regions of mRNAs and consist of evolutionarily conserved sensing domains, which selectively bind specific ligands, and variable expression platforms, which contain RNA elements that control gene expression. Riboswitches demonstrate highly specific binding to their cognate ligands, which range from rather simple molecules, such as amino acids (lysine and glycine) and nucleobases (adenine, guanine, and pre-queuosine1 or preQ1), to larger and chemically more complex coenzymes (S-adenosylmethionine or SAM, thiamine pyrophosphate or TPP, adocobalamin, and flavin mononucleotide) (for review see ref. 4). The binding of a riboswitch to its target ligand stabilizes the metabolite-bound conformation of the sensing domain that, in turn, triggers the adoption of a particular conformation of the expression platform. In the absence of the ligand, a segment of the sensing domain engages in the formation of an alternative conformation, thereby changing the folding of the expression platform. The sequence of the expression platform determines the mechanism of gene regulation, which, in most cases, can take place either at the level of transcription or translation. Transcriptional control is dictated by the formation of a terminator or antiterminator hairpin, while translational regulation is directed by the formation or disruption of a hairpin bearing translation initiation signals. However, the primary event, specific recognition of the ligand by riboswitch RNA, and the subsequent step, stabilization of the ligand-bound riboswitch conformation, are crucial for deciding whether gene expression should be turned on or off. Therefore, the three-dimensional structure of the ligandbound sensing domain should provide details critical for understanding the riboswitch mechanism at the molecular level. Recent studies have also uncovered another interesting feature of riboswitches, their ability to bind metabolite-like antibiotics (5–7). These studies, together with the identification of the antibiotic resistance mutations within riboswitch sequences (for review see ref. 5), suggest that riboswitches, along with other cellular targets (8), may be involved in the mechanism of antibiotic action. The finding of riboswitches that control essential genes in pathogenic bacteria (5) and the likely absence of riboswitches in humans prompts the search for other metabolite-like molecules capable of recognizing riboswitches and suppressing bacterial growth. The three-dimensional structures of riboswitches would undoubtedly help in the design of such metabolite analogs and would aid in the further evaluation of riboswitches as potential drug targets. Since riboswitches that recognize different compounds do not bear much similarity, the ligand-binding principles uncovered in one riboswitch structure can hardly be applied to understand the rules governing ligand recognition by other riboswitch types. Therefore, it is necessary to determine the structures of the metabolite-sensing domains of different riboswitches. The fact that such

Preparation and Crystallization of Riboswitch–Ligand Complexes

117

structural studies require significant quantities of riboswitch– ligand complexes for experimentation demands a universal procedure to obtain sufficient yields of highly pure RNA. Since the length of most of the metabolite-sensing domains exceeds 60 nucleotides, structure determination by Nuclear Magnetic Resonance spectroscopy is rendered extremely challenging. Therefore, this chapter is primarily focused on the protocols necessary to design and prepare riboswitches for tertiary structure solution using X-ray crystallography. Nevertheless, the procedures described herein can be successfully utilized for any other application that requires large quantities of highly pure RNA.

2. Materials 2.1. Construct Preparation 2.1.1. Oligonucleotide Synthesis and Purification

1. CPG 2000 columns, 0.2 µM scale (Glen Research, Sterling, VA). 2. DNA synthesizer (Applied Biosystem, Foster City, CA). 3. 8-mL Glass vials (Wheaton Science Products, Millville, NJ). 4. ~30% Ammonium hydroxide. 5. Nylon syringe filters: 0.2-µm pore size, 25-mm diameter (Fisher Scientific, Pittsburgh, PA).

2.1.2. Polyacrylamide Gel Electrophoresis

1. RNase ERASE (MP Biomedicals, USA). 2. Polyacrylamide gel solution: 10–15% acrylamide/bis-acrylamide solution 29:1 (w/w) (Bio-Rad Laboratories, Hercules, CA). Store in a dark place. 3. 8 M Urea solution (Sigma-Aldrich, St. Louis, MO). Store in a dark place. 4. 1× Tris/borate/EDTA (TBE) buffer: 89 mM Tris base, 89 mM boric acid, 2 mM ethylenediaminetetraacetic acid (EDTA). 5. Denaturing loading buffer: 1× TBE, 8 M urea, spatula tip of xylene cyanol and bromophenol blue. 6. 10% Ammonium persulfate (APS) solution. Store at +4°C in a dark place for 2–4 weeks. 7. Tetramethylethylenediamine (TEMED). 8. Thin layer chromatography (TLC) plates, 20 × 20 cm (Selecto Scientific, Suwanee, Georgia). 9. Elutrap systems, each with four elutrap chambers, assembled with BT1 and BT2 membranes (Schleicher & Schuell, Keene, NH). 10. 20 mM Tris–acetate, pH 8.0. 11. Ethanol precipitation reagents: 3 M sodium acetate, pH 5.2, ethanol, 80% (v/v) ethanol solution.

118

Pikovskaya et al.

2.1.3. Cloning

1. 0.1 M MgCl2. 2. StuI, 20,000 U/mL, and HindIII, 20,000 U/mL, restriction enzymes (New England Biolabs, Ipswich, MA), supplied with 10× NE buffer 2 (NEB 2): 10 mM Tris–HCl, pH 7.9, 10 mM MgCl2, 50 mM NaCl. 3. 10× Ligation buffer: 500 mM Tris–HCl, pH 7.5, 100 mM MgCl2, 100 mM dithiothreitol, 10 mM ATP. 4. T4 DNA ligase (1,000 U/mL) (New England Biolabs). 5. Escherichia coli XL1-Blue and DH5a competent cells (Invitrogen, Carlsbad, CA). 6. 100 mg/mL ampicillin. Dissolve in 50% ethanol and store at −20°C. 7. Miniprep and Gigaprep plasmid DNA purification kits (QIAGEN, Valencia, CA). 8. LB/Amp medium: 10 g tryptone, 5 g yeast extract, 10 g NaCl. Dissolve in 1 L of water, autoclave for 30 min, and add ampicillin to 100 mg/mL.

2.2. RNA Preparation

1. 1 M Tris–HCl, pH 8.0.

2.2.1. DNA Template Preparation

2. Phenol/chloroform mixture (1:1): phenol, pH 8.0 (ACROS Organics, Geel, Belgium), chloroform (Fisher Scientific).

2.2.2. In Vitro Transcription

1. Transcription mixture components: 1 M K-HEPES, pH 7.9, 1 M DTT, 250 mM spermidine, 100 mM of each ribonucleotide triphosphate (ATP, GTP, CTP, and UTP), 2 M MgCl2, and T7 RNA polymerase (10 U/mL). 2. Amicon cell and ultrafiltration membranes with a molecular weight cutoff of 3,000 Da (Millipore, Billerica, MA).

2.2.3. RNA Chromatography

1. Mono Q column connected to the AKTA system (Amersham, Piscataway, NJ). 2. 0.5 M NaOH, 10% acetic acid. 3. Buffer A: 20 mM Tris–HCl, pH 7.0. 4. Buffer B: 20 mM Tris–HCl, pH 7.0, 1 M NaCl.

2.3. Complex Formation

1. NMR buffer: 10% D2O, 50 mM deuterated potassium acetate, pH 6.7 (Cambridge Isotope Laboratories, Andover, MA). 2. NMR tubes (Shigemi, Tokyo, Japan), glass Pasteur pipettes.

2.4. Initial Crystallization Screening

1. Commercial sparse matrix kits: Crystal Screen, Crystal Screen 2, and Natrix from Hampton Research (Aliso Viejo, CA), and kits from QIAGEN (Valencia, CA), Emerald Biosystems (Bainbridge, WA), and Axygen Scientific (Union City, CA). 2. Pregreased crystallization plates with 24 wells (Hampton Research, Aliso Viejo, CA).

Preparation and Crystallization of Riboswitch–Ligand Complexes

119

3. 22 mm Siliconized glass circle cover slides (Hampton Research). 4. MicroDuster III: compressed gas can (VWR). 5. Stereomicroscope 15:1 zoom range with polarizing filters (e.g., Nikon SMZ1500).

3. Methods 3.1. Construct Preparation 3.1.1. Construct Design

The choice of the riboswitch sequence is critical for the success of structural studies. Although in all cases it is necessary to test riboswitches from several organisms, we usually limit ourselves to 3–5 typical riboswitches, which contain most of the evolutionarily conserved nucleotides and the shortest predicted helices with the least number of bulges and other irregularities. Most often, such riboswitches can be found in thermophilic organisms. To make sure that the selected riboswitch sequence does not produce multiple folds, we verify the formation of the secondary structure in silico using Mfold, a DNA- and RNA-folding algorithm developed by M. Zuker and coworkers (9). To prepare RNA for structural studies, we employ in vitro transcription using bacteriophage T7 RNA polymerase. T7 RNA polymerase, however, can add extra nucleotides to the 3¢ end of the transcript. In many riboswitches, the 3¢ end is paired with the 5¢ end, producing a helix, which can form stacking interactions with other molecules in the crystal lattice. Since heterogeneity at the 3¢ end may interfere with the potential formation of crystal contact(s), a self-cleaving hammerhead ribozyme is joined to the 3¢ end of the riboswitch sequence (10). Examples of sensing domain constructs and a schematic of a typical hammerhead ribozyme are shown in Fig. 1a, b, respectively. T7 RNA polymerase requires a ~17-bp double-stranded promoter for transcription initiation. Since the template sequence need not be double-stranded, it is easiest to obtain the DNA template by hybridizing two oligonucleotides: the long one, which contains the promoter and the riboswitch-ribozyme sequence, and the short one, complementary to the promoter region. In our hands, transcription efficiency varies inversely with the length of the single-stranded sequence; therefore, we utilize such templates exclusively for the transcription of very short RNA species. The transcription yields for longer RNAs can be increased using double-stranded DNAs as templates. These templates can be prepared by PCR or by annealing two long complementary DNA oligonucleotides encoding both the promoter region and the sequence of interest. Since the chemical synthesis of long DNA oligonucleotides and preparative PCR is not practical for largescale transcription, we, therefore, clone the double-stranded tem-

120

Pikovskaya et al.

a U A U

U GG

A G

U C

U

U

U

A

A

C CC G C G GG C A U G C C U G U A A A A

G C A C G C A A G U GC G C U A

U A U A C U A G G 5’

U A U G G U C 3’

U A U

GG

A G

U C

U

U

A

A

C CC U C A A G A G U G U U C U C A A A A

G U U U G G G A U A A U C C U A

C

xpt G-riboswitch

U A U A C U U C G G 5’

U A U G A A G U C 3’

U

add A-riboswitch

b cleavage site 5’

...G G A C U

A A U G C C A G C G U A G GG A A GU UCG A C GG A U C 3’

c

U C GA A U C CC U U CA A C U GCC U A G A CU GA A U G AG C G HindIII A U G C G C A G G U

U A

StuI

T7 promoter

Hind III Transcription Start

TAATACGACTCACTAT AGGCCT ACCAGGGGATCCTCTAGAGTCGACCTGCAGGCATGC AAGCTT

M13 Forward Primer

M13 Reverse Primer

pUT7

Ampr

Fig. 1. Design of riboswitch constructs for crystallization. (a) Sequences of the sensing domains of purine-sensing mRNAs: Bacillus subtilis 68-mer xpt G-riboswitch (left ) and Vibrio vulnificus 71-mer add A-riboswitch (right ). (b) An example of RNA obtained after transcription from the HindIII-linearized pUT7 plasmid. A hammerhead ribozyme is joined to the 3¢ end of the sensing domain and its cleavage site is marked. Five of the six bases comprising the former HindIII site are boxed in the RNA sequence. (c) The pUT7 plasmid used for cloning and transcription of the riboswitch. The Stu I and Hind III restriction sites are boxed and the T7 promoter is indicated. M13 forward and reverse primers can be used to sequence the plasmid.

Preparation and Crystallization of Riboswitch–Ligand Complexes

121

plate into a plasmid under the control of a T7 promoter and amplify plasmid DNA in Escherichia coli culture. The riboswitchribozyme fragment for cloning is produced by annealing two chemically synthesized complementary oligonucleotides, resulting in a blunt end on the 5¢ end of the riboswitch sequence and half of the HindIII restriction site on the 3¢ end of the ribozyme sequence. The double-stranded fragment is then ligated at the StuI and HindIII restriction sites of the pUT7 plasmid (Fig. 1c) (11), and after amplification and linearization with the HindIII enzyme, the template can be used for the transcription of milligram amounts of RNA. 3.1.2. Oligonucleotide Synthesis and Purification

1. Synthesize complementary oligonucleotides in a DNA synthesizer (see Note 1) using 0.2-µm scale CPG 2000 columns (see Note 2). 2. Dry the columns in the lyophilizer and transfer the resin from the columns into 8-mL glass vials. Add 2 mL of ~30% ammonium hydroxide to each vial, close the vials tightly, seal them with parafilm, and deprotect the oligonucleotides at 55°C overnight. 3. Chill the vials and filter the contents of the vials using syringes and 0.2-µm, 25-mm nylon syringe filters into 50-mL conical tubes. Dry the samples by blowing air into the tubes under the chemical hood and dissolve the dried oligonucleotides in 1–3 mL of denaturing loading buffer in preparation for large-scale gel purification.

3.1.3. Large-Scale PAGE for Oligonucleotides

1. These instructions pertain to the use of a custom-made largescale two-chamber gel system, which is similar to other commercial gel systems, such as Life Systems Model S2. It is essential that the glass plates for the gels are scrubbed clean with a rinsable detergent, such as RNase ERASE, rinsed thoroughly with distilled water, and sprayed with ethanol. 2. Take a large glass plate (40 cm in length, 36 cm in width), position three spacers (0.4-cm thick) along the sides and bottom, place another glass plate on top, and clamp the plates together. Prepare 700 mL of 10% gel (see Note 3). Mix 50 mL of gel with 300 µL of APS solution and 30 µL of TEMED, pour the gel, and allow the gel to polymerize with the top slightly elevated. Mix the remaining 650 mL of gel with 5 mL of APS solution and 300 µL of TEMED and pour it. Insert and clamp the combs (as many as six oligonucleotides can be purified on one gel) into the gel and allow the gel to polymerize for 45 min. 3. Clamp a metal plate against the surface of one glass plate to ensure the even distribution of heat and preheat the gel in

122

Pikovskaya et al.

1× TBE buffer at 55 W for ~60 min. Load the oligonucleotides and run the gel for ~15 h. 4. Place the gel assembly on the bench, remove the top glass plate, wrap the gel and the remaining glass plate with Saran Wrap, and flip and place them onto fluorescent TLC plates. Remove the glass plate. Use an ultraviolet lamp with l = 365 nm to locate the dark bands of DNA and cut them out of the gel (see Note 4). The gel slices can be stored at −20°C. 5. Assemble an elutrap chamber by placing a BT1 trapezoidal impermeable membrane at each end of the chamber between 2 U-shaped inserts (so that the higher corner faces the arrow on the chamber). Place a BT2 rectangular semipermeable membrane between 2 U-shaped inserts closer to the middle of the chamber, dividing the chamber into two sections: a larger one for gel slices and a smaller one for eluate collection. Tighten both ends of the chambers with a piece of plastic, place the gel slices into the chamber, and place the chamber into a tray in the elutrap system that can accommodate up to four chambers. Add 20 mM Tris–acetate buffer to the chambers (enough to cover the gel) and to the system. Perform electrophoresis for each system with four chambers in the cold room at no more than 16–18 W to prevent overheating. 6. Collect eluate every 2–3 h and estimate the DNA concentration using UV spectroscopy at l = 260 nm. If necessary, continue electrophoresis overnight at 70 V. Precipitate the eluted oligonucleotides with ethanol in 50-mL conical tubes: add one-tenth volume of 3 M sodium acetate and 3.3 volumes of 100% ethanol to each oligonucleotide solution, chill the mixture at −20°C overnight, and centrifuge it for 60 min at >4,000 × g. Discard the supernatant, wash the pellet with ~10 mL of 80% ethanol, and centrifuge for another 15 min. Air dry the pellet and dissolve it in water. 3.1.4. Cloning

1. Prepare the double-stranded DNA fragment to be cloned by annealing complementary oligonucleotides in a total volume of 10 µL. Mix the oligonucleotides (final concentration of each oligonucleotide is 50–150 pM/µL), add 3 mM MgCl2, incubate the mixture at 95°C for 3 min, and chill it on ice. Verify the formation of the DNA duplex by electrophoresis on a 2.5% agarose gel. 2. Prepare the vector DNA by digesting the pUT7 plasmid with the StuI and HindIII restriction enzymes in NEB 2. Purify the vector using 1% agarose gel electrophoresis. 3. Ligate 1 µL of fragment (5–50 pM/µL) with 100 ng of the vector in 10 µL mixture supplemented with T4 DNA ligase buffer and 1 U of T4 DNA ligase at 14°C overnight.

Preparation and Crystallization of Riboswitch–Ligand Complexes

123

4. Transform 4 µL of the ligation mixture into E. coli strain XL1Blue and select the clones on LB plates supplemented with 100 µg/mL ampicillin. 5. Grow 3-mL bacterial cultures in LB/ampicillin medium and purify plasmid DNA using a Miniprep kit. The inserts can be identified by a comparison of the EcoRI- and HindIIIdigested ligated and control plasmids on a 2.5% agarose gel. The clones with inserts of the correct size must be verified by DNA sequencing. 3.2. RNA Preparation 3.2.1. DNA Template Preparation

1. For large-scale plasmid DNA purification, grow bacteria in 3 L of LB/ampicillin medium at 37°C overnight (see Note 5) and purify plasmids using a Gigaprep kit. Expect 5–15 mg of plasmid DNA. 2. Prepare the DNA template for in vitro transcription by restricting 5 mg of plasmids with 1,000 U of HindIII in 5 mL of NEB 2 buffer supplemented with 0.25 mL of 1 M Tris–HCl, pH 8.0, at 37°C for ~4 h. Verify the completion of the reaction using 1% agarose gel electrophoresis. 3. Perform phenol/chloroform extraction by shaking the restricted DNA in the presence of 2.5 mL of phenol/chloroform mixture for several minutes in 50-mL conical tubes (see Note 6) and then centrifuging at 4,000 × g for 5 min. Carefully collect the upper aqueous phase and repeat the extraction with chloroform alone. Precipitate the plasmid DNA with ethanol, as described in Subheading 3.1.3, step 6.

3.2.2. In Vitro Transcription

1. Prior to large-scale in vitro transcription, perform small-scale transcription. A typical reaction is done in a 100-µL mixture of 100 mM K-HEPES buffer, pH 7.9, 30 mM DTT, 2 mM spermidine, 4 mM of each ribonucleotide triphosphate, 6 µg of DNA template, 15 mM MgCl2, and 2 units of T7 RNA polymerase (see Note 7). The mixture should be preheated at 37°C prior to the addition of DNA template and polymerase. The transcription of each DNA template should be routinely tested in various concentrations of MgCl2 (10–20 mM), DNA template (40–100 µg/mL), and polymerase (10–200 U/mL). Incubate the reactions at 37°C for ~2–3 h; they should get cloudy from the precipitation of magnesium pyrophosphate. Adjust the final concentration of MgCl2 to 30 mM and incubate the reactions for another 30 min to complete ribozyme cleavage from the RNA. Verify the transcription yield and the efficiency of ribozyme cleavage using a 15% small-scale PAGE. 2. Using the optimal conditions found during analytical transcription, perform large-scale in vitro transcription in a total volume of 50 mL for ~4 h at 37°C. Incubate for another 45 min after

124

Pikovskaya et al.

adjusting the MgCl2 concentration to ensure complete ribozyme cleavage and store the reactions at −20°C. For a more efficient reaction, the reagents and water should be preheated. To avoid polymerase aggregation, polymerase should be added drop by drop while swirling the mixture. 3. Thaw the transcription mixture and centrifuge at 4,000 × g for ~20 min. Isolate the supernatant and repeat the previous step one more time before filtering the mixture through 0.45µm and 0.22-µm disposable filtration systems (see Note 8). Immediately concentrate the mixture using a 50-mL Amicon cell with an ultrafiltration membrane to a total volume of ~3–5 mL at 4°C. Mix the RNA with an equal volume of the loading buffer and purify it on 2–4 large polyacrylamide gels, followed by electroelution and ethanol precipitation, as described in Subheading 3.1.3. 3.2.3. RNA Chromatography

1. Dissolve RNA in ~5 mL of water and filter it through a 0.2-µm, 25-mm nylon syringe filter to remove residual acrylamide particles. 2. To remove urea and other contaminants, purify the RNA using anion exchange chromatography on a Mono Q column. We routinely clean the column before loading the sample by injecting ~2 mL of 0.5 M NaOH followed by ~2 mL of 10% acetic acid into the column and then washing it with a step gradient of 100% buffer B. Once the column is clean, load no more than 5–8 mg of RNA per run. After loading the RNA, wash the column with 5 column volumes (CV) of buffer A and elute RNA using a 10 CV gradient of 0–100% buffer B. Typically, only half of the RNA can be eluted after the first gradient; therefore, repeat elution several times using a step gradient of 100% buffer B (see Note 9). 3. Precipitate eluted RNA with ethanol, as described in Subheading 3.1.3. Dry the pellet in the lyophilyzer for 5 min, dissolve it in ~0.5 mL of water, and determine the RNA concentration.

3.3. Complex Formation

Riboswitch–ligand complexes can be prepared by mixing RNA with the ligands in a one-to-one ratio at various conditions. The formation of the complexes can be monitored by different techniques, including gel-shift assay and Nuclear Magnetic Resonance (NMR) Spectroscopy. Since it is difficult to discriminate between unbound RNA and RNA–ligand complexes using a gel-shift assay, the use of NMR spectroscopy is more preferable. 1. Prepare a 250-µL NMR sample of 0.2 mM RNA in NMR buffer. Using an RNase-free glass Pasteur pipette, place each sample into an RNase-free NMR tube and carefully insert a plunger into the tube while avoiding the formation of air bubbles. Secure the tip of the plunger and NMR tube with Parafilm. 2. Record the NMR spectrum of the free RNA sample using jump-and-return water suppression for detection. RNA base

Preparation and Crystallization of Riboswitch–Ligand Complexes

125

pairing is typically characterized by peaks in the region from 9 to 14 ppm. 3. To prepare a riboswitch–ligand complex, mix the free RNA NMR sample with the desired ligand and repeat the NMR experiment. In most cases, binding affects RNA base pairing, which is characterized by changes in the NMR spectra. To precisely achieve one-to-one binding within the complex, the RNA may be titrated with ligand in increments of 10–25% until spectra changes are no longer observed. 3.4. Initial Crystallization Screening

Riboswitch crystals are grown by the vapor diffusion method in the hanging drop format. To set crystallization drops manually, combine the sample with the reservoir solution, obtained from commercial sparse matrix kits, in a one-to-one ratio (see Note 10). To increase the chances of successful crystallization, hanging drops are typically set at two concentrations (0.1–0.3 and 0.4–1.0 mM) and two temperatures (+4 and +20°C). 1. To employ the hanging drop format, add 1 mL of reservoir solution to each well in the 24-well crystallization plate. For the first screening, we typically use Crystal Screen, Crystal Screen 2, and Natrix kits. 2. Take a circular glass slide and blow away dust particles from the surface using the Microduster. 3. Place a 1-µL drop of reservoir solution on the glass slide. Add 1 µL of RNA–metabolite complex sample to the reservoir solution drop and mix without expelling any air bubbles into the drop. Up to two additional drops with different sample concentrations can be prepared on the same glass slide. 4. Flip the glass slide and position it over the well. Gently press down on the glass slide to make sure that the edges of the glass slide are fully in contact with the grease. 5. Repeat steps 2–4 for each of the 24 wells in the crystallization plate and place the plate in the incubator at +20°C. Set another identical crystallization plate in the cold room (see Note 11). 6. Allow the drops in the plates to equilibrate for at least 1 day before checking them. Check the plates using a stereomicroscope at a magnification range of ~5–10×. The use of polarizing filters is highly recommended for locating small crystals. A given drop can remain unchanged or it can contain crystals, a phase separation, and/or precipitation after initial crystallization screening (Fig. 2a–c). 7. If a drop contains crystals or crystal-like material, optimize the crystallization conditions by, for instance, varying the ratio between the reservoir and sample solutions, the concentration of the precipitant and RNA-metabolite complex, and the pH of the buffer (Fig. 2d–f).

126

Pikovskaya et al.

Fig. 2. Crystal formation of ligand–riboswitch complexes after initial crystallization screening and optimization. (a) Microcrystals of the preQ1 riboswitch surrounded by heavy precipitate. (b) Crystals of the adenine riboswitch grown from a phase separation. (c) Needle-like crystals of the preQ1 riboswitch. (d) Macrocrystals of the TPP riboswitch surrounded by precipitate. (e) Crystal conglomerate of the TPP riboswitch. (f) Macrocrystals of the guanine riboswitch.

4. Notes 1. Oligonucleotides up to 220 nt can be synthesized using an ABI synthesizer. Alternatively, purified oligonucleotides can be purchased from different companies.

Preparation and Crystallization of Riboswitch–Ligand Complexes

127

2. 0.2-mm Scale columns were found most reliable for the synthesis of long oligonucleotides. 3. Use an 8% gel for oligonucleotides over 150 nt and a 12.5% gel for oligonucleotides up to 80 nt. 4. Any UV lamp can be used for the “UV shadow” technique. To prevent DNA damage, minimize the exposure time and cover the bands that are not immediately cut with aluminum foil. 5. To achieve better yields in large-scale DNA purification, it is preferable to transform the desired plasmid into E. coli DH5α cells. 6. Make sure that the tubes withstand centrifugation in the presence of the phenol/chloroform mixture. 7. T7 RNA polymerase can be purified from E. coli using the plasmid described in (12). 8. Ethanol precipitation of the transcription mixture at this step may lead to the loss of RNA due to the formation of partially insoluble precipitate. The removal of magnesium pyrophosphate can be achieved by pyrophosphatase treatment. 9. Some RNAs can be more efficiently eluted if 0.5 mL of either 10% acetic acid or 0.5 M NaOH is injected into the column prior to elution with the step gradient. To ensure that the eluted RNA sample has a neutral pH, verify pH using litmus paper; adjust pH with acetic acid, if necessary. 10. Hanging and sitting drops can also be prepared using crystallization robots, such as Mosquito (TTP LabTech, UK), which allow much smaller drops (0.2–0.3 µL) to be set in a 96-well plate format. 11. As an alternative, it is possible to set the “+4°C plate” at room temperature if the plate is kept on ice and if prechilled ice packs are placed on the glass slides with prepared drops until the entire tray is completed.

Acknowledgment This research was supported by NIH GM073618.

References 1. Breaker, R. R. (2006). Riboswitches and the RNA World, in The RNA World (Gesteland, R. F., Cech, T. R., and Atkins, J. F., Eds.) pp. 89–108, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

2. Nudler, E., and Mironov, A. S. (2004). The riboswitch control of bacterial metabolism. Trends Biochem. Sci. 29, 11–17. 3. Winkler, W. C., and Breaker, R. R. (2005). Regulation of bacterial gene expression by

128

4.

5.

6.

7.

8.

Pikovskaya et al. riboswitches. Annu. Rev. Microbiol. 59, 487–517. Serganov, A., and Patel, D. J. (2007). Ribozymes, riboswitches and beyond: regulation of gene expression without proteins. Nat. Rev. Genet. 8, 776–790. Blount, K. F., and Breaker, R. R. (2006). Riboswitches as antibacterial drug targets. Nat. Biotechnol. 24, 1558–1564. Blount, K. F., Wang, J. X., Lim, J., Sudarsan, N., and Breaker, R. R. (2007). Antibacterial lysine analogs that target lysine riboswitches. Nat. Chem. Biol. 3, 44–49. Sudarsan, N., Cohen-Chalamish, S., Nakamura, S., Emilsson, G. M., and Breaker, R. R. (2005). Thiamine pyrophosphate riboswitches are targets for the antimicrobial compound pyrithiamine. Chem. Biol. 12, 1325–1335. Ataide, S. F., Wilson, S. N., Dang, S., Rogers, T. E., Roy, B., Banerjee, R., Henkin, T. M., and Ibba, M. (2007). Mechanisms of resistance to

9.

10.

11.

12.

an amino acid antibiotic that targets translation. ACS Chem. Biol. 2, 819–827. Zuker, M. (2003). Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 31, 3406–3415. Price, S. R., Ito, N., Oubridge, C., Avis, J. M., and Nagai, K. (1995). Crystallization of RNA–protein complexes. I. Methods for the large-scale preparation of RNA suitable for crystallographic studies. J. Mol. Biol. 249, 398–408. Serganov, A., Rak, A., Garber, M., Reinbolt, J., Ehresmann, B., Ehresmann, C., GrunbergManago, M., and Portier, C. (1997). Ribosomal protein S15 from Thermus thermophilus – cloning, sequencing, overexpression of the gene and RNA-binding properties of the protein. Eur. J. Biochem. 246, 291–300. Studier, F. W., Rosenberg, A. H., Dunn, J. J., and Dubendorff, J. W. (1990). Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185, 60–89.

Chapter 10 Crystallization of the glmS Ribozyme-Riboswitch Daniel J. Klein and Adrian R. Ferré-D’Amaré Summary Procedures that were critical for crystallization of the glmS ribozyme-riboswitch RNA domain from the thermophilic Gram-positive bacterium Thermoanaerobacter tengcongensis are described. Experimental design based on screening multiple variant RNA sequences and techniques used to identify initial crystallization conditions were similar to those employed for most RNAs. However, serendipitous in-drop digestion of one RNA construct at a specific internucleotide linkage was crucial for the growth of high-quality glmS ribozyme crystals. Biochemical analysis of crystalline RNA identified the site of scission and guided design of an optimized bimolecular RNA construct. Finally, modifications of ionic strength and pH of solutions used for stabilization of the crystals were essential for optimal diffraction and binding of the activator glucosamine-6-phosphate, respectively. Although their details are specific to the glmS ribozyme, these general strategies may be useful for analyzing and improving crystals of other RNAs. Key words: Ribozyme, Riboswitch, In-drop digestion, RNase T1 mapping, Crystal dehydration

1. Introduction X-ray crystallography is the most powerful tool available to elucidate the structure of biological RNAs. Success with this method depends foremost on production of high-quality crystals, which can be challenging for many RNAs (1). As with proteins, orthologous RNA sequences, particularly those from thermophiles, exhibit marked differences in their thermodynamic stabilities and propensities for adopting the native conformation (2). This feature can be exploited when searching for optimal RNA constructs for crystallization. However, even optimal RNA constructs can present problems for crystallization because of the relative instability of RNA phosphodiester linkages (3) over the time course

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_10 © Humana Press, a part of Springer Science + Business Media, LLC 2009

129

130

Klein and Ferré-D’Amaré

sometimes required for crystal growth. Therefore, new crystals should always be analyzed to determine the RNA chain(s) present and their covalent structures. Here we describe the steps that led to our discovery of highquality crystals of the glmS ribozyme from Thermoanaerobacter tengcongensis (4). Three steps were crucial to crystal production. First, in-drop digestion of our initial single-chain RNA construct occurred during crystallization trials and, serendipitously, was required to produce high-quality crystals. Analysis of the molecular contents of these crystals by RNase T1 mapping was utilized to pinpoint the labile internucleotide linkage, allowing design of a two-chain RNA construct for more rapid and reproducible crystal growth. Second, limited dehydration following crystal growth often produced cracks, but concomitantly improved diffraction quality. Although the mother liquor was already cryoprotective, the resolution limit improved from ~3 Å to at least 1.7 Å when precipitant concentrations were briefly increased immediately prior to flash freezing. Third, a change from acidic to basic pH following crystal growth was necessary for high-occupancy binding by glucosamine-6-phosphate (GlcN6P) (5), the natural metabolite ligand of the glmS ribozyme (6). Unlike many protein crystals, which are highly sensitive to pH (7), crystals of the glmS ribozyme tolerated substantial changes to pH without noticeable effects on diffraction.

2. Materials 2.1. Design and Synthesis of glmS Ribozyme RNA

1. DNA oligonucleotides for making RNA constructs by PCR. 2. Taq DNA polymerase, 5,000 U/mL (New England Biolabs, Ipswich, MA). 3. T7 RNA Polymerase, 50,000 U/mL (New England Biolabs, Ipswich, MA). 4. Diethylpyrocarbonate (DEPC)-treated water (see Note 1).

2.2. Crystallization Strategies

1. 15- to 24-well culture plates, e.g., “vapor diffusion crystalplate®” available from MP Biomedicals (formerly ICN Biomedicals, Irvine, CA). 2. Siliconized glass cover slips (8).

2.3. Crystal Analysis

1. 10× kinase buffer: 700 mM Tris–HCl, pH 7.6, 100 mM MgCl2, 50 mM dithiothreitol (DTT), store at −20°C. 2. CEU buffer: 10.25 mM citric acid, 14.75 mM sodium citrate, 1 mM ethylene diamine tetraacetic acid (EDTA), 7 M urea, adjust to pH 5.0 with HCl, store at −20°C. 3. CU buffer: 10.25 mM citric acid, 14.75 mM sodium citrate, 10 M urea, adjust to pH 5.0 with HCl, store at −20°C.

Crystallizatiion of the glmS Ribozyme-Riboswitch

131

4. Na2CO3/EDTA buffer: 50 mM Na2CO3, pH 11.7, 1 mM EDTA, store at −20°C. 5. FA gel-loading buffer: 90% (v/v) deionized formamide, 10% (v/v) 10× TBE, 0.025% (w/v) bromophenol blue, 0.025% (w/v) xylene cyanol. 10× TBE is prepared by mixing 108 g Tris base, 55 g boric acid, 40 mL, pH 8.0, 0.5 M sodium EDTA, adjusting to one liter with water and autoclaving for 20 min. 6. RNase T1, 100,000 U/mL (Roche, Basel, Switzerland). 7. Microspin G-25 columns (GE Healthcare, Piscataway, NJ). 8. T4 Polynucleotide Kinase, 10,000 U/mL (New England Biolabs, Ipswich, MA). 9. Urea-PAG: polyacrylamide gel containing 8 M urea. 2.4. Postcrystallization Treatments

1. Low pH cryodehydration buffer: 25% PEG 4000, 1.7 M LiCl, 30 mM MgCl2, 100 mM MES, pH 5.5. 2. High pH cryodehydration buffer: 25% PEG 4000, 1.7 M LiCl, 30 mM MgCl2, 20 mM GlcN6P, 100 mM Tris–HCl, pH 8.5. 3. Nylon loops: 20-mm monofilament with a loop diameter ranging from 0.1 to 0.5 mm, depending on the size of the crystal (Hampton Research, Aliso Viejo, CA). 4. 1 mM EDTA. 5. [g-32P]-ATP, 170 mCi/mL (Amersham, Piscataway, NJ). 6. X-ray film or imaging plate.

3. Methods 3.1. Design and Synthesis of glmS Ribozyme RNA

Several strategies for production of milligram quantities of homogenous RNA have been extensively described previously, such as the use of bacteriophage T7 RNA polymerase and cis-cleaving ribozymes (9–11). These strategies have been instrumental to high-resolution structural studies of RNAs over the past decade. We designed glmS ribozyme constructs derived from RNA sequences of several bacteria (Fig. 1). In each case, we engineered a binding site for the U1A spliceosomal protein in the functionally dispensable, variable loop of helix P1 (6), a strategy that has proven successful with several other RNAs (12–15).

3.2. Crystallization Strategies

The basic strategies to screen conditions for macromolecular crystallization have changed little over the past few decades. In short, a solution containing the macromolecule of interest is slowly brought to a state of supersaturation to promote crystal nucleation. Numerous techniques for conducting crystallization screens have been extensively reviewed (7). Considerations specific

Fig. 1. Various RNA constructs of the glmS ribozyme derived from several Gram-positive bacteria and their associated crystals. (a) No crystals were obtained for B. subtilis ribozyme. Crystals shown for the B. cereus and F. nucleatum ribozymes (b and c) diffracted X-rays to low resolution. Clusters of needle crystals of the T. tengcongensis ribozyme grew overnight, while larger rod crystals appeared in the same drop after ~3–4 weeks (d). Bars are equal to 200 mm.

Crystallizatiion of the glmS Ribozyme-Riboswitch

133

to the crystallization of RNAs have also been reviewed (16). Once initial conditions for crystallization have been identified, further optimization of the variables associated with crystal growth (e.g., temperature, additives, macromolecular concentrations) is conducted to improve the quality of the crystals. We performed an extensive search for crystallization conditions for each of the four initial constructs (Fig. 1). For one of these constructs we failed to identify any conditions for crystallization, while others produced crystals that exhibited poor diffraction quality, even after optimization. Our crystallization experiments with the T. tengcongensis glmS ribozyme were the most noteworthy, as multiple crystal forms appeared in the same drop (Fig. 1d). Initial analysis of the slower growth crystals of the T. tengcongensis glmS ribozyme demonstrated promising diffraction quality that warranted further investigation (see later). 3.3. Crystal Analysis

In macromolecular crystallography it is prudent to question the protein and/or nucleic acid composition of all newly discovered crystals. Usually, the goal is to establish that a new crystal contains the desired macromolecule(s) or is of an undesired low molecular weight salt. In other cases, unintended modifications of the covalent structure of the macromolecule(s) occur during crystallization. This is especially true of RNA, which has the tendency to undergo cleavage of the phosphodiester backbone. In these cases, it is good practice to learn as much as possible about the macromolecular composition of a new crystal before embarking on intensive crystallographic analysis. Our experience with the T. tengcongensis glmS ribozyme provides an example of serendipity in RNA crystallization. We searched for crystallization conditions for a 151-nucleotide (nt) RNA construct that represented the postcleavage state of the T. tengcongensis glmS ribozyme. This RNA was engineered to contain a binding site for the RNA-binding domain of the U1A spliceosomal protein in the loop of helix P1. Therefore, we set out to find crystallization conditions for the U1A-glmS ribozyme RNA complex. The most promising crystals grew after 3–4 weeks and diffracted to ~3 Å resolution. However, analysis of the RNA content of these crystals by urea-PAGE revealed the presence of two RNA species that were ~125-nt and ~25-nt in length. Both RNAs could be radiolabeled using T4 Polynucleotide Kinase, suggesting each RNA contained a free 5¢-OH (Fig. 2a). The larger ~125-nt RNA was purified and used for RNase T1 digestion and alkaline hydrolysis (17). As a control, the 151-nt RNA construct was also subjected to RNase T1 digestion. By analyzing the pattern RNA fragments produced from RNase T1 digestion of the ~125 nt and 151 nt RNAs we deduced that the ~125 nt RNA contained exactly 26 fewer nucleotides at its 5¢-end, as compared to the full-length 151 nt RNA construct. This indicates that a phosphodiester bond in the U1A-binding loop

134

Klein and Ferré-D’Amaré

Fig. 2. (a) 5¢-End labeling of RNA harvested from crystals of the T. tengcongensis glmS ribozyme. Two distinct RNA species, each containing a free 5¢-OH, are present in the crystal. (b) RNase T1 analysis of purified 5¢-[32P]-labeled T. tengcongensis glmS ribozyme ~125-mer RNA harvested from crystals (lanes 1–3). RNase T1 digestion is also shown for the full-length crystallization construct (151 nucleotides) before in-drop digestion (lane 4).

was cleaved. Measurements of X-ray fluorescence using crystals grown in the presence of selenomethionine-labeled U1A protein suggested the absence of the protein in the slower-growing crystal form. Based on this finding, we designed an RNA construct that spanned nucleotides 26–151 of the original construct (see Note 2). This RNA was transcribed in vitro by T7 RNA polymerase and purified by urea-PAGE. A 26 nt RNA, which comprises the 5¢end of the original construct, was synthesized by Dharmacon (Lafayette, CO). Annealing of these two RNAs reproduced the

Crystallizatiion of the glmS Ribozyme-Riboswitch

135

nicked RNA that produced the original crystals. Crystals of the bimolecular RNA constructs grew readily in the absence of U1A protein, as described previously (4). 3.3.1. Preparation and 32 P-Labeling of RNA from Crystals

1. Using a nylon loop, remove crystal(s) from mother liquor and dissolve in 20 mL of 1 mM EDTA. 2. Heat sample to 95°C for 2 min, then chill on ice for 5 min. Centrifuge briefly to collect condensation. 3. Combine the following: 14.5 mL of crystal RNA, 2 mL of 10× kinase buffer, 2 mL of 100 mM DTT, 1 mL of T4 PNK, 0.5 mL of [g-32P]-ATP. Incubate at 37°C for 1.5 h (see Note 3). 4. Remove unincorporated [g-32P]-ATP with a G25 spin column. Dilute radiolabeled RNA sample to a final volume of 50 mL with DEPC-treated water. 5. Analyze RNA by electrophoresis on a 12% urea-PAG. 6. Autoradiograph and excise bands corresponding to RNA from the crystal. 7. From gel slices, passively elute RNA into DEPC-treated water at 4°C with gentle rocking for ~12 h. 8. Ethanol-precipitate labeled RNA and dry completely using a centrifugal vacuum concentrator (e.g., Speedvac). 9. Resuspend RNA in ~20 mL of DEPC-treated water and store at −20°C.

3.3.2. RNase T1 Mapping

1. Mix 3 mL of [g-32P]-labeled RNA with 24 mL of CEU buffer. 2. Incubate mixture at 50°C for 5 min. 3. Add 3 mL of RNase T1 (1 U/mL). 4. Withdraw 10 mL aliquots of digestion mixture at varying time points (e.g., 4, 8, 15 min) and quench by adding 20 mL of FA gel-loading buffer.

3.3.3. Alkaline Hydrolysis

1. Mix 1 mL of [g-32P]-labeled RNA with 3.6 mL of Na2CO3/ EDTA buffer. 2. Incubate mixture at 95°C for 75–120 s, then place on ice for 1 min. 3. Add 8.4 mL of CU buffer. 4. Add 20 mL of FA gel-loading buffer. 5. Load samples from alkaline hydrolysis and RNase T1 digestion onto a 12% urea-PAG and perform electrophoresis. 6. Autoradiograph gel with X-ray film or with imaging plates.

3.4. PostCrystallization Treatment

Once crystals of the T. tengcongensis glmS ribozyme were reproduced employing the redesigned two-strand RNA construct (without U1A protein), a subsequent challenge was optimizing

136

Klein and Ferré-D’Amaré

their diffraction limit. For many RNA crystals this can be a slow process of trial and error that often requires modifying the RNA sequence and/or conditions for crystal growth (e.g., (8, 16)). However, there are numerous examples of protein crystals for which diffraction quality has been dramatically improved after crystal growth by simple treatments, such as annealing, dehydration, and soaking with ligands (18). We found that the diffraction limit of our glmS ribozyme crystals improves from ~3.0 Å to at least 1.7 Å resolution by briefly soaking in a cryodehydration solution. Shrinkage of the unit cell volume by ~9% concomitant with rearrangements of crystal contacts occurred upon dehydration (Fig. 3a, b). Not surprisingly, given the large changes in crystalline packing,

Fig. 3. Dehydration of crystals of the T. tengcongensis glmS ribozyme. (a) Crystal contacts between glmS ribozymes in two adjacent asymmetric units of the unit cell. (b) Same view of the crystal contacts after dehydrating cryoprotection. (c–e) A cluster of crystals is shown in mother liquor (c), after transfer to the cryodehydration solution for 5 min (d), note appearance of small cracks), and after physical separation of a large piece of the crystal (e). Bar equals 200 mm.

Crystallizatiion of the glmS Ribozyme-Riboswitch

137

soaking often produced visible physical damage, including cracks in the crystal (Fig. 3c, d). Although the appearance of crystal cracks typically indicates a decrease in the crystalline order, we instead observed small crystal cracks to be a sign of successful dehydration. Moreover, cracking aided the separation of single crystal segments from larger clusters, thus facilitating data collection (Fig. 3e). These results underscore the importance of testing simple postcrystallization treatments before making conclusions about the diffraction quality and utility of a given crystal form. The pH of the cryodehydration buffer was found to influence markedly the binding of GlcN6P to crystals of the glmS ribozyme (see Note 4). When the low-pH cryodehydration buffer is used, GlcN6P does not appear to bind, as evidenced by a complete lack of electron density corresponding to the small molecule. However, diffraction data collected from crystals soaked in the high-pH cryodehydration buffer revealed clear electron density corresponding to GlcN6P. Importantly, this property appears to be specific to GlcN6P, as binding of glucose-6-phosphate is evident at both pH 5.5 and pH 8.5 (4, 5). For dehydration of the glmS ribozyme we performed the following manipulations. 1. Using a nylon loop, transfer a crystal(s) from its mother liquor to 10 mL of cryodehydration buffer, and incubate for 5–10 min on a glass coverslip positioned on the microscope stage. 2. Small cracks occasionally become visible indicating shrinkage of the unit cell. 3. Physically separate an undamaged segment of the crystal by using a cat’s whisker. 4. Using a nylon loop, plunge the crystal into liquid nitrogen.

4. Notes 1. 1 mL of DEPC is added to 4-L of deionized water, mixed vigorously, and autoclaved for 1 h to destroy excess DEPC. DEPC is toxic and should be handled in a fume hood. Autoclaving destroys DEPC, converting it into CO2 and ethanol. 2. A DNA template encoding the trans-acting form of the T. tengcongensis glm S ribozyme encompassed nucleotides 26–151 of the original RNA (Fig. 1d). This construct was flanked by cis-cleaving hammerhead and HDV ribozymes at its 5¢ and 3¢ ends, respectively. A DNA template was created by PCR from overlapping DNA oligonucleotides, and cloned into pUC19 using the EcoRI and BamHI sites of the vector. Cotranscriptional cleavage by the hammerhead and HDV ribozymes produces a glmS RNA with a 5¢-OH and 2¢,3¢cyclic phosphate.

138

Klein and Ferré-D’Amaré

3.

32

P is a high-energy beta radiation emitter and personal protective equipment should be used at all times when conducting experiments involving this isotope.

4. The pH of 5.2–6.0 in crystal growth solutions must be elevated to pH ~8.5 during the cryodehydration step in order for GlcN6P binding to be detectable by X-ray crystallography.

Acknowledgments A.R.F. is an Investigator of the Howard Hughes Medical Institute and was a Distinguished Young Scholar in Medical Research of the W.M. Keck Foundation. D.J.K. was a Damon Runyon Fellow supported by the Damon Runyon Cancer Research Foundation (DRG-1863-05). This work was also supported by grants from the NIH (GM63576 to A.R.F. and GM084076 to D.J.K.) and the W.M. Keck Foundation. References 1. Ferré-D’Amaré, A. R., Zhou, K., and Doudna, J. A. (1998). A general module for RNA crystallization. J. Mol. Biol. 279, 621–631. 2. Fang, X. W., Golden, B. L., Littrell, K., Shelton, V., Thiyagarajan, P., Pan, T., and Sosnick, T. R. (2001). The thermodynamic origin of the stability of a thermophilic ribozyme. Proc. Natl Acad. Sci. U. S. A. 98, 4355–4360. 3. Holley, R. W., Apgar, J., and Merrill, S. H. (1961). Evidence for the liberation of a nuclease from human fingers. J. Biol. Chem. 236, PC42–PC43. 4. Klein, D. J., and Ferré-D’Amaré, A. R. (2006). Structural basis of glmS ribozyme activation by glucosamine-6-phosphate. Science 313, 1752–1756. 5. Klein, D. J., Wilkinson, S. R., Been, M. D., and Ferré-D’Amaré, A. R. (2007). Requirement of helix P2.2 and nucleotide G1 for positioning of the cleavage site and cofactor of the glmS ribozyme. J. Mol. Biol. 373, 178–189. 6. Winkler, W. C., Nahvi, A., Roth, A., Collins, J. A., and Breaker, R. R. (2004). Control of gene expression by a natural metabolite-responsive ribozyme. Nature 428, 281–286. 7. McPherson, A. (1999). Crystallization of biological macromolecules, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

8. Rupert, P. B., and Ferré-D’Amaré, A. R. (2004). Crystallization of the hairpin ribozyme: illustrative protocols. Methods Mol. Biol. 252, 303–311. 9. Milligan, J. F., and Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods Enzymol. 180, 51–62. 10. Price, S. R., Ito, N., Oubridge, C., Avis, J. M., and Nagai, K. (1995). Crystallization of RNA-protein complexes I. Methods for the large-scale preparation of RNA suitable for crystallographic studies. J. Mol. Biol. 249, 398–408. 11. Ferré-D’Amaré, A. R., and Doudna, J. A. (1996). Use of cis- and trans-ribozymes to remove 5¢ and 3¢ heterogeneities from milligrams of in vitro transcribed RNA. Nucleic Acids Res. 24, 977–978. 12. Ferré-D’Amaré, A. R., Zhou, K., and Doudna, J. A. (1998). Crystal structure of a hepatitis delta virus ribozyme. Nature 395, 567–574. 13. Ferré-D’Amaré, A. R., and Doudna, J. A. (2000). Crystallization and structure determination of a hepatitis delta virus ribozyme: use of the RNA-binding protein U1A as a crystallization module. J. Mol. Biol. 295, 541–556. 14. Rupert, P. B., and Ferré-D’Amaré, A. R. (2001). Crystal structure of a hairpin ribozyme-inhibitor

Crystallizatiion of the glmS Ribozyme-Riboswitch complex with implications for catalysis. Nature 410, 780–786. 15. Adams, P. L., Stahley, M. R., Kosek, A. B., Wang, J., and Strobel, S. A. (2004). Crystal structure of a self-splicing group I intron with both exons. Nature 430, 45–50. 16. Ferré-D’Amaré, A. R., and Doudna, J. A. (2000). Methods to crystallize RNA, in Current Protocols in Nucleic Acid Chemistry (Beaucage, S. L., Bergstrom, D. E., Glick, G.

139

D., and Jones, R. A., Eds.), pp. 7.6.1-7.6.10, Wiley, New York, NY. 17. Ziehler, W. A., and Engelke, D. R. (2000). Probing RNA structure with chemical reagents and enzymes. Curr. Protocols Nucleic Acid Chem. 2, 6.1.1–6.1.21. 18. Heras, B., and Martin, J. L. (2005). Post crystallization treatments for improving diffraction quality of protein crystals. Acta Crystallogr. D Biol. Crystallogr. 61, 1173–1180.

Chapter 11 Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering Jan Lipfert, Daniel Herschlag, and Sebastian Doniach Summary Riboswitches are functional RNA molecules that control gene expression through conformational changes in response to small-molecule ligand binding. In addition, riboswitch 3D structure, like that of other RNA molecules, is dependent on cation–RNA interactions as the RNA backbone is highly negatively charged. Here, we show how small-angle X-ray scattering (SAXS) can be used to probe RNA conformations as a function of ligand and ion concentration. In a recent study of a glycine-binding tandem aptamer from Vibrio cholerae, we have used SAXS data and thermodynamic modeling to investigate how Mg2+-dependent folding and glycine binding are energetically coupled. In addition, we have employed ab initio shape reconstruction algorithms to obtain low-resolution models of the riboswitch structure from SAXS data under different solution conditions. Key words: RNA, Riboswitches, Small-angle X-ray scattering, RNA folding, RNA aptamers

1. Introduction Since the discovery of catalytic RNA by Cech and Altman in the early 1980s it has become increasingly clear that RNA, like proteins, can fold into intricate 3D structures to carry out cellular functions. The RNA backbone, unlike proteins, is highly negatively charged and RNA folding and interactions are therefore dependent on the presence of positively charged cations (1, 2). Ion-dependent RNA folding has been studied by a number of techniques, including footprinting methods (3, 4), gel mobility, NMR spectroscopy (5), and fluorescence resonance energy transfer (6, 7).

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_11 © Humana Press, a part of Springer Science + Business Media, LLC 2009

141

142

Lipfert, Herschlag, and Doniach

SAXS has been an important technique to monitor ioninduced RNA folding (8–15). In particular, SAXS measurements have shown that RNA molecules undergo rapid compaction on the millisecond time scale upon addition of ions, driven primarily by electrostatic shielding. Formation of native tertiary interactions and folding to the catalytically active conformation, in contrast, can be much slower. Recently, the discovery of riboswitches has further increased the interest in RNA structure and conformations. Riboswitches are RNA molecules located in the 5¢-UTR of mRNA that are capable of binding small-molecule metabolites. Small-molecule binding to the aptamer domain causes conformational changes that directly regulate gene expression, at the transcriptional or translational level, without the need for protein cofactors. For recent reviews of riboswitch mechanisms and structures see refs. 16–22. The glycine-binding VCI-II aptamer from Vibrio cholerae is part of a particular intriguing riboswitch (23). It features two glycine-binding sites that bind glycine cooperatively (23). We have used SAXS in combination with hydroxyl radical footprinting to follow the conformations of the VCI-II tandem aptamer as a function of glycine and Mg2+ concentration (24). The results show that, on going from low salt concentrations to high Mg2+ (10 mM) conditions, the molecule undergoes a partial folding transition characterized by significant conformational changes and compaction. Addition of glycine in the presence of millimolar Mg2+ leads to a further conformational change and compaction upon glycine binding. Thermodynamic modeling has indicated that the second transition from the conformation in high Mg2+ alone to the glycinebound state requires the association of additional Mg2+ ions (24). In a follow-up study, we have probed the structure of the VCI-II tandem aptamer in the absence and presence of glycine for a range of different mono- and divalent ions. The results indicate that partial folding in the absence of glycine can be induced by any of the tested ions and is likely dominated by unspecific electrostatic contributions of the ion atmosphere. Glycine binding, in contrast, is not observed in monovalent and certain divalent ions (Sr2+, Ba2+) and likely requires specific ion binding (Jan Lipfert, Adelene Y.-L. Sim, Daniel Herschlag, and Sebastian Doniach, in preparation). In this chapter, we describe in detail how SAXS measurements can be used to study the conformations of functional RNAs as a function of salt and ligand concentration by using the VCI-II tandem aptamer as a representative example. Particular emphasis is given to the use of ab initio 3D structure reconstruction algorithms. Programs to obtain low-resolution shapes from SAXS data have first been developed for applications to proteins and protein complexes (25–27). Recently, we have shown that

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering

143

these programs can be successfully applied to nucleic acids (28). In the case of the VCI-II tandem aptamer, we have obtained lowresolution reconstructions of all three thermodynamic states (the low salt or unfolded state, the compact intermediate, and the fully folded and glycine-bound state) by using the 3D reconstruction algorithm DAMMIN (24).

2. Materials

2.1. In Vitro RNA Transcription

All chemicals are purchased from Sigma-Aldrich, Co., unless otherwise noted. Chemicals are dissolved in deionized, RNase-free Milli-Q water; stocks are prepared every 1–3 months and stored at −20°C. 1. 10× Transcription buffer (quantities for 50 mL stock): 20 mL 1 M Tris–HCl, pH 8.1, 12.5 mL 1 M magnesium chloride, 1 mL 1 M spermidine, 0.5 mL 10% Triton X-100, 16 mL deionized, RNase-free Milli-Q water. 2. 1 M Dithiothreitol (DTT). 3. 10 mM NTP stock: 10 mM each of ATP, CTP, GTP, and UTP. Stocks are adjusted to pH 7.0 by addition of sodium hydroxide. Concentrations are determined from the absorbance at 259, 272, 252, and 262 nm, respectively. The absorption coefficients are a259(ATP) = 15.3 × 103 (M cm)−1, a272(CTP) = 9.6 × 103 (M cm)−1, a252(GTP) = 13.7 × 103 (M cm)−1, and a262(UTP) = 9.9 × 103 (M cm)−1. 4. T7 RNA polymerase, »50,000 U/mL. 5. DNA template with T7 promoter (T7 promoter sequence: GCGCTTAATACGACTCACTATA). Short DNA templates (up to 100 bp) can be purchased from Integrated DNA Technologies (http://www.idtdna.com). The DNA template for the VCI-II riboswitch was created by PCR ligation (29). 6. 0.5 M Sodium ethylenediamine tetraacetic acid (EDTA), adjusted to pH 8. 7. 3 M Na-acetate, pH 5.4. 8. Cold ethanol, stored at −20°C. 9. Oakridge 50-mL centrifuge tubes.

2.2. RNA Gel Purification

1. 10× Tris–borate–EDTA (TBE) buffer. For 2 L stock, use 242 g Tris base (1 M final concentration), 102.7 g boric acid (0.83 M final concentration), and 7.44 g EDTA. Adjust volume to 2 L with RNase-free Milli-Q water and autoclave. Final 1× TBE buffer has pH ~8.0.

144

Lipfert, Herschlag, and Doniach

2. 8% Acrylamide gel mix. For 1 L stock, dissolve 420 g urea in 800 mL water (for 7 M urea final concentration). Add 200 mL 40% acrylamide (29:1 acrylamide:bis-acrylamide) solution. 3. Tetramethylethylenediamine (TEMED). 4. 10% Ammonium persulfate (APS). 5. Loading dye solution for gel purification. For 10 mL stock, add 0.4 mL 0.5 M EDTA solution to 9.6 mL formamide (for a final EDTA concentration of 20 mM). Add »2 mg/mL xylene cyanol and bromophenol blue. 6. Tris–EDTA–NaCl (TEN) buffer. For 250 mL stock, use 2.5 mL 1 M Tris–HCl (final concentration: 10 mM), pH 8, 417 mL 600 mM EDTA (final concentration: 1 mM), and 25 mL 3 M NaCl (final concentration: 300 mM). Add water to 250 mL final volume and filter through a 0.2-mm pore size syringe filter. 7. 50-mL Falcon tubes (BD Biosciences, San Jose, CA). 2.3. SAXS Measurements

1. 10× (500 mM) Na–MOPS buffer, pH 7.0. 2. 10× (100 mM) glycine stock. 3. 10× MgCl2 stocks: 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1.0, 2.0, 5.0, 10.0, 20.0, 50.0, 100.0, and 200.0 mM. 4. 50 kDa Cut-off microcone centrifuge filters (Millipore).

2.4. SAXS Molecular Weight Standard

1. 100 mM Na-acetate buffer, pH 5.4. 2. Guanidine hydrochloride (GdnHCl). 3. Horse heart cytochrome c, lyophilized. 4. 0.2-mm Pore size syringe filters and sterile syringes.

3. Methods 3.1. How Much RNA Is Required for a SAXS Measurement? 3.1.1. RNA Concentration Requirements

The forward scattering I(0) from an RNA or protein sample in solution in a SAXS experiment is given by I(0) = Kc(vΔr)2 (MW)2,

(1)

where c is the concentration, MW is the molecular weight of the macromolecule, Dr is the average electron contrast of the molecule, n the partial specific volume, and K is a constant that depends on parameters of the measurement setup and is typically determined from comparison with a molecular weight standard. Eqn. 1 is valid only if interparticle interference effects are negligible, which is typically the case in SAXS measurements of biological samples

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering

145

(see later). Equation 1 provides a useful guideline to determine appropriate sample concentrations. The scattering contrast (Dr)2 is approximately constant for all RNA molecules (and about five times larger for RNA compared to proteins (12)). The expected scattering signal is therefore linear in the (molar) concentration and increases quadratically with the molecular weight of the sample. The VCI-II tandem aptamer comprises 226 residues and has a molecular weight of »70 kDa. With the SAXS set up at beam line 12-ID at the Advanced Photon Source, high-quality scattering profiles can be obtained with an RNA concentration of 20 mM for this construct. For smaller (larger) RNA constructs, the concentration should be increased (decreased) according to the ~(MW)2 proportionality. Control measurements should be carried out with at least twofold higher and twofold lower concentrations to check for aggregation or interparticle interference effects (see later). In the case of the VCI-II glycine-aptamer, we have obtained scattering profiles with RNA concentrations in the range of 5–50 mM and observed no systematic dependence of the profile shape on RNA concentration (24). 3.1.2. Sample Volume Requirements

The intense X-ray radiation used in SAXS measurements at stateof-the-art synchrotron sources generates radicals that quickly degrade biological samples (see later). RNA samples for SAXS measurements should therefore be used only for one measurement. The sample requirement per measurement is determined by the sample cell used in the setup. We have developed a sample cell and cell holder for biological SAXS measurements with 16 mL sample volume (30). Allowing for pipetting and loading losses, this corresponds to »20 mL per measurement. The number of measurements required depends, naturally, on the purpose of the study. Obtaining high-quality scattering profiles for an RNA under particular solution conditions typically requires triplicate repeat measurements at different RNA concentrations, i.e., about ten individual measurements. A thorough investigation of an RNA construct, which includes characterization under different solution conditions and Mg2+ and ligand titrations, requires at least similar to 50 individual measurements or about 1 mL RNA solution at the desired measurement concentration. This corresponds to »1–2 mg RNA for RNA constructs with a few hundred residues.

3.2. In Vitro RNA Transcription

Production of milligram quantities of RNA requires transcription reactions with ³10 mL final volume. As the amount of reagents required represents a significant cost, it is advisable to test reagents first by running test reaction with 50–100 mL final volume and to check reaction products on a denaturing gel. In all steps of RNA transcription, purification, and preparation for the measurements great care must be taken to avoid contamination with RNases.

146

Lipfert, Herschlag, and Doniach

1. For a 10-mL transcription reaction, mix 1 mL 10× transcription buffer, 0.4 mL 1 M DTT, 1 mL 10 mM NTP stock, 0.1 mL DNA template (100 mg/mL), 7.4 mL water, and 0.1 mL » 50,000 units/mL T7 polymerase in an Oakridge 50-mL centrifuge tube. 2. Vortex the transcription solution and incubate for 3–4 h in a 37°C waterbath. 3. Add 1 mL 0.5 M EDTA, 1 mL 3 M Na-acetate, pH 5.4, and 30 mL cold ethanol. Precipitate RNA for at least 4 h at −20°C. 4. Centrifuge for 40 min at 16,000 rpm (24,000 × g). Carefully pour out the supernatant; a translucent RNA pellet should be visible. Dry and resuspend the RNA pellet in 0.5 mL water. 3.3. RNA Gel Purification

1. Add 75 mL TEMED and 750 mL 10% APS to 100 mL 8% denaturing acrylamide gel mix to initiate the polymerization reaction. Immediately pour gels using large spacers to make 1–2 large wells per gel for loading. Let the acrylamide polymerize for at least 1 h. 2. Run gels with 1× TBE buffer and use a metal plate attached to the glass gel plates to avoid temperature gradients. Before loading the RNA constructs, prerun the gel at 55 W for at least 30 min. 3. Add 1/2 volume loading dye to the resuspended RNA solution and load the gel. Run the purification gel at 55 W for 3 h (adjust the time for shorter or longer RNA constructs). 4. Take down the gel and wrap in Saran wrap. Image gel bands by UV shadowing and mark desired bands. Excise gel bands with a sterile scalpel. Crush gel pieces with a pestle in a 50-mL Falcon tube (BD Biosciences), add »5 mL of TEN buffer. 5. Freeze gel pieces on dry ice and thaw at room temperature. Repeat three freeze–thaw cycles. Agitate for at least 4 h at 4°C. 6. Spin in swinging bucket centrifuge for 20 min at 4,000 rpm (1,500 × g). Pour out supernatant and filter solution through a 0.2-mm syringe filter. 7. Add 1/10 volume 3 M Na-acetate, pH 5.4, and 3 volumes cold ethanol. Precipitate RNA for at least 4 h at −20°C. 8. Centrifuge for 40 min at 16,000 rpm (24,000 × g). Carefully pour out the supernatant; a translucent RNA pellet should be visible. Wash pellet with 200 mL of 70% ethanol/30% water solution and air dry the pellet. Resuspend RNA in 100–400 mL water (see Note 1). Determine the RNA concentration photospectrometrically by absorbance at 260 nm. A typical absorption coefficient for RNA

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering

147

is a260 » 2.2 × 106 (M cm)−1 (i.e., a reading of 1 absorption unit with a 1-cm cuvette corresponds to an RNA concentration of 0.45 mM or 32 mg/mL VCI-II RNA) (see Note 2). 3.4. SAXS Measurements 3.4.1. Sample Preparation for SAXS Measurements RNA Sample Preparation

Buffer subtraction for background correction is essential for SAXS measurements of biological samples. As the scattering signal from RNA samples is comparable in strength to the buffer signal, especially at high momentum transfer q (cf. the sample and buffer profiles in Fig. 1), it is crucial to match the buffer closely to the sample. One strategy to achieve good agreement between buffer and sample conditions is to prepare a stock solution of buffer, to take aliquots for the sample and buffer measurements, and to add the same relative volume of RNA solution and water to the sample and buffer aliquots, respectively. Another possibility is to exchange the buffer repeatedly (3–4 times) using centrifuge spin columns. As the RNA concentration in a typical SAXS measurement is relatively high, it is important to take into account finite concentration effects. For example, 100 mM of a 100-residue RNA can be associated with up to approximately 5 mM of Mg2+ ions. For accurate Mg2+ titrations, it is, therefore, necessary to prepare the samples by buffer exchange. Similarly, titrations for ligands with micromolar- or submicromolar-binding affinities require buffer exchange. For the VCI-II tandem aptamer, Mg2+ titrations in the absence and presence of 10 mM glycine are prepared as follows: 1. Prepare 1 mL buffer aliquots for different Mg2+ concentrations. Mix 0.1 mL 500 mM Na–MOPS buffer, 0.1 mL 100 mM glycine stock, 0.1 mL 10× MgCl2 stock at the desired concentration, and 0.7 mL water. For the titration in the absence of glycine, omit the glycine stock and use 0.8 mL water. 2. Prepare RNA aliquots with 0.8 nmols of RNA (to obtain 40-mL aliquots of 20 mM RNA solutions) in 50-kDa cut-off microcone centrifuge filters. Exchange ³ 100 mL buffer at least three times in the microcone filters. 3. Elute the RNA aliquots into 40 mL final volume. This sample volume is sufficient for two measurements. Equilibrate RNA samples for 20 min in a 50°C water bath. 4. Spin buffer and RNA sample solutions for 10 min at 13,000 rpm (16,000 × g) immediately before the SAXS measurement. As wet lab equipment and time are limited during synchrotron beam times, it can be advantageous to prepare aliquots in advance. In this case, Eppendorf tubes with the RNA aliquots should be flash frozen by immersion in liquid nitrogen and stored and shipped on dry ice (see Note 3).

148

Lipfert, Herschlag, and Doniach 1 0.6

0.8

I

0.4 0.6

I

0.2 0.4

0 0

0.1

0.2

0

0.2

0.3

q (A−1)

0

0.1

0.2

q

0.3

(A−1)

Fig. 1. The effect of radiation damage on scattering profiles. Raw scattering profiles for a 24-bp DNA duplex (main graph) and for the VCI-II glycine riboswitch construct (inset). Scattering profiles from five subsequent 1.0-s exposures of the DNA/RNA samples are shown for each nucleic acid. Corresponding buffer profiles measured immediately before and after the DNA/RNA samples are virtually identical and shown in the bottom part of the graphs. The 24-bp DNA duplex sample was prepared as described (31, 32) and measured at a DNA concentration of 0.52 mM in 1 mM Na– MOPS buffer, pH 7.0, with 10 mM MgCl2. The VCI-II RNA was prepared as described in Subheading 3 and measured at an RNA concentration of 20 mM in 50 mM Na– MOPS, pH 7.0, with 10 mM MgCl2 and 10 mM glycine. The effect of radiation damage is clearly visible as an increase in forward scattering for subsequent exposures in the 24-bp DNA duplex data set. All five profiles were found identical in the experiments with riboswitch RNA Note that both samples were exposed to the same radiation dose. The fact that the DNA duplex measurement, but not the VCI-II RNA, shows strong signs of radiation damage is due to the high DNA and low buffer concentration used. Measurements of the same DNA sample in 50 instead of 1 mM Na–MOPS show no signs of radiation damage (data not shown).

Sample Preparation for a SAXS Molecular Weight Standard

1. Dissolve 2.38 g GdnHCl in 50 mL Na-acetate buffer (for a final concentration of 0.5 M GdnHCl). Filter buffer solution through 0.2-mm pore size syringe filter. 2. Weight out 8 mg of cytochrome c and dissolve in 1 mL buffer. Filter protein solution through 0.2-mm pore size syringe filter. 3. Prior to the SAXS measurement, spin buffer and protein sample solutions for 10 min at 13,000 rpm (16,000 × g). Cytochrome c is a 11.7-kDa globular protein, its radius of gyration is 13.8 Å, and in solution it has an intense brown color. It serves as a convenient measurement and molecular weight standard. The buffer used for the cytochrome c standard with its GdnHCl content is also useful to wet new sample cells by repeated loading.

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering 3.4.2. SAXS Measurements of Riboswitch RNA

149

We use a sample cell specifically designed for biological SAXS measurements (30) and a Hamilton syringe for rinsing and loading the cell. We recommend measuring a matching buffer profile before and after each RNA sample measurement. 1. Rinse the cell with deionized, RNase-free water. 2. Rinse the cell three times with the desired buffer. Load the cell with buffer solution. 3. Measure buffer scattering profile. 4. Load cell with the RNA solution in the same buffer. 5. Measure RNA scattering profile. 6. Repeat from step 1. The exposure time needs to be adjusted to achieve a good signalto-noise ratio without causing radiation damage (see later).

3.5. Data Analysis 3.5.1. SAXS Data Reduction

3.5.2. Radiation Damage

The raw data from the X-ray detector have to be processed to obtain a one-dimensional scattering profile. The details of this procedure depend on the detector and measurement setup. Most SAXS beam lines make efficient procedures for circular averaging of the detector signal (in the case of 2D CCD detectors) available to their users. We calibrate the scattering angle by comparison to silver behenate, a scattering standard (33), and correct for incident beam intensity. The intense X-ray beams available at synchrotrons generate radicals that degrade biological samples such as RNA. One particular effect of ionizing radiation is the formation of intermolecular crosslinks that cause aggregation. It is important to limit the exposure time such that radiation-induced changes of the sample are negligible, as otherwise the signal becomes uninterpretable. The standard procedure to gauge the effects of radiation exposure on the sample is to record several scattering profiles in short succession and to compare the subsequent measurements of the same sample. Systematic changes of the SAXS profile in a series of exposures are a sign of radiation damage. As crosslinked and aggregated species scatter strongly in the forward direction (due to the ~(MW)2 dependence in Eq. 1), radiation damage typically leads to an increase in the forward scattering (Fig. 1). RNA samples are more resistant to radiation damage than proteins. At beam line 12-ID of the Advanced Photon Source, we routinely expose protein samples for 2 × 0.5 s and RNA samples 4–5 × 1.0 s without incurring significant radiation damage. However, the acceptable radiation dose will vary depending on sample and buffer conditions, and we compare subsequent exposure on the same sample routinely for all our SAXS measurements. In general, higher RNA concentration increases the probability of intermolecular crosslinks and therefore increases

150

Lipfert, Herschlag, and Doniach

the effects of radiation damage. In contrast, we have observed that higher buffer concentrations (MOPS or Tris) tend to confer some protection from radiation damage (Fig. 1). 3.5.3. Buffer Subtraction

If the scattering profiles obtained from subsequent exposures of the same sample show no systematic differences, they can be averaged to improve signal quality. The next step is then to subtract the buffer scattering signal for background correction. We recommend comparing buffer profiles measured before and after the RNA sample measurement (see, e.g., the buffer profiles in Fig. 1). If the two buffer profiles show no systematic differences, their average is subtracted from the RNA sample scattering profile. If the two buffer profiles exhibit systematic differences, it is useful to compare to similar buffer profiles to decide which measurements should be discarded as outliers.

3.5.4. Concentration Dependence

Equation 1 predicts a linear dependence of the scattering profile on RNA concentration for a monodisperse and dilute sample. An important test of sample quality is therefore to compare measurements at different RNA concentrations after rescaling of the scattering profiles by RNA concentration. Changes in the shape of the scattering profile with increasing RNA concentration can have several reasons. An increase in the scattering at low q with increasing concentration (in addition to the linear dependence predicted from Eq. 1) is often a result of sample aggregation and radiation damage. However, even monodisperse samples in the absence of radiation damage exhibit a nontrivial concentration dependence at high concentrations, as the measured scattering profile is a product of the particle form factor P(q) (which describes the scattering pattern of an ideal infinite dilute monodisperse solution of the particles) and the solution structure factor S (q,c) (34). I(q,c) = cP(q)S(q,c).

(2)

The structure factor S(q,c) describes the interactions of RNA molecules in solution. S(q,c) deviates from 1 most strongly at low q. S(q,c) < 1 at low q is indicative of repulsive interactions, a situation frequently observed for measurements of RNA samples under low salt conditions where strong electrostatic repulsion dominates the interparticle potential. S(q,c) > 1 at low q is characteristic of attractive interparticle interactions (but in practice very difficult to distinguish from aggregation and radiation damage). For sufficiently dilute solutions S(q,c) » 1 for all values of q. For most purposes it is preferable to measure in this dilute solution regime, even though it is in principle possible to compute S(q,c) from solution theory (34).

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering 3.5.5. Guinier Analysis

151

The scattering intensity at low q in the dilute limit behaves as (34, 35). lim I (q) = I (0) exp( − Rg2 q 2 / 3). q→0

(3)

Equation 3 is called the Guinier approximation (see Note 4). Rg is the radius of gyration and I(0) is the forward scattering intensity. Rg and I(0) are determined from a linear fit to a plot of ln(I(q)) as a function of q2. It is important to choose an appropriate q range for the Guinier fits. For the lowest q values, a beam stop obscures the detector and blocks the transmitted beam. In close vicinity to the beam stop strong X-ray reflections, so called parasitic scattering, will dominate the signal and it is best to discard data points below a certain value qmin. The value of qmin is specific to a particular setup and can be determined from repeated Guinier analysis of a scattering standard with different choices of qmin. Equation 3 is only valid for small q. The upper limit of the range of q values used in the Guinier fit, qmax, should be chosen adaptively such that qmaxRg < 1.3 for approximately spherical objects and smaller for elongated shapes (34). The value of I(0) obtained from Guinier fits (Eq. 3) can be related to the molecular weight of the sample (Eq. 1) by comparison with a molecular weight standard of known concentration c and molecular weight MWS: 2

MW =

I (0) cS ⎛ vS ΔrS ⎞ MWS2 . I S (0) c ⎜⎝ vΔr ⎟⎠

(4)

If a nucleic acid weight standard is used (e.g., a short DNA duplex (24)) (nSDrS/nDr)2 » 1.0, if a protein weight standard is used (nSDrS/nDr)2 » 0.4 for RNA samples. In practice, the molecular weight determination by Eq. 4 is often limited by the accuracy of the concentration measurements (36). Use of Eq. 4 provides, nonetheless, a good check on sample quality. 3.5.6. Thermodynamic Modeling of Scattering Data

SAXS profiles provide a powerful tool to follow structural transitions of RNA molecules as a function of solution conditions. It is often desirable to describe transitions, at least as a first pass, by simple thermodynamic models. In such models, the measured scattering profiles are decomposed into linear combinations of scattering profiles representing contributions from different structural states. In the case of the VCI-II glycine-riboswitch construct, for example, we have modeled the SAXS data using a three-state model with an unfolded state, a partially folded state in the absence of glycine, and a glycine-bound conformation. In general, models can be fit either to the Rg data, using the relationship

152

Lipfert, Herschlag, and Doniach N

Rg,2 k = ∑ fi , k Rg,2 i

(5)

i =1

or to the full scattering profiles using N

I k (q ) = ∑ fi , k I i (q ).

(6)

i =1

where Rg,k and Ik(q) are the radius of gyration and scattering profile obtained for solution condition k (where different solution conditions can correspond to different concentrations of Mg2+, glycine, to different time points, etc.) and Rg,i and Ii(q) are the radius of gyration and scattering profile of the thermodynamic state i, which are properties of the state and do not depend on the solution conditions. For example, if an RNA exhibits two-state folding behavior as a function of Mg2+ concentration with an unfolded state U and a folded state F, 2 2+ 2+ 2 2+ 2 Eq. 5reads Rg ([Mg ]) = fU ([Mg ]) Rg,U + fF ([Mg ]) Rg, F and 2+ 2+ 2+ Eq. 6 simplifies to I (q)([Mg ]) = fU ([Mg ])IU (q) + fF ([Mg ])I F (q). The fi,k are the fractional occupancies of the states i under condition k. The dependence of the fi on solution conditions can be modeled using simple empirical relationships, such as the Hill equation. Often it is possible to determine Ii(q) and Rg,i directly, e.g., by considering the highest and lowest Mg2+ points in a twostate RNA folding transition. Analysis of SAXS profiles by singular value decomposition can be very useful to determine whether a two-state model is sufficient and to model partially populated intermediates (see later).

3.5.7. Analysis of Scattering Profiles by Singular Value Decomposition

In analyzing titration series, e.g., series of scattering profiles obtained as a function of salt or ligand concentration or time, two problems commonly arise: how to determine the minimum number of states that are required to fit the data; and, if there are intermediate states present, how to determine the scattering profiles and fractional occupancies of intermediates that cannot be observed in isolation. Analysis of the scattering profiles by singular value decomposition (SVD) can help to address both these questions (37). SVD constructs an optimal and orthogonal basis for the space of scattering profiles. Inspection of the singular values and basis functions using the criteria of Henry and Hofrichter (38) permits to determine the number of state required to fit the data. Furthermore, determination of the scattering profiles of intermediates states is greatly facilitated by fitting the experimental data in the space of SVD basis functions (rather than by treating each q channel as an independent fitting parameter). Several articles provide excellent introductions to the details of SVD analysis and fitting (38–40).

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering

153

3.5.8. Graphical Representation of Scattering Profiles

At first sight, SAXS profiles appear relatively featureless. To aid the comparison of scattering profiles, it is useful to consider several representations of the scattering data that visually emphasize particular aspects of the data. Figure 2 presents SAXS profiles for the VCI-II tandem aptamer under different solution conditions. A linear I(q) plot (Fig. 2a) emphasizes the low q region, where the scattering intensity is largest. This representation is useful to discern radiation damage (Fig. 1) or concentration dependencies. Features at intermediate or high q that are related to the shape of the molecule are, however, difficult to distinguish in the I(q) plot. A logarithmic plot (Fig. 2b) emphasized the low and high q data about equally and is a useful representation to provide an overview of scattering data. The so-called Kratky representation of q2 × I(q) as a function of q (Fig. 2c) emphasized intermediate to high values of q and is particularly useful to distinguish different conformations of the same construct. Folding transitions are best monitored in the Kratky representation (37, 41). In the case of the VCI-II construct the unfolded, partially folded, and glycine-bound conformations are most readily compared in the Kratky plots (Fig. 2c). In some cases representation of the data as q × I(q) as a function of q, a so-called Holtzer plot, has been proven to be most useful (31).

3.6. Ab Initio Structure Reconstructions

The distribution function of intramolecular distances P(r) can be obtained from a regularized Fourier transform of the scattering profile I(q). A number of software packages are available for calculation of the P(r) function. We use the program GNOM (42) with default parameters to compute P(r) distributions. The value of the input parameter Dmax, the maximum intramolecular distance, is determined by varying Dmax in steps of 2–5 Å. The appropriate value of Dmax yields a solution that (a) fits the experimental scattering profile I(q) with a (b) smooth and (c) strictly positive P(r) distribution. The output files of the program GNOM can serve as input to a suite of programs developed by Svergun and coworkers, see later. A perl script to run series of GNOM transformations with different values for Dmax is available online at http://drizzle.stanford.edu/scripts.html or can be obtained from the authors upon request.

3.6.1. Indirect Fourier Transforms

3.6.2. Structure Reconstruction

Early SAXS work on RNA was primarily limited to the use of Rg and Dmax to follow structural transitions and to manual model building by considering the calculated P(r) functions (43). However, it is possible to determine the low-resolution shape of RNA molecules from SAXS data by using ab initio structure reconstruction algorithms (28). The basic principle behind these algorithms is to represent the molecular shape by a set of dummy atoms or beads (25–27). For each configuration of beads a

B

1

C 0

log(I/I(0))

I / I(0)

A

Lipfert, Herschlag, and Doniach

0.5

q2× I/I(0)

154

−1

10

x 10−4

5

−2 0

0

0.1

q

0.2

(A−1)

0.3

0

0.1

q

0.2

(A−1)

0.3

0

0.1

q

0.2

0.3

(A−1)

Fig. 2. Scattering profiles for the unfolded (thin solid line), high Mg2+ and no glycine (thin dashed line), and glycine-bound (thick gray line) conformations of the VCI-II tandem glycine aptamer. Data are shown in (a) as linear I(q) plots, (b) as logarithmic log(I(q)) plots, and (c) in Kratky representation [q2 × I(q) as a function of q].

theoretical scattering profiles is computed and compared to the experimentally measured I(q). The programs use nonlinear optimization strategies to iteratively update the bead models by adding, removing, and translating beads to improve the fit to the experimental data. Here, we describe use of the software DAMMIN (26), which uses simulated annealing to update the model and imposes a compactness criterion. DAMMIN uses the output of the program GNOM (see earlier), the default model search volume is a sphere with a diameter given by Dmax in the GNOM file. We use DAMMIN with default parameters and a Dmax value that is 10–20 Å larger than the optimal value determined in the GNOM analysis, to ensure a sufficiently large search volume. As RNA molecules tend to adopt elongated shapes (44), it can in some cases increase the model resolution to run additional reconstructions starting from a cylindrical search volume. DAMMIN computes the fit of the model to the data and it is advisable to inspect the agreement between the scattering profile of the reconstruction and the experimental I(q). As the relationship between 1D scattering profiles and 3D models is not unique, it is advantageous to test the robustness of the shape reconstructions for a particular scattering profile by running 10–15 reconstructions with different initial random seeds. We compare the resulting models by computing pairwise Normalized Spatial Discrepancy (NSD) values with the program SUPCOMB (45). Models that have NSD values £1 are considered similar and indicate repeatable reconstruction runs. In this case, SUPCOMB can be used to generate an averaged model from the individual reconstructions that corresponds to the union of all models and a “filtered” model that corresponds to the consensus model from all runs. Finally, we recommend to compute series of reconstructions and to compare the consensus models for several repeat measurements of the same sample. In the study of the VCI-II glycineriboswitch construct, we have obtained NSD values <1 both for

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering

155

the comparison of models from repeat DAMMIN runs against the same scattering profile and for reconstructions using different experimental profiles for the same solution condition (24). The resulting models provide low-resolution electron densities maps for all three thermodynamic states of the riboswitch (Fig. 3). The programs GNOM, DAMMIN, and SUPCOMB are available from the website of the Svergun group in Hamburg (http://www.embl-hamburg.de/ExternalInfo/Research/Sax/ software.html). Perl scripts to conveniently run series of DAM MIN reconstructions and to average the results are available online at http://drizzle.stanford.edu/scripts.html or can be obtained from the authors upon request.

A

B

C

90°

90°

Fig. 3. Low-resolution structural models for the VCI-II riboswitch tandem aptamer under different solution conditions. Average unfolded conformation (column A ), conformation in the presence of 10 mM magnesium and absence of glycine (column B ), and glycine-bound structure (column C ). The first three rows show the “filtered” structure (see Subheading 3.6 for details of the reconstruction procedure) for each of the conformations in three different orientations. Black scale bars in each column correspond to 20 Å, the diameter of an A-form RNA helix. The rendered densities were generated by convoluting the bead models with a Gaussian kernel using the program Situs (46, 47). The last row shows the “filtered” models as beads and the convex hull of all bead models for a given conformation as a transparent surface. Reproduced from ref. 24 with permission from Elsevier Limited.

156

Lipfert, Herschlag, and Doniach

3.6.3. Electrostatic Calculations Using Bead Models as Input

Recently, there has been substantial progress in describing ion-RNA interactions and ion-dependent RNA folding transitions from first principles using electrostatic theories such as Poisson–Boltzmann (PB) (1, 48–51). Electrostatic modeling requires a representation of the molecular geometry of the RNA molecule of interest. Most often, atomic resolution models derived from crystallography are used (48–51). However, this approach is problematic for molecules that have not been crystallized or for conformations for which no high-resolution structure is available. An attractive alternative approach is to use the bead models derived from ab initio SAXS shape reconstructions to define the molecular geometry for electrostatic calculations. We have recently shown that SAXS-derived bead models with uniformly assigned charges can be used in PB calculations to predict ion association to nucleic acids (28). The accuracy of predictions based on bead models is similar to those obtained with atomic resolution models (28). Application of bead model electrostatic calculations to the modeling of the salt dependence of RNA folding is currently under way (Jan Lipfert, Adelene Y.-L. Sim, Daniel Herschlag, and Sebastian Doniach, in preparation). A script to convert DAMMIN models to the PQR format appropriate for PB calculations with the Adaptive Poisson Boltzmann Solver (52) is available online at http://drizzle.stanford.edu/scripts. html or can be obtained from the authors upon request.

4. Notes 1. After in vitro transcription and gel purification, the purity of the RNA construct should be confirmed minimally on a denaturing acrylamide gel, using “Stains All” solution (SigmaAldrich). Additional tests using native gels or a functional assay, if available, are desirable. 2. Absorption coefficients can be calculated from the base sequence (see, e.g., http://www.idtdna.com/analyzer/Applications/OligoAnalyzer/Default.aspx). 3. Airlines no longer allow dry ice on board of aircraft, so if travel to the synchrotron source is by plane, the samples need to be shipped on dry ice in advance of the beam time. As storage in Mg2+-containing buffer has the potential to degrade RNA, it might be preferable to ship the RNA on dry ice in water or low salt buffer and to prepare aliquots immediately before measurement in the case of sensitive constructs or international shipping.

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering

157

4. Two quantities are commonly used for the momentum transfer, q = 4psin(θ)/l and s = 2sin(θ)/l, where 2θ is the total scattering angle and l the X-ray wavelength. The quantities q and s have the same units, length−1, typically in Å−1 or nm−1, and they are related by 2ps = q. Some of the equations commonly used in the analysis of SAXS data (such as the Guinier equation 3 and the Debye formula (34)) can be written more compactly if q is used. Use of s, in contrast, has the advantage that features at a particular value of s correspond directly to the length scale given by s−1. In this chapter, we consistently use q = 4psin(θ)/l.

Acknowledgments We thank Yu Bai, Rhiju Das, Nathan Boyd, Adelene Y.-L. Sim, Mona Ali, Vincent B. Chu, Rebecca Fenn, Sönke Seifert, and the members of the Herschlag group for useful discussions and for help with data collection and sample preparation. This research was supported by the National Institutes of Health Grant PO1 GM0066275. Use of the Advanced Photon Source was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC0206CH11357. This research used resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. References 1. Draper, D.E., Grilley, D., and Soto, A.M. (2005) Ions and RNA folding. Annu. Rev. Biophys. Biomol. Struct. 34, 221–243. 2. Woodson, S.A. (2005) Metal ions and RNA folding: a highly charged topic with a dynamic future. Curr. Opin. Chem. Biol. 9, 104–109. 3. Sclavi, B., Woodson, S., Sullivan, M., Chance, M., and Brenowitz, M. (1998) Following the folding of RNA with time-resolved synchrotron X-ray footprinting. Methods Enzymol. 295, 379–402. 4. Brenowitz, M., Chance, M.R., Dhavan, G., and Takamoto, K. (2002) Probing the structural dynamics of nucleic acids by quantitative time-resolved and equilibrium hydroxyl radical “footprinting”. Curr. Opin. Struct. Biol. 12, 648–653.

5. Furtig, B., Buck, J., Manoharan, V., Bermel, W., Jaschke, A., Wenter, P., Pitsch, S., and Schwalbe, H. (2007) Time-resolved NMR studies of RNA folding. Biopolymers 86, 360–383. 6. Walter, N.G. (2001) Structural dynamics of catalytic RNA highlighted by fluorescence resonance energy transfer. Methods 25(1), 19– 30. 7. Kim, H.D., Nienhaus, G.U., Ha, T., Orr, J.W., Williamson, J.R., and Chu, S. (2002) Mg2+ dependent conformational change of RNA studied by fluorescence correlation and FRET on immobilized single molecules. Proc. Natl Acad. Sci. U. S. A. 99(7), 4284–4289. 8. Russell, R., Millett, I.S., Doniach, S., and Herschlag, D. (2000) Small angle X-ray scattering

158

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Lipfert, Herschlag, and Doniach reveals a compact intermediate in RNA folding. Nat. Struct. Biol. 7(5), 367–370. Russell, R., Millett, I.S., Tate, M.W., Kwok, L.W., Nakatani, B., Gruner, S.M., Mochrie, S.G., Pande, V., Doniach, S., Herschlag, D., and Pollack, L. (2002) Rapid compaction during RNA folding. Proc. Natl Acad. Sci. U. S. A. 99(7), 4266–4271. Takamoto, K., Das, R., He, Q., Doniach, S., Brenowitz, M., Herschlag, D., and Chance, M.R. (2004) Principles of RNA compaction: insights from the equilibrium folding pathway of the P4-P6 RNA domain in monovalent cations. J. Mol. Biol. 343, 1195–1206. Russell, R., Zhuang, X., Babcock, H.P., Millett, I.S., Doniach, S., Chu, S., and Herschlag, S. (2002) Exploring the folding landscape of a structured RNA. Proc. Natl Acad. Sci. U. S. A. 99(1), 155–160. Fang, X.-W., Littrell, K., Yang, X., Henderson, S.J., Seifert, S., Thiyagarajan, P., Pan, T., and Sosnick, T.R. (2000) Mg2+-dependent compaction and folding of yeast tRNA and the catalytic domain of the B. subtilis RNase P RNA determined by small-angle X-ray scattering. Biochemistry 39, 11107–11113. Fang , X.-W. , Golden , B.L., Littrell, K., Shelton, V., Thiyagarajan, P., Pan, T., and Sosnick, T.R. (2001) The thermodynamic origin of a thermophilic ribozyme. Proc. Natl Acad. Sci. U. S. A. 98(8), 4355–4360. Chauhan, S., Caliskan, G., Briber, R.M., Perez-Salas, U., Rangan, P., Thirumalai, D., and Woodson, S.A. (2005) RNA tertiary interactions mediate native collapse of a bacterial group I ribozyme. J. Mol. Biol. 353(5), 1199–1209. Kwok, L.W., Shcherbakova, I., Lamb, J.S., Park, H.Y., Andresen, K., Smith, H., Brenowitz, M., and Pollack, L. (2006) Concordant exploration of the kinetics of RNA folding from global and local perspectives. J. Mol. Biol. 355(2), 282–293. Mandal, M. and Breaker, R.R. (2004) Gene regulation by riboswitches. Nat. Rev. Mol. Cell. Biol. 5, 451–463. Winkler, W.C. and Breaker, R.R. (2005) Regulation of bacterial gene expression by riboswitches. Annu. Rev. Microbiol. 59, 487–517. Soukup, J.K. and Soukup, G.A. (2004) Riboswitches exert genetic control through metabolite-induced conformational change. Curr. Opin. Struct. Biol. 14, 344–349. Vitreschak, A.G., Rodionov, D.A., Mironov, A.A., and Gelfand, M.S. (2004) Riboswitches: the oldest mechanism for the regulation of gene expression. Trends Genet. 20, 44–50.

20. Coppins, R.L., Hall, K.B., and Groisman, E.A. (2007) The intricate world of riboswitches. Curr. Opin. Microbiol. 10, 176–181. 21. Schwalbe, H., Buck, J., Furtig, B., Noeske, J., and Wohnert, J. (2007) Structures of RNA switches: insight into molecular recognition and tertiary structure. Angew. Chem. Int. Ed. Engl. 46, 1212–1219. 22. Edwards, T.E., Klein, D.J., and FerreD’Amare, A.R. (2007) Riboswitches: smallmolecule recognition by gene regulatory RNAs. Curr. Opin. Struct. Biol. 17, 273–279. 23. Mandal, M., Lee, M., Barrick, J.E., Weinberg, Z., Emilsson, G.M., Ruzzo, W.L., and Breaker, R.R. (2004) A glycine-dependent riboswitch that uses cooperative binding to control gene expression. Science 306, 275–279. 24. Lipfert, J., Das, R., Chu, V.B., Kudaravalli, M., Boyd, N., Herschlag, D., and Doniach, S. (2007) Structural transitions and thermodynamics of a glycine-dependent riboswitch from Vibrio cholerae. J. Mol. Biol. 365, 1393–1406. 25. Chacon, P., Moran, F., Diaz, J.F., Pantos, E., and Andreu, J.M. (1998) Low-resolution structures of proteins in solution retrieved from X-ray scattering with a genetic algorithm. Biophys. J. 74, 2760–2775. 26. Svergun, D.I. (1999) Restoring low resolution structure of biological macromolecules from solution scattering using simulated annealing. Biophys. J. 76, 2879–2886. 27. Walther, D., Cohen, F.E., and Doniach, S. (2000) Reconstruction of low resolution three-dimensional density maps from onedimensional small angle X-ray scattering data for biomolecules in solution. J. Appl. Cryst. 33, 350–363. 28. Lipfert, J., Chu, V.B., Bai, Y., Herschlag, D., and Doniach, S. (2007) Low resolution models for nucleic acids from small-angle X-ray scattering with applications to electrostatic modeling. J. Appl. Cryst. 40, S229–S235. 29. Hartmann, H.R., Bindereif, A., Schön, A., and Westhof, E. (2005) Handbook of RNA Biochemistry, Wiley-VCH, Weinheim. 30. Lipfert, J., Millett, I.S., Seifert, S., and Doniach, S. (2006) A sample holder for small-angle X-ray scattering static and flow cell measurements. Rev. Sci. Inst. 77, 46108. 31. Chu, V.B., Yu, B., Lipfert, J., Pande, V.S., Herschlag, D., and Doniach, S. (2008) Critical assessment of nucleic acid electrostatics via experimental and computational investigation of an unfolded state ensemble. J. Am. Chem. Soc. 130, 12334–12341 32. Bai, Y., Das, R., Millett, I.S., Herschlag, D., and Doniach, S. (2005) Probing counterions

Riboswitch Conformations Revealed by Small-Angle X-Ray Scattering

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

modulated repulsion and attraction between nucleic acid duplexes in solution. Proc. Natl Acad. Sci. U. S. A. 102(4), 1035–1040. Huang, T.C., Toraya, H., Blanton, T.N., and Wu, Y. (1993) X-ray-powder diffraction analysis of silver behenate, a possible low-angle diffraction standard. J. Appl. Cyrst. 26, 180–184. Svergun, D.I. and Koch, M.H.J. (2003) Small-angle scattering studies of biological macromolecules in solution. Rep. Prog. Phys. 66, 1735–1782. Guinier, A. (1939) La diffraction des rayons X aux très petits angles: Application à l’étude de phénomènes ultramicroscopiques. Ann. Phys. (Paris), 12, 161–237. Mylonas, E. and Svergun, D.I. (2007) Accuracy of molecular weight determination of proteins in solution by small-angle X-ray scattering. J. Appl. Cyrst. 40, S245–S249. Doniach, S. (2001) Changes in biomolecular conformations seen by small angle X-ray scattering. Chem. Rev. 101, 1763–1778. Henry, E.R. and Hofrichter, J. (1992) Singular value decomposition: application to analysis of experimental data. Methods Enzymol. 210, 129–192. Segel, D.J., Fink, A.L., Hodgson, K.O., and Doniach, S. (1998) Protein denaturation: a small-angle X-ray scattering study of the ensemble of unfolded states of cytochrome c. Biochemistry 37, 12443–12451. Lipfert, J., Columbus, L., Chu, V.B., and Doniach, S. (2007) Analysis of small-angle X-ray scattering data of protein–detergent complexes by singular value decomposition. J. Appl. Cryst. 40, S235–S239. Dantas, G., Watters, A.L., Lunde, B., Eletr, Z., Isern, N., Lipfert, J., Doniach, S., Kuhlman, B., Stoddard, B.L., Varani, G., and Baker, D. (2006) Mistranslation fragment of an in silico designed novel-fold protein forms and exceptionally stable symmetric homodimer with a high-affinity interface. J. Mol. Biol. 362, 1004–1024. Svergun, D.I. (1992) Determination of the regularization parameter in indirect-transform

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

159

methods using perceptual criteria. J. Appl. Cryst. 25, 495–503. Lipfert, J. and Doniach, S. (2007) Small-angle X-ray scattering from RNA, proteins, and protein complexes. Ann. Rev. Biophys. Biomol. Struct. 36, 307–327. Hyeon, C., Dima, R.I., and Thirumalai, D. (2006) Size, shape, and flexibility of RNA structures. J. Chem. Phys. 125(19), 194905. Kozin, M.B. and Svergun, D.I. (2001) Automated matching of high- and low-resolution structural models. J. Appl. Cryst. 34, 33–41. Wriggers, W., Milligan, R.A., and McCammon, J.A. (1999) Situs: a package for docking crystal structures into low-resolution maps from electron microscopy. J. Struct. Biol. 125, 185–195. Wriggers, W. and Chacón, P. (2001) Using situs for the registration of protein structures with low-resolution bead models from X-ray solution scattering. J. Appl. Cryst. 34, 773–776. Misra, V.K. and Draper, D.E. (2000) Mg2+ binding to tRNA revisited: the nonlinear Poisson–Boltzmann model. J. Mol. Biol. 299, 813–825. Grilley, D., Soto, A.M., and Draper, D.E. (2006) Mg2+ RNA interaction free energies and their relationship to the folding of RNA tertiary structures. Proc. Natl Acad. Sci. U. S. A. 103, 14003–14008. Chu, V.B., Bai, Y., Lipfert, J., Herschlag, D., and Doniach, S. (2007) Evaluation of ion binding to DNA duplexes using a size-modified Poisson–Boltzmann theory. Biophys. J. 93(9), 3202–3209. Bai, Y., Travers, K., Chu, V.C., Lipfert, J., Doniach, S., and Herschlag, D. (2007) Quantitative and comprehensive decomposition of the ion atmosphere around nucleic acids. J. Am. Chem. Soc. 129, 12427–12438. Baker, N.A., Sept, D., Joseph, S., Holst, M.J., and McCammon, J.A. (2001) Electrostatics of nanosystems: applications to microtubules and the ribosome. Proc. Natl Acad. Sci. U. S. A. 98, 10037–10041.

Chapter 12 Time-Resolved NMR Spectroscopy: Ligand-Induced Refolding of Riboswitches Janina Buck, Boris Fürtig, Jonas Noeske, Jens Wöhnert, and Harald Schwalbe Summary A detailed understanding of cellular mechanisms requires knowledge of structure and dynamics of the involved biomacromolecules at atomic resolution. NMR spectroscopy uniquely allows determination of static and dynamic processes at atomic level, including structured states often represented by a single state as well as by unstructured conformational ensembles. While a high-resolution description of structured states may also be obtained by other techniques, the characterization of structural transitions occurring during biomolecular folding is only feasible exploiting NMR spectroscopic methods. The NMR methodical strategy includes the fast initiation of a folding reaction in situ and the possibility to detect the induced process with sufficient time resolution on the respective NMR time scale. In the case of ligand-induced structural transitions of RNA, the initiation of the folding reaction can be achieved by laser-triggered deprotection of a photolabile caged ligand whose release induces folding of a riboswitch RNA. The strategy discussed here is general and can also be transferred to other biological processes, where at least one key reagent or substrate, e.g., ions, ligands, pH, or one specific conformational state, can be photochemically caged. The rates of reversible and irreversible reactions or structural transitions that can be covered by real-time NMR methods range from milliseconds up to hours. In this chapter, we discuss the application of a time-resolved NMR strategy to resolve the ligand-induced folding of the guanine-sensing riboswitch aptamer domain of the B. subtilis xpt-pbuX operon. Key words: NMR spectroscopy, Riboswitch RNA, Photolabile caged compound, RNA folding, Time-resolved NMR

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_12 © Humana Press, a part of Springer Science + Business Media, LLC 2009

161

162

Buck et al.

1. Introduction Deciphering cellular mechanisms requires insight into the static conformations of the involved biomacromolecules and into the structural transitions that occur upon a functional modification due to stimuli including intramolecular interactions, pH, or temperature jumps, changes of ionic strength or irradiation with light to name a few. NMR spectroscopy has the capacity to resolve both static and dynamic processes in atomic detail, including their detection on a variety of different time scales. Resolving folding events that take place on the time scale of milliseconds up to hours can be characterized at atomic resolution by NMR spectroscopic methods as real-time one-dimensional (1D), two-dimensional (2D), and three-dimensional (3D), diffusion and photo-CIDNP experiments (1). This span of time scales comprises biological processes such as conformational transitions, folding events, and macromolecular assembly. Molecular size may be a limitation of NMR studies as residue-specific analysis of kinetic data can only be obtained provided that sufficient spectral resolution can be achieved. The real-time NMR studies discussed here have been applied to resolve the ligand-induced folding of the guaninesensing riboswitch aptamer domain (GSRapt) of the B. subtilis xpt-pbuX operon (2). The RNA construct comprises 73 nucleotides. As this size of an RNA is at the upper limit for NMR analysis, the chemical shift dispersion had to be improved using selectively isotope labeled RNA (14N-guanosine and 15N-uridine labeled RNA) prepared by enzymatic in vitro transcription techniques (3). The signals of imino-hydrogen atoms of guanine and uracil residues resonate in the spectral region of 10–15 ppm (1H chemical shift). These NMR signals are direct NMR spectroscopic reporters of base-pairing interactions in RNA as they are only detectable when protected from exchange with the solvent water (3). Therefore, these signals are used to probe the ligandinduced folding of the GSRapt-RNA. Since the chemical shift dispersion of the signals of unlabeled GSRapt-RNA in one-dimensional NMR spectra was not sufficient, selectively 15N-uridine labeled RNA was prepared. The introduction of an additional NMR active heteronucleus to the RNA combined with NMR filter techniques (4) allowed editing of the iminoproton NMR resonances according to nucleotide type (see Fig. 1). Thus, the alternate detection of either the 14N-bound proton signals (here: guanosine residues) or the 15N-bound proton signals (here: uridine residues) is possible, and therefore the spectral resolution in the one-dimensional spectra could be improved (see Fig. 2). The principle of “caging” molecules namely blocking a functional group with a photolabile protecting group is widely applied to a number of different biological systems (5). This approach

Time-Resolved NMR Spectroscopy: Ligand-Induced Refolding of Riboswitches

163

Fig. 1. NMR spectroscopic setup. (a) Schematic representation of the NMR/laser coupling setup (6). Using coupling optics and an optical quartz fiber, the Ar-ion laser beam (7–9 W, 340–365 nm) is transferred into the NMR sample situated directly in the active site of the NMR spectrometer. Length of laser illumination is controlled by the spectrometer that is connected with a shutter. Maximum dispersion of the laser beam corresponding to maximal uniform illumination of the NMR sample is obtained using a quartz tip insert. (b) Pulse sequence of the NMR X-filter kinetic experiment; f1: x,x,y,y,−x,−x,−y,−y; f2: −x,−x,−y,−y,x,x,y,y; f*: x,−x (incremented interleaved in L1); frec: x,x,−x,−x;D = 1/2JNH; cos b = exp(−T/T1Imino); tL: laser duration; t: JR-delay (experimental conditions for kinetic studies of GSRapt: d1: 0.125 s, acquisition time: 0.0984 s, b: 78°, T1Imino: 0.142 s).

allows for fast light-induced initiation of the reaction by in situ release of the biological active molecule and has been applied in real-time NMR studies to investigate protein (6) as well as RNA folding (7). The regulatory principle of riboswitch RNAs is based on the conformational differences of the receptor region (aptamer domain) in the presence and absence of the respective ligand which further influences the 3¢-downstream sequence (expression platform) that is responsible for gene regulation in the cellular context. The guanine-sensing riboswitch aptamer domain is known to be a specific receptor region for the ligands guanine (Kd ~ 5 nM) and hypoxanthine (Kd ~ 50 nM) (8). The binding mode of the RNA–ligand complex is defined through basepairing interactions of almost all functional groups of the ligand to the RNA. The crucial interaction defining the specificity of the molecular recognition is given by a Watson–Crick base-pairing interaction between the purine ligand and a cytidine residue of the RNA (9–11). The strategy to resolve the ligand-induced conformational changes of the GSRapt presented here is based on the introduction

164

Buck et al.

Fig. 2. NMR iminoproton spectra of GSRapt. (a) Unlabeled free RNA before laser pulse. Inset, chemical structure of photocaged hypoxanthine (DMNPE-hypoxanthine). (b) Unlabeled RNA–ligand complex after laser pulse. Inset, chemical structure of hypoxanthine. (c and d) 15N-uridine labeled RNA–ligand complex with annotated NMR resonance assignment of resolved residues (17): (c) 1H{15N}-detection results in spectrum of uridine resonances, and (d) 1H{14N}-detection results in spectrum of guanosine resonances.

of a covalently attached photolabile group blocking the Watson– Crick edge of the ligand and therefore preventing the formation of the RNA–ligand complex (see Fig. 2). The experimental setup for in situ laser-induced deprotection of the caged ligand within the NMR spectrometer (6) is realized by direct coupling of laser optics (see Fig. 1). Fast initiation and subsequent NMR spectroscopic detection of the binding event with residue-specific resolution is thus feasible. Information obtained from time-resolved NMR studies may include both the overall kinetic behavior of the biological system and the residue-specific kinetic characterization of the folding event. In addition, thermodynamic parameters and detailed analysis of reaction mechanisms can be derived (1). In the presented example (2), incorporation of the time-resolved NMR data into experimentally restrained molecular dynamics simulations provided insight into the RNA structural ensembles involved in each step during the ligand-induced conformational transition (see Fig. 3).

Time-Resolved NMR Spectroscopy: Ligand-Induced Refolding of Riboswitches

165

2. Materials 2.1. RNA Preparation

1. Transcription reaction: 200 mM Tris–glutamic acid buffer, pH 8.1; 20 mM dithiothreitol (in water, store at −20°C); 2 mM spermidine (in water, store at −20°C); T7 RNA polymerase (12); magnesium acetate (concentration range: 5–30 mM); linearized DNA template (concentration range: 25–30 nM); rNTP-mix (concentration range: 10–40 mM); for 15N-uridine labeled RNA: unlabeled rGTP, rATP, rCTP (Sigma-Aldrich, Munich, Germany), and 15N-rUTP (Silantes, Munich, Germany). 2. DEAE-Sepharose column (Amersham Pharmacia Biotech, Uppsala, Sweden). 3. DEAE equilibration solution: 0.1 M Na-acetate, pH 5.5; gradient solutions: 0.6–3 M Na-acetate, pH 5.5. 4. Ethanol (precooled to −80°C). 5. HPLC column: Vydac-RP18, 10 × 250 mm; temperature: 60°C; flow rate: 5 mL/min. 6. HPLC buffer A: 50 mM potassium phosphate, pH 5.9, 2 mM tetrabutyl-ammoniumhydrogen sulfate; buffer B: buffer A supplemented with 60% (v/v) acetonitrile. 7. Micro concentrators (Vivaspin, Sartorius, Göttingen, Germany). 8. LiClO4/acetone solution: 2% (w/v) LiClO4 in acetone. 9. NMR buffer: 25 mM potassium phosphate, 50 mM potassium chloride, pH 6.2.

2.2. NMR Spectroscopy

1. The NMR experiments are recorded on a Bruker NMR spectrometer (Bruker, Rheinstetten, Germany) equipped with a 5-mm z-axis gradient TXI-HCN cryogenic probe (AV800MHz, S/N ~ 1,000:1 on standard sucrose samples). 2. Data acquisition and processing is done with the software TOPSPIN (Bruker, Karlsruhe, Germany). 3. Data analysis is done with the software felix2000 (Accelrys, San Diego, CA) and SigmaPlot 9.0. 4. All NMR spectra are recorded in H2O/D2O (9:1) at 283 K (NMR samples: shigemi tube with sample volume of 220 mL) using pulse sequences with WATERGATE water suppression (13) or jump-return-echo pulse sequences (14). 5. The laser setup includes the direct coupling of a quartz fiber from a CW argon ion laser (Spectra-Physics, Darmstadt, Germany) into the NMR shigemi tube that is equipped with a quartz tip insert (see Fig. 1) (6). In order to illuminate the sample with a laser pulse of defined length, the setup includes

166

Buck et al.

a shutter that is controlled by an electronic signal generated at distinct time points during the NMR experiment and controlled by NMR pulse program software. 6. NMR samples containing the synthesized (2) photolable caged compound (DMNPE-hypoxanthine) are prepared in the dark.

3. Methods 3.1. RNA Preparation

The GSRapt-RNA construct comprises 73 nucleotides (see Fig. 3). In order to obtain site-resolved kinetic information, selectively 15N-uridine labeled RNA is prepared in adequate amounts by enzymatic in vitro transcription using T7 RNA polymerase. The linearized DNA plasmid, including the T7 RNA polymerase

Fig. 3. Time-resolved NMR results. (a) GSRapt secondary and (b) tertiary structure including the time-resolved NMR results. dark gray: half-life values (t1/2 (s)) in time range 18.9–23.6 s; light gray: half-life values in time range 27.1–30.7 s; asterisk: overlaid signal; hyp hypoxanthine; helices P1, P2, P3 and loop regions L2 and L3 are labeled according to Breaker et al. (8). Gray solid lines: Watson–Crick base-pairing interactions; gray dashed lines: noncanonical base-pairing interactions. (c–e) Normalized integrals of individual iminoproton signals during time course of reaction shown with monoexponential (c and d) and linear (e) fits (solid line). (c) Core region signal of U51/U67 depicted in dark gray. (d) Loop region signal of G37/G38/G45 shown in light gray. (e) Signal of U81 that is part of helix P1 (remains constant in the free and the ligand-bound state of the RNA) shown in black color.

Time-Resolved NMR Spectroscopy: Ligand-Induced Refolding of Riboswitches

167

promoter sequence, serves as DNA template. The in vitro transcription reaction is firstly purified by a DEAE-Sepharose anionexchange column where most of the not-incorporated nucleotide triphosphates, the polymerase, and abortive products can be removed. To finally isolate the product RNA, preparative ionexchange reverse-phase HPLC is performed. RNA-containing fractions are freeze dried, desalted, and finally folded into a homogenous monomeric conformation. Correct folding can be verified by native gel electrophoresis. Before using the RNA in the NMR studies, it is exchanged into NMR buffer. 1. The transcription reaction is mixed using the standard reaction conditions and the optimized reaction conditions. Experiments optimizing the magnesium acetate concentration, the DNA template concentration, and the concentration of the rNTP-mix (including the four rNTPs in relative amounts) may be performed in transcription reactions at analytical scale before (see Note 1). The transcription reaction mixture is stirred at 37°C until the reaction yield is not further increasing (reaction time course is followed by analytical HPLC, see Note 2). Magnesium phosphate precipitation is removed by centrifugation at the end of the reaction. 2. The first purification step is done by a DEAE-Sepharose column. After equilibration of the DEAE-Sepharose with the equilibration solution, the transcription reaction is loaded and the RNA is eluted using a 0.6–3.0 M Na-acetate gradient. Here, the riboswitch RNA fractions are typically eluted at ~2 M Na-acetate. The fractions are collected and analyzed using UV spectroscopy at 260 nm and denaturing polyacrylamide gel electrophoresis. RNA-containing fractions are combined and the RNA is precipitated at −20°C overnight by adding cooled ethanol (−80°C). Finally, the precipitated RNA is centrifuged at 8,000 × g; the pellet is air dried and dissolved in water. 3. The RNA is then purified using ion-exchange reverse-phase HPLC (buffers A and B). An acetonitrile gradient (0–60%) is applied and, e.g., riboswitch RNA is eluted at ~25% acetonitrile. RNA-containing fractions are characterized by their UV absorption properties at 260 nm and collected accordingly (see Note 3). 4. After freeze drying and desalting using micro concentrators according with manufacturers’ instructions, the RNA is precipitated with 5 volumes of a LiClO4/acetone solution at −20°C overnight. The RNA pellet is then collected by centrifugation at 8,000 × g and subsequently dissolved in 1 mL water. The RNA is folded into its monomeric conformation by denaturation at 95°C for 5 min and a fast cooling step (1:10 dilution with ice-cooled water and incubation on ice for 1 h) (see Note 4). Finally, the RNA is exchanged into NMR buffer using micro concentrators.

168

Buck et al.

3.2. NMR Spectroscopy

The real-time experiments are designed as pseudo-3D NMR experiments, composed of a series of consecutive one-dimensional NMR spectra (see Fig. 1). Thereby, the dimensions of the pseudo-3D dataset correspond to the chemical shift and the real time, where one plane is recorded before and k planes after the laser trigger. Investigation of the laser-induced time-resolved NMR studies requires the optimization of a number of experimental parameters. 1. The molecular ratio between RNA and ligand needs to be defined precisely and reproducibly. Determination of the photolabile caged ligand concentration is done by the ERETIC NMR experiment (15) using GTP solutions of different concentration as reference for calibration. 2. The experimental constraint to have a short dead time of the reaction, namely a short laser pulse, together with a quantitative photolytic deprotection of the ligand has to be optimized. Due to signal overlap in the presence of RNA, the efficiency of caged ligand release is determined on an NMR sample of caged hypoxanthine alone, but under identical conditions. Thus, ligand concentration, buffer conditions, and temperature should correspond to the conditions of the time-resolved NMR experiments in the presence of RNA. The rate of photolytic deprotection is analyzed integrating the proton signals (H2/H8) of the ligand before and after laser irradiation, respectively. Here, an optimum of photolytic deprotection of (DMNPE) hypoxanthine of 80% could be achieved with a laser irradiation of 1.5 s at a ligand concentration of 250 mM (see Note 5). 3. The time resolution of the kinetic experiments is limited by the conditions that restrict a sufficient signal-to-noise ratio of every single one-dimensional spectrum. Therefore, the acquisition parameters (~experiment time) have to be optimized with regard to maximal temporal resolution. This optimization mainly includes the relaxation delay (see Note 6), the number of scans, and the concentration of the sample (see Note 7). The kinetic experiments to characterize RNA folding are carried out as follows: the first 128 data points (8 scans, each point in an interval of 2.1 s) are recorded prior to illumination resulting in the determination of the reaction baseline. Reaction initiation is then induced by release of hypoxanthine (laser pulse: 1.5 s, 350 nm, 4 W) followed by another 128 data points (2.1 s per data point) that contain the kinetic information. After processing of the resulting pseudo-3D data-set and Fourier transformation in w3, including exponential multiplication with a line-broadening factor of 20 Hz (line-broadening factor equal 1.3 of the line width at half height for optimal signal-to-noise), phase correction, and baseline correction in w3, 3–5 individual experiments (recorded under identical conditions) with 256 1D spectra, respectively, are coadded to improve the sensitivity of the timeresolved NMR studies.

Time-Resolved NMR Spectroscopy: Ligand-Induced Refolding of Riboswitches

169

4. The spectral resolution of the one-dimensional NMR studies is limited by the size of the analyzed RNA and therefore the number of NMR signals. Although the iminoproton region of RNA in general represents a region of high intrinsic spectral resolution, it is still restricted to a few resolved residues of the 73-nt GSRaptRNA. The use of selectively 15N-uridine labeled RNA in combination with the application of NMR X-filter techniques increases the resolution and allows detection of site-resolved information by one-dimensional real-time NMR experiments. Detection of the 15N-bound and the 14N-bound iminoprotons during the kinetic experiment in a collective manner is realized by selection of the different spin systems through introduction of two 15N-90°pulses in the NMR pulse sequence (see Fig. 1), where the phase of the first one is incremented interleaved in the number of the subsequent 1D spectra (f* = x,−x). Thus, the different spin systems (1H–14N and 1 H–15N) show opposite signs in even data points while both have the same sign in odd data points. The 1H,14N-edited kinetic (here: kinetics of the guanosine residues) may now be extracted by addition of the odd and even data points, while the difference between odd and even data points results in determination of the 1H,15N-edited kinetic (here: kinetics of uridine residues). With an initially recorded kinetic spectrum of 256 data points in total, this transformation results in two separated subspectra with 128 data points (64 data points prior to and 64 data points following illumination), respectively. 5. Following processing of the real-time NMR spectra, the time constants (k (s−1)) of the individual iminoproton signals may be extracted by integration of their line width at half height. The analysis starts with the 64th data point (last data point prior to illumination) of each of the two subspectra of guanosine and uridine residues, respectively. The integrals of the upcoming signals are corrected by the zero point of the kinetics and normalized by the averaged values of the last 20 residual data points. The resulting values are fitted with a monoexponential function (k is the single fit parameter). The half-life values (t1/2 (s)) are obtained using the formula for a first-order process: t1/2 = ln 2/k.

4. Notes 1. Transcription parameters: it is best to optimize concentrations of the variable parameters in the following order: magnesium acetate, rNTP-mix and DNA-template and keep all other parameters constant, respectively. 2. The transcription time course may also be followed by analytical denaturing gel electrophoresis.

170

Buck et al.

3. The HPLC gradient has to be optimized for each RNA construct. Instead of using preparative HPLC, isolation of the appropriate RNA may also be possible by preparative denaturing polyacrylamide gel electrophoresis. 4. There is no general effective RNA folding procedure. The conditions therefore have to be optimized for each RNA construct individually. However, denaturation followed by fast or slow cooling steps or by dilution in water or optimized buffer may be useful starting points to obtain a homogenous monomeric RNA conformation. 5. Increase of laser irradiation time may tend to induce local heating effects of the sample, an effect that can be monitored on the lock signal. 6. The sensitivity of a single one-dimensional spectrum of the kinetic NMR experiment may be optimized by selective excitation of iminoproton spin magnetization under Ernst angle conditions (16) in order to improve time resolution (see Fig.1). 7. The signal-to-noise ratio can be improved by coadding several separately recorded spectra (improvement of factor √ n, where n is the number of summed spectra). The reproducibility of the laser-induced deprotection allows for preparation and detection of several samples under identical conditions.

Acknowledgments The work was supported by the DFG (SFB 579: “RNA-LigandInteraction”), the state of Hesse (Center for Biomolecular Magnetic Resonance (BMRZ)), the “Studienstiftung des Deutschen Volkes” (predoctoral fellowship for B.F.) and the E.U. (“EUNMR-European Network of Research Infrastructures for Providing Access and Technological Advancement in Bio-NMR”, FP-2005-RII3-Contract-no. 026145).

References 1. Fürtig, B., Buck, J., Manoharan, V., et al. (2007). Time-resolved NMR studies of RNA folding. Biopolymers 86, 360–383. 2. Buck, J., Fürtig, B., Noeske, J., Wöhnert, J., Schwalbe, H. (2007). Time-resolved NMR methods resolving ligand-induced RNA folding at atomic resolution. Proc. Natl Acad. Sci. U. S. A. 104, 15699–15704.

3. Fürtig, B., Richter, C., Wöhnert, J., Schwalbe, H. (2003). NMR spectroscopy of RNA. Chembiochem 4, 936–962. 4. Otting, G., Wüthrich, K. (1989). Extended heteronuclear editing of 2D 1H NMR spectra of isotope-labeled proteins, using the X (w1, w2) double half filter. J. Magn. Reson. 85, 586–594.

Time-Resolved NMR Spectroscopy: Ligand-Induced Refolding of Riboswitches 5. Mayer, G., Heckel, A. (2006). Biologically active molecules with a “light switch”. Angew. Chem. Int. Ed. Engl. 45, 4900–4921. 6. Kühn, T., Schwalbe, H. (2000). Monitoring the kinetics of ion-dependent protein folding by time-resolved NMR spectroscopy at atomic resolution. J. Am. Chem. Soc. 122, 6169–6174. 7. Wenter, P., Fürtig, B., Hainard, A., Schwalbe, H., Pitsch, S. (2005). Kinetics of photoinduced RNA refolding by real-time NMR spectroscopy. Angew. Chem. Int. Ed. Engl. 44, 2600–2603. 8. Mandal, M., Boese, B., Barrick, J. E., Winkler, W. C., Breaker, R. R. (2003). Riboswitches control fundamental biochemical pathways in Bacillus subtilis and other bacteria. Cell 113, 577–586. 9. Serganov, A., Yuan, Y. R., Pikovskaya, O., et al. (2004). Structural basis for discriminative regulation of gene expression by adenine- and guanine-sensing mRNAs. Chem. Biol. 11, 1729–1741. 10. Noeske, J., Richter, C., Grundl, M. A., Nasiri, H. R., Schwalbe, H., Wöhnert, J. (2005) An intermolecular base triple as the basis of ligand specificity and affinity in the guanine- and adenine-sensing riboswitch RNAs. Proc. Natl Acad. Sci. U. S. A. 102, 1372–1377. 11. Batey, R. T., Gilbert, S. D., Montange, R. K. (2004). Structure of a natural guanine-responsive

12.

13.

14.

15.

16.

17.

171

riboswitch complexed with the metabolite hypoxanthine. Nature 432, 411–415. Davanloo, P., Rosenberg, A. H., Dunn, J. J., Studier, F. W. (1984). Cloning and expression of the gene for bacteriophage T7 RNA polymerase. Proc. Natl Acad. Sci. U. S. A. 81, 2035–2039. Liu, M., Mao, X., Ye, C., Huang, H., Nicholson, J. K., Lindon, J. C. (1998). Improved WATERGATE pulse sequences for solvent suppression in NMR spectroscopy. J. Magn. Reson. 132, 125–129. Sklenar, V., Bax, A. (1987). A new water suppression technique for generating pure-phase spectra with equal excitation over a wide bandwidth. J. Magn. Reson. 75, 378–383. Akoka, S., Barantin, L., Trierweiler, M. (1999). Concentration measurement by proton NMR using the ERETIC method. Anal. Chem. 71, 2554–2557. Ernst, R. R., Bodenhausen, G., Wokaun, A. (1994) Principles of Nuclear Magnetic Resonance in One and Two Dimensions. Oxford University Press, Oxford. Noeske, J., Buck, J., Fürtig, B., Nasiri, H. R., Schwalbe, H., Wöhnert, J. (2007). Interplay of “induced fit” and preorganization in the ligand induced folding of the aptamer domain of the guanine binding riboswitch. Nucleic Acids Res. 35, 572–583.

Chapter 13 Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing and Selective 2´-Hydroxyl Acylation and Primer Extension Catherine A. Wakeman and Wade C. Winkler Summary RNA sequences fold upon themselves to form complex structures. Functional analysis of most biological RNAs requires knowledge of secondary structure arrangements and tertiary base interactions. Therefore, rapid and comprehensive methods for assessing RNA structure are highly desirable. Computational tools are oftentimes employed for prediction of secondary structure. However, a greater degree of accuracy is achieved when these methods are combined alongside structure probing experimentation. Multiple probing techniques have been developed to assist identification of base-paired regions. However, most of these techniques investigate only a subset of RNA nucleotides at a time. A combination of structure probing approaches is thus required for analysis of all nucleotides within a given RNA molecule. Therefore, methods that investigate local structure for all positions in a sequence-independent manner can be particularly useful in characterizing secondary structure and RNA conformational changes. This chapter outlines protocols for two techniques, in-line probing and Selective 2¢-Hydroxyl Acylation Analyzed by Primer Extension (SHAPE), which largely accomplish this goal. Key words: Riboswitch, RNA folding, RNA secondary structure, RNA structure analysis, In-line probing, SHAPE

1. Introduction Riboswitches are RNA molecules that function as direct sensors for intracellular metabolites and metal ions (1–5). They are typically located within the 5¢-untranslated region (5¢-UTR) of the mRNA that hosts them and are responsible for regulation of the downstream genes in response to fluctuations of their metabolite or metal ion ligand(s). Specifically, association of the target metabolite Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_13 © Humana Press, a part of Springer Science + Business Media, LLC 2009

173

174

Wakeman and Winkler

(or metal ion) to the ligand-binding portion (aptamer) stabilizes an RNA conformation that either induces or represses expression of downstream gene(s) (2–5). Functional analyses of riboswitch RNAs have greatly benefited from identification of secondary structure and tertiary base interactions. This has been largely accomplished through a combination of bioinformatics-based approaches and structural probing techniques. These methods have directly assisted in the analyses of nucleotides that are involved in ligand recognition, estimation of apparent KD values, and in characterizing ligand-induced structural rearrangements (6, 7). Moreover, these predictions of secondary structure have largely been validated by the three-dimensional analyses of riboswitch aptamer domains by X-ray crystallography and NMR (2–5). In total, there are a variety of routinely used RNA structure probing techniques, each exhibiting particular strengths and weaknesses. One method of characterizing secondary structure is to partially digest the RNA substrate with RNase enzymes that exhibit specific substrate preferences (8). For example, nucleotides that are single stranded can be inferred by partial digestion of the target RNA by RNase S1, which prefers such sites. Similarly, unpaired Gs can be identified by partial digestion with RNase T1. Conversely, base-paired and stacked nucleotides can be indicated via RNase V1 cleavages. Combined, these individual enzymatic reactions can assist in the identification of helices or unpaired nucleotides. Chemical agents may also be employed for structural probing purposes, such as cleavage of single-stranded regions via lead probing or methylation of unpaired guanine N1 and N2 positions by kethoxal (9, 10). Finally, nucleotide analog interference mapping (NAIM) is a powerful method of rapidly investigating the effect(s) of substituting specific nucleotide functional groups (11). However, methods that investigate local structure in a sequence-independent manner are particularly useful in predicting RNA secondary structure and for analyzing conformational changes. Two such methods, briefly described herein, are referred to as in-line probing and Selective 2¢-Hydroxyl Acylation Analyzed by Primer Extension (SHAPE) (12–14). In-line probing provides a measurement of the relative flexibility of specific RNA linkages due to the inherent differences in their rates of spontaneous cleavage (12, 15). Linkages that exhibit greater ²flexibility ²(i.e., unpaired status) exhibit increased rates of spontaneous cleavage. Therefore, comparative analysis of the cleavage rates for all linkages of a given RNA molecule can be used to identify unpaired and base-paired regions. SHAPE also probes the relative flexibility of RNA linkages. Specifically, the nucleophilic character of the 2¢ ribose hydroxyl is influenced by its proximity to the adjacent 3¢-phosphodiester anion, which is in turn governed by the local structural context. Therefore, reactivity of the 2¢ ribose hydroxyl to an electrophile, e.g., N-Methylisatoic anhydride (13), is sensitive to local structure. Similar to in-line probing, SHAPE

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

175

can assist in monitoring the local structure for every position in a given RNA molecule in a sequence-independent manner.

2. Materials 2.1. Preparation of RNA by In Vitro Transcription

1. T7 RNA polymerase; store at −20° C (see Note 1). 2. DNA template generated by PCR amplification using a forward primer that incorporates the T7 promoter sequence (TAATACGACTCACTATAGGG). 3. 10 × transcription buffer: 300 mM Tris–HCl pH 8.0, 100 mM DTT, 1% Triton X-100, 1 mM spermidine, 400 mM MgCl2. 4. 25 mM NTP mix. 25 mM rATP, rCTP, rGTP, and rUTP (Roche), store at −20° C. 5. Optional. Yeast inorganic pyrophosphatase (Sigma) (see Note 2). 6. 2 ´ urea loading buffer: 10 M urea, 1.5 mM EDTA pH 8.0, 10 mM Tris–HCl pH 8.0, 0.05% bromophenol blue, 0.05% xylene cyanol. 7. Crush-soak solution: 200 mM NaCl, 10 mM Tris–HCl pH 7.5, and 1 mM EDTA. Filter-sterilize the solution. 8. This section assumes the use of fluor-coated thin layer chromatography (TLC) plates and a hand held device for shortwave UV light (254 nm).

2.2. Radiolabeling of Nucleic Acids at the 5 ¢ Terminus

1. ~10–50 pmol synthetic RNA per ~13.3 pmol ATP [g-32P]. 2. Calf intestinal alkaline phosphatase (CIP) (New England Biolabs, Ipswich, MA). 3. T4-polynucleotide kinase (PNK) (New England Biolabs). 4. Adenosine 5¢-triphosphate (ATP) [g-32P], 6,000 Ci/mmole on reference date. 5. 5 × kinase buffer: 25 mM MgCl2, 125 mM CHES pH 9.0, 15 mM DTT (see Note 3). 6. Phenol:Chloroform:Isoamyl alcohol at 25:24:1 (v/v/v). 7. Chloroform. 8. Optional. Glycogen (20 mg/mL) (Roche, see Note 4). 9. 3 M sodium acetate pH 5.2. 10. Autoradiography film.

2.3. In-Line Probing

1. 5¢-terminus radiolabeled RNA; store at −20° C for up to 2 weeks (see Note 5). 2. 2 × in-line/SHAPE buffer: 100 mM Tris–HCl pH 8.3, 200 mM KCl, 40 mM MgCl2 (see Notes 6 and 7).

176

Wakeman and Winkler

3. RNase T1 (Sigma) diluted to 4 units/mL in H2O; store at 4 C. 4. 10 × T1 buffer: 0.25 M sodium citrate, pH 5.0. 5. 10 × OH buffer: 0.5 M Na2CO3, pH 9.0, 10 mM EDTA, pH 8.0. 2.4. SHAPE Probing

1. Nonradioactively labeled RNA; store at −20° C. 2. 5¢-radiolabeled DNA oligonucleotide (~15–30 nucleotides) that is complementary to the 3¢-end of the target RNA; store at −20 C for up to 2 weeks. 3. 65 mM N-methylisatoic anhydride (NMIA) (Invitrogen, Carlsbad, CA) solution dissolved in dimethyl sulfoxide (DMSO) (Sigma); prepare fresh (see Note 8). 4. Transcriptor reverse transcriptase (Roche); store at −20° C (see Note 9). 5. dNTPs solution: 10 mM dATP, dCTP, dGTP, and dTTP (Roche); store at −20° C. 6. 4 M NaOH. 7. 10 ´ TBE: 108 g Tris, 55 g boric acid, and 3.725 g EDTA in 1 L H2O. 8. Acid stop buffer. Make acid stop dye (85% formamide (v/v), 0.5 TBE, 50 mM EDTA, pH 8.0, 0.05% bromophenol blue, and 0.05% xylene cyanol). Mix the acid stop dye 25:4 (v/v) with 1 M unbuffered Tris base; store at −20° C. 9. Thermosequenase cycle sequencing kit (USB, Cleveland, Ohio) or other method for generation of DNA sequencing ladders via dideoxynucleotide incorporation. 10. 2 × in-line/SHAPE buffer. 100 mM Tris-HCl pH 8.0, 200 mM KCl, 40 mM MgCl2 (see Notes 6 and 7)

2.5. Polyacrylamide Gel Electrophoresis (PAGE)

1. 37% Acrylamide/bis-acrylamide 29:1 (w/w): dissolve 89.5125 g acrylamide and 3.0875 g bis-acrylamide in water, and bring up to a final volume of 250 mL. Filter through Whatman paper. Store in amber bottles at 4° C (see Note 10). 2. Urea. 3. Running buffer: 1 × TBE (10 × TBE diluted 1:10 in H2O). 4. Ammonium persulfate (APS): make a 10% solution (w/v) with H2O; store at 4° C for up to a month. 5. N,N,N¢,N¢-Tetramethylethylenediamine (TEMED); store at 4° C. 6. This section assumes the use of glass plates that are 32.5 cm × 41 cm with 0.75-mm spacers and 24-well combs for in-line probing and SHAPE gels. We typically use 28 cm × 16.5 cm glass plates with 0.75-mm spacers and 4–8 well combs for RNA preparative techniques (see Note 11). 7. Vacuum pump and gel dryer.

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

177

8. Whatman paper, 3-mm chromatography paper. 9. 35 cm × 43 cm phosphor screens (Amersham). 10. Phosphor imaging instrumentation, e.g.,Typhoon 9200 Variable Mode Imager, and software (Molecular Dynamics, Sunnyvale, CA).

3. Methods RNA phosphodiester linkages are spontaneously cleaved as the result of an SN2 attack by the ribose 2¢ oxygen on the adjacent phosphorus (12). This attack results in a 5¢ cleavage fragment with a 2¢,3¢-cyclic phosphate and a 3¢ cleavage fragment with a 5¢ hydroxyl terminus. The rate at which this reaction occurs is dependent upon the degree of “in-line” positioning of the 2¢ oxygen, phosphorus, and 5¢ leaving group oxygen atoms for a given RNA internucleotide linkage (16–19). Maximal rates are achieved when the 2¢ oxygen, the phosphorus center, and the 5¢ oxygen leaving group form a perfect 180° . This arrangement is required for a productive nucleophilic attack by the 2¢ oxygen. Linkages for nucleotides that participate in stable base-pairs exhibit lower rates of spontaneous cleavage than for nucleotides that are located in relatively unstructured regions (12, 18). Thus, spontaneous cleavage of phosphodiester linkages is highly dependent upon the local structural context. To exploit this for the prediction of RNA secondary structure a radiolabeled phosphorus group is placed at either the 5¢ or 3¢ terminus of the target RNA. These radiolabeled RNAs are then incubated under desired reaction conditions for a period of time (generally ~40 h) that is long enough for spontaneous self-cleavage to occur for a small subset of the total population of RNA molecules (12). The resulting RNA fragments are then resolved by denaturing gel electrophoresis alongside appropriate sequencing ladders and analyzed by phosphor imaging instrumentation. Bands of greater intensity correlate with “flexible” RNA linkages, which are more capable of randomly adopting a near perfect in-line conformation. Conversely, portions of the gel that appear to lack bands correlate with positions that are likely to be involved in base-pairing interactions. The reaction components for in-line probing can be significantly varied with only modest effects upon the assay results. Temperature of the reactions can also be modestly varied, although the overall incubation times must be adjusted accordingly (12, 20). Changes in magnesium concentration can modestly affect the band intensity for all reactive positions. However, magnesium can be removed from the reactions with a modest reduction in spontaneous cleavage (12). In prior studies, our laboratory has titrated magnesium into in-line probing reactions while maintaining

178

Wakeman and Winkler

high monovalent cation concentrations in order to characterize a magnesium-sensing riboswitch RNA (21). Under these conditions the influence of magnesium on the rates of RNA spontaneous cleavage is greatly minimized. In the past, experimentation very similar to in-line probing as described herein has been interpreted as demonstrating that addition of magnesium to RNA leads to high rates of spontaneous cleavage for backbone linkages adjacent to specific magnesium-binding sites. We recommend using caution when arriving at these conclusions. Sites of spontaneous cleavage observed by in-line probing reactions (in the presence of magnesium) do not necessarily correlate well with highly occupied magnesium-binding pockets that have been identified by structure-determining methods such as X-ray crystallography. In contrast, the in-line character of RNA linkages in three-dimensional structures has been shown to agree well with the relative rates of RNA cleavages observed during in-line probing (12). Additionally, as mentioned earlier, magnesium can be removed from in-line probing reactions under certain conditions with only a minor influence on overall spontaneous cleavage rates (12). More detailed discussion on any potential correlation between metal-binding sites and cleavages at specific linkages, a subject that is not without debate, is beyond the scope of this protocol outline. Similar to in-line probing, SHAPE is an unbiased method for distinguishing base-paired and single-stranded regions of RNA in a sequence-independent manner. Originally developed by Kevin Weeks’ laboratory, several versions of the SHAPE probing protocol have been published elsewhere (13, 14, 22). SHAPE probing is built upon the observation that the reactivity (i.e., nucleophilic character) of the ribose 2¢-hydroxyl is influenced by local nucleotide flexibility. For nucleotides constrained by base-pairing interactions, close proximity of the 3¢ phosphodiester anion reduces reactivity of the 2¢-hydroxyl. This corresponds with lower rates of reactivity to an appropriate electrophile, such as N-methylisatoic anhydride (NMIA), for formation of a 2¢-O-adduct. However, flexible linkages allow for conformations that exhibit increased reactivity to NMIA. Sites of 2¢-O-adduct formation can then be detected via disruption of cDNA synthesis by reverse transcriptase. Importantly, NMIA reactions are self-quenching by virtue of the fact that NMIA undergoes a parallel self-inactivating reaction (13). For these reasons, SHAPE is an experimentally easy, robust, and rapid method for prediction of paired and single-stranded regions within a target RNA. Both in-line probing and SHAPE analysis have been employed to investigate riboswitch RNAs in the presence and absence of their ligands. These analyses have revealed the overall secondary structure patterns for riboswitch RNAs and highlighted ligandinduced structural changes. When repeated at a range of ligand concentrations, both methods can be used to determine the apparent dissociation constants (KD) for RNA–ligand interactions.

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

179

Not restricted to metabolite-binding RNAs, these methods are generally applicable for analysis of RNA structure and the study of RNA–ligand interactions. 3.1. Preparation of RNA by In Vitro Transcription

1. For a yield of > 200 pmols of RNA, combine ~20–50 pmol DNA template, 2.5 mL 10 ¥ transcription buffer, 2.5–5 mM each NTP, ~50 mg/mL T7 RNA polymerase, and 0.0025– 0.01 U inorganic pyrophosphatase (optional) in a final volume of 25 mL. These reactions can be scaled appropriately for recovery of the desired quantity of synthetic RNA. The reaction should be incubated for 2–3 h at 37° C (see Note 12) and terminated by the addition of an equal volume of 2 ¥ urea loading buffer (see Note 13). The transcription reaction can be stored at −20° C at this point. 2. The RNA should be resolved by, and excised from, a denaturing polyacrylamide gel (as detailed in Subheading 3.5) (see Note 14). The polyacrylamide percentage (6–15%) should be chosen as deemed appropriate for the length of the RNA substrate. 3. Once the bromophenol blue dye marker has run ~2/3 the length of the plates, electrophoresis can be terminated. At this point, the glass plates containing the gel should be removed from the gel rig. To remove the gel from the glass plates, slide out the spacers, lay the glass plates flat on the bench top, and carefully pry them apart. The gel will typically preferentially stick to one plate. Flip the sandwich over so that the gel-associated plate is on the bottom and continue with the separation. Once one plate is removed, place plastic wrap over the exposed side of the gel. Flip the gel and gently separate the plastic wrapassociated gel from the remaining plate. Cover the newly exposed side of the gel with plastic wrap. 4. The synthetic RNA can then be visualized by UV shadowing. Place the plastic wrap-enclosed gel sandwich over a TLC plate and briefly expose to shortwave UV light. UV-absorbing material such as RNA polymers and free nucleotides will appear as dark shadows. Outline the topmost shadow with a fine-point marker as it should correspond to the target RNA. The lowest migrating band will likely represent the free nucleotides. 5. Excise the circled region with a razor blade, remove the outer layer of plastic wrap, and cut the gel slice into ~1-mm squares. 6. Place the gel bits into a 1.5-mL microcentrifuge tube and add approximately two volumes of crush-soak solution (typically ~400–600 mL). Incubate on a tube rotator at room temperature for 2 h or at 4° C overnight. Remove and save supernatant. 7. Ethanol precipitate the RNA by adding ~2.5 volumes cold 100% ethanol and incubating at −20° C for 30 min. Pellet the RNA by centrifuging at 20,000 × g for 15 min. Wash pellet

180

Wakeman and Winkler

with 200 mL 70% ethanol and centrifugation at 20,000 × g for 5 min. Carefully remove supernatant and air dry the pellet air for 1–5 min or by speedvac. 8. Resuspend the RNA in ~20–100 mL H2O and quantify RNA yield via calculation of extinction coefficient values and measurement of A260. Store at −20°C until use. 3.2. Radioactive Labeling of Nucleic Acids at the 5 ¢ Terminus

1. Prior to radioactive labeling, the RNA must first be dephosphorylated at the 5¢ terminus. The 5¢ terminus of DNA oligonucleotides produced by solid-phase synthesis typically does not contain phosphate groups. Therefore, skip to step 4 for radiolabeling of commercially obtained DNA oligonucleotides. Combine ~10– 40 pmols RNA and the commercial buffer for CIP in a 10 mL final volume and incubate at 50° C for 15 min. 2. Remove CIP by phenol:chloroform:isoamyl alcohol extraction. First, increase the volume of the reaction to 200 mL with H2O. Add 200 mL phenol:chloroform:isoamyl alcohol and shake or vortex for ~5 s to mix completely. Centrifuge at 20,000 × g for 5 min to separate into two phases. Remove the top, aqueous phase, which contains the RNA, and discard the bottom phase containing the protein. Repeat this procedure with 200 mL chloroform to remove traces of phenol. 3. Concentrate the RNA via ethanol precipitation. Add 1/10 volume (20 mL) 3 M sodium acetate and 1 mL glycogen to the RNA and mix (see Note 4). Add 2.5 volume (500 mL) 100% ethanol, mix by inversion, and incubate at −20° C for 30 min. Pellet the RNA by centrifugation at 20,000 × g for 15 min. The pellet should then be washed by the addition of 200 mL 70% ethanol and centrifuged at 20,000 × g for 5 min. Discard supernatant, air dry or speedvac the pellet for 1–5 min, and resuspend the RNA in 10 mL H2O. 4. Once the RNA has been dephosphorylated, a radiolabeled phosphate group can be transferred to the 5¢ terminus using T4 PNK. Per 20 mL kinase reaction employ 5 mL of CIPtreated RNA, 4 mL 5 kinase buffer, 4–12 mL ATP [g-32P], and 2 mL T4 polynucleotide kinase (PNK) at 10 U/mL. Increase the final volume to 20 mL with H2O. Incubate the reaction for 35 min at 37°C. For 5¢-radiolabeling of DNA oligonucleotides set up identical reactions except with 10–50 pmol DNA. 5. The resulting 5¢-radiolabeled nucleic acids can then be resolved by denaturing polyacrylamide gel electrophoresis. Use 6% PAGE for nucleic acids greater than 75 nucleotides and 10% for nucleic acids less than 75 nucleotides (as detailed in Subheading 3.5 and step 3 of Subheading 3.1). Prior to discarding, buffers in the upper and lower reservoirs of the gel rig should be checked for radioactivity.

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

181

6. Remove the polyacrylamide gel from the glass plates and place between layers of plastic wrap. Radioactively labeled bands can then be identified via brief exposure to autoradiography film. The gel should be secured inside the film cassette so that it cannot move and can be reproducibly positioned against the autoradiography film. A sheet of autoradiography film should be exposed to the gel for ~1 min and developed. Outline the region of the gel sandwich that contains the radiolabeled RNA as identified by dark band(s) on the autoradiography film. 7. Excise the RNA from the gel using the procedure detailed in Subheading 3.1, steps 5–8. 3.3. In-Line Probing

1. In a final volume of 10 mL, combine 2 ´ in-line/SHAPE buffer, ~75–200 kcpm 5¢-radiolabeled RNA, and the desired amount(s) of any other substance to be included in the assays (e.g., RNA-binding protein or metabolite). Incubate at 22–25° C for ~40 h (see Note 15). This will allow for spontaneous cleavage of a subpopulation of the RNAs via single-hit kinetics (12). Terminate the reaction by addition of 10 mL of 2 ´ urea loading buffer and immediate storage at −20° C. Resolve the RNA fragments by denaturing polyacrylamide gel electrophoresis alongside size marker ladders. There are many different size markers that could be used. Our laboratory typically includes lanes containing RNAs that have been partially digested with RNase T1 to visualize guanosine residues and a lane for RNAs that were briefly exposed to increased pH and high temperature for visualization of cleavage at all nucleotide positions. We refer to these size marker control reactions herein as “T1” and “–OH” ladders, respectively. 2. Prepare the T1 ladder by mixing 1 mL 5¢-radiolabeled RNA (~100 kcpm – similar in quantity to the experimental lanes) with 1 mL RNase T1 (4 U/mL) and 1 mL 10 T1 buffer and increase the volume to 10 mL with 2 urea loading buffer. Incubate at 50° C for 20 min. The most appropriate reaction times will need to be optimized per target RNA. Add 3 mL 2 ´ urea loading buffer and 7 mL H2O and store at −20° C prior to resolution by denaturing PAGE. 3. Prepare the –OH ladder by mixing ~100 kcpm 5¢-radiolabeled RNA with 1 mL of 10 OH buffer and increase the volume to 10 mL with H2O. Incubate at 95°C for 3–8 min to induce scission after every base (the appropriate time interval required for these reactions will need to be optimized per target RNA). Stop reaction with 10 mL 2 ´ urea loading buffer and immediately store at −20° C. 4. Resolve the in-line probing reactions and size marker ladders by denaturing 6–20% polyacrylamide gel electrophoresis (the most appropriate polyacrylamide percentage should depend

182

Wakeman and Winkler

on the size of the target RNAs) next to ~100 kcpm nonreacted (NR) radiolabeled RNA in 10 mL 2 urea loading buffer, following the protocol described in Subheading 3.5. 5. Results from these assays should resemble representative reactions shown in Fig.1a and c. When analyzing the data, consider bands to be regions of the RNA that are flexible/unstructured such as loops and bulges. Clearings on the gel can be considered regions of RNA, such as helices, that are structurally constrained by base interactions (see Note 16). The location of these regions can be mapped onto the RNA sequence using the T1 and OH ladders as shown in Fig. 1b and d. The T1 ladder will reveal the location for guanosine residues within the RNA, while the OH ladder displays bands for every nucleotide. The nonreacted RNA demonstrates the quality of the RNA prior to treatment and should display minimal to no banding. For more detail on data analysis, see Subheading 3.6. 3.4. SHAPE Probing

1. In a final volume of 18 mL, combine 10 mL 2 ´ in-line/SHAPE buffer, 2 pmol RNA, and the desired concentration of potential ligand to be tested (see Note 17). Incubate for 1 h at room temperature to allow the RNA to equilibrate with its potential ligand and for RNA folding (see Note 18). 2. At this point, the mixture should be split into two separate 9-mL reactions. Add 1 mL NMIA solution to one reaction aliquot and 1 mL DMSO to the remaining aliquot. The reaction containing DMSO alone serves as a control. Incubate these reactions at 37° C for 45 min (see Note 19). 3. Increase the volume to 200 mL with H2O and terminate the reactions by ethanol precipitation of the RNA as described in step 3 of Subheading 3.2. 4. The pellets should be resuspended in 11 mL H2O that also contains ~150–200 kcpm of the 5¢-radiolabeled DNA oligonucleotide (reverse complement to the 3¢-end of the substrate RNA). Incubate the reactions at 65°C for 5 min followed by 20 min at 37° C. 5. Add 2 mL 10 mM dNTPs, 2 μL H2O, and 4 μL commercial 5° Transcriptor buffer and incubate at 52° C for at least 1 min but no more than 5 min. 6. Add 1 μL Transcriptor reverse transcriptase to initiate cDNA synthesis. Each reaction should be incubated at 52° C for exactly 5 min. For these reactions, oligonucleotide primers are typically 18–25 nucleotides in length and contain a G or C at the 3¢ terminus. 7. Terminate the cDNA synthesis reactions through addition of 1 mL 4 M NaOH and incubation at 95° C for 5 min. This should completely hydrolyze the substrate RNA.

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

183

Fig. 1. Representative in-line probing of metabolite-binding RNA elements. (a) In-line probing of the thiamine pyrophosphate (TPP)-binding riboswitch aptamer of the Mycoplasma gallisepticum hatABC leader region. These reactions were performed in the presence or absence of thiamine (T) or TPP, the preferred ligand, and were resolved by denaturing 10% PAGE. Binding of T induces conformational changes as evidenced by changes in intensity for multiple bands within the individual lanes. The gray curved lines denote the regions of this RNA that were analyzed by line traces in panel e of this figure. T1 and –OH lanes represent size marker ladders resulting from partial digestion of the RNA molecules by RNase T1, which cleaves after G residues, and alkali, which cleaves after every nucleotide, respectively. The NR lane includes nonreacted RNA to demonstrate the quality of the RNA prior to in-line probing. (b) Probing changes induced by addition of 1 mM TPP have been mapped onto the RNA secondary structure diagram. Open circles indicate regions of the RNA displaying a constant level of scission under all reaction conditions. Light gray and dark gray circles indicate regions of the RNA displaying decreased or increased levels of scission upon addition of TPP, respectively. RNA nucleotides that are likely to be involved in recognition of the diphosphate portion of TPP are denoted by arrows. (c) Representative in-line probing of a Streptomyces coelicolor S-adenosylmethionine (SAM)-binding riboswitch in the presence or absence of SAM. (d) Probing changes induced by SAM are summarized in relation to the secondary structure diagram. (e) Representative data showing detection of ligand recognition determinants using in-line probing. Line traces are shown for regions of the Mycoplasma TPP riboswitch highlighted in panel a. The light gray line results from the no-ligand control lane, while the dark gray and black lines result from the lanes containing 1 mM T and 1 mM TPP, respectively. While both T and TPP induce a similar overall conformational change, in-line probing patterns revealed several bands that are reduced in intensity for reactions that included TPP, but not T (27). These differences are highlighted by normalized line graphs shown beneath representative gel slices. Because T and TPP differ only by a diphosphate, these data suggest that the TPP-specific positions are likely to be required for recognition of ligand phosphates. Indeed, threedimensional structures of TPP riboswitches confirmed these biochemical data (2, 3, 7).

184

Wakeman and Winkler

Fig. 2 Representative SHAPE probing for a riboswitch RNA. (a) SHAPE probing of a B. subtilis M-box RNA (magnesiumsensing riboswitch) at a range of magnesium concentrations. Bands of higher intensity correspond to RNA linkages that exhibit greater flexibility than for linkages constrained by base pairing. The +DMSO reactions did not include any NMIA and therefore show spontaneous RNA breakdown products and structure-induced reverse transcriptase pausing. These control reactions were performed at magnesium concentrations both above and below the presumed EC50 for the structural rearrangement. Sequencing ladders are resolved adjacent to the SHAPE reactions. (b) Results of SHAPE analysis are mapped onto the B. subtilis M-box RNA aptamer sequence. White circles indicate regions that exhibit constantly high levels of NMIA modification in all magnesium concentrations. Light gray and dark gray circles indicate regions that exhibit decreased (become more structured) or increased NMIA modification upon magnesium binding, respectively. (c) SHAPE probing of a thiamine pyrophosphate (TPP)-binding aptamer in the presence or absence of 10 mM TPP. Representative gels were loaded with 5 or 20 mL aliquots to show reduced individual band resolution with large loading volume. TPP-induced NMIA modifications are coded similar to panel b.

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

185

8. Neutralize this reaction by adding 29 mL acid stop buffer and store at −20° C for up to 2 weeks prior to resolution by denaturing PAGE. 9. Prepare DNA sequencing ladders using standard manufacturers instructions and ~50–100 kcpm of the 5¢-radiolabeled DNA oligonucleotide that was employed for cDNA synthesis after NMIA modification. 10. Incubate DNA sequencing ladders and reaction samples to 95° C for 5 min prior to resolution by denaturing 10% PAGE. 11. The results from these assays should resemble representative reactions shown in Fig. 2a and c. When analyzing the data, consider bands to be regions of the RNA that are flexible/unstructured such as loops and bulges. Clearings on the gel can be considered regions of RNA such as helices that are structured. The location of these regions can be mapped onto the RNA sequence using the DNA sequencing ladders. Reactions that included NMIA should be directly compared to control reactions lacking NMIA. The first ~20 nucleotides, proximal to the 5¢-terminus, will be difficult to analyze due to significant pausing exhibited by reverse transcriptase enzyme during initiation. For more details on data analysis, see Subheading 3.6. 3.5. Polyacrylamide Gel Electrophoresis

1. Prepare 10% gel solution. In a 500-mL final volume, combine 240 g urea, 50 mL 10 ´ TBE, 135 mL 37% acrylamide/ bis-acrylamide solution. Once the powder is dissolved, filter the solutions through Whatman paper. This solution can be stored in amber bottles at room temperature for up to a month. Adjust the volume of acrylamide/bis-acrylamide solution to achieve the desired polyacrylamide percentage. 2. To initiate polymerization of 100 mL of PAGE solution, gently add and mix 0.8 mL 10% APS and 0.04 mL TEMED, pour immediately, and slide in the comb. Allow the gel to polymerize for >30 min. 3. Gently remove the comb and rinse out the wells. 4. Assemble the gel electrophoresis rig and fill the upper and lower reservoirs with running buffer. 5. Prerun the gel for 15 min. For glass plates that are 32.5 cm × 41 cm with 0.75-mm spacers we typically conduct electrophoresis at constant 60 Watts. For glass plates that are 28 cm × 16.5 cm with 0.75-mm spacers we typically electrophorese samples at constant 40 Watts. 6. Prior to loading the samples, rinse the wells thoroughly with running buffer. 7. For probing reactions, continue electrophoresis until the bromophenol blue indicator dye is ~1 in. from the bottom of

186

Wakeman and Winkler

the plate. For gels employed for preparative purposes we typically continue electrophoresis until the bromophenol blue dye has run ~2/3 the length of the plate, in order to retain free nucleotides within the gel. 8. For gels that are to be dried, remove one of the glass plates, allowing the gel to remain attached to the second plate. Press a sheet of Whatman paper on top of the exposed gel. The gel will adhere to the Whatman paper and can then be easily peeled away from the remaining glass plate. Cover the exposed side of the gel with plastic wrap and place in a gel dryer under vacuum pressure at 80° C for 2–3 h. 9. Expose the dried gel to a phosphor screen, which can then be analyzed via standard phosphor imaging instrumentation. 3.6. Analysis

1. This section assumes use of software resembling ImageQuant (Molecular Dynamics). 2. Each lane of the gel will exhibit a distinctive banding pattern. For in-line probing, the bands correspond to RNA transcripts where the 3¢ terminus has been generated by spontaneous cleavage of a flexible RNA linkage. Regions of the gel lacking bands are thereby indicative of RNA linkages that exhibit poor rates of spontaneous cleavage (e.g., base-paired positions). For SHAPE, the bands correspond to cDNAs containing a 3¢ terminus at the site of NMIA modification. Since NMIA reactivity is dependent upon RNA linkage flexibility the intensity of individual bands correlates with RNA linkage flexibility. Changes to the banding pattern between lanes are therefore suggestive of RNA conformational changes. Bands may darken, lighten, or disappear below detection. Dramatic banding changes can be easily observed by eye (e.g., Figs. 1 and 2). This information can easily be structurally related to the RNA primary sequence through the aid of adjacent control reactions. For example, every SHAPE reaction should be resolved alongside DNA sequencing reactions, performed using the oligonucleotide primer also employed for cDNA synthesis of the NMIA-modified RNAs. For in-line probing, we find it convenient to include control reactions for cleavages at every position (alkaline-induced hydrolysis of the RNA backbone; “-OH”) and cleavages at G residues (partial digestion by RNase T1; “T1”). Subtle changes in the banding pattern can be qualitatively analyzed by comparing the relative intensity profile of each lane. Specifically, a line trace can be drawn over the desired region of the gel and graphed, thereby generating a lane profile. Peaks and valleys correspond to bands and cleared regions, respectively. These data can be exported to spreadsheet analyses software such as Excel or SigmaPlot and carefully plotted and analyzed. Similar line traces copied onto control lanes, such as the T1 and –OH reactions, can be used to correlate line trace data to the overall RNA sequence. The line

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

187

profile data can be normalized by converting the region of the each lane with the highest counts to 1 and the region with the lowest counts to 0. This type of analysis is shown in Fig. 1e. 3. Alternatively, a box can be drawn around individual bands and the relative intensity for the area within can be obtained by standard analyses by software such as ImageQuant. A similarly sized control box for background subtraction should be placed on an area of the gel that lacks obvious bands. 4. These assays can also be useful for characterization of ligand– RNA interactions by setting up reactions with a range of ligand concentrations. To account for subtle differences in loading, the relative intensity for the area within an individual box can be divided by the relative intensity for the area within a boxed region that encompasses the entire lane. If the lower ligand concentrations and upper ligand concentrations are below and at ligand saturation, respectively, these experimental data may be used for estimation of EC50 values or estimates of cooperativity. For this type of analysis, multiple individual bands that display increased and decreased intensity in response to ligand interactions or conformational changes should be directly compared with one another. The easiest and most rapid method is to normalize each box series to the boxed band with the highest and lowest intensity measurements. If these values are normalized to maximal and minimal values of 1 and 0, respectively, then multiple band series can be compared to one another despite potentially significant differences in their overall relative intensity. A highly recommended alternative to these standard types of data analyses is a very useful software program called semi-automated footprinting analysis (SAFA) (23). The latter software package significantly shortens the gel quantification process while reducing systematic error introduced during data analysis.

4. Notes 1. While T7 RNA polymerase is commonly used for high yield production of RNA by in vitro transcription, other RNA polymerases such as SP6 can also be used. However, when other RNA polymerases are used in these reactions, their individual promoter preferences onto DNA templates should be incorporated into the DNA template. 2. Inorganic pyrophosphatase is a useful addition to in vitro transcription reactions in order to improve the yield of RNA. During transcription, pyrophosphates will be released into solution which can result in the chelation of Mg2+ and subsequently

188

Wakeman and Winkler

the reduction of RNA polymerase function. Inorganic pyrophosphatase will reduce the accumulation of pyrophosphates. 3. Use of a high pH CHES buffer (5 ´ : 25 mM MgCl2, 125 mM CHES of pH 9, 15 mM DTT) can aid in the removal of oligonucleotide structure near the 5¢ portion of an RNA molecule which can interfere with the ability of PNK to phosphorylate the 5¢ termini. The 10 ´ buffer supplied by New England Biolabs can be used when structure is not a problem. 4. Glycogen is an optional but useful addition during ethanol precipitation because it precipitates the RNA and allows for easy visualization of the pellet. It is inert and its presence should not affect subsequent reactions. 5. Radiolysis of labeled RNAs can be a source of background noise in probing reactions. We recommend storing 5¢-radiolabeled RNA at <100 kcpm/ml to reduce this problem. 3¢-terminus radiolabeled RNA could also be used as substrates for in-line probing, although methods of 3¢-labeling are not discussed herein. 6. In-line probing can be also used to study magnesium-dependent folding pathways under high monovalent conditions. When employing in-line probing to investigate divalent ion-induced RNA folding pathways, MgCl2 should be eliminated from the 2 ´ in-line/ SHAPE buffer and it is recommended to increase monovalent concentrations (see next note). The individual components of the 2 ´ SHAPE buffer can also be significantly varied (13, 14). We typically perform SHAPE reactions with a final concentration of monovalent ions ranging from 0 to 200 mM and concentrations of magnesium ranging from 0 to 40 mM. Typically both in-line probing and SHAPE reactions are performed at pH 7.5–8.3 (14). However, our standard SHAPE and in-line probing reactions are conducted at pH 8.0 and 8.3, respectively. 7. If high monovalent ions do not disrupt the molecular interactions being studied, in-line probing reactions can be performed in the presence of 1–3 M monovalent cations. Under these conditions, MgCl2 can be eliminated from the reactions and varied as desired with little effect on the general efficiency of backbone self-cleavage. Additionally, monovalent ions at high concentrations are likely to outcompete the loosely associated “atmosphere” of divalent ions (24, 25). At lower monovalent cation concentrations, variations of MgCl2 in in-line probing reactions can lead to modest fluctuations in banding intensities for all bands within the individual lanes. High monovalent cation conditions have also been successfully used for SHAPE analysis of magnesium-induced folding of RNAs (21). 8. While we have found that a 65-mM NMIA solution (6.5 mM final concentration) is ideal for RNA constructs ranging from 100 to 300 nucleotides, NMIA concentration is a parameter

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing

189

that should be varied during troubleshooting (13, 14). For RNA constructs shorter than 100 nucleotides in length, the NMIA solution should be increased (e.g., 130 mM). For RNA constructs longer than 300 nucleotides, a 30-mM NMIA solution is likely to be preferable. If low signal is observed in the +NMIA lanes but an intense band representing full-length product is seen, this is an indication of insufficient modification of the RNA and higher concentrations of NMIA should be tested. If low molecular weight intense bands are observed and little to no full-length product can be seen, this is an indication of excessive modification of the RNA and lower concentrations of NMIA should be tested (14). 9. Superscript III (Invitrogen) or other commercial reverse transcriptase enzymes can be used with similar results, although the reaction temperatures may need to be individually optimized. 10. A dust mask, gloves, and eye protection should be worn when handling powdered acrylamide due to the fact that acrylamide is extremely neurotoxic until polymerized. 11. Upon purchase, the outside of the glass plates should be permanently marked so that proper orientation of the plates can be maintained in all subsequent runs. 12. Formation of a white precipitate during in vitro transcription is indicative of the accumulation of high levels of pyrophosphates. This will lead to a reduction in transcription; therefore, the reaction can be stopped prior to the full 2–3 h incubation if significant precipitation is observed. The optional addition of inorganic pyrophosphatase will prevent the formation of this precipitate and increase the RNA yield. 13. Alternatively, the transcription reaction can be stopped by the addition of phenol:chloroform:isoamyl alcohol – see methods from Subheading 3.2, steps 2 and 3. This method is recommended if the transcription reaction has been scaled up in volumes greater than 100 mL to increase the yield of RNA. 14. Size exclusion chromatography can also be used to purify the RNA of interest if an FLPC and an appropriate gel filtration column is available. The Superdex 200 10/300 GL column (GE Healthcare) can be used successfully to purify RNA molecules ranging in size from 10 to 60 kDa, while a Superdex 75 10/300 GL column will resolve molecules approximately sized between 3 and 70 kDa. After the transcription reaction has been resolved by size-exclusion chromatography, RNAcontaining fractions should be pooled and concentrated (by either ethanol precipitation or centrifugation-based concentrator devices) (see Subheading 3.2, step 3). 15. For in-line probing, shorter incubation times result in fainter bands but longer incubation times result in higher background. A highly recommended discussion on the rates of spontaneous

190

Wakeman and Winkler

cleavage of RNA linkages and effects of pH and metal ions can be found elsewhere (12, 20). 16. While bands on an in-line probing gel are typically indicative of flexible RNA linkages, certain bands may result from nucleotides that are positioned such that the 2¢ hydroxyl and oxyanion leaving group most closely approximate the in-line conformation required for efficient phosphodiester scission (12). 17. Although the volume at this point of the SHAPE protocol is 18 mL, the final volume of the reaction will be a total of 20 mL, so this should be taken into account when calculating the amount of ligand to add. 18. Prior to addition of the 2 ´ in-line/SHAPE buffer, it is recommended to resuspend the RNA in a low ionic strength buffer such as 0.5 ´ TE, heating to 95° ´ C for 2 min, and snap cooling by placing on ice for 2 min (14). However, this step may be unnecessary, depending on the particular RNA molecule being investigated. 19. While 37° C is a useful temperature for many applications of SHAPE technology, this method can be used at temperatures ranging from 20 to 75° C (14, 26). For this method to be used at different temperatures, the NMIA modification step must be adapted. The NMIA should be incubated with the RNA for 5 half-lives of hydrolysis. The half-life of hydrolysis for NMIA at 37° C is ~8.3 min; therefore, at 37° C the NMIA modification step is incubated for ~45 min. However, the half-life of NMIA hydrolysis changes at different temperatures and can be calculated using the equation: half-life (min) = 360 ´ exp[−0.102 ´ temperature ( °C)].

References 1. Winkler, W.C. and Breaker, R.R. (2005). Regulation of bacterial gene expression by riboswitches. Annu. Rev. Microbiol. 59, 487–517. 2. Schwalbe, H., Buck, J., Furtig, B., Noeske, J., and Wohnert, J. (2007). Structures of RNA switches: insight into molecular recognition and tertiary structure. Angew. Chem. Int. Ed. Engl. 46, 1212–1219. 3. Wakeman, C.A., Winkler, W.C., and Dann, C.E., III. (2007). Structural features of metabolite-sensing riboswitches. Trends Biochem. Sci. 32, 415–424. 4. Edwards, T.E., Klein, D.J., and Ferré-DÕAmaré, A.R. (2007). Riboswitches: small-molecule

recognition by gene regulatory RNAs. Curr. Opin. Struct. Biol. 17, 273–279. 5. Gilbert, S.D. and Batey, R.T. (2006). Riboswitches: fold and function. Chem. Biol. 13, 805–807. 6. Winkler, W., Nahvi, A., and Breaker, R.R. (2002). Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression. Nature 419, 952–956. 7. Serganov, A., Polonskaia, A., Phan, A.T., Breaker, R.R., and Patel, D.J. (2006). Structural basis for gene regulation by a thiamine pyrophosphate-sensing riboswitch. Nature 441, 1176–1171.

Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing 8. Ehresmann, C., Baudin, F., Mougel, M., Romby, P., Ebel, J.P., and Ehresmann, B. (1987) Probing the structure of RNAs in solution. Nucleic Acids Res. 15, 9109–9028. 9. Werner, C., Krebs, B., Keith, G., and Dirheimer, G. (1976). Specific cleavages of pure tRNAs by plumbous ions. Biochim. Biophys. Acta. 432, 161–175. 10. Brunel, C. and Romby, P. (2000). Probing RNA structure and RNA–ligand complexes with chemical probes. Methods Enzymol. 318, 3–21. 11. Ryder, S.P. and Strobel, S.A. (1999). Nucleotide analog interference mapping. Methods 18, 38–50. 12. Soukup, G.A. and Breaker, R.R. (1999). Relationship between internucleotide linkage geometry and the stability of RNA. RNA 5, 1308–1325. 13. Merino, E.J., Wilkinson, K.A., Coughlan, J.L., and Weeks, K.M. (2005). RNA structure analysis at single nucleotide resolution by selective 2¢-hydroxyl acylation and primer extension (SHAPE). J. Am. Chem. Soc. 127, 4223–4231. 14. Wilkinson, K.A., Merino, E.J., and Weeks, K.M. (2006). Selective 2¢-hydroxyl acylation analyzed by primer extension (SHAPE): quantitative RNA structure analysis at single nucleotide resolution. Nat. Protoc. 1, 1610–1616. 15. Nahvi, A., Sudarsan, N., Ebert, M.S., Zou, X., Brown, K.L., and Breaker, R.R. (2002). Genetic control by a metabolite-binding mRNA. Chem. Biol. 9, 1043. 16. Westheimer, F.H. (1968). Pseudo-rotation in the hydrolysis of phosphate esters. Acc. Chem. Res. 1, 70–78. 17. Usher, D.A. (1969). On the mechanism of ribonuclease action. Proc. Natl. Acad. Sci. USA 62, 661–627. 18. Usher, D.A. and McHale, A.H. (1976). Hydrolytic stability of helical RNA: a selective

19.

20.

21.

22.

23.

24.

25.

26.

27.

191

advantage for the natural 3¢, 5¢-bond. Proc. Natl. Acad. Sci. USA 73, 1149–1153. Dock-Bregeon, A.C. and Moras, D. (1987). Conformational changes and dynamics of tRNAs: evidence from hydrolysis patterns. Cold Spring Harb. Symp. Quant. Biol. 52, 113–121. Li, Y. and Breaker, R.R. (1999). Kinetics of RNA degradation by specific base catalysis of transesterification involving the 2¢-hydroxyl group. J. Am. Chem. Soc. 121, 5364–5372. Dann, C.E. III, Wakeman, C.A., Sieling, C.L., Baker, S.C., Irnov, I., and Winkler, W.C. (2007). Structure and mechanism of a metalsensing regulatory RNA. Cell 130, 878–892. Mortimor, S.A. and Weeks, K.M. (2007). A fast-acting reagent for accurate analysis of RNA secondary and tertiary structure by SHAPE chemistry. J. Am. Chem. Soc. 129, 4144–4145. Das, R., Laederach, A., Pearlman, S.M., Herschlag, D. and Altman, R.B. (2005). SAFA: a semi-automated footprinting analysis software for high-throughput quantification of nucleic acid footprinting experiments. RNA 11, 344–354. Das, R., Travers, K.J., Bai, Y., and Herschlag, D. (2005). Determining the Mg2+ stoichiometry for folding an RNA metal ion core. J. Am. Chem. Soc. 127, 8272–8273. Draper, D.E., Grilley, D., and Soto, A.M. (2005). Ions and RNA folding. Annu. Rev. Biophys. Biomol. Struct. 34, 221–243. Wilkinson, K.A., Merino, E.J., and Weeks, K.M. (2005). RNA SHAPE chemistry reveals nonhierarchical interactions dominate equilibrium structural transitions in tRNAAsp transcripts. J. Am. Chem. Soc. 127, 4659–4667. Winkler, W., Nahvi, A., and Breaker, R.R. (2002). Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression. Nature 419, 952–956.

Chapter 14 Identification of Metabolite–Riboswitch Interactions Using Nucleotide Analog Interference Mapping and Suppression Juliane K. Soukup and Garrett A. Soukup Summary Riboswitches are RNA elements capable of modulating gene expression through interaction with cellular metabolites. One member of the riboswitch family, the glmS riboswitch, is unique among riboswitches in that it modulates gene expression by undergoing self-cleavage in the presence of its metabolite glucosamine-6-phosphate (GlcN6P). In order to investigate the interactions between the glmS RNA and GlcN6P we performed nucleotide analog interference mapping (NAIM) and suppression (NAIS). These techniques have been previously used to identify important functional groups in and tertiary contacts necessary for self-splicing and self-cleaving by catalytic RNAs, RNA–protein complexes, RNA folding, and RNA–metal ion interactions. Described here are the details of NAIM and NAIS experiments we have utilized to investigate RNA–ligand interactions between the glmS riboswitch and GlcN6P. These techniques can be employed to study a wide variety of RNA–small molecule interactions. Key words: Riboswitch, Ribozyme, glmS, Glucosamine-6-phosphate, Nucleotide analog interference mapping, Nucleotide analog interference suppression

1. Introduction Structural investigations of RNA have led to a vast knowledge of RNA function. Biophysical techniques such as X-ray crystallography and NMR provide detailed analyses at the atomic level. Such techniques are however typically performed in isolation from a functional assay of the RNA, thereby producing a static picture of the molecule’s three-dimensional structure. In contrast, Nucleotide Analog Interference Mapping (NAIM) and Suppression (NAIS) techniques provide atomic-level resolution of functional groups and tertiary interactions through an activity assay. Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_14 © Humana Press, a part of Springer Science + Business Media, LLC 2009

193

194

Soukup and Soukup

Essential components for successful NAIM and NAIS experiments include (a) the ability to incorporate phosphorothioatecontaining nucleotide analogs (Fig. 1) into the functional RNA of interest; and (b) optimization of a selective activity assay that physically distinguishes functional RNA molecules from inactive molecules (Fig. 2). NAIM is generalizable to a wide variety of nucleic acids and can be applied to the analysis of a diversity of chemical groups. Using NAIM, the biochemical contribution of a single chemical group within a nucleic acid catalyst can be defined using a nucleotide analog that modifies the atom(s) of interest (1–3). To provide a general means of identification, the analogs are chemically tagged with a 5¢-phosphorothioate, a derivative that replaces the 5¢ nonbridging RP oxygen atom with a sulfur atom. The

Fig. 1. Nucleotide analogs. General chemical structures of the three subsets of phosphorothioate nucleotide analogs: parental, 2¢ functional group modifications and nucleobase functional group modifications. Altered functional groups are highlighted in gray.

Identification of Metabolite–Riboswitch Interactions

195

Fig. 2. NAIM and NAIS biochemical techniques applied to investigation of the glmS ribozyme. A population of RNAs is prepared containing a phosphorothioate-tagged nucleotide analog (N) randomly incorporated at a low level throughout the RNA. NAIM assay (left gel lane of bottom panel of figure). The population is selected for function in an assay in which active molecules are those that undergo self-cleavage in the presence of glucosamine-6-phosphate. Active and inactive fractions are separated by size. The phosphorothioate linkage is cleaved by the addition of iodine. The cleavage products are resolved by polyacrylamide gel electrophoresis. Sites of analog interference are identified by loss of bands in the sequencing ladder of the cleaved fraction compared to the uncleaved fraction (middle gel lane). NAIS assay (right gel lane of bottom panel of figure). The population is selected for function in an assay in which active molecules are those that undergo self-cleavage in the presence of glucosamine. Active and inactive fractions are separated by size. The phosphorothioate linkage is cleaved by the addition of iodine. The cleavage products are resolved by polyacrylamide gel electrophoresis. Sites of analog suppression are identified as a reappearing band in the sequencing ladder of the cleaved fraction. Uncleaved, inactive fractions are separated by size and used as a control for analog incorporation (middle gel lane of bottom panel of figure).

phosphorothioate linkage serves as a chemical marker because it is selectively cleaved by iodine (4). Phosphorothioate analogs are randomly incorporated into the RNA by in vitro transcription at a level of one or two analogs per molecule (1–3). In the example shown in Fig. 2, functional RNAs are distinguished

196

Soukup and Soukup

from nonfunctional by their ability to undergo self-cleavage in the presence of a specific ligand. When an essential nucleotide functional group is modified, that subset of RNA variants is underrepresented in the active fraction. Therefore positions at which nucleotide analogs cause interference result in the absence of corresponding bands on a polyacrylamide sequencing gel. We have thus used this technique to identify essential functional groups within the glucosamine-6-phosphate (GlcN6P)-dependent glmS self-cleaving ribozyme. Our use of NAIS is unique in that it combines ligand analog substitution with interference modification to potentially identify specific tertiary interactions between the RNA and ligand. The assay is based on the principle that if a tertiary interaction is disrupted by deletion or alteration of one functional group in an interacting pair, then no additional penalty will result from deletion or alteration of the second functional group (5–7). Sites of interference suppression exist where modifications that previously caused interference with the normal metabolite are tolerated with ligand analogs. In this manner NAIS can provide information about contacts between the RNA and its ligand. In our studies we utilized glucosamine as ligand analog (which lacks the phosphate group of the natural ligand) in order to identify contacts to or steric boundaries of the phosphate “binding pocket” within the glucosamine-6-phosphate binding site of the glmS ribozyme (8). The NAIM and NAIS experimental details described later can be modified for the study of other RNA–small molecule interactions.

2. Materials 2.1. Transcription and Gel Purification of Riboswitch RNAs Containing Nucleotide Analogs

1. A DNA template for transcription of the glmS riboswitch RNA prepared by primer extension and PCR amplification using synthetic DNA corresponding to the glmS ribozyme sequence from Bacillus cereus. 2. Autoclaved stock solution of 1 M Na+–HEPES, pH 7.5. 3. Sterile filtered stock solutions of 1 M MgCl2, 1 M DTT, and 1 M spermidine. 4. 0.1 M stocks of ATP, CTP, GTP, and UTP, pH adjusted to 7.5 with NaOH. 5. 10% Triton X-100. 6. Stock solutions of each nucleotide analog studied; most analogs can be purchased from Glen Research and come as 10× stocks.

Identification of Metabolite–Riboswitch Interactions

197

7. T7 RNA polymerase can be purchased commercially or purified from Escherichia coli strain (see Note 1). 8. Mutant T7 RNA polymerase (Y639F mutation) purified from E. coli strain (see Note 1). 9. 2× Formamide-loading buffer (sterile filtered): 95% formamide, 25 mM EDTA, 0.025% bromophenol blue, 0.025% xylene cyanol. 10. 10× TBE: 0.9 M Tris–borate, pH 8.0, 10 mM EDTA. 11. 20% Acrylamide stock solution (American Bioanalytical, Natick, MA) containing 1× TBE, 8 M urea and 0.08% TEMED. Store at 4°C. 12. 10% Ammonium persulfate (APS) solution (w/v) in water. Store at +4°C for 3–4 weeks. 13. 1.5-mm thick Teflon spacers and comb for gel preparation (21-cm long). 14. Glass plates (one plate 20 cm wide × 18 cm long, second plate 20 cm wide × 20 cm long). 15. Materials for UV shadowing: razor blades, plastic wrap, handheld shortwave UV lamp, fluorescent thin layer chromatography plate (TLC) (EM Science, Gibbstown, NJ). 16. Crush soak buffer: 10 mM Na+–HEPES, pH 7.5, 1 mM EDTA, 0.2 M NaCl. 17. Ice-cold 100% ethanol. 2.2. glmS Ribozyme Self-cleavage Reaction for NAIM

1. 100 mM Stock of D-glucosamine-6-phosphate sodium salt (Sigma). 2. Sterile filtered stock solution of 1 M MgCl2. 3. Freshly prepared 100 mM Mn(II) acetate. 4. 2× Formamide-loading buffer (sterile filtered) with EDTA: 95% formamide (v/v), 25 mM EDTA, 0.025% (w/v) bromophenol blue, 0.025% (w/v) xylene cyanol, 20 mM EDTA. 5. 0.8-mm thick Teflon spacers and comb for gel preparation (21-cm long).

2.3. glmS Ribozyme Self-cleavage Reaction for NAIS

1. 100 mM Stock of D-glucosamine hydrochloride (Sigma).

2.4. Radiolabeling of Reacted and Unreacted glmS RNAs

1. 10× Reaction buffer: 200 mM Tris–HCl pH 7.5, 20 mM MgCl2, 500 mM NaCl. 2. Calf intestinal alkaline phosphatase: 1 U/mL (Invitrogen, Carlsbad, CA). 3. 20 mM EDTA.

198

Soukup and Soukup

4. T4 polynucleotide kinase: 10 U/mL (Invitrogen) supplied with 10× buffer. 5. [g-32P]ATP: ~6,000 Ci/mmol, 10 mCi/mL (GE Healthcare, Piscataway, NJ). 6. Glass plates (one plate 20 cm wide × 39 cm long, second plate 20 cm wide × 42 cm long). 0.8-mm thick Teflon spacers and comb for gel preparation (42-cm long). 7. Autoradiography supplies: autoradiography cassette, Kodak X-Omat photographic film, film processor, push pins. 8. Ultima gold scintillation fluid (PerkinElmer, Waltham, MA) and 7-mL polypropylene screw top scintillation tubes. 2.5. Gel Separation of Cleaved and Uncleaved RNAs for NAIM and NAIS

1. Glass plates (one plate 33 cm wide × 39 cm long, second plate 33 cm wide × 42 cm long). 0.4-mm thick Teflon spacers, and comb for gel preparation (40-cm long). 2. A stand for gel electrophoresis such as a gel box with aluminum backing from Gibco BRL. 3. Freshly prepared 100 mM iodoethane (0.013 g iodine dissolved in 1 mL 100% ethanol). 4. Supplies for gel analysis: phosphorImager screen (GE Healthcare), plastic wrap, Whatman filter paper, gel dryer.

2.6. Quantitation of NAIM and NAIS Data

1. Phosphorimager and ImageQuant software (GE Healthcare).

3. Methods Detailed as follows are the steps necessary to perform NAIM and NAIS on the glmS self-cleaving riboswitch. Briefly, in this procedure nucleotide analogs are incorporated into RNA molecules using in vitro transcription at 37°C. It should be noted that analog and NTP concentrations may need to be adjusted depending on the length and sequence of the RNA of interest. Most nucleotide analogs are incorporated using bacteriophage T7 RNA Polymerase (2, 3). However, many nucleotide analogs that contain modifications on the minor groove face of the nucleotide need to be incorporated with a mutant form of T7 RNA Polymerase (Y639F) (2, 3). Once nucleotide analogs are incorporated into an RNA population, a selective assay is performed. To be able to utilize the NAIM and NAIS techniques one must be able to separate active RNA molecules from inactive. For the glmS riboswitch, active RNA molecules are distinguished as those that undergo self-cleavage,

Identification of Metabolite–Riboswitch Interactions

199

removing 14 nucleotides from the 5¢ end of the RNA (8–10). Whereas inactive molecules remain full length, active RNAs can be separated by size on a denaturing polyacrylamide gel. A critical feature of successful NAIM and NAIS experiments is the optimization of a selective activity assay. The assay conditions must be stringent in order to identify functional groups most important for RNA structure and function. In the case of catalytically active RNA molecules, selective conditions are typically those that provide 50% reaction at the Km for substrate interaction or Kd for ligand binding. In our analysis of the glmS ribozyme we performed NAIM analyses using the natural ligand GlcN6P (8). Using glmS RNA with parental phosphorothioate analogs (Fig. 1) incorporated, we perform a selective self-cleavage assay in the presence of magnesium. Such experiments identify essential phosphate oxygens within the RNA. Subsequently thiophilic metal ions (e.g., manganese) are included in the assay (in place of magnesium) to determine if phosphorothioate interferences are rescued, indicating that such sites are involved in metal ion coordination (11–13). Additional NAIM experiments are performed with base and sugar analogs using GlcN6P and divalent metal ion (8). In order to perform NAIS experiments we utilize the ligand analog glucosamine (GlcN) (8). Specifically, utilizing glucosamine as ligand analog (which lacks the phosphate group of the natural ligand), we identified potential contacts to or steric boundaries of the phosphate “binding pocket” within the glucosamine-6-phosphate binding site of the glmS ribozyme. Using the same nucleotide analog-containing RNAs described for NAIM, we perform selective activity assays with GlcN and divalent metal ion, increasing reaction time when needed to achieve 50% reaction (8). Once the activity assay is performed (whether it is for NAIM or NAIS experiments) the active and inactive glmS RNA populations are radiolabeled and purified. The samples are prepared by iodine cleavage of phosphorothioate linkages, using equivalent radioactive counts of each RNA. Quantitation and analysis of the NAIM and NAIS data reveals essential functional groups and important tertiary contacts. These methods can be modified to investigate other RNA–small molecule interactions. 3.1. Transcription and Gel Purification of Riboswitch RNAs Containing Nucleotide Analogs

1. Transcription reactions are prepared to synthesize nucleotide analog containing RNAs as well as a control, unmodified glmS RNA. Set up 100 mL transcription reactions in 1.5-mL sterile microcentrifuge tubes using T7 RNA polymerase (0.08 m g/ m L). Final concentrations of reagents should be as follows: 25 pmol DNA template; 50 mM Na+–HEPES of pH 7.5; 15 mM MgCl2; 2 mM spermidine; 5 mM DTT; 1 mM each of ATP, CTP, GTP, and UTP; and phosphorothioate nucleotide analog at a concentration previously determined

200

Soukup and Soukup

to yield ~5% incorporation (2, 3). The ratio of nucleotide analog to parental nucleotide may need to be optimized for some RNAs. Transcription of ribozymes containing nucleotide analogs with a modification on the minor groove face of the nitrogenous base or at the 2¢ position of the ribose sugar may need to be performed with a mutant form of T7 RNA polymerase, the Y639F variant (2, 3). These reactions should contain the following reagents: 25 pmol DNA; 40 mM Na+–HEPES of pH 7.5; 15 mM MgCl2; 4 mM spermidine; 10 mM DTT; 0.05% (v/v) Triton X-100; 1 mM each of ATP, CTP, GTP, and UTP; and phosphorothioate nucleotide analog at a concentration previously determined to yield ~5% incorporation (2, 3). 2. Incubate the transcription reactions at 37°C for ³1 h. 3. During the incubation time, prepare the correct percentage polyacrylamide gel mix by placing ~80 mL of gel mix into a 250-mL flask (see Note 2). Add ~600 mL of 10% APS to the flask, and swirl to mix well. Pour the gel and let it polymerize for ³30 min. The gel spacers and comb should be 1.5-mm thick. 4. When the transcription reactions are completed, add 100 mL of 2× formamide-loading buffer to stop the reaction. Heat the samples at 95°C for 2 min before loading. 5. Prepare the gel for loading using 1× TBE as a running buffer. Load the samples and run the gel at 25 W until the RNA of interest migrates approximately halfway down the gel. Use dye migration as a guide to determine when to stop running the gel (14). 6. Remove the gel from the glass plate assembly, placing plastic wrap on one side of the gel. Place the gel on a fluorescent TLC plate and shine a handheld UV lamp on the gel. Transcribed RNA will absorb UV light while the TLC plate will fluoresce, thereby making it easy to visualize the RNA product as a “shadow.” Using a clean razor blade for each RNA sample, cut the gel slices out. Dice the gel slice into ~1-mm-sized pieces to facilitate elution. Place the gel pieces into 1.5-mL microfuge tubes. The gel pieces should not occupy more than half the volume of the microfuge tube. Fill the rest of the volume of the tube with crush soak buffer. As a rule, the volume of buffer should exceed the volume of gel by twofold. Elute the sample overnight at 4°C (see Note 3 for alternate elution time). 7. After elution, move the liquid eluant from each tube to new 1.5-mL microfuge tube(s). Do not put more than 400 mL of eluant in each tube. Add 1.0 mL of cold 100% ethanol to each tube. Incubate the sample(s) on ice or dry ice for 15 min. To pellet the RNA, centrifuge the tubes at 19,700 × g for 15

Identification of Metabolite–Riboswitch Interactions

201

min. Remove the ethanol and wash the pellet with cold 70% ethanol. Add 500 mL of cold 70% ethanol and centrifuge the tubes at 19,700 × g for 5 min. Remove the ethanol and let the pellet air dry for a few minutes. Resuspend the RNA pellet in ~20 mL of water (see Note 4). 3.2. glmS Ribozyme Self-cleavage Reaction for NAIM

1. Pour a 10% denaturing polyacrylamide gel for separation of the cleaved and uncleaved RNAs. The gel spacers and comb should be 0.8-mm thick. 2. Prepare reactions for the selective assay by combining 10 mL of the RNA prepared in Subheading 3.1 (RNA should be ~2 mM) into a reaction with 50 mM Na+–HEPES of pH 7.5, 200 mM GlcN6P, and 5 mM MgCl2. The reactions are incubated at room temperature (23°C) for 45 s. Reaction conditions represent selective assay conditions with ligand at a concentration near the apparent Kd thereby achieving ~50% ribozyme self-cleavage. Reaction conditions must be optimized for each nucleotide analog utilized to ensure the ~50% self-cleavage. Reactions are stopped by adding an equal volume of 2× formamide-loading buffer with EDTA. 3. In the case of manganese-rescue experiments the reactions are prepared as described earlier except that 5 mM Mn(II) acetate is substituted for MgCl2. No change in reaction time is typically needed to achieve 50% ribozyme self-cleavage. Reactions are stopped as described earlier. 4. Load the samples on the gel and run at 25 W for ~2 h (see Note 5). When the gel is finished, excise the “cleaved” and “uncleaved” bands, elute the samples, and precipitate RNA as described in Subheading 3.1, steps 6–7. Resuspend the “uncleaved” and “cleaved” samples in 8 and 12 mL of water, respectively.

3.3. glmS Ribozyme Self-cleavage Reaction for NAIS

1. The protocol for this procedure is similar to that for NAIM. Pour a 10% denaturing polyacrylamide gel. The gel spacers and comb should be 0.8-mm thick. 2. Prepare reactions for the selective assay by combining 10 mL of the RNA prepared by the protocol in Subheading 3.1 (RNA should be ~2 mM). The reactions contain the RNA, 50 mM Na+–HEPES of pH 7.5, 5 mM GlcN, and 5 mM MgCl2 (8). The reactions are performed at room temperature (23°C) for 35 min. Reaction conditions need to be optimized each time a new ligand is tested in order to determine apparent Kd and thus the concentration of ligand needed for NAIS. In addition, as stated earlier, reaction conditions must be optimized for each nucleotide analog utilized in order to ensure that ~50% self-cleavage is achieved. Reactions were stopped by adding an equal volume of 2× formamide-loading buffer with EDTA.

202

Soukup and Soukup

3. Load the samples on the gel and run at 25 W for ~2 h (see Note 5). When the gel is finished, excise the “cleaved” and “uncleaved” bands, elute the samples, and precipitate RNA as described in Subheading 3.1, steps 6–7. Resuspend the “uncleaved” and “cleaved” samples in 8 and 12 mL of water, respectively. 3.4. Radiolabeling of Reacted and Unreacted glmS RNAs

1. The purified, uncleaved RNAs from NAIM and NAIS experiments above (8 mL) are 5¢-dephosphorylated. The uncleaved RNAs are incubated for 30 min at 37°C in 1× Reaction buffer with 0.1 U/mL of calf intestinal alkaline phosphatase (CIAP). 2. CIAP is inactivated by incubation in 3.3 mM EDTA at 95°C for 15 min. 3. The uncleaved and cleaved RNAs are then 5¢-end labeled. The RNAs are incubated at 37°C for 30 min with 3 mL of [g-32P] ATP, 1× T4 Kinase buffer, and 1 U/mL T4 kinase. 4. Pour a long 10% denaturing polyacrylamide gel for purification of the radiolabeled RNAs. The gel spacers and comb should be 0.8-mm thick. 5. When the kinase reactions are finished, add an equal volume of 2× formamide-loading buffer. Load the samples on the 10% gel and run the gel at ~35 W for 1 h. 6. When the gel is finished, remove it from the glass plates and wrap the gel in plastic wrap. Put the gel in an autoradiography cassette and tape in place. Take the cassette to a darkroom along with photographic film. In the dark open the cassette and place a piece of photographic film on the gel. Be sure to identify the orientation of the film by cutting one corner of the film. Close the cassette and allow the film to expose for 1–2 min. This is usually enough time to get a nice, sharp image of the RNA bands. Open the cassette and remove the film and place it in a film developer. Be careful to not disturb the placement of the gel in the cassette. 7. When the film has been developed, identify the RNA bands of interest. You may also see unused [g-32P]ATP exposed on the film. Using a pushpin, poke holes in the film on either side of the gel bands of interest. Carefully place the film back on the gel in the exact same spot that the exposure is performed. Using a fine point marker, mark through the holes onto the plastic wrap surrounding the gel. Lift the film and be sure you can see the marks. Remove the gel from the cassette and place it on a clean glass plate. Using a fresh razor blade for each RNA sample, excise the RNA bands from the gel. As a rule if you make a horizontal cut ~1 mm above and ~1 mm below your mark and then vertical cut through the marks, you should retrieve most of the RNA from the gel. Place the gel slices into clean 1.5-mL microfuge tubes and add enough

Identification of Metabolite–Riboswitch Interactions

203

crush soak buffer to cover the gel slice (usually 400 mL). Elute the sample overnight at 4°C. 8. Remove the eluant from the tube with the gel chunk and proceed with an ethanol precipitation as described in Subheading 3.1, step 7. Resuspend the samples in 11 mL of water. 3.5. Gel Separation of Reacted and Unreacted RNAs to Map Nucleotide Analog Interferences and Suppressions

1. In order to load equivalent amounts of each RNA sample on the gel, measure the extent of radiolabeling using a scintillation counter. Prepare scintillation tubes by adding 3 mL of Ultima gold scintillation fluid to each sample tube. Add 1 mL of each radioactive RNA to individual tubes. Mix well. Using a scintillation counter, determine the cpm (counts per minute) value for each sample and calculate the volume of each RNA for loading an equal number of counts (~100,000 counts) per lane. The sample volumes should all be the same (usually 10 mL); adjust with sterile water if necessary. Add 10 mL of 2× formamide-loading buffer. Split the sample into two tubes, one labeled “+iodine” and one labeled “−iodine.” The “−iodine” sample is ready for loading on the gel. To the “+iodine” samples add 1 mL of 100 mM iodoethane. Heat all samples at 95°C for 2 min before loading on the gel. 2. Pour a 10% denaturing polyacrylamide gel (sequencing size) for separation of the radiolabeled RNAs. The gel spacers and comb should be 0.4-mm thick. Avoid bubbles while pouring the gel. Once the gel is polymerized, prerun at 75 W for 15 min (see Note 6). Load 10 mL of each RNA sample on the gel and run the gel at 75 W for the time necessary to separate the region of interest (see Note 5). 3. When the gel is finished, take it from the gel stand and remove one of the glass plates. In order to pry the gel plates apart, it is sometimes necessary to use a razor blade, as a spatula is too thick. Transfer the gel to filter paper. Be careful so as not to rip the gel when pulling it off the glass plate. Once the gel is on the filter paper, cover it with plastic wrap. Trim the excess filter paper and plastic wrap from around the gel. Place the gel in a gel dryer for 1 h at 80°C. Expose the gel to a phosphorimager screen overnight.

3.6. Quantitation of NAIM and NAIS Data

1. Scan the phosphorimager screen (see representative NAIM gel in Fig. 3 and NAIS gel in Fig. 4). 2. Use ImageQuant software to quantitate band intensities. 3. Export ImageQuant data into Excel for calculation of interference (k) values. 4. Interference (k) values for the parental analogs (phosphorothioate effects) are calculated by dividing the band intensities for the cleaved RNAs by the band intensities for the uncleaved RNAs.

204

Soukup and Soukup

Fig. 3. Phosphorothioate NAIM of the B. cereus glmS ribozyme. Autoradiograms depict examples of phosphorothioate nucleotide analog (NaS) incorporation in uncleaved and cleaved ribozymes. Nucleotide positions of interest are indicated to the left of each gel. Sites of phosphorothioate interference are denoted at the left of corresponding bands with a filled circle. Sites of manganese rescue are denoted at the left of corresponding bands with an asterisk.

Fig. 4. NAIS at position G3 within the B. cereus glmS ribozyme. Autoradiograms depict examples of phosphorothioate nucleotide analog (NaS) incorporation in uncleaved and cleaved ribozymes. Ribozymes are prepared by reaction with GlcN6P or GlcN under selective conditions in the presence of magnesium ions. The nucleotide position that exhibits phosphorothioate interference with GlcN6P is denoted at the left of the corresponding band with a filled circle. The site of interference suppression with GlcN is denoted by an arrow. Identification of positions at the 5¢ end of the glmS ribozyme is aided by the inclusion of alkaline hydrolysis ladders (OH).

5. Subsequent interference values for base and sugar analogs are calculated using the following equation which normalizes for phosphorothioate effects (2, 3, 8, 15): 6. Interference (k) = (parental cleaved/analog cleaved)/(parental uncleaved/analog uncleaved). 7. The k values can be further normalized for variations in analog incorporation, loading differences, and extent of reaction between samples.

Identification of Metabolite–Riboswitch Interactions

205

4. Notes 1. T7 RNA polymerase and the mutant Y639F polymerase were purified in our laboratory as His-tagged proteins. 2. Use ref. 14 to choose the correct polyacrylamide gel percentage to maximize separation of your nucleic acid of interest. 3. An alternative to overnight elution is to elute the gel bands in crush soak buffer at room temperature for ³1 h. The room temperature elution time will need to be optimized for your specific nucleic acid. Longer nucleic acid samples will need longer elution times. 4. The resuspension volume for your nucleic acid of interest will need to be optimized for your specific experiment. 5. Use ref. 14 to estimate the running time for your gel electrophoresis experiments. The migration of bromophenol blue and xylene cyanol can be used to determine running time. 6. For electrophoresis of the large gel at 75 W, we use a specialized aluminum-backed gel box. If such an apparatus is not available, you should attach an aluminum plate to your gel plate assembly to allow for consistent heating across the gel and to avoid cracking the glass plates. References 1. Strobel, S. and Shetty, K. (1997). Defining the chemical groups essential for Tetrahymena group I intron function by nucleotide analog interference mapping. Proc. Natl. Acad. Sci. U. S. A. 94, 2903–2908. 2. Ryder, S. and Strobel, S. (1999). Nucleotide analog interference mapping. Methods 18, 38–50. 3. Ryder, S., Ortoleva-Donnelly, L., Kosek, A. and Strobel, S. (2000). Chemical probing of RNA by nucleotide analog interference mapping. Methods Enzymol. 317, 92–109. 4. Gish, G. and Eckstein, F. (1988). DNA and RNA sequence determination based on phosphorothioate chemistry. Science 240, 1520–1522. 5. Strobel, S., Ortoleva-Donnelly, L., Ryder, S., Cate, J. and Moncoeur, E. (1998). Complementary sets of noncanonical base pairs mediate RNA helix packing in the group I intron active site. Nat. Struct. Biol. 5, 60–66. 6. Szewczak, A., Ortoleva-Donnelly, L., Ryder, S., Moncouer, E. and Strobel, S. (1998). A minor groove triple helix in the active site of the Tetrahymena group I intron. Nat. Struct. Biol. 5, 1037–1042.

7. Soukup, J., Minakawa, N., Matsuda, A. and Strobel, S. (2002). Identification of A-minor tertiary interactions within a bacterial group I intron active site by 3-deazaadenosine interference mapping. Biochemistry 41, 10426–10438. 8. Jansen, J.A., McCarthy, T.J., Soukup, G.A. and Soukup, J.K. (2006). Backbone and nucleobase contacts to glucosamine-6-phosphate in the glmS ribozyme. Nat. Struct. Mol. Biol. 13, 517–523. 9. Winkler, W.C., Nahvi, A., Roth, A., Collins, J.A. and Breaker, R.R. (2004) Control of gene expression by a natural metabolite-responsive ribozyme. Nature 428, 281–286. 10. McCarthy, T.J., Plog, M.A., Floy, S.A., Jansen, J.A., Soukup, J.K. and Soukup, G.A. (2005). Ligand requirements for glmS ribozyme selfcleavage. Chem. Biol. 12, 1221–1226. 11. Christian, E. and Yarus, M. (1993). Metal coordination sites that contribute to structure and catalysis in the group I intron from Tetrahymena. Biochemistry 32, 4475–4480. 12. Chanfreau, G. and Jacquier, A. (1994). Catalytic site components common to both splicing steps of a group II intron. Science 266, 1383–1387.

206

Soukup and Soukup

13. Hardt, W., Warnecke, J., Erdmann, V. and Hartmann, R. (1995). Rp-phosphorothioate modifications in RNase P RNA that interfere with tRNA binding. EMBO J. 14, 2935–2944. 14. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: A Laboratory

Manual. (Nolan, C., ed.), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 15. Ortoleva-Donnelly, L., Szewczak, A.A., Gutell, R.R. and Strobel, S. (1998). The chemical basis of adenosine conservation throughout the Tetrahymena ribozyme. RNA 4, 498–519.

Chapter 15 RNA-Dependent RNA Switches in Bacteria Tina M. Henkin Summary The versatility of RNA as a regulatory molecule has become increasingly apparent in recent years. RNA elements within a transcript can sense a large variety of physiological signals, resulting in modulation of expression of the gene(s) encoded on that transcript by effects on transcript synthesis, mRNA stability, and translation. The response of the transcript to the signal can result in direct sequestration of elements involved in various steps of gene expression, or promotion of a structural rearrangement of the target mRNA that affects the activity of elements important for expression. The regulatory signal can be as simple as a change in temperature, or as complex as a translating ribosome, the processivity of which is affected by the appropriate cellular parameter. Among these many regulatory mechanisms are those in which the regulatory signal is a specific RNA that binds to the target mRNA. Systems of this type are widespread in bacteria, and they fall into several classes based on the type of RNA that acts as the regulator, and the mode of interaction of the regulatory RNA with its target. These include systems in which the regulatory RNA is encoded on the opposite strand of its target, systems in which the regulatory RNA is encoded elsewhere in the genome, and systems in which RNAs (such as tRNAs) that are normally used for a different cellular process are borrowed for use as regulatory molecules. This review will summarize these classes of regulatory mechanisms and their use in bacterial systems. Key words: Regulatory RNA, sRNA, tRNA, Antitermination, mRNA stability, Translation initiation

1. Introduction Mechanisms in which a cis-acting element within an mRNA serves as the target for response to a regulatory signal have emerged as a common theme for gene regulation in bacteria (1). The target element is commonly positioned in the 5¢ region of the transcript, upstream of the start of the regulated coding sequence. This region, designated the leader RNA, may contain a signal for premature termination of transcription, the activity of which Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_15 © Humana Press, a part of Springer Science + Business Media, LLC 2009

207

208

Henkin

is modulated by binding of the regulatory factor. Alternatively, the translation initiation region of the regulated coding sequence may have the potential to be sequestered in a structural element that prevents binding of the translation initiation complex. A third possibility is that binding of the regulatory factor affects the stability of the mRNA transcript, thereby affecting its availability to the translational machinery. The target element can also be located within the 3¢-untranslated region of the transcript or an intron, where it can affect mRNA stability or splicing in eukaryotic systems. In each case, the regulatory element in the mRNA specifically recognizes its cognate signal to allow changes in gene expression in response to changes in cell physiology. The first systems of this type that were described utilize translation of a short peptide encoded within the leader region of a transcript to modulate expression of the downstream coding sequence. Processivity of the ribosome during leader RNA translation is modulated by the availability of a specific aminoacylated tRNA (e.g., tRNATrp for the E. coli tryptophan biosynthesis operon) or by binding of a small molecule to the elongating ribosome (e.g., tryptophan for the E. coli tryptophanase operon). The position of the stalled ribosome over the leader peptide coding region can affect folding of the transcript to affect formation of the helix of an intrinsic transcriptional terminator or a competing antiterminator, or can block binding of the Rho transcription termination factor at Rho-dependent terminators. Ribosome stalling can also be triggered by binding of an antibiotic, as in a set of chloramphenicol resistance genes from Gram-positive bacteria, and the resulting effect on transcript structure can affect accessibility of the Shine–Dalgarno sequence necessary for binding of the 30S ribosomal subunit and translation of the downstream coding sequence, and can also affect transcript stability. In each of these systems, the translating ribosome senses an informational signal and transmits that signal to the mRNA that it is translating, resulting in effects on the fate of that mRNA. Environmental information can also be transmitted to mRNA targets by RNA-binding proteins, the activity of which is affected by an appropriate signal. These systems are similar to classical DNA-binding activators and repressors but act at a level of gene expression subsequent to transcription initiation. In simpler autoregulation systems, the availability of excess levels of the RNA-binding protein represents the regulatory signal. Binding of the protein to the transcript can affect premature termination of transcription, translation initiation, and transcript stability, thereby affecting expression of the target genes. RNAs can also directly sense environmental parameters, without the intervention of proteins or ribosomes. In these systems, the RNA folds into a structure that is competent for specific binding of the signal molecule, often a cellular metabolite,

RNA-Dependent RNA Switches in Bacteria

209

and binding results in a change in the RNA structure that impacts expression of the genes encoded on that transcript. Regulation commonly occurs at the levels of premature termination of transcription or initiation of translation, and can also affect mRNA stability. The simplest of these systems are the RNA thermosensors, in which structure sequestering the translation initiation region is sensitive to variation in temperature. Metabolite-binding RNAs utilize complex RNA structures to recognize the cognate ligand. The base-pairing properties of RNA can also be used to allow the target RNA to recognize a second RNA molecule that serves as the regulatory signal. There are three basic classes of RNA-dependent regulatory systems that rely on base pairing between the regulatory RNA and its mRNA target: (1) cis-encoded antisense RNA systems; (2) trans-encoded small RNA (sRNA) systems; and (3) tRNA-dependent systems. In each case, the regulatory RNA binds specifically to its target, and gene expression can be affected at the levels of transcription termination, translation initiation, and mRNA stability. There are a number of additional RNA-mediated regulatory mechanisms in bacterial cells, but this review will focus on those that involve RNA base-pairing interactions between the regulatory RNA and its target.

2. Cis-Encoded Antisense RNA Systems

Cis-encoded antisense RNAs are short transcripts that are transcribed from the opposite strand of same DNA region that encodes their target “sense” RNA transcript, and are therefore perfectly complementary to their target (Fig. 1a). sRNAs of this type were first identified in bacteria in genetic elements including plasmids, bacteriophages, and transposable elements, and have recently been uncovered in chromosomal genes (2). Unlike many trans-encoded sRNAs (see later), the cis-encoded sRNAs usually have a single target. These RNAs are often highly structured, and pairing between the sRNA and the target is usually initiated within loop sequences, followed by partial melting of the sRNA structure to allow an extended base-paired region. Some of the cis-encoded antisense sRNAs require a helper protein to stabilize the RNA–RNA interaction, while most function in the absence of a protein, probably due to the extensive complementarity between the sRNA and its target. Antisense sRNAs can affect gene expression at multiple levels. Termination of transcription can be induced in systems where the target RNA can fold into mutually exclusive terminator and antiterminator structures. In the plasmid systems studied to date,

210

Henkin

Fig. 1. Regulation by antisense sRNAs. (a) Cis-encoded antisense sRNAs. The regulatory sRNA (antisense sRNA) is derived from transcription of the same DNA region as the target gene (sense mRNA), but from the opposite strand; the two RNAs are therefore perfectly complementary. (b) Trans-encoded antisense sRNAs. The regulatory sRNA (antisense sRNA) is derived from a different DNA region than the target gene and is only partially complementary to the target RNA (sense mRNA). The promoter regions are labeled (−35 and −10 are the positions of interaction with RNA polymerase, and +1 is the transcription start site), and the bent arrow indicates the position and direction of transcription initiation. The transcription termination sites are labeled (T). Dotted lines represent the sense and antisense RNA transcripts, and hatch marks indicate base pairing between the regulatory sRNA and the target mRNA.

the transcript normally folds into the antiterminator structure, which allows RNA polymerase to continue transcription through the termination site. Binding of the sRNA destabilizes the antiterminator, usually by sequestering sequences that form the 5¢ side of this element. This frees sequences that form the 3¢ side of the antiterminator, allowing them to instead form the terminator helix, which results in premature termination of transcription and downregulation of the downstream target gene. Antisense RNAs can also act at the level of translation initiation, by binding to the region of the target transcript that must be recognized by the translation initiation machinery. Binding therefore prevents translation. A similar interaction can affect the stability of the target mRNA, by protecting the mRNA from the action of endoribonucleases or stimulating cleavage. RNase III has been implicated in several cases as the endoribonuclease responsible for degradation of the RNA–RNA complex, consistent with its preference for cleavage of extended RNA helices. Cis-encoded antisense RNAs also play important roles in plasmid replication

RNA-Dependent RNA Switches in Bacteria

211

control independent of effects on gene expression (e.g., for the ColE1 replication system). In some cases, synthesis of the antisense RNA is constitutive, and its regulatory effect is sensitive to the antisense sRNA:target RNA ratio. In contrast, regulation of chromosomal genes by antisense sRNAs usually involves regulation of the synthesis of the antisense RNA.

3. Trans-Encoded sRNA Systems As suggested by the name, trans-encoded sRNAs are encoded in a locus distinct from the DNA region that encodes the target (Fig. 1b). These sRNAs are only partially complementary to their target mRNAs, and a single sRNA may have multiple targets (3). The limited complementarity between the sRNA and its targets is likely to be responsible for the requirement in many cases for the Hfq protein, an RNA chaperone that can bind to and stabilize sRNAs and can also facilitate complex formation of the sRNA with its mRNA target. Regulation mediated by trans-encoded antisense sRNAs commonly operates at the levels of translation initiation or mRNA stability. Translational repression usually occurs by binding of the sRNA to a region of the target mRNA that includes the binding site for the translation initiation complex, so that sRNA–mRNA complex formation prevents binding of the 30S ribosomal subunit. A less common mechanism utilizes an mRNA target that contains an intrinsic structural element that occludes the ribosome-binding site, so that translation is repressed in the absence of the sRNA. Binding of the sRNA to a region of the transcript that would otherwise pair with the ribosome-binding site region results in unfolding of the mRNA and increased translation. It is interesting to note that the same sRNA can in some cases (e.g., E. coli DsrA) have opposite effects on different target mRNAs, depending on the inherent structural conformation of the target transcript (e.g., available or sequestered ribosome-binding site) and the position of sRNA base pairing relative to the ribosome-binding site. Inhibition of translation initiation can result in decreased stability of a transcript within the cell because the decreased occupancy of the mRNA by elongating ribosomes allows increased access to endoribonucleases that initiate the mRNA degradation pathway. sRNA binding can also directly stimulate transcript degradation, as described earlier for cis-encoded antisense RNAs, by enhancing cleavage by a specific endoribonuclease. Whereas the extended duplex formed as a consequence of binding of a cis-encoded antisense sRNA often results in cleavage by RNase III, degradation promoted by trans-encoded sRNAs appears in at least in some cases (e.g., E. coli RyhB) to utilize RNase E.

212

Henkin

4. tRNA-Dependent Regulation: The T Box Mechanism

The T box regulatory mechanism is used for a large number of amino acid-related genes, primarily in Gram-positive bacteria. Genes in this family exhibit a set of conserved primary sequence and structural elements in the leader region of the transcript. Most genes in this group are regulated at the level of premature termination of transcription, and formation of the terminator helix can be prevented by binding of a specific uncharged tRNA to the nascent RNA transcript. tRNA binding promotes stabilization of an antiterminator element that includes sequences necessary for formation of the terminator helix, so that antiterminator formation precludes termination (4, 5). Regulation at the level of translation initiation is predicted for a subset of T box genes. Binding of a specific tRNA to the leader RNA is dependent on an appropriately positioned triplet sequence that represents a specific codon, the identity of which matches the amino acid specificity of the gene encoded downstream of the regulatory domain (e.g., a tyrosyl gene contains a UAC tyrosine codon, a phenylalanyl gene contains a UUC phenylalanine codon, etc.). The codon, designated the “Specifier Sequence,” directs pairing with the anticodon of the matching tRNA. A second pairing between the four unpaired residues at the 3¢ end of the tRNA and four residues within a bulged domain of the antiterminator element is required for stabilization of the antiterminator and readthrough of the termination site. Both uncharged and charged tRNAs are competent for codon–anticodon pairing at the Specifier Sequence. However, only uncharged tRNA can permit pairing of the acceptor end with the antiterminator, as aminoacylation of the tRNA places the amino acid at the 3¢ end of the tRNA where it interferes with interaction with the antiterminator (Fig. 2). This system therefore monitors the ratio between two cellular RNAS (uncharged and charged species of a specific tRNA) rather than the absolute amount of a single RNA species, as is the case for the antisense sRNAs. tRNAGly-dependent antitermination of the Bacillus subtilis glyQS T box leader region terminator can occur in a purified in vitro transcription system, indicating that no cellular factors other than the tRNA are required. Direct interaction of purified glyQS leader RNA and tRNAGly could also be demonstrated, and structural mapping verified the predicted model for the leader RNA and revealed structural changes in both RNAs in the complex. These and other studies showed that while the two regions of RNA–RNA base pairing (codon–anticodon and antiterminatortRNA 3¢ end) are required, other interactions between the two RNAs are important for complex formation and tRNA-dependent antitermination. In contrast to the antisense sRNAs described,

RNA-Dependent RNA Switches in Bacteria

213

Fig. 2. tRNA-dependent regulation. (a) Binding of a specific aminoacylated tRNA, directed by pairing between the codon located at the Specifier Sequence (S) and the tRNA anticodon, allows formation of the terminator helix (T), and transcription of the downstream coding sequence is inhibited. (b) Binding of the cognated uncharged tRNA at both the Specifier Sequence and the antiterminator (AT) stabilizes the antiterminator and promotes readthrough of the termination site and transcription of the downstream coding sequence.

the initial pairing at the ends of the tRNA does not result in extensive unfolding of the RNA structure, as the overall arrangement of both the leader RNA and the tRNA is maintained in the complex. Changes in the Specifier Sequence (e.g., UAC tyrosine codon to UUC phenylalanine codon) are sometimes sufficient to allow recognition of a different tRNA, but the response to a noncognate tRNA is always less efficient than the interaction with the cognate tRNA, both in vivo and in vitro. These studies showed that while the two regions of RNA–RNA base pairing (codon– anticodon and antiterminator-tRNA 3¢ end) are required, other interactions between the two RNAs are important for complex formation and tRNA-dependent antitermination. These are likely to involve non-Watson–Crick interactions and subtle conformational variations between different tRNA classes that have evolved for adequate discrimination by aminoacyl-tRNA synthetases.

5. Conclusion RNA has demonstrated a remarkable ability to interact specifically with a large variety of ligands, including its natural cellular partners (e.g., ribosomes and aminoacyl-tRNA synthetases), RNA-binding proteins, and small molecules. These interactions have been appropriated by a large variety of regulatory mechanisms to control gene expression in response to a wide range of physiological signals. The recognition of this class of mechanism

214

Henkin

in recent years has opened up a new field of investigation, and it is likely that many more variations on this basic theme will be uncovered, both in bacteria and in higher organisms.

Acknowledgments This work was supported by NIH R01 GM47823 and NIH R01 GM63615.

References 1. Grundy, F. J. and Henkin, T. M. (2006). From ribosome to riboswitch: control of gene expression in bacteria by RNA structural rearrangements. Crit. Rev. Biochem. Mol. Biol. 41, 329–338. 2. Brantl, S. (2007). Regulatory mechanisms employed by cis-encoded antisense RNAs. Curr. Opin. Microbiol. 10, 102–109. 3. Storz, G., Altuvia, S., and Wassarman, K. M. (2005). An abundance of RNA regulators.Annu. Rev. Biochem. 74, 199–217.

4. Henkin, T. M. (1994). tRNA-directed transcription antitermination. Mol. Microbiol. 13, 381–387. 5. Henkin, T. M. and Grundy, F. J. (2006). Sensing metabolic signals with nascent RNA transcripts: the T box and S box riboswitches as paradigms. Cold Spring Harb. Symp. Quant. Biol. 71, 231–237.

Chapter 16 Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes and Lead(II) Clément Chevalier, Thomas Geissmann, Anne-Catherine Helfer, and Pascale Romby Summary Enzymatic probing and lead(II)-induced cleavages have been developed to study the secondary structure of RNA molecules either free or engaged in complex with different ligands. Using a combination of probes with different specificities (unpaired vs. paired regions), it is possible to get information on the accessibility of each nucleotide, on the binding site of a ligand (noncoding RNAs, protein, metabolites), and on RNA conformational changes that accompanied ligand binding or environmental conditions (temperature, pH, ions, etc.). The detection of the cleavages can be conducted by two different ways, which are chosen according to the length of the studied RNA. The first method uses end-labeled RNA molecules and the second one involves primer extension by reverse transcriptase. We provide here an experimental procedure that was designed to map the structure of mRNA and mRNA–sRNA interaction in vitro. Key words: RNA, RNA–RNA interaction, Secondary structure, RNA structure probing, Ribonuclease, Lead(II)-induced cleavages

1. Introduction The diversity of RNA functions is intimately linked to the capability of the RNA to adopt different conformations. RNA switches have the obvious advantage to promote diverse points of contacts for the selective recognition of various RNA ligands. For instance large RNA–protein complexes such as the ribosome require numerous proteins that coordinate RNA conformational changes during the assembly (1, 2). More recently it was shown Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_16 © Humana Press, a part of Springer Science + Business Media, LLC 2009

215

216

Chevalier et al.

that regulatory regions of bacterial mRNAs undergo conformational changes in respond to the environmental cues (thermosensors) (3) or to metabolite concentration changes (riboswitches) (4, 5). In addition, many novel small noncoding RNAs (sRNA) have been identified in bacterial genomes (6). A large class of sRNA represses or activates gene expression through direct pairings with target mRNA (7). Mechanistic studies reveal that the interaction between sRNA and mRNA is usually constrained by the structures of both RNAs, and that fast pairings are required for efficient regulation (8, 9). Over the years the determination of high-resolution RNA and ribonucleoprotein particle (RNP) structures by X-ray analysis and NMR provided a guide for further ongoing mechanistic and functional studies (4, 10). While a huge progress has been made in this field with the discovery of many novel RNA structural motifs (11), it is however still difficult to predict accurately the structure of long RNA molecules such as mRNA simply based on the primary structure (12). Therefore enzymatic and chemical probing remains a useful approach to monitor the RNA structure of any size under a large variety of experimental conditions. Structure mapping in solution provides the accessibility or the reactivity of each nucleotide toward chemicals or enzymes, and identifies without ambiguity the unpaired RNA regions. The elaboration of a secondary structure model is obtained by coupling the mapping data with the help of several computer folding programs based on energy minimization (13), statistics (14), or stochastic simulations (15, 16), and if available with phylogenetic and sequence comparison (see Chapter “Structural Probing of RNA Thermosensors”). The effect of mutations at strategic positions and base compensatory changes can further validate the existence of helices. Enzymes are also extensively used to map the binding site of a ligand (noncoding RNAs, protein, metabolites), to study RNP assembly, and to monitor the mRNA conformational changes that accompanied ligand binding or biochemical activity. Ribonucleases were first adapted for RNA sequencing methodology (17). Most of the enzymes induce cleavages within unpaired RNA regions (18), whereas RNase V1 is the only probe that provides positive signal for the existence of helical regions (19, 20) (see Table 1). The enzymes are easy to handle and are useful to identify secondary structure RNA elements such as hairpin or pseudoknot motifs. Due to the bulky size of the RNases, they are however sensitive to steric hindrance. Another probe which provides information on unpaired regions is the divalent cation lead(II). Since stable RNA helices are resistant to lead(II)induced cleavages, the probe was largely used to map prokaryotic mRNA–sRNA interaction in vitro (21–24). Lead(II) was also

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

217

Table 1 Properties of the probes. MW is for molecular weight, G for guanine, A for adenine, C for cytosine, U for uracile, N for nucleotide Probes

MW

specificity

Product

Special considerations

RNase T1

11,000

Unpaired G

…Gp3¢

Active under a wide range of conditions: pH, T between 4 and 55°C, with or without magnesium ions and monovalent ion, active in urea

RNase T2

36,000

Unpaired A > C, G, U

…Ap3¢

Active under different conditions: T between 4 and 40°C, works with or without magnesium ions

RNase A

12,500

Unpaired C and U

…C/Up3¢

Active under a wide range of conditions: T°C between 4 and 55°C, with or without magnesium ion and monovalent ion, active in urea

Nuclease S1

32,000

Unpaired N

…NOH3¢

Optimum pH at 4.5, requires Zn2+ for its activity

RNase V1

15,900

Paired or stacked N

…NOH3¢

Requires divalent cations and active under a wide range of T (from 4 to 45°C) and pH (4–9)

Pb(II) acetate

200

Specific divalent ion sites

…Np3¢

Cleaves under a large variety of experimental conditions. Ratio Mg2+/Pb2+ influences the reaction Appropriate for footprinting, can be used in vivo

Unpaired regions

found to be appropriate to map the structure of mRNAs and regulatory RNAs in living bacterial cells since it easily penetrates the membrane (25, 26). Thus, the comparison between in vivo and in vitro mapping determines the functional RNA structure. In the present review, we provide an experimental guide of the most commonly used enzymes and lead(II)-induced cleavages for mapping RNA structure in vitro. Other detailed protocols used for chemical probing in vitro will be found in this issue (see Chapters “Structural Probing of RNA Thermosensors” and “Analysis of the RNA Backbone: Structural Analysis of Riboswitches by In-Line Probing and Selective 2¢-Hydroxyl Acylation and Primer Extension”) and have been previously reported (27, 28).

218

Chevalier et al.

2. Materials 2.1. RNA Preparation and Renaturation

1. Safety rules have to be applied for handling radioactive materials (see Note 1). 2. In vitro transcribed RNA of interest (e.g., Staphylococcus aureus regulatory RNAIII and its target mRNA), purified and dephosphorylated at 5¢ end. 3. T4 polynucleotide kinase (PNK), 10 U/mL, supplied with 10× T4 PNK buffer (Fermentas Vilnius, Lithuania). 4. T4 RNA ligase (Ambion, Austin, TX). 5. Radiochemicals: [g-32P]ATP (3,000 Ci/mmol); [5¢-32P]pCp (3,000 Ci/mmol). 6. RNA elution buffer: 500 mM ammonium acetate, pH 6.5, 1 mM ethylenediaminetetraacetic acid (EDTA). 7. Phenol saturated with 0.1 M Na-acetate, pH 6.5. 8. Phenol/chloroform/isoamyl alcohol (25:24:1) mixture, pH adjusted to 8.0 with Tris–HCl.

2.2. Enzymatic and Lead(II) Hydrolysis

1. Basic laboratory materials and equipment are required such as microcentrifuge, vortex, thermoblock, water bath, radioactivity counter, autoradiography films, intensifying screens. 2. The buffer conditions for enzymatic hydrolysis and lead(II) have to be adapted according to the mRNA and the nature of the ligand. Composition of buffers is given at the final concentration in the assays. Stock solutions, that are usually prepared, are 5–10 times concentrated. All solutions, water, Eppendorf tubes and tips should be RNase free. The buffers are stored at −20°C. 3. Buffer N1: 50 mM Tris–HCl, pH 7.5, 10 mM MgCl2, 100 mM KCl. 4. Buffer N2: 50 mM Tris–HCl, pH 7.5, 10 mM MgCl2, 100 mM KCl, 1 mM ZnCl2. 5. Buffer N3: 50 mM Tris–acetate, pH 7.5, 5 mM Mg-acetate, 100 mM K-acetate (see Note 2). 6. Several ribonucleases are commercially available. RNase T1 (specific for unpaired guanines) can be purchased from Fermentas (ref. EN0541, 1,000 U/mL), RNase T2 (specific for unpaired residues with a preference for adenine) is from Invitrogen (Carlsbad, CA) (ref. 18031-013, 20 U/mL), RNase V1 (specific for double stranded regions) is from Pierce (Rockford, IL) (ref. MB092700, 0.9 U/mL) or from Ambion (Billerica, MA) (ref. 2275, 0.1 U/mL), nuclease S1 (specific for unpaired regions) is from (Madison, WI) (M5761, 100 U/mL), and RNase A (specific for unpaired C and U residues) is from Ambion (AM2274, 1 mg/mL).

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

219

7. Lead(II) acetate (ref. 31723, Acros organics, Geel, Belgium). 8. 1 mg/mL total yeast tRNA (Sigma). 9. Precipitation/inactivation buffer (Ambion). 10. 0.3 M Na-acetate, pH 6.0. 2.3. Fractionation of End-Labeled RNA Fragments

1. Buffer DT1: 20 mM sodium citrate of pH 4.5, 1 mM EDTA, 7 M urea, 0.02% (w/v) xylene cyanol, 0.02% (w/v) bromophenol blue. 2. Ladder Buffer: 0.1 M Na2CO3/0.1 M NaHCO3, pH 9. 3. RNA-loading buffer: 0.02% xylene cyanol, 0.02% bromophenol blue in 8 M urea.

2.4. Detection of Cleavages by Primer Extension

1. 1× RTB buffer: 50 mM Tris–HCl of pH 8.0, 10 mM MgCl2, 50 mM KCl, 1 mM dithiothreitol (DTT). 2. dNTP mix: 2.5 mM of dATP, dGTP, dCTP, dTTP (Amersham). 3. Avian myeloblastosis virus reverse transcriptase (AMV RT) purchased from MP Biochemicals (France) or Life Sciences (USA). 4. DNA-loading buffer: 1 mM EDTA, 0.02% xylene cyanol, 0.02% bromophenol blue in formamide. 5. Oligodeoxyribonucleotide primer 5¢-AGGGAATGTTTTACAGTAT-3¢. 6. Micro Bio-spin 6 chromatography column (Biorad). 7. RNA hydrolysis buffer: 50 mM Tris–HCl of pH 7.5, 7.5 mM EDTA, 0.5% SDS. 8. Individual dideoxyribonucleotide nucleotide triphosphates (ddNTP).

2.5. Fractionation of Cleaved Fragments by PolyacrylamideUrea Gel Electrophoresis

1. Electrophoresis apparatus for slab gels (30 × 40 cm) and generator (2,000 V). 2. 1× TBE buffer: 0.09 M Tris–borate of pH 8.3, 1 mM EDTA. 3. 25% Polyacrylamide gel in 8 M urea: dissolve 480 g urea in 625 mL Rotiphorese 40 solution (acrylamide/bis-acrylamide ratio 19:1) (Roth, Karlsruhe, Germany), adjust the volume to 1 L with bidistillated water, filtrate the solution. 4. Other reagents for PAGE: 8 M urea, N, N, N¢, N¢-tetramethyleethylenediamine (TEMED); ammonium persulfate (APS) should be prepared as a 10% (w/v) solution in water (see Note 3). 5. Gel-fixing solution: 10% ethanol, 6% acetic acid in water. Prepare 2 L before gel fixing.

220

Chevalier et al.

3. Methods 3.1. Setting the Experimental Conditions

1. Probing the conformation of RNAs with different enzymes requires defined buffer conditions (pH, ionic strength, magnesium concentration, temperature). The optimal conditions vary with the enzymes, and subtle conformational changes of the RNA may occur under different experimental conditions (Table 1). For instance, nuclease S1 requires Zn2+ for its activity and its optimal pH is 4.5. Some of the enzymes, as RNases T1 and A, do not require magnesium for hydrolysis and can work at different temperatures (29). Thus, the influence of monovalent or divalent ion (such as magnesium) can be tested on the RNA folding, and thermal transition of RNA molecules can be obtained by varying the temperature (see Note 4). Such experiments provide information on the stability of the secondary structure domains. They also allow the identification of tertiary elements since these interactions are the first to break during the melting of an RNA structure. The functioning of some of the riboswitches is dependent on the kinetics of ligand binding (30). In addition, the kinetics for complex formation between sRNA and target mRNAs is essential for regulation (9). In most of the cases, footprinting assays using enzymes are conducted under equilibrium conditions. In order to get a dynamic view of complex formation, time-resolved kinetic footprinting assays have been developed with probes that generate hydroxyl radicals in the ms range (2, 31, 32). Recent methods make use of a quench-flow apparatus and exploit reactions that are faster than the interactions between mRNA with their ligands (33). Such a method can be certainly adapted with lead(II) ion. 2. The probe to RNA ratio must be adapted so that the experiments are conducted under limited and statistical conditions in order to get less than one cut per molecule. For the first experiment, different concentrations of the probes and a timescale dependence should be performed. This is also required when the commercial source of the probe has been changed. As mentioned earlier, defined mild buffer conditions have to be used for all the probes in order to be close to the in vivo conditions (neutral pH, presence of divalent and monovalent ions, temperature). In that respect, RNase T1 (specific for unpaired guanines), RNase A (specific for unpaired cytosines and uridines), RNase V1 (specific for paired regions), and lead(II) (specific for unpaired regions) can be used under strictly identical experimental conditions and in addition provide complementary information. Since the purification of RNase T2 will be no more commercially available, nuclease S1

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

221

can provide equivalent data. However the activity of nuclease S1 is considerably reduced at pH 7.5, and hence large amount of enzyme should be added in the assays. Reducing agents (DTT, or b-mercaptoethanol) should be included in the footprinting assays if the RNA ligand is a protein (see Note 5). 3. The protocols presented here have been used for the analysis of the Staphylococcus aureus regulatory RNAIII free or bound to target mRNAs (22, 34). This RNA is the intracellular effector of the quorum-sensing system and regulates in a coordinated way numerous virulence factors (35, 36). Typical experiments on end-labeled RNAIII are shown in Fig. 1 (see Note 6). The probing experiments were useful to delimitate the base pairings between the S. aureus RNAIII and the target mRNA (see Note 7). They also suggested that noncanonical base pairs could form within the duplex and thus would contribute to enhance the stability of the pairings. Interestingly, in several cases the structures of the interacting RNAs could impose topological constraints that limit the propagation of the intermolecular base pairings even if the extent of base pairing between the two RNAs is predicted to be rather long (34, 37). Furthermore, probing the RNA structure was helpful to design mutations at strategic positions of the RNAs to evaluate their effects on in vivo regulation. 3.2. Detection Methods

1. The identification of the cleavages can be done by two different methodologies depending on the length of the RNA molecule. The first method, which uses end-labeled RNA, is limited to molecules containing less than 200 nucleotides due to the gel resolution limitation (17). This method can only detect cleavages and is well appropriate for many of the bacterial sRNAs, which have a size below 300 nucleotides. However, for long RNAs such as S. aureus RNAIII (514 nucleotides long), the primer extension approach is preferable (38). The latter method detects stops of reverse transcription (RT) at the residue preceding a cleavage or a modification at a Watson– Crick position. Thus, this approach can be used for many enzymes and chemical probes and can be applied to RNA of any size. The length of the primer varies usually from 12 to 18 nucleotides. For long RNA, primers are selected every 200 nucleotides. Before probing the RNA structure, assays should be performed to define the best concentration of the RNA, the choice of the primer sequence, and the hybridization conditions in order to get an efficient primer extension.

3.3. RNA Preparation

RNA is typically transcribed in vitro with T7 RNA polymerase from a plasmid template carrying the T7 promoter fused to the gene of interest (39) (see Note 8). The RNA is then separated

3.3.1. RNA Transcription

435

441

445

457

480

496 490

510

c Lead(II) T1 T2 V1 strong V1 weak

* ** ** *

CCC U C A A A C A A A A U A 500 A U UC C C C A A U G C A 420 U A C A RNAIII A U C G A G C U A U A U C G 490 G C U A U G U A G C 460 470 U A C G 480 U A 510 U G U A G C 5'-- AA A U U C U U U U U --3' AU A AC U G U A A A A CA U U C C C U U A A U A A U A A G U A U G

430

410

420

SD 5’--UAUUUGAGGAUAGGUGUUAUUAAUUAUGAAAAAGAAUUUUAUU-- mRNA 3’--UAGAUAAAAACC--CCUACAAUAAUUAAUACUUUUUUU-AAAAUAA-- RNAIII

Fig. 1. Enzymatic probing and lead(II)-induced cleavages on 5¢ end-labeled Staphylococcus RNAIII (3¢ domain) free or bound to SA1000 mRNA. (a, b) Fractionation on 15% (a) and 12% (b) polyacrylamide-urea 8 M gel electrophoresis of 5¢ end-labeled RNA fragments generated by hydrolysis with RNase T1 (T1), RNase T2 (T2), RNase V1 (V1), and lead(II) (Pb). The experiments were carried out on free RNA (lane 3) or in the presence of increasing concentrations of the mRNA (100 nM, lane 4; 400 nM, lane 5). Incubation controls done on the free RNA (lane 1) or bound to the mRNA (lane 2). (T, L) RNase T1 and alkaline ladders, respectively. Changes induced by mRNA binding are shown by bars on the left side of the gels. The concentrations of enzymes used per assay are: RNase T1, 0.02 U; RNase T2, 0.05 U; RNase V1, 0.05 U; and Pb2+, 40 mM. (c) Secondary structure model of the 3¢ domain of RNAIII showing the RNase and lead(II) cleavages (legend in the inset). Squared nucleotides in RNAIII are complementary to the mRNA. Effect of mRNA binding: protections are given by black dot and enhanced cleavages are denoted by asterisk. The duplex formed between SA1000 mRNA and RNAIII is shown with the RNase V1 cuts that appear only upon duplex formation. RNAIII masks the ribosome-binding site blocking the access of the 30S ribosomal subunit at the initiation step (34). SD is for Shine and Dalgarno sequence. Noteworthy is that the 3¢ side of helix 13 of RNAIII became more accessible toward single-strand-specific probes when bound to the mRNA.

413

428

433

445 441

457

480

Pb T2 V1 1 2 3 4 3 4 3 4 TL

* ** ** * * * * * *

496

b

*

*

T1 T2 V1 1 2 3 3 4 5 3 4 5 3 4 5 TL

*

a

* *

*

** * *

222 Chevalier et al.

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

223

from shorter RNA fragments, DNA template, and the excess of NTP by using either gel filtration column (40), Mono Q column (41), or denaturing polyacrylamide-urea gel electrophoresis (PAGE) (39). If the RNA is not homogeneous after transcription due to the presence of abortive transcription or cleavage products, it is worthwhile to use PAGE instead chromatography for the purification procedure. However the elution process of the RNA from the gel may not be highly efficient for long RNAs (>500 nucleotides). 3.3.2. End Labeling of RNA

1. For 5¢ end labeling, the RNA should be dephosphorylated at its 5¢ end, and labeled using [g-32P]ATP and T4 PNK according to the Ambion protocol (http://www.ambion.com/ techlib/misc/RNA5_labeling.html). To avoid the 5¢ dephosphorylation of RNA, which is not highly efficient for structured RNAs, in vitro RNA transcription can be carried out in the presence of ApG. 2. The 3¢ end labeling is performed with [5¢-32P]pCp and T4 RNA ligase as previously described (42). A detailed protocol is also given by Ambion (http://www.ambion.com/techlib/ misc/RNA3_labeling.html). The labeled RNAs should be purified by denaturing PAGE on 8% polyacrylamide (0.5% bis-acrylamide)/8 M urea slab gels (for gel preparation see Subheading 3.5.2, step 1). Gel filtration column can also be used to separate the labeled RNA from the excess of unincorporated radioisotope. However it has to be checked that the RNA is homogeneous and is not cleaved after the labeling procedures.

3.3.3. Purification and Renaturation of RNA

1. After PAGE purification, labeled or cold RNAs are eluted from gel slices covered with the RNA elution buffer in the presence of 10% (v/v) phenol by gentle mixing at 4°C overnight. Add equal volume of phenol/chloroform/isoamyl alcohol mixture, mix the samples for 1 min, and centrifuge 1 min at high speed. Take carefully the aqueous phase containing the RNA, transfer the solution into a new sterile 1.5-mL micro tube, and add 2.5 volumes of cold ethanol for RNA precipitation. Incubate the mixture at −20°C overnight and collect RNA by centrifugation at 13,000 × g for 30 min. After two washing steps with 200 mL of 70% cold ethanol, the pellet is vacuum dried and dissolved in sterile water. Since the RNA is purified under denaturing conditions, it is worth spending effort to carry out a renaturation process before the probing experiments (see Note 9). One protocol is as follows: the RNA is preincubated 1 min at 90°C in sterile water, quickly cooled on ice for 1 min and incubated at 20°C or at 37°C in the appropriate buffer containing MgCl2 for 20 min.

224

Chevalier et al.

3.4. Enzymatic Hydrolysis and Lead(II)-Induced Cleavages

All reactions are conducted in a total volume of 10 mL. Appropriate dilutions of enzymes and of lead(II)-acetate are done in sterile water or in the commercial buffers just before use. The dilutions of enzymes are extemporaneously prepared. For footprinting assays, the mRNA–RNAIII complex is preformed before the enzymatic or Pb(II) reaction in the appropriate buffer optimal for binding. Incubation controls in the absence of the probes and in the presence (or in the absence) of the RNA ligand are always performed in order to detect nonspecific cleavages in RNA or RT pauses (see Note 10). In these controls, the enzyme or lead(II) are replaced by sterile water or buffer used for the dilution of enzymes.

3.4.1. Enzyme Hydrolysis

1. Labeled RNAIII or mRNA (50,000 cpm, 1 mL), or the cold RNA species (1 pmol, 1 mL) are denatured in 4 mL sterile water at 90°C for 1 min and then cooled on ice for 2 min. Centrifuge briefly and keep the tubes in ice. 2. 2 mL 5× buffer N1 is added and the samples are incubated at 20°C (or 37°C) for 15 min for renaturation. 3. mRNA–RNAIII duplex formation is performed at 25°C (or 37°C) for 10 min in the presence of increasing concentrations of RNAIII or mRNA (final concentrations 100, 200, 400 nM). The samples are mixed and centrifuged briefly. 4. 1 mL total tRNA is added to all samples. 5. Enzymatic hydrolysis is performed by addition of 1 mL RNase as follows: – RNase T1 (0.2 U/mL), 10 min at 20°C or 5 min at 37°C – RNase T2 (0.05 U/mL), 10 min at 20°C or 5 min at 37°C – RNase A (1 mg/mL), 10 min at 20°C or 5 min at 37°C – RNase V1 (0.05 U/mL), 5 min at 25°C or 2 min at 37°C 6. The RNases are added on one side of the Eppendorf tube and reactions are initiated after brief mixing and centrifugation (1 s) of the samples. 7. For nuclease S1, cleavage reaction is carried out in Buffer N2 in the presence of 50 U of enzyme for 10 min at 20°C (or 5 min at 37°C). 8. In order to define the best conditions for the hydrolysis, it is important to try initially three different concentrations of the enzymes: RNase T1 (0.1, 0.2, 0.5 U), RNase T2 (0.01, 0.05, 0.1 U), RNase A (0.25, 1, 2 mg), RNase V1 (0.01, 0.05, 0.1 U), and nuclease S1 (25, 50, 100 U).

3.4.2. Lead(II)-Induced Cleavages

1. Labeled mRNA or RNAIII (1 mL, 50,000 cpm), or the cold RNA species (1 pmol, 1 mL) are denatured (see Subheading 3.4.1, step 1) and renatured in the presence of 2 mL of buffer N3 (5×) at 20°C (or 37°C) for 15 min. 2. mRNA–RNAIII duplex formation is performed at 25°C (or 37°C) for 10 min in the presence of increasing concentrations

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

225

of RNAIII or mRNA (final concentrations: 100, 200, 400 nM). Mix and centrifuge briefly the samples. 3. 1 mL of total tRNA is added to all samples. 4. Hydrolysis is initiated with 2.5 mL of different concentrations of lead(II)-acetate from 12, 40, 80 to 120 mM for 10 min at 20°C or 5 min at 37°C. Mix and centrifuge briefly the samples (see Note 11). The best results in our hands were with 40 mM. 3.4.3. Reaction Termination 3.4.3.1. Enzymatic Hydrolysis

1. Enzymatic hydrolysis are arrested by adding 20 mL of the precipitation/inactivation buffer. 2. After vigorous mixing, the samples are left in a dry-ice-ethanol bath for 10 min and centrifuged at 13,000 × g at 4°C for 15 min. 3. The supernatants are discarded (if using end-labeled RNA, check that no radioactivity is present) and the pellets are washed twice with 200 mL of 70% cold ethanol. After a short centrifugation at 13,000 × g for 5 min at 4°C, the supernatants are discarded and the pellets are vacuum dried (not more than 5 min). 4. End-labeled RNA fragments are dissolved in 6 mL of RNAloading buffer, whereas cold RNA fragments are dissolved in 4 mL of sterile water. 5. If a protein is used as a ligand in footprinting assays, it is advised to make a phenol extraction. In that case, add to all samples, 40 mL of 0.3 M Na-acetate of pH 6.0 and 50 mL of phenol/ chloroform/isoamyl alcohol mixture. Mix the samples for 1 min and centrifuge 1 min at high speed. Take carefully the aqueous phase containing the RNA, transfer the solution into a new sterile 1.5-mL micro tube, and add 2.5 volumes of cold ethanol (~150 mL) for RNA precipitation. The samples are then precipitated and treated as in Subheading 3.4.3.1, steps 2–4.

3.4.3.2. Lead(II)-Induced cleavages

3.5. Fractionation of End-Labeled RNA Fragments 3.5.1. Ladders for Cleavage Assignments

1. The reactions are stopped by adding 5 mL of 0.1 M EDTA. 2. Add 50 mL of 0.3 M Na-acetate of pH 6.0 and 150 mL of cold ethanol to all samples. After a vigorous mix, the samples are transferred in a dry ice-ethanol bath for 10 min. The samples are then precipitated and treated as in Subheading 3.4.3.1, steps 2–4. 1. RNase T1 ladder: labeled mRNA (25,000 cpm) is preincubated at 50°C for 5 min in 5 mL of the Buffer DT1 containing 1 mg total tRNA. Reaction is then performed at 50°C for 10 min in the presence of 1 mL of RNase T1 (0.5 U). 2. Alkaline ladder: labeled mRNA (100,000 cpm) is incubated at 90°C for 3 min in the presence of total tRNA (2 mg) in 5 mL of the Ladder Buffer. 3. Both ladders can be prepared for several experiments and stored at −20°C. Do not heat the samples before loading on PAGE.

226

Chevalier et al.

3.5.2. Separation of End-Labeled RNA Fragments by PAGE

1. The end-labeled RNA fragments are separated by electrophoresis on 12 or 15% polyacrylamide-(0.5% bis) – 8 M urea slab gels (0.35 mm × 30 cm × 40 cm) in 1× TBE. To prepare 100 mL of 15% gel, mix 60 mL 25% polyacrylamide – 8 M urea solution, 10 mL 10× TBE buffer, 30 mL 8 M urea, 75 mL TEMED, and 750 mL 10% APS. The gel solution is poured slowly between two glass plates that are separated by one spacer on each side and placed horizontally on the bench. After polymerization (~30 min), the comb is removed and the wells are washed carefully. 2. Prerun the gel at 75 W for 30 min using 1× TBE as running buffer. 3. The samples are heated (except the RNase T1 and alkaline ladders) for 3 min at 90°C, centrifuged briefly and 3-mL aliquots are loaded per well. Before loading, be aware that each sample contains the same amount of radioactivity (except for the ladder that should have twice more radioactivity). Load in the following order: incubation controls, reactions on free RNA and in the presence of increasing concentrations of ligand, RNase T1 ladder, and alkaline ladder. 4. Run PAGE at 75 W to heat the gel and to avoid band compression. The migration time must be adapted to the length of the RNA, knowing that on 15% polyacrylamide gel, xylene cyanol migrates as 39 nucleotide- and bromophenol blue as 9 nucleotide-long RNA. Short migration on 15% gel is convenient to fractionate small-size fragments (1–50 nucleotide-long RNA fragments). For a 250 nucleotide-long RNA, a longer migration on 12% PAGE is necessary to assign the cleavages on the whole RNA molecule (see Fig. 1). 5. At the end of the run, remove carefully the upper glass, fix the 12% gel for 5–30 min in Fixing gel solution, transfer to Whatman 3-MM paper, and dry for 30 min at 80°C. The 15% gel can be transferred without drying on an old autoradiography film and wrapped with a plastic film. Overnight exposure is done at −80°C using an intensifying screen. 6. Several enzymatic properties have to be taken into account when reading the gels (see Notes 12 and 13). 7. Several technical problems may be revealed during PAGE (see Notes 14–17).

3.6. Detection of Cleavages by Primer Extension Analysis 3.6.1.5¢ End Labeling of Oligodeoxyribonucleotide Primer

1. The following reagents are mixed: 10 mL of 5 mM oligodeoxyribonucleotide primer, 5 mL of [g-32P]ATP, 2 mL of 10× T4 PNK buffer, 2.5 mL of sterile water, 0.5 mL T4 polynucleotide kinase. Incubate at 37°C for 1 h. 2. To the sample, add 30 mL sterile water, load on a Micro Biospin 6 chromatography column (Biorad), and centrifuge at 1,000 × g for 5 min at 20°C. The volume is then adjusted with

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

227

sterile water to get 100,000 cpm/mL of end-labeled DNA primer. This rapid purification protocol is sufficient to remove the excess of radioactive ATP. 3.6.2. Hybridization

1. To the 4 mL of the cleaved mRNA (1 pmol), add 1 mL of 5¢ end-labeled DNA primer (around 100,000 cpm). 2. The samples are heated 1 min at 90°C and quickly cooled on ice after a brief centrifugation. 3. 1 mL 5× RTB buffer is added and the samples are incubated for 15 min at 20°C.

3.6.3. Primer Extension

1. The reaction is done in 15 mL. 2. To the hybridization mix, add 2 mL 5× RTB, 2 mL dNTP mix (2.5 mM of each dNTP), 4 mL sterile water, 1 mL RT (2 U/mL, diluted freshly in the commercial buffer). The samples are incubated for 30 min at 37°C. 3. To all samples, add 50 mL of 0.3 M Na-acetate of pH 6.0 and 200 mL of cold ethanol. The samples are then precipitated and treated as in Subheading 3.4.3.1, steps 2–3. The endlabeled DNA fragments are dissolved in 6 mL of DNA-loading buffer. All samples are adjusted to the same amount of radioactivity per 1 mL. 4. To improve the quality of the gels, the RNA template may be hydrolyzed by alkaline treatment. Just after primer extension, add 20 mL of the RNA hydrolysis buffer and 3.5 mL of 3 M KOH. The samples are heated at 90°C for 3 min and at 37°C for at least 1 h. To all samples, add 6 mL 3 M acetic acid, 100 mL 0.3 M Na-acetate of pH 6.0, and 300 mL of cold ethanol. After precipitation (see Subheading 3.4.3.1, steps 2–3), the pellets are washed twice with 70% ethanol, vacuum dried, and dissolved in 6 mL of DNA-loading buffer.

3.6.4. Gel Fractionation of Labeled cDNA Fragments

1. The cleavage positions are identified by running in parallel a sequencing reaction (see Chapter “Structural Probing of RNA Thermosensors”). The elongation step is performed as described in Subheading 3.6.3 except that in the presence of one of the dideoxyribonucleotide ddXTP (2.5 mM), the corresponding deoxyribonucleotide dXTP (25 mM) and the three other deoxyribonucleotides (100 mM) are added. 2. All samples are heated at 90°C for 3 min, and centrifuged briefly. 3. 3-mL Aliquots are loaded per well on 8% polyacrylamide-(0.4% bis)/8 M urea slab gels in 1× TBE. The migration conditions must be adapted to the size of the fragments to be analyzed, knowing that on 8% polyacrylamide gel, xylene cyanol migrates as 81 nucleotide- and bromophenol blue as 19 nucleotide-long RNA. After migration, the gels are dried, and exposed with an X-ray film and intensifying screen overnight at −80°C.

228

Chevalier et al.

4. Technical problems that are revealed by the autoradiography may occur during the handling process (see Notes 16–20).

4. Notes 1. Safety rules: for manipulating radioactivity, work behind a plexiglass screen, and wear glasses and gloves. Perform phenol extraction under a fume hood. 2. For lead(II)-induced cleavages, buffers with chloride ions should be avoided since PbCl2 may precipitate in solution. 3. To keep high resolution of the gels, acrylamide, urea solutions, and in particular ammonium persulfate should be prepared freshly. 4. RNase T1 and RNase A hydrolysis can be conducted under a variety of experimental conditions. It is essential however to adapt the enzymatic hydrolysis for each condition in order to have less than one cut or modification per molecule, i.e., more than 80% of the RNA should not be modified or cleaved. 5. If no effect of ligand binding is observed in footprinting experiments, it is essential to define the experimental conditions (buffer, temperature, ions) that are required for efficient binding by other methodologies such as bandshift analysis. The footprinting experiments should also be performed in the presence of increasing concentrations of the ligand (see Fig. 1). Such experiments might reveal different ligand-binding sites on the RNA molecule. 6. Each experiment should be repeated at least twice, and only the reproducible cleavages will be considered. As mentioned previously, the elaboration of a secondary structure RNA model requires data from enzymes of complementary specificities. Only the combination will help to define helical and loop regions. 7. Results should be interpreted with care because protection does not necessarily result from a direct shielding effect, but could be due to a steric hindrance effect (particularly observed with the bulky RNases) or to a conformational change of the RNA. Contrarily, enhanced cleavages result from RNA conformational changes (Fig. 1). 8. In vitro RNA transcripts generated by T7 RNA polymerase suffer from a considerable degree of heterogeneity at 3¢ end and sometimes from heterogeneity at the 5¢ end. This can be revealed by doubled cleavages using end-labeled RNA. The heterogeneity poses a significant problem when analyzing the cleavage sites using long end-labeled RNAs. Heterogeneity can

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

229

be eliminated by addition of self-cleaving ribozyme sequences at the 5¢- and/or 3’¢-end of the target RNA sequence (43). PAGE is typically the method of choice for separation of ribozymes from the RNA of interest. For long RNA molecules, electro-elution might help to increase the elution efficiency. 9. During the purification, RNA can be partially denatured; therefore, it is essential to design “renaturation” protocols in order to obtain “conformationally homogeneous” RNA population. It is also necessary to test whether this conformation is biologically relevant (enzymatic activity for ribozyme, efficient ligand binding). Alternative RNA conformations may coexist, and can be revealed by the simultaneous presence of single-stranded and double-stranded specific cleavages. By varying the concentration of MgCl2, one of the two conformers might be stabilized. 10. Appropriate incubation controls are essential to identify cleavages that are induced during the incubation treatments, and the pauses of reverse transcriptase that are due to stable secondary structures or cuts. Nucleotides for which strong bands are visible in the control lanes are not considered for interpretation. Incubation control in the presence of the ligand has to be done since the ligand might be contaminated with traces of RNase during its purification. If too many bands are observed in the incubation controls of the end-labeled RNA, repurify the RNA and prepare new sterile buffers. If too many RT pauses are observed in the incubation controls, it can be due to RNase contamination, strong secondary structure of RNA (the extension can be done at 42°C), or primer location. For this experiment, AMV (Avian Myeloblastosis Virus) RT should be used rather than MMLV (Moloney Murine leukemia Virus) RT, the latter being much more sensitive to RNA secondary structure. 11. Since lead(II) competes with Mg2+ for RNA binding, the efficiency of cleavages will depend on the Pb2+/Mg2+ ratio. 12. The RNase cleavages in the RNA can induce conformational rearrangements of the cleaved RNA that can potentially provide new targets for secondary cleavages. Thus, these secondary cleavages do not reflect the native structure of the RNA. Usually these cleavages are weak, are not reproducibly found in all experiments, and occur when the RNA digestion is too strong. They can be distinguished from the primary cuts by comparing the hydrolysis patterns obtained from the 5¢ or 3¢ end-labeled RNA. 13. RNase V1 and nuclease S1 hydrolysis generate RNA fragments, which end up with 3¢ OH and 5¢-P groups in contrast to alkali and most of the RNases (Table 1). Therefore, 5¢ end-labeled

230

Chevalier et al.

fragments generated by alkali will migrate faster than the RNase V1/nuclease S1 fragments, and conversely the 3¢-endlabeled RNA generated by alkali will migrate slower than the RNase V1/nuclease S1 fragments. This difference is only observed for the shortest RNA fragments (see Fig. 1). 14. If no full-length RNA is observed, the RNA hydrolysis is too strong. Time of the hydrolysis and/or enzyme concentrations have to be reduced. 15. Compression of bands due to stable secondary structure (in general rich in G–C base pairs) can be observed using endlabeled RNA. Heat the samples before loading on the gel and the gels should be warm before sample loading and during the migration. 16. If end-labeled RNA aggregates in the gel pockets and only fragments of small sizes can be visualized on the gel, the RNA pellet was not correctly dried after ethanol precipitation. Do not interpret the data. 17. Samples may not migrate correctly during electrophoresis due to the presence of salt. Add several washing steps with 80% ethanol at the end of the procedure. 18. Absence of radioactive signal after primer extension could mean that the modified RNA did not efficiently precipitate, or the hybridization conditions are not optimized. Since the modified RNA is not labeled, particular caution should be taken to prevent the loss of the pellet. 19. Reverse transcriptase stops at the nucleotide preceding the cleaved nucleotide. Thus, the resulting cDNA is one nucleotide shorter than the cDNA corresponding to the sequencing lane. 20. Too many reverse transcriptase stops in the control lanes may be due to several reasons: degradation of the RNA template, pauses of the enzyme due to stable RNA secondary structure (increase the temperature of elongation, change the primer), the conditions of elongation are not sufficiently optimized (adjust the enzyme or dNTP concentration).

Acknowledgments This work was supported by the Centre National de la Recherche Scientifique (UPR 9002 CNRS), the University Louis Pasteur of Strasbourg, the Ministère de la Recherche (ANR05-MIIME, ANR07-BLANC), and the European Community (FOSRAK, EC005120; BacRNA EC018618).

Probing mRNA Structure and sRNA–mRNA Interactions in Bacteria Using Enzymes

231

References 1. Kaczanowska, M., and Ryden-Aulin, M. (2007). Ribosome biogenesis and the translation process in Escherichia coli. Microbiol. Mol. Biol. Rev. 71, 477–494. 2. Nguyenle, T., Laurberg, M., Brenowitz, M., and Noller, H. F. (2006). Following the dynamics of changes in solvent accessibility of 16 S and 23 S rRNA during ribosomal subunit association using synchrotron-generated hydroxyl radicals. J. Mol. Biol. 359, 1235–1248. 3. Narberhaus, F., Waldminghaus, T., and Chowdhury, S. (2006). RNA thermometers. FEMS Microbiol. Rev. 30, 3–16. 4. Serganov, A., and Patel, D. J. (2007). Ribozymes, riboswitches and beyond: regulation of gene expression without proteins. Nat. Rev. Genet. 8, 776–790. 5. Barrick, J. E., and Breaker, R. R. (2007). The distributions, mechanisms, and structures of metabolite-binding riboswitches. Genome Biol. 8, R239. 6. Gottesman, S., McCullen, C. A., Guillier, M., Vanderpool, C. K., Majdalani, N., Benhammou, J., Thompson, K. M., FitzGerald, P. C., Sowa, N. A., and FitzGerald, D. J. (2006). Small RNA regulators and the bacterial response to stress. Cold Spring Harb. Symp. Quant. Biol. 71, 1–11. 7. Storz, G., Altuvia, S., and Wassarman, K. M. (2005). An abundance of RNA regulators. Annu. Rev. Biochem. 74, 199–217. 8. Wagner, E. G. H., Altuvia, S., and Romby, P. (2002). Antisense RNAs in bacteria and their genetic elements. Adv. Genet. 46, 361–398. 9. Brantl, S. (2007). Regulatory mechanisms employed by cis-encoded antisense RNAs. Curr. Opin. Microbiol. 10, 102–109. 10. Chen, Y., and Varani, G. (2005). Protein families and RNA recognition. FEBS J. 272, 2088–2097. 11. Leontis, N. B., Lescoute, A., and Westhof, E. (2006). The building blocks and motifs of RNA architecture. Curr. Opin. Struct. Biol. 16, 279–287. 12. Jossinet, F., Ludwig, T. E., and Westhof, E. (2007). RNA structure: bioinformatic analysis. Curr. Opin. Microbiol. 10, 279–285. 13. Zuker, M. (2003). Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 31, 3406–3415. 14. Do, C. B., Woods, D. A., and Batzoglou, S. (2006). CONTRAfold: RNA secondary structure prediction without physics-based models. Bioinformatics 22, e90–e98. 15. Xayaphoummine, A., Bucher, T., and Isambert, H. (2005). Kinefold web server for RNA/DNA folding path and structure prediction including pseudoknots and knots. Nucleic Acids Res. 33, W605–W610.

16. Gultyaev, A. P., van Batenburg, F. H., and Pleij, C. W. (1995). The computer simulation of RNA folding pathways using a genetic algorithm. J. Mol. Biol. 250, 37–51. 17. Donis-Keller, H., Maxam, A. M., and Gilbert, W. (1977). Mapping adenines, guanines, and pyrimidines in RNA. Nucleic Acids Res. 4, 2527–2538. 18. Ehresmann, C., Baudin, F., Mougel, M., Romby, P., Ebel, J. P., and Ehresmann, B. (1987). Probing the structure of RNAs in solution. Nucleic Acids Res. 15, 9109–9128. 19. Lockard, R. E., and Kumar, A. (1981). Mapping tRNA structure in solution using double-strandspecific ribonuclease V1 from cobra venom. Nucleic Acids Res. 9, 5125–5140. 20. Favorova, O. O., Fasiolo, F., Keith, G., Vassilenko, S. K., and Ebel, J. P. (1981). Partial digestion of tRNA – aminoacyl-tRNA synthetase complexes with cobra venom ribonuclease. Biochemistry 20, 1006–1011. 21. Kolb, F. A., Malmgren, C., Westhof, E., Ehresmann, C., Ehresmann, B., Wagner, E. G. H., and Romby, P. (2000). An unusual structure formed by antisense-target RNA binding involves an extended kissing complex with a four-way junction and a side-by-side helical alignment. RNA 6, 311–324. 22. Huntzinger, E., Boisset, S., Saveanu, C., Benito, Y., Geissmann, T., Namane, A., Lina, G., Etienne, J., Ehresmann, B., Ehresmann, C., Jacquier, A., Vandenesch, F., and Romby, P. (2005). Staphylococcus aureus RNAIII and the endoribonuclease III coordinately regulate spa gene expression. EMBO J. 24, 824–835. 23. Darfeuille, F., Unoson, C., Vogel, J., and Wagner, E. G. H. (2007) An antisense RNA inhibits translation by competing with standby ribosomes. Mol. Cell 26, 381–392. 24. Sharma, C. M., Darfeuille, F., Plantinga, T. H., and Vogel, J. (2007). A small RNA regulates multiple ABC transporter mRNAs by targeting C/A-rich elements inside and upstream of ribosome-binding sites. Genes Dev. 21, 2804–2817. 25. Lindell, M., Romby, P., and Wagner, E. G. H. (2002). Lead(II) as a probe for investigating RNA structure in vivo. RNA 8, 534–541. 26. Ivanova, N., Lindell, M., Pavlov, M., Holmberg Schiavone, L., Wagner, E. G. H., and Ehrenberg, M. (2007). Structure probing of tmRNA in distinct stages of trans-translation. RNA 13, 713–722. 27. Huntzinger, E., Possedko, M., Winter, F., Moine, H., Ehresmann, C., and Romby, P. (2005). Probing RNA structures with enzymes and chemicals in vitro and in vivo, in Handbook of RNA Chemistry (Hartmann, R. K.,

232

Chevalier et al.

Bindereif, A., Schön, A., and Westhof, E., eds), Wiley-VCH, Weinheim, pp. 151–171. 28. Marchand, V., Mougin, A., Méreau, A., and Branlant, C. (2005). Study of RNA–protein interactions and RNA structure in ribonucleoprotein particles, in Handbook of RNA Chemistry (Hartmann, R. K., Bindereif, A., Schön, A., and Westhof, E., eds), Wiley-VCH, Weinheim, pp. 172–228. 29. Waldminghaus, T., Heidrich, N., Brantl, S., and Narberhaus, F. (2007). FourU: a novel type of RNA thermometer in Salmonella. Mol. Microbiol. 65, 413–424. 30. Coppins, R. L., Hall, K. B., and Groisman, E. A. (2007). The intricate world of riboswitches. Curr. Opin. Microbiol. 10, 176–181. 31. Fabbretti, A., Pon, C. L., Hennelly, S. P., Hill, W. E., Lodmell, J. S., and Gualerzi, C. O. (2007). The real-time path of translation factor IF3 onto and off the ribosome. Mol. Cell 25, 285–296. 32. Shcherbakova, I., Mitra, S., Beer, R. H., and Brenowitz, M. (2006). Fast Fenton footprinting: a laboratory-based method for the timeresolved analysis of DNA, RNA and proteins. Nucleic Acids Res. 34, e48. 33. Fabbretti, A., Milon, P., Giuliodori, A. M., Gualerzi, C. O., and Pon, C. L. (2007). Realtime dynamics of ribosome–ligand interaction by time-resolved chemical probing methods. Methods Enzymol. 430, 45–58. 34. Boisset, S., Geissmann, T., Huntzinger, E., Fechter, P., Bendridi, N., Possedko, M., Chevalier, C., Helfer, A. C., Benito, Y., Jacquier, A., Gaspin, C., Vandenesch, F., and Romby, P. (2007). Staphylococcus aureus RNAIII coordinately represses the synthesis of virulence factors and the transcription regulator Rot by an antisense mechanism. Genes Dev. 21, 1353–1366. 35. Novick, R. P. (2003). Autoinduction and signal transduction in the regulation of sta-

phylococcal virulence. Mol. Microbiol. 48, 1429–1449. 36. Toledo-Arana, A., Repoila, F., and Cossart, P. (2007). Small noncoding RNAs controlling pathogenesis. Curr. Opin. Microbiol. 10, 182–188. 37. Kolb, F. A., Engdahl, H. M., Slagter-Jäger, J. G., Ehresmann, B., Ehresmann, C., Westhof, E., Wagner, E. G. H., and Romby, P. (2000). Progression of a loop–loop complex to a four-way junction is crucial for the activity of a regulatory antisense RNA. EMBO J. 19, 5905–5915. 38. Qu, H. L., Michot, B., and Bachellerie, J. P. (1983). Improved methods for structure probing in large RNAs: a rapid ‘heterologous’ sequencing approach is coupled to the direct mapping of nuclease accessible sites. Application to the 5¢ terminal domain of eukaryotic 28S rRNA. Nucleic Acids Res. 11, 5903–5920. 39. Milligan, J. F., and Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods Enzymol. 180, 51–62. 40. Romaniuk, P. J., de Stevenson, I. L., and Wong, H. H. (1987). Defining the binding site of Xenopus transcription factor IIIA on 5S RNA using truncated and chimeric 5S RNA molecules. Nucleic Acids Res. 15, 2737–2755. 41. Jahn, M. J., Jahn, D., Kumar, A. M., and Söll, D. (1991). Mono Q chromatography permits recycling of DNA template and purification of RNA transcripts after T7 RNA polymerase reaction. Nucleic Acids Res. 19, 2786. 42. England, T. E., Bruce, A. G., and Uhlenbeck, O. C. (1980). Specific labeling of 3¢ termini of RNA with T4 RNA ligase. Methods Enzymol. 65, 65–74. 43. Walker, S. C., Avis, J. M., and Conn, G. L. (2003). General plasmids for producing RNA in vitro transcripts with homogeneous ends. Nucleic Acids Res. 31, e82.

Chapter 17 Structural Probing of RNA Thermosensors Claude Chiaruttini, Frédéric Allemand, and Mathias Springer Summary Chemical probing of RNA structure has become one of the most popular approaches to map the conformation of RNA molecules of various sizes under well-defined experimental conditions. The method monitors the sensitivity of each nucleotide to various chemicals, which reflects its hydrogenbonding environment within the RNA molecule. The goal of this chapter is to provide the reader with an experimental guide to mapping the secondary structure of RNA thermosensors in vitro with the most suitable chemical probes. Key words: RNA, Thermosensor, Secondary structure, RNA structure probing, Modifying probes, Modification mapping

1. Introduction 1.1. Definition and Examples of Thermosensors

One of the most fascinating recent discoveries in the field of bacterial gene expression is that mRNAs can function as direct sensors of both the physical and the metabolic environment of the cell in the absence of trans-acting factors such as proteins or small regulatory RNAs. Metabolic sensors are called riboswitches, the main subject of this book, and the temperature sensors are called either RNA thermosensors or RNA thermometers (1). These latter devices generally respond to a rise in temperature by an increase in expression attributed to a conformational change in the translation initiation region of a given mRNA that allows the ribosome to bind to its recognition site. This is the case for bacterial heat-shock genes, such as the Escherichia coli rpoH gene, encoding the heat-shock sigma factor (2), for many genes controlled by the ROSE element acting in cis (1), and possibly for the

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_17 © Humana Press, a part of Springer Science + Business Media, LLC 2009

233

234

Chiaruttini, Allemand, and Springer

dnaK gene from several bacterial species (see ref. 1). The expression of genes regulating virulence is also often controlled by thermosensors, such as for the Listeria monocytogenes prfA (3) or Yersinia pestis lcrF genes (4). In some cases, a decrease in temperature has been proposed to cause an increase of expression also via conformational changes, such as for the E. coli gene cspA encoding the cold-shock protein A (5) and the bacteriophage lambda cIII gene, involved in the control of the lysis-lysogeny decision (6). The secondary structures associated with the temperature sensing generally straddle the translation initiation signals but can also penetrate into the coding regions, as is the case for E. coli rpoH. These structures generally consist of several stemloop structures, with the exception of L. monocytogenes prfA thermosensor that is made from a single hairpin. Understanding how RNA thermosensors work requires a detailed knowledge of their secondary structure. The most complete and precise way to get this structural information is obviously to derive it from NMR or crystallographic studies. However, the use of these techniques demands a large amount of material and is often excessively time consuming. In vitro probing approach allows the precise and relatively rapid characterisation of RNA secondary structures. However, this approach should not be used alone, but coupled with other methods that are briefly reviewed in this chapter before we focus on in vitro probing techniques. 1.2. In Vivo Determination of RNA Secondary Structure 1.2.1. In Vivo Chemical Probing

1.2.2. Mutational Analysis

One of the main concerns expressed when RNA structure is probed in vitro is how the RNA is folded in a more complex environment such as in living cells. In this respect, structurespecific chemical probes can be used to map RNA structure in vivo under different cell growth conditions. However, the number of suitable probes is severely restricted by the fact they must obey certain size, structure, and charge constraints to be capable to get through the cell wall and membrane. So far, the most commonly used reagents for in vivo RNA secondary structure probing studies are the two base-modifying reagents, dimethyl sulphate (DMS) (7, 8) and β-ethoxy-α-ketobutyraldehyde (kethoxal) (9), and the RNA cleavage-inducing reagent, lead (II) (10). Nevertheless, despite the limitations that are inherent to the use of in vivo probing, the comparison of in vivo and in vitro probing data provides complementary data for determining functional RNA structure. As discussed in Subheading 1.4.1, the probing approach alone does not allow identification of the interacting nucleotides and the data that are derived from these studies have to be considered as constraints to “direct” computer RNA secondary structure folding programs. Genetic analysis is also a very powerful method to detect and study RNA structural elements. Classical genetic selections can be

Structural Probing of RNA Thermosensors

235

used if the lack of expression of the thermosensor under study is associated with a phenotype. In such cases, mutations can readily be selected by plating cells at the restrictive temperature (either high or low temperature depending on the temperature response of the specific thermosensor). If no selectable phenotype is available, mutants can be selected or screened for using genetic fusions in which the RNA fragment under study is fused to a reporter gene. The classical tools for such studies are fusions to lacZ, the gene encoding β-galactosidase, but other reporter genes are also available (chloramphenicol acetyl-transferase encoding cat gene, kanamycin-resistance encoding kan gene, GFP encoding gene, etc.). If, in the case of a heat-inducible thermosensor, the expression of a translational fusion to lacZ confers a Lac− phenotype at low temperature, cis-acting mutations can be selected or screened for that increase expression at this restrictive temperature. Such mutations may allow identification of nucleotides responsible for the lack of translation at low temperatures and revertant mutations can be searched for with the aim of identifying possible complementary interacting partners. If the biological system does not allow for either direct or fusion-mediated selections or even screening, the classical approach consists of using secondary structure folding programs (see Subheading 1.3.2) to draw a secondary structure around the translation initiation site of the gene under study. The validity of this structure can then be tested experimentally by introducing mutations that destabilise or stabilise the predicted structure and analysing their effect on the expression of the gene at different temperatures. 1.3. In Silico Determination of RNA Secondary Structure 1.3.1. Comparative Sequence Analysis

1.3.2. Secondary Structure Folding Programs

This is a very powerful method to determine RNA secondary structure that is based on the collection and comparison of similar sequences from related organisms to extract information about the secondary and tertiary structures of RNAs. This method was originally used to characterise the cloverleaf structure of tRNA and has been extensively used for ribosomal RNAs, the RNA moiety of RNase P, and group I and II introns (11). Its applicability is based on phylogenetic analysis, and depends on the number of sequences known and, in our specific case, the extent with which a given thermosensor is conserved in evolution. Although human expertise still remains essential for comparative sequence analysis, automated processes are largely used to look for covariation with possible implication for structure (11–13). The secondary structure of RNA can also be directly predicted with computer programs based on either the thermodynamics and/or the kinetics of RNA folding. The most popular programs Mfold (http://frontend.bioinfo.rpi.edu /applications/ mfold/cgi-bin/rna-form1.cgi) and RNAfold (http://rna.tbi. univie.ac.at/cgi-bin/RNAfold.cgi) both use the same algorithm

236

Chiaruttini, Allemand, and Springer

for computing minimal free-energy structures assuming nearest neighbour models. Some secondary structure algorithms, such as CONTRAfold (http://contra.stanford.edu/contrafold/server. html), use statistical sampling of known RNA structures to create a model that can be used to predict secondary structure. Other programs, such as RNAalifold (http://rna.tbi.univie.ac.at/cgibin/alifold.cgi), are based on both thermodynamics and comparative sequence analysis using algorithms similar to that of Mfold or RNAfold to compute a consensus sequence from RNA alignments. However, an RNA structure might assume different conformations during transcription or as a result of its function. Programs such as KINEfold (http://kinefold.curie.fr/cgi-bin/ form.pl) implement stochastic simulations of folding kinetics. Others, such as STAR (14), include results of evolving intermediate structural states. 1.4. Chemical Probing In Vitro 1.4.1. Principle and Detection

RNA structure probing monitors the sensitivity of each nucleotide to chemicals, which reflects its electrostatic and hydrogenbonding environment within the RNA molecule in solution. Chemical RNA-modifying reagents are ideally suited for the study of double-stranded RNA structures like thermosensors mainly, because they are much smaller than ribonucleases and can reach almost every part of the RNA molecule. Applicable reagents react with a specific moiety present in the RNA. If this moiety is directly involved in an interaction such as hydrogen bond formation, the sensitivity to the reagent is strongly reduced or even abolished. The RNA is first subjected to limited chemical modification using reagents that specifically probe unpaired nucleotides. It is important that modifications are introduced at a statistically low level (less than one modification per molecule). Controls are incubated in parallel under the same conditions but in the absence of the chemical. Modified nucleotides are then mapped on RNA using the primer extension technique, which allows the detection of prematurely terminated cDNA chains at modified nucleotides. More precisely, the RNA is hybridized with a radiolabelled oligodeoxyribonucleotide complementary to a chosen sequence in the RNA, which serves as primer for cDNA extension by reverse transcriptase, thus allowing the probing of internal regions of large RNA molecules. The cDNA extension reaction proceeds from the 3′-end of the primer in the 3′-direction, in the presence of the four deoxyribonucleotides (dNTPs). When using modified RNA as template, prematurely terminated cDNAs are produced instead of full-length cDNA chains, which are normally synthesized using unmodified RNA. These cDNA chain abortions are the result of chemical modifications at one of the Watson–Crick positions, which blocks the progress of reverse transcription at the nucleotide preceding the modified nucleotide. Finally, the resulting cDNA chains are sized at primary

Structural Probing of RNA Thermosensors

237

structure resolution by electrophoresis on a sequencing gel in parallel to sequencing reactions performed on unmodified RNA using the same primer. At this point, it is important to stress that the data provided by the probing approach are not sufficient per se to model an RNA secondary structure since, even though they do provide unambiguous information on the hydrogen-bonding status of each nucleotide, they do not allow identification of the pairing partners. Therefore, even if the probing approach is a true first-intention method to determine RNA structure, it must be coupled with the investigation approaches that have been described earlier, in particular in silico tools like computer RNA folding programs. 1.5. Validation of an RNA Secondary Structure Model

The biological relevance of the resulting RNA secondary structure model needs to be addressed. This requires in vivo validation of the model by first analysing the effect of single mutations introduced by site-directed mutagenesis in both partners of a predicted base pair and, second, the effect of compensatory base changes which restore the Watson–Crick interaction. Note that mutagenesis analysis should be associated with the probing approach to analyse the effect of mutations on the stability of the RNA structure.

2. Materials It is anticipated that the reader is familiar with standard biochemical and molecular biology practice, such as gel electrophoresis, chromatography, and handling of radioactive materials, including protection against radioactive emission. Importantly, in order to avoid RNase contamination, it is strongly recommended that all aqueous solutions and 1.5-mL polypropylene tubes are sterilized by autoclaving. 2.1. RNA Modification Reactions

1. The choice of the chemical probes is dictated by the conditions required for RNA secondary structure probing, which are to be as mild as possible, in order to preserve the structural integrity of the RNA and to obtain a low level of RNA modification: pH at or close to physiological value, moderate temperature, and short incubation times. In this respect, DMS, 1-cyclohexyl-3(2-morpholinoethyl)-carbodiimide metho-ptoluene sulphonate (CMCT) and kethoxal, the specificities of which will be briefly described hereafter, are the most adapted chemical probes as they will only react with certain nitrogen atoms in the base moiety of nucleotides when these are not involved in hydrogen bond formation.

238

Chiaruttini, Allemand, and Springer

2. Specificities of the probes used. At neutral pH, DMS methylates RNA at the N7 position of guanine and at the N1 and N3 Watson–Crick positions of adenine and cytosine, respectively. At pH 8.0, CMCT alkylates RNA at the N3 and N1 Watson– Crick positions of uridine and guanine, respectively. At neutral pH, kethoxal reacts with the N1 and N2 Watson–Crick positions of guanine, giving a cyclic adduct between these two positions and its two carbonyls, which is stabilized by borate ions (see Note 1). Methylation at N7-G by DMS cannot be detected by the cDNA extension technique. Reaction of DMS with N1-A is stronger than with N3-C and reaction of CMCT with N3-U is stronger than with N1-G. Certain guanines and uridines are occasionally stabilized in an enol-tautomer form due to a specific local environment and are therefore reactive to DMS at their N1 or N3 positions. In order to avoid overinterpretation of the data, it should be reminded that reactivity towards chemical probes is influenced significantly by the electrostatic environment of the nucleotides. Strictly speaking, it is therefore incorrect to assume that chemical probing maps stereochemical accessibilities. Chemical probing is a method of choice to unravel the existence of specific structural features and noncanonical base pairs since most of the time they involve Watson–Crick positions and the N7 of adenine. In particular, chemical probing easily detects noncanonical base pairs like sheared A-G base pair or reverse Hoogsteen A-U base pair, which are widespread in RNA molecules. 3. 5× DMS buffer: 250 mM sodium cacodylate of pH 7.5, 50 mM magnesium acetate, 300 mM ammonium chloride. 4. 5× CMCT buffer: 250 mM sodium borate of pH 8.0, 50 mM magnesium acetate, 300 mM ammonium chloride. CMCT buffer: 5× CMCT buffer diluted 5 times with water (see Note 2). 5. Chemical probes: DMS (Aldrich) is diluted 1:30 in 100% ethanol; CMCT (Aldrich) is dissolved at 100 mg/mL in CMCT buffer, and kethoxal (ICN Biochemicals) is diluted at 40 mg/ mL in 20% ethanol (see Note 3). Fresh working solutions are prepared just before use. 6. RNA transcripts are synthesized from plasmid templates by in vitro transcription using T7 RNA polymerase according to published protocols and adjusted to a concentration of 2 μM in water. Store at −20°C. 7. Carrier RNA: tRNA, 2 mg/mL. 8. Sodium borate solution: 100 mM sodium borate, pH 8.0.

Structural Probing of RNA Thermosensors

2.2. 5 ′-Radiolabelling and Purification of the Deoxyribonucleotide Primer

239

1. [γ-32P]ATP (GE Healthcare Life Sciences or ICN), specific activity of 3,000 Ci/mmol and concentration 3.33 μM. Store at 4°C. 2. T4 DNA polynucleotide kinase (New England Biolabs), 10 U/μL. Store at −20°C. 3. HPLC-purified oligodeoxyribonucleotide primer, 50 μM in water (see Note 4). Store at −20°C. 4. T4 PNK buffer (10×): 500 mM Tris–HCl of pH 7.5, 100 mM magnesium chloride, 50 mM dithiothreitol (DTT), 1 mM spermidine–HCl, 1 mM ethylenediamine tetraacetic acid (EDTA). Store at −20°C. 5. Phenol/chloroform/isoamyl alcohol mixture at a ratio 25:24:1 (v/v/v).

2.3. Primer Extension on RNA with Reverse Transcriptase

1. Four dNTPs (Promega), 100 mM each. Store at −20°C. 2. Avian myeloblastosis virus (AMV) reverse transcriptase (Finnzymes), 20 U/μL. Store at −20°C. 3. 10× extension reaction buffer: 250 mM Tris–HCl of pH 8.3, 50 mM magnesium chloride, 500 mM potassium chloride, 20 mM DTT. Store at −20°C. 4. Stop solution: 87.5% formamide, 0.05% (w/v) xylene cyanol, 0.05% (w/v) bromophenol blue, 5 mM EDTA. Note that formamide should be deionized by stirring with AG 501-X8 resin (Biorad). Store at 4°C.

2.4. RNA Sequencing Reaction

1. Individual ddNTPs (GE Healthcare Life Sciences), 5 mM each. Store at −20°C.

2.5. Sequencing Gel Analysis

1. 10× TBE buffer: 890 mM Tris–borate, 20 mM EDTA. Store at room temperature. 2. 40% Acrylamide/bisacrylamide solution (19:1) (MP Biomedicals) and N,N,N¢ ,N¢-tetramethyl-ethylenediamine (TEMED) (Promega). Store both solutions at 4°C. Note that acrylamide/bisacrylamide solution is a potent neurotoxic when unpolymerized and so care should be taken against exposure to skin. 3. Ammonium persulphate: prepare 10% (w/v) solution in water. Store at 4°C up to 1 month. 4. Fixing solution: 50% ethanol, 15% acetic acid. Prepare just before use. 5. 3MM chromatography paper (Whatmann). 6. Gel electrophoresis apparatus and gel dryer such as Model 583 (Biorad).

240

Chiaruttini, Allemand, and Springer

3. Methods Precipitation of RNA with ethanol is carried out by addition of one-tenth volume of 3 M sodium acetate of pH 5.2, and 3 volumes of 100% ethanol followed by precipitation at −80°C for 15 min. RNA is recovered by centrifugation at 16,000 × g for 15 min at 4°C. RNA pellets are then subjected to washing with 200 μL of 70% ethanol followed by centrifugation at 16,000 × g for 5 min at 4°C and dried under vacuum for 30 min. 3.1. RNA Modification Reactions

1. Prepare an RNA solution by mixing 110.25 μL water with 1.75 μL 2 μM RNA. Incubate at 80°C for 3 min and immediately dip the tubes into a dry ice/ethanol bath for 1 min. Thaw by gently flicking the tube. Spin the tube at 400 × g for 2 s to recover the microdrops of water that have evaporated on the internal side of the cap and keep the tube on ice. 2. Prepare the “DK mix” by transferring 72 μL RNA solution into a tube containing 18 μL 5× DMS buffer. 3. Prepare the “C mix” by transferring 36 μL RNA solution into a tube containing 9 μL 5× CMCT buffer. 4. Incubate both mixes at 25°C for 10 min and add 4.5 and 2.25 μL carrier RNA to the “DK mix” and “C mix”, respectively. Keep at room temperature before modification reaction. 5. DMS reaction: combine 20 μL “DK mix” and 2 μL working DMS solution; the control reaction contains 20 μL “DK mix” and 2 μL 100% ethanol; incubate at 25°C for 5 min. Stop the reaction by ethanol precipitation as described earlier. Resuspend the final RNA pellet in 10 μL water. The final RNA concentration is ~50 nM. 6. Kethoxal reaction: combine 20 μL “DK mix” and 2 μL working kethoxal solution; the control reaction contains 20 μL “DK mix” and 2 μL 20% ethanol; incubate at 25°C for 5 min and add 10 μL 100 mM sodium borate, pH 8.0. Stop the reaction by ethanol precipitation as described earlier. Resuspend the final RNA pellet in 10 μL 12.5 mM sodium borate solution. The final RNA concentration is ~50 nM. 7. CMCT reaction: combine 20 μL “C mix” and 2 μL of working CMCT solution; the control reaction contains 20 μL of “C mix” and 2 μL of CMCT buffer; incubate at 25°C for 10 min. Stop the reaction by ethanol precipitation as described earlier. Resuspend the final RNA pellet in10 μL of water. The final RNA concentration is ~50 nM.

Structural Probing of RNA Thermosensors

3.2.5 ′-Radiolabelling and Purification of the Deoxyribonucleotide Primer

241

1. Combine the following reagents in the order indicated: 10.75 μL water, 2.5 μL 10× T4 PNK buffer, 0.5 μL oligodeoxyribonucleotide primer, 10 μL [γ-32P]ATP, 1.25 μL T4 polynucleotide kinase. Incubate at 37°C for 1.5 h. 2. The following manipulations, which are aimed to remove unreacted ATP, are carried out at room temperature. Add 25 μL water to the reaction mixture and extract with 50 μL of a phenol/chloroform/isoamyl alcohol solution by vortexing for 30 s. Centrifuge at 800 × g for 1 min and transfer the upper aqueous phase into a fresh tube. Wash the organic phase with 50 μL water by vortexing for 30 s and centrifuge as indicated earlier. Combine the aqueous phases and precipitate the oligodeoxyribonucleotide with ethanol. The pellet is resuspended in 100 μL of water and the oligodeoxyribonucleotide is precipitated by ethanol one more time. Finally, the pellet is washed and dried as described earlier and resuspended in 12.5 μL water at an estimated final concentration 2 μM (see Note 5).

3.3. Primer Extension on RNA with Reverse Transcriptase

1. Prepare the “oligo mix” by combining the following solutions: 3.5 μL 10× extension reaction buffer, 3.5 μL 5′-[32P] labelled oligodeoxyribonucleotide. This mixture is sufficient for one round of probing experiments with the three modifying reagents followed by cDNA extension with one primer. 2. Prepare the “dNTPs/DTT mix” by combining the following solutions: 64 μL water, 20 μL 10× extension reaction buffer, 8 μL of 100 mM DTT, 2 μL of each dNTP. Store at −20°C. 3. Prepare the “RT mix” by combining the following solutions: 34.3 μL “dNTPs/DTT mix”, 0.7 μL AMV reverse transcriptase. 4. Combine 4 μL of either modified or unmodified RNA at ~50 nM from Subheading 3.1 and 1 μL of the “oligo mix”. Incubate at 80°C for 3 min and immediately dip the tubes into a dry ice/ethanol bath for 1 min. Thaw by gently flicking the tubes. Spin the tubes at 400 × g for 2 s and keep the tubes on ice. 5. Add 5 μL of the “RT mix” and incubate at 38°C for 20 min (see Note 6). 6. Stop the reaction with 10 μL stop solution. The samples can be stored at −20°C. For gel analysis, load 5–10 μL of the sample.

3.4. RNA Sequencing Reactions

1. Combine the following solutions: 6.75 μL water, 0.75 μL 2 μM RNA, 1 μL 10× extension reaction buffer, 1.5 μL 2 μM 5′-labelled oligodeoxyribonucleotide. Incubate at 80°C for 3 min and immediately dip the tubes in a dry ice/ethanol bath

242

Chiaruttini, Allemand, and Springer

for 1 min. Thaw by gently flicking the tube. Spin the tubes at 400 × g for 2 s. 2. Add 1 μL AMV reverse transcriptase, vortex and keep on ice. The final volume of this “pre-reaction mix” is 11 μL. 3. Prepare four “ dNTP/ddNTP mixes” as follows. The “dATP/ ddATP mix” is prepared by combining 30 μL water, 50 μL “dNTPs/DTT mix”, 20 μL 5 mM ddATP; the “dCTP/ ddCTP mix”: 40 μL water, 50 μL “dNTPs/DTT mix”, 10 μL 5 mM ddCTP; the “dGTP/ddGTP mix”: 30 μL water, 50 μL “dNTPs/DTT mix”, 20 μL 5 mM ddGTP; the “dTTP/ ddTTP mix”: 30 μL water, 50 μL “dNTPs/DTT mix”, 20 μL 5 mM ddTTP. Keep the mixes on ice. 4. Label four tubes and add 2.5 μL of the corresponding “dNTP/ ddNTP mix”. Keep the tubes on ice. 5. Add 2.5 μL “pre-reaction mix” to each tube and incubate at 38°C for 30 min. 6. Stop the reaction with 10 μL stop solution. The samples can be stored at −20°C. For sequencing gel analysis, load 1 μL of the sequencing sample combined with 9 μL of a stop solution/water (1:1) mix. 3.5. Sequencing Gel Analysis

1. We use sequencing gels that are 0.4-mm thick, 400-mm long, and either 200- or 350-mm wide, depending on the number of samples to be analysed. 2. Prepare a 6% gel solution as follows: for a 200-mm wide gel, dissolve 20 g urea in 6.3 mL acrylamide/bisacrylamide solution, 4.2 mL 10× TBE buffer, 15.8 mL water, and 220 μL 10% ammonium persulphate. For 350-mm wide gel, scale up all components. Filter the gel solution by aspiration through two layers of 3MM chromatography paper. 3. Transfer 1 mL gel solution into a plastic tube and add 1 μL TEMED. Inject the solution at the bottom of the pair of glass plates using a 1-mL pipetter. Tilt the plates at an angle of about 45°C. Polymerization takes ~15 min. 4. Add 13 μL TEMED to the remaining gel. Slowly pour the gel solution between the glass plates using a 50-mL syringe and insert the comb carefully not to allow air bubbles to become trapped under the wells. Polymerisation of the gel takes ~1 h. 5. Prepare the running buffer by diluting 100 mL 10× TBE buffer with 900 mL water. 6. After the gel has set, carefully remove the comb and wash the wells with the running buffer. Pre-run the gel at constant power (55 W for a 200-mm wide or 90 W for a 350-mm wide gel) for 30 min. 7. Heat the samples at 80°C for 3 min and cool immediately on ice for at least 1 min.

Structural Probing of RNA Thermosensors

243

8. Load the samples on the gel and perform electrophoresis at constant power for ~1.5 h. Under these conditions, the bromophenol blue dye runs off the gel at the end of the run. 9. After the run is completed, immerse the gel attached to a glass plate in 1.5 L of the fixing solution for 15 min. Carefully lay a sheet of 3MM chromatography paper cut to the appropriate dimensions on the gel and transfer the gel to the paper by blotting the residual fixing solution three times with a fresh double layer of kitchen paper. Dry the gel at 80°C for 1.5 h. Expose the gel to X-ray film or PhosphorImager for 24 h. Modification blocks the progress of reverse transcriptase at the nucleotide preceding the modified nucleotide. Thus, the resulting cDNA is one nucleotide shorter than the cDNA corresponding to the relevant nucleotide on the sequencing lane (see Note 7).

4. Notes 1. It is imperative to obey strict safety rules when using RNAmodifying reagents because these reagents are potential carcinogens. It is therefore strongly recommended to carry out modification experiments in a fume hood and to wear protective gloves. After use, it is imperative to destroy the reagents by discarding working solutions in bottles containing 500 mL 1N sodium hydroxide for DMS and kethoxal and 500 mL 10% acetic acid for CMCT. The supernatants derived from the precipitation step following RNA modification, and tubes that have been in contact with the chemicals must be treated the same way. The tubes should be disposed of in an appropriate can according to the lab safety rules. 2. A strong indication for the existence of an RNA helix is to monitor its stability by carrying out modification reactions in “semi-native” conditions. “Semi-native” means that the RNA secondary structure is forced to “breathe” by replacing magnesium in the reaction buffers with EDTA. Under such conditions, nucleotides involved in Watson–Crick interactions become more reactive towards the probes. This can be done by using 5× SND (250 mM sodium cacodylate of pH 7.5, 5 mM EDTA) buffer instead of 5× DMS buffer when probing with DMS and kethoxal and 5× SNC (250 mM sodium borate of pH 8.0, 5 mM EDTA), or SNC (50 mM sodium borate of pH 8.0, 5 mM EDTA) buffers instead of 5× CMCT and CMCT buffers, respectively, when probing with CMCT. Note that omitting magnesium also results in weakening the stability of tertiary interactions which are the first to break.

244

Chiaruttini, Allemand, and Springer

3. Kethoxal is a highly viscous solution and should therefore be weighed instead of measured by pipetting. Our current practice is to tare a tube, then transfer some kethoxal in with a 200 μL pipettor, and finally weigh the tube again. 4. Oligodeoxyribonucleotide primers should be 15–25 nucleotides long; shorter primers may not hybridize efficiently under the conditions described in Subheadings 3.3 and 3.4. Note that natural RNA can possess post-transcriptional modifications (such as m2G, m62A), which may interfere with reverse transcription: care should be taken in choosing appropriate oligonucleotide. 5. Purification of 5′-labelled-oligodeoxyribonucleotide by gelfiltration chromatography using spin columns (GE Healthcare Life Sciences, Microspin G-50 column) works as well as the repeated ethanol-precipitation method that we currently use. 6. The major difficulty in primer extension is encountering regions of RNA that cause the reverse transcriptase to pause or even terminate. The resulting stops show up as bands of intermediate sizes on the gel, both confusing interpretation and reducing the yield of fully extended primer. Such termination sites may be caused by extensive GC-rich stretches of RNA or by secondary structures within the RNA. We use the following measures to reduce such artefacts. The primers are chosen so that the extension products are <200 nucleotides, to reduce the distance traversed by reverse transcriptase, thus minimizing the likelihood of encountering a significant secondary structure. We use nucleotide concentrations well in excess over the apparent Km of the enzyme to reduce pausing by the enzyme. We perform extension reactions at 38°C. However, in case the secondary structure causes the enzyme to pause, a higher incubation temperature (42°C) is recommended to minimize this effect. 7. If the bands are blurred, clean the extension products with a precipitation/washing step after the extension reaction. Resuspend the dry final pellet in 20 μL of a stop solution/ water (1:1); mix and analyse as indicated.

Acknowledgments We thank Ciaran Condon for critical reading of the manuscript. This work was supported by the CNRS (UPR9073), the University of Paris Diderot-Paris VII, and the ANR (ANR-05BLAN-0159-01).

Structural Probing of RNA Thermosensors

245

References 1. Narberhaus, F., Waldminghaus, T. and Chowdhury, S. (2006). RNA thermometers. FEMS Microbiol. Rev. 30, 3–16. 2. Morita, M. T., Tanaka, Y., Kodama, T. S., Kyogoku, Y., Yanagi, H. and Yura, T. (1999). Translational induction of heat shock transcription factor sigma32: evidence for a built-in RNA thermosensor. Genes Dev. 13, 655–665. 3. Johansson, J., Mandin, P., Renzoni, A., Chiaruttini, C., Springer, M. and Cossart, P. (2002). An RNA thermosensor controls expression of virulence genes in Listeria monocytogenes. Cell 110, 551–561. 4. Hoe, N. P. and Goguen, J. D. (1993). Temperature sensing in Yersinia pestis: translation of the LcrF activator protein is thermally regulated. J. Bacteriol. 175, 7901–7909. 5. Yamanaka, K., Mitta, M. and Inouye, M. (1999). Mutation analysis of the 5′ untranslated region of the cold shock cspA mRNA of Escherichia coli. J. Bacteriol. 181, 6284–6291. 6. Altuvia, S., Kornitzer, D., Teff, D. and Oppenheim, A. B. (1989). Alternative mRNA structures of the cIII gene of bacteriophage lambda determine the rate of its translation initiation. J. Mol. Biol. 210, 265–280. 7. Mayford, M. and Weisblum, B. (1989). Conformational alterations in the ermC transcript in vivo during induction. EMBO J. 8, 4307–4314.

8. Benito, Y., Kolb, F. A., Romby, P., Lina, G., Etienne, J. and Vandenesch, F. (2000). Probing the structure of RNAIII, the Staphylococcus aureus agr regulatory RNA, and identification of the RNA domain involved in repression of protein A expression. RNA 6, 668–679. 9. Balzer, M. and Wagner, R. (1998). A chemical modification method for the structural analysis of RNA and RNA–protein complexes within living cells. Anal. Biochem. 256, 240–242. 10. Lindell, M., Romby, P. and Wagner, E. G. (2002). Lead(II) as a probe for investigating RNA structure in vivo. RNA 8, 534–541. 11. Michel, F. and Costa, M. (1998). in RNA Structure and Function, eds. Simons, R. W. and Grunberg-Manago, M. (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY), pp. 175–202. 12. Eddy, S. R. (2002). A memory-efficient dynamic programming algorithm for optimal alignment of a sequence to an RNA secondary structure. BMC Bioinformatics 3, 18. 13. Xu, X., Ji, Y. and Stormo, G. D. (2007). RNA Sampler: a new sampling based algorithm for common RNA secondary structure prediction and structural alignment. Bioinformatics 23, 1883–1891. 14. van Batenburg, F. H., Gultyaev, A. P. and Pleij, C. W. (1995). An APL-programmed genetic algorithm for the prediction of RNA secondary structure. J. Theor. Biol. 174, 269–280.

Chapter 18 Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting Translational Regulators in Bacteria Pierre Fechter, Clément Chevalier, Gulnara Yusupova, Marat Yusupov, Pascale Romby, and Stefano Marzi Summary Toeprinting was developed to study the formation of ribosomal initiation complexes in bacteria. This approach, based on the inhibition of reverse transcriptase elongation, was used to monitor the effect of ribosomal components and translational factors on the formation of the active ribosomal initiation complex. Moreover, this method offers an easy way to study in vitro how mRNA conformational changes alter ribosome binding at the initiation site. These changes can be induced either by environmental cues (temperature, ion concentration), or by the binding of metabolites, regulatory proteins, and trans-acting RNAs. An experimental guide is given to follow the different steps of the formation of ribosomal initiation complexes in Escherichia coli and Staphylococcus aureus, and to monitor the mechanism of action of several regulators on translation initiation in vitro. Protocols to prepare the ribosome and the subunits are also given for Thermus thermophilus, Staphylococcus aureus, and Escherichia coli. Key words: Ribosome, Ribosome purification, mRNA, Translation initiation, Toeprinting, Translational regulator, Escherichia coli, Staphylococcus aureus, Thermus thermophilus

1. Introduction Translational regulation in bacteria results in fast adaptation of protein synthesis to environmental conditions. In Escherichia coli, the synthesis of many essential proteins that bind RNA, proteins involved in metabolism and in stress-related responses, are regulated at the translational level. Regulation occurs by a variety of mechanisms that influence the binding of the mRNA to the 30S ribosomal subunit at a step preceding the assembly of the active initiation complex, i.e. formation of the first codon-anticodon Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_18 © Humana Press, a part of Springer Science + Business Media, LLC 2009

247

248

Fechter et al.

interaction. Many of these regulatory events involve structured elements in the 5¢ untranslated regions of mRNAs that are specifically recognized by a large variety of trans-acting factors ranging from metabolites and non-coding RNAs to proteins (1, 2). In other cases, structured mRNA regions can directly sense external signals like temperature (3), or the intracellular concentration of magnesium without the help of a protein (4). Conformational switches within mRNA can lead to translation activation by facilitating the accessibility of the ribosome to its Translation Initiation Region (TIR), or to repression by promoting the formation of an inhibitory structure that masks the TIR. In most of the known cases, competitive binding of the repressor and the ribosome to overlapping/adjacent mRNA-binding sites leads to the formation of mutually exclusive repressor–mRNA and ribosome–mRNA complexes (5, 6). However, in several examples, the repressor and the ribosome bind to distinct sites, raising the question of how repression may occur. Simultaneous binding of the repressor and the ribosome to the same mRNA was demonstrated for the translational regulation of the two ribosomal proteins S4 and S15 in E. coli (6, 7). The formation of the mRNA–repressor–ribosome complex causes ribosome stalling at the pre-initiation stage and prevents the formation of the active complex. Recent data show that repressor protein traps the mRNA structure on the platform of the 30S subunit outside the normal mRNA path and prevents the accommodation step required to form the anticodon–codon interaction in the peptidyl-tRNA (P site)-binding site (8). More recent data indicate that sRNAs bind far upstream of the TIR and prevent the formation of the active ribosomal initiation complex without the strong alteration of mRNA structure (9). Therefore, in addition to in vivo genetics and expression studies (10), it is of interest to monitor in vitro the step of the initiation process at which regulation occurs. An easy way to inspect ribosome binding to mRNAs at the initiation step was devised by L. Gold and coworkers (11). This method, called toeprinting, involves cDNA synthesis by reverse transcriptase on a template mRNA to which the ribosome, together with the initiator tRNA, is bound. The position of the reverse transcriptase termination (toeprint) gives the exact position of the 3¢ edge of the mRNA with respect to the bound ribosome and the tRNA species (12). In the presence of the initiator tRNA, the toeprint was identified for different mRNAs to be located at position +16 from the AUG codon. This 3¢ boundary position of the ribosome was later on validated by footprinting assays (13, 14) and by X-ray analysis (15–17). Toeprinting assays were further used to show that the selection of the correct initiator tRNA is made by the three initiation factors on the 30S subunit (12). The ribosome–mRNA binary complex was also detected while being less stable than the ternary tRNA–mRNA–ribosome

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting

249

(18, 19). Thus, different steps of the formation of the initiation complex can be probed. Furthermore, a comparative study with different mRNAs indicated that the nucleotide distance between the Shine–Dalgarno sequence (SD), the P site, and the 3¢ edge of the mRNA is fixed while secondary structure elements can be accommodated within the mRNA track (16, 20). More recently, toeprinting was applied to monitor the effect of a novel ribosomal elongation factor LepA during the elongation process (21). Hence, it was obvious to apply this strategy to unravel various regulatory mechanisms. The technique was successfully used to prove the temperature dependence of ribosome binding to thermosensor mRNAs (22, 23) and to follow the formation of binary and ternary ribosome complexes in the presence of regulatory proteins (8) and trans-acting sRNAs (e.g. (9, 24, 25). We present here an experimental guide for toeprinting assays and provide typical examples obtained on several mRNAs regulated by proteins or trans-acting RNAs. In addition, protocols for the purification of homogenous ribosomes are given for Escherichia coli, Staphylococcus aureus, and Thermus thermophilus.

2. Materials 2.1. Cell Culture and Lysis

1. French Press (Thermo Spectronic) and fermentor (Infors HT) are required for high-scale purification of ribosomes. 2. For S. aureus, ribosomes are obtained from the strain RN6390 that carries a deletion of the rub gene (see Note 1). S. aureus is grown in Brain Heart Infusion broth (AES laboratoire, Bruz, France). All bacterial cultures are stored in 25% glycerol at −80°C. 3. E. coli MRE600 cells. 4. T. thermophilus HB8 cells. 5. LB medium: 1% (w/v) trypton, 0.5% (w/v) yeast extract, 1% (w/v) NaCl. The pH is adjusted to 7.5 with KOH. 6. BHI medium (AES laboratoire). 7. Blood-agar plates (Merck). 8. T. thermophilus medium: 0.5% (w/v) polypeptone; 0.2% (w/v) yeast extract; 0.2% (w/v) NaCl; vitamin-mineral solution (USBiological, Swampscott, MA), 1 mL/L media. The pH is adjusted with KOH to 7.2–7.4.

2.2. Ribosome Purification

All solutions and buffers are sterilized or filtrated before use. Buffers for toeprinting and reverse transcription can be stored at −20°C.

250

Fechter et al.

1. Gradient maker (SG15, Amersham) connected to a peristaltic pump (Biorad). 2. Capillary tubes used to collect gradient (e.g. Corning). 3. FPLC system with fraction collector (Amersham). 4. Low-speed centrifuge (Sigma) and ultracentrifuge with rotors Ti45, Ti70.1, SW32Ti, and SW28 (Beckmann). 5. Buffer AE: 20 mM MgCl2, 200 mM NH4Cl, 20 mM Tris–HCl of pH 7.5, 0.1 mM ethylenediamine tetraacetic acid disodium salt (EDTA), 6 mM 2-mercaptoethanol. 6. Cushion 1E: 30% sucrose, 10 mM MgCl2, 500 mM NH4Cl, 20 mM Tris–HCl of pH 7.5, 0.1 mM EDTA, 6 mM 2-mercaptoethanol. 7. Buffer BE: 10 mM MgCl2, 50 mM NH4Cl, 20 mM Tris– HCl of pH 7.5, 0.1 mM EDTA, 6 mM 2-mercaptoethanol. 8. Buffer DE: 3 mM MgCl2, 300 mM NH4Cl, 20 mM Tris–HCl of pH 7.5, 0.15 mM Na2EDTA, 2 mM dithiothreitol (DTT). 9. Solution SE: 50% sucrose. To prepare 500 mL of 50% sucrose solution, dissolve 250 g of sucrose in sterile water and add one spoon of Bentonite (Sigma). Leave the solution stirring at 20°C for 1 h, and then remove Bentonite by centrifugation at 20,000 × g for 20 min. 10. Buffer S1: 20% sucrose (use solution SE) in Buffer DE (see Note 2). 11. Buffer S2: 5% sucrose (use solution SE) in Buffer DE. 12. Loading buffer for agarose gel: 40 mM Tris–acetate of pH 8.0, 1 mM EDTA, 0.25% bromophenol blue, 30% glycerol. 13. Running buffer for agarose gel: 40 mM Tris–acetate of pH 8.0, 1 mM EDTA. 14. Buffer CE: 10 mM MgCl2, 50 mM KCl, 20 mM Tris–HCl of pH 7.5, 0.1 mM EDTA, 1 mM DTT. 15. Buffer AS: 20 mM Tris–HCl of pH 7.3, 200 mM NH4Cl, 20 mM MgCl2, 1 mM EDTA, 2 mM β-mercaptoethanol. 16. Buffer BS: 20 mM Tris–HCl of pH 7.3, 500 mM NH4Cl, 10 mM MgCl2, 1 mM EDTA, 2 mM β-mercaptoethanol. 17. Buffer CS: 20 mM Tris–HCl of pH 7.3, 300 mM NH4Cl, 3 mM MgCl2, 0.15 mM EDTA, 2 mM β-mercaptoethanol. Buffer CS0.1: buffer CS with 0.1 mM MgCl2. 18. Buffer DS: 20 mM Tris–HCl of pH 7.3, 50 mM NH4Cl, 5 mM MgCl2, 0.15 mM EDTA, 2 mM β-mercapthoethanol. 19. Buffer AT: 150 mM MgCl2, 500 mM NH4Cl, 40 mM Tris–HCl of pH 7.5, 1.5 mM EDTA, 1 mM DTT. 20. Buffer BT: 50 mM MgCl2, 150 mM NH4Cl, 20 mM Tris–HCl of pH 7.5, 0.5 mM EDTA, 1 mM DTT.

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting

251

21. Cushion 1T: 1.5 M sucrose, 0.68 M CsCl, 150 mM MgCl2, 20 mM Tris–HCl of pH 7.5, 1.5 mM EDTA, 1 mM DTT. 22. Cushion 2T: 1.8 M sucrose, 0.8 M CsCl, 150 mM MgCl2, 20 mM Tris–HCl of pH 7.5, 1.5 mM EDTA. 23. Buffers CT0, CT0.8 and CT1: 10 mM MgCl2, 400 mM NaCl, 20 mM Tris–HCl of pH 7.5, 0.5 mM EDTA, 1 mM DTT and 0.0, 0.8, and 1.0 M (NH4)2SO4, respectively. 24. YM100 Centricon cartridges (Millipore). 25. DNase I RNase free (Roche Diagnostics GmbH, Germany). 26. Phenylmethanesulphonylfluoride (PMSF): dissolve in ethanol. Handle under hood with extreme care. 27. 10 mM Tris–HCl, pH 7.5. 28. Staphylococcus aureus lysostaphin (Sigma). 29. Buffer TE: 10 mM Tris–HCl of pH 8.0, 1 mM EDTA. 30. 1 M MgCl2. 31. 200-mL Toyopearl Butyl 650S column. 32. 3 M Na-acetate, pH 5.2. 2.3. mRNA Preparation

2.4. Toeprinting Assays

The RNA is transcribed in vitro with T7 RNA polymerase from a linearized plasmid template carrying the T7 promoter fused with the gene of interest (26). The RNA is then purified, as described in Chapter “Probing mRNA Structure and sRNA– mRNA Interactions in Bacteria Using Enzymes and Lead(II)” by Chevalier et al., by using gel filtration column (27) or Mono Q column (28). If RNA is not homogenous after purification due to the presence of abortive transcripts or cleavage products, a further purification using 8% polyacrylamide-(0.5% bis)/8 M urea slab gels in 1× TBE is recommended. 1. The symbols + and − in the buffer names refer to the presence or absence of MgCl2, respectively. Buffer TP5X+: 100 mM MgCl2, 100 mM Tris–HCl of pH 7.5, 300 mM NH4Cl, 5 mM DTT. 2. Buffer TP5X−: 100 mM Tris–HCl of pH 7.5, 300 mM NH4Cl, 5 mM DTT. 3. MgCl2100 solution: 100 mM MgCl2. 4. dNTP500 mix: 500 μM of each dATP, dGTP, dCTP, dTTP. 5. DNA primer for RT: the length of the primer usually varies from 12 to 18 nucleotides. The primer is complementary to a region of the mRNA located around 40 nucleotides downstream of the initiation codon. It is worthwhile to make efforts on establishing the optimal conditions for hybridization and extension.

252

Fechter et al.

6. DNA-loading buffer: 1 mM EDTA, 0.02% xylene cyanol, 0.02% bromophenol blue in formamide. 7. T4 polynucleotide kinase (PNK), 10 U/μL (Fermentas, Glen Burnie, MD) supplied with 10× PNK buffer. 8. [γ-32P]ATP, 3,000 Ci/mmol (Amersham, France). 9. tRNAfMet (Sigma). 10. Reverse transcriptases (RT): MMLV RT (Gibco), AMV RT (MP Biochemicals or Life Sciences). 11. Total tRNA, 2 μg/μL (Sigma). 12. DNA elution buffer: 500 mM ammonium acetate of pH 6.0, 1 mM EDTA. 13. Phenol: saturate with 20 mM Tris–HCl, pH 8.0. 14. Chloroform/isoamylic alcohol (24:1) mixture. 2.5. Buffers and Solutions for PAGE

1. Electrophoresis apparatus for slab gels (30 cm × 40 cm, BRL) and generator (200 W). 2. 1× TBE: 0.09 M Tris–borate of pH 8.3, 1 mM EDTA. 3. 25% Polyacrylamide/8 M urea solution: dissolve 480 g urea (Merck) in 625 mL Rotiphorese 40 gel solution with acrylamide/bis-acrylamide ratio 19:1 (Roth, Karlsruhe, Germany), adjust the volume to 1 L with bidistillated water, and filter the solution. Use this stock solution to prepare 8% polyacrylamide gel/8 M urea/1× TBE gels.

3. Methods 3.1. Setting the Experimental Conditions 3.1.1. Formation of Binary Complexes

1. Efficient initiation of translation in bacteria requires the formation of a stable initiation complex where the initiation codon (mainly AUG) interacts with the initiator fMet – tRNAfMet in the ribosomal P site. This interaction is the result of the mRNA accommodation step that follows the mRNA–30S binding step (29). Prior to the toeprinting assay, the formation of the complex between the mRNA and the 30S subunit in the absence of the initiator tRNA can be monitored by different approaches, such as filterbinding assays on nitrocellulose membrane, analytical sucrose gradients, and gel filtration chromatography. These methods require labelled mRNA and can detect the formation of a stable binary mRNA–30S complex. The toeprinting assay, however, can additionally map the 3¢ edge of the mRNA segment shielded by the bound ribosome. It was previously shown that the reverse transcriptase stops close

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting

253

to the SD sequence of the unstructured mRNA bound to the ribosome (18). The formation of the SD–aSD helix is the most stable interaction that usually takes place in the binary complex. This complex can be detected using the reverse transcriptase (RT) from the Avian Myeloblastosis Virus (AMV) or from the Moloney Murine Leukemia Virus (MMLV). However, these RTs have different properties. Under suboptimal experimental conditions, MMLV RT is much more sensitive to the secondary structures of the mRNA than AMV RT and can detect rather weak mRNA– 30S subunit interactions. The AMV RT is more processive than MMLV RT and is usually used if the binary complex is stable. An example is given in Fig. 1a with the rpsO mRNA. The weak interaction between the pseudoknot structure of the mRNA and the ribosome was detected only at low concentrations of MMLV RT. Hence, the use of MMLV at suboptimal conditions can be a valuable tool to monitor the effect of the structured regulatory region of the mRNA on 30S recognition in the absence of the initiator tRNA (Fig. 1a) (8). 3.1.2. Formation of Ternary Complexes

1. The presence of the initiator tRNAfMet is necessary to form an irreversible and active mRNA–tRNA–30S initiation complex. In this complex, mRNA is placed into the 30S mRNA channel (15), and the mRNA–30S interactions are strengthened by the codon–anticodon interaction in the P site. In this case, both AMV and MMLV RT can be used to monitor the mRNA–ribosome interaction. The toeprint is observed at position +16 with AMV RT and usually at position +17 with MMLV RT (Fig. 1).

3.1.3. Formation of Regulatory Complexes

1. The regulatory protein traps the ribosome in the inactive pre-initiation complex. The transition from the pre-initiation complex to the active initiation complex is a key step under tight regulation. Structured regions present in the leader regions of mRNAs may delay the mRNA accommodation step. trans-Acting regulators or physical cues (pH, temperature) may prevent this transition from occurring and cause the accumulation of the pre-initiation complex. Ribosomal protein S15 regulates the translation of its own mRNA by stalling the 30S subunit in the pre-initiation complex (7). In this system, S15 stabilizes a pseudoknot structure that is also recognized by the 30S subunit and prevents the conformational changes that are required to accommodate the initiation codon in the P site. The toeprinting assay performed with MMLV RT shows that in the presence of S15, the ternary complex 30S–mRNA–tRNA

254

Fechter et al.

Fig. 1. Toeprinting on several mRNAs and effects of trans-acting regulators. (a) Effect of E. coli ribosomal protein S15 on the formation of the binary and ternary ribosomal initiation complexes: appearance of reverse-transcriptase stop at position +10 (entrance of the pseudoknot structure stabilized by S15) is the signature for the stalled S15–mRNA–30S complex (entrapment complex). The signal at position +17 corresponds to the active mRNA–tRNA–30S complex (initiation complex). Toeprinting experiments were conducted with E. coli rpsO mRNA transcript, E. coli 30S subunit, in the absence or presence of S15 and tRNAfMet. The initiator tRNA was always added after S15. (Lanes 1 and 2) Incubation controls of free mRNA or bound to S15, respectively; (lane 3) mRNA + 30S; (lane 4) mRNA + 30S + tRNAfMet; (lane 5) mRNA + 30S + S15; (lane T and C) sequencing DNA ladders. Primer extension was carried out with various concentrations of MMLV RT. At 4 U of MMLV, the signature of the pseudoknot structure in the binary complex cannot be detected (lane 3). (b) Effect of S. aureus RNAIII on the formation of the ternary 30S–mRNA–tRNA initiation complex using either E. coli or S. aureus 30S. (Lanes 1 and 2) Incubation controls of free mRNA or mRNA bound to RNAIII, respectively; (lane 3) mRNA + 30S + tRNAfMet; (lanes 4–7) toeprinting performed in the presence of wild-type RNAIII (lanes 4, 100 nM; lanes 5, 200 nM) or mutant RNAIII lacking the hairpin 13 (lanes 6, 100 nM; lanes 7, 200 nM). tRNA was added after RNAIII and the 30S subunits. The ternary mRNA–tRNA–30S complex shows a toeprint at position +16 while the mRNA–RNAIII duplex induces an RT stop at position +3. SD is for Shine–Dalgarno sequence.

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting

255

is not formed anymore (toeprint at +17 disappears) while the binary 30S–mRNA complex is stabilized (toeprint at +10 appears) (Fig. 1a). These effects can be observed only if the initiator tRNA is added after the repressor protein; therefore, it is essential to mix the reaction compounds in a defined order. 2. Repression through direct masking of the RBS by regulatory RNA. Many translational repressors bind the 5¢ leader region of the mRNA that overlaps with the RBS. Such regulation is documented for mRNAs targeted by the regulatory RNAIII from Staphylococcus aureus. The RNAIII is the intracellular effector of the quorum sensing system that regulates numerous virulence factors in a coordinated way (30). The RNAIII base pairs with target mRNAs to form extended duplexes that hinder the SD sequence of the mRNA. This duplex is also a substrate for the doublestrand-specific endoribonuclease III, which initiates the rapid degradation of the repressed mRNA (25). Since the mRNA signal disappears rapidly as soon as the RNAIII is produced, it is quite difficult to prove by in vivo experiments whether the RNAIII inhibits gene expression at the transcriptional or at the post-transcriptional level. On the other hand, a toeprinting assay allows one to monitor in vitro whether the formation of the RNAIII–mRNA duplex is sufficiently stable to prevent the formation of the binary and ternary complexes. A typical example is given for the 5¢ regulatory region of SA2353 mRNA that is recognized by RNAIII (Fig. 1b). S. aureus ribosomes provide a toeprint at the expected +16 position, which disappears upon the addition of RNAIII. Concomitantly, a toeprint appears at position +3 that corresponds to the formation of the RNAIII–mRNA duplex. A mutant of RNAIII that is not able to bind rapidly to the mRNA has no effect on ribosome binding. Again, efficient competition between RNAIII and the ribosome is only observed if the initiator tRNA is added after the regulatory RNA. 3. Translation initiation in Gram-positive and Gram-negative bacteria may slightly differ due to some variations in mRNAs and ribosomes. Would these differences influence the regulation of translation initiation? Toeprinting assays may help to answer this question if one can study heterologous complexes. Here we show that S. aureus SA2353 mRNA is equally well recognized by S. aureus and E. coli ribosomes and that the competition between ribosomes and RNAIII does not depend on the nature of the ribosome (Fig. 1b). Thermus thermophilus ribosomes also bind efficiently to E. coli mRNAs (results not shown).

256

Fechter et al.

3.2. Ribosome Preparation 3.2.1. Preparation of the E. coli 70S Ribosome and 30S Subunits

1. Prepare a 1-L preculture of E. coli MRE600 cells by growing bacteria overnight at 37°C in LB medium using 1 mL of the frozen bacterial stock as a starting culture. Transfer the preculture to 100 L fermentor and continue growth with aeration in the same medium at 37°C. Collect cells at OD600 = 1. 2. Wash 200 g cells in 1 L of buffer AE, and resuspend in 200 mL of the same buffer. Add DNase I to 1 U/mL and PMSF to 1 μg/mL. 3. Disrupt cells in French Press and remove debris by centrifugation for 30 min at 30,000 × g at 4°C. 4. Collect ribosomes by centrifugation at 150,000 ´ g in a Ti 45 rotor for 4 h at 4°C. The ribosome pellet forms two layers. On the bottom of the tube, ribosomes form a glass-type pellet. A loose ribosome pellet is formed on the top of the glass-type pellet. Wash upper layer from the surface of glass-type pellet with buffer AE. 5. Resuspend ribosomes in 50 mL buffer AE and load on the top of high-density Cushion 1E buffer in Ti 45 centrifuge tube (25 mL of ribosome suspension is loaded onto 25 mL of Cushion 1E) and centrifuged at 90,000 ´ g for 19 h at 4°C. Ribosomes form pellet on the bottom of centrifuge tube. 6. Resuspend ribosomes in buffer BE to final concentration 50 mg/mL (~ 20 μM) and keep at −80°C (see Note 3) in small aliquots or in 1 mL aliquots for 30S subunit preparation. 7. Dialyse extensively the 70S ribosomes (1 mL from 50 mg/ mL solution) in Buffer DE at 4°C. 8. Prepare sucrose gradient directly in ultra-clear Beckman tubes (25 × 89 mm) for the SW28 rotor. Mix 18 mL of Buffer S1 and 18 mL of Buffer S2 using a gradient maker in order to obtain a 36 mL 20% (bottom) to 5% (top) sucrose gradient. Six tubes with gradients are usually prepared. The gradients are kept on ice for 2 h prior to ultracentrifugation. 9. Carefully layer the dialysed 70S ribosomes on the top of the tube (no more than 8 mg/per gradient) and perform ultracentrifugation in SW28 rotor at 70,000 g for 10 h at 4°C. 10. Fractionate the sucrose gradients and check the lighter fractions containing 30S subunits by electrophoresis. Take small aliquots and mix them with the same volume of loading buffer for agarose gel. Load the samples on 1% agarose gel and carry out electrophoresis in running buffer for agarose gel (see Note 4). 11. Dialyse the fractions containing pure 30S subunits in Buffer CE at 4°C overnight. Concentrate the 30S subunits using

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting

257

YM100 Centricons to a final concentration of 10 mg/mL (~12 μM) and store at −80°C in 5-μL aliquots. 3.2.2. Preparation of the S. aureus 70S Ribosome and 30S Subunits

1. Grow preculture of S. aureus cells RN6390 in 10 mL of BHI medium at 37°C overnight using a single colony grown on a blood-agar plate. Transfer the preculture to 1 L of BHI medium and continue growth at 37°C. Collect cells at OD600 = 5 and wash them with 10 mM Tris– HCl, pH 7.5. 2. Lyse 5 g of S. aureus cells for 30 min at 37°C in 10 mL of TE supplemented with 200 μg of lysostaphin and 100 U of DNase I. Pellet cell debris by centrifugation for 20 min at 30,000 × g at 4°C. 3. Load the supernatants onto 0.5 mL 35% sucrose in buffer AS, in Ti70.1 centrifuge tubes. Pellet the ribosomes for 5 h at 100,000 × g at 4°C. The ribosomes form a glass-type pellet covered by cell debris. Carefully wash the pellet with 200 μL of buffer BS to remove the thin upper layer. 4. Resuspend ribosomes in 2 mL buffer 2 with smooth agitation at 4°C overnight. Load the solution onto 0.5 mL 35% sucrose in buffer BS, in Ti70.1 centrifuge tubes. Centrifuge for 4 h at 100,000 × g at 4°C. 5. Carefully resuspend the translucent pellets in 7 mL buffer CS and incubate for 30 min at 37°C to dissociate ribosomes that are not tightly formed. Precipitate the ribosomes by addition of 170 μL 1 M MgCl2 and 2/3 volumes of ethanol. Centrifuge for 30 min at 3,500 × g, and resuspend the pellets in 1 mL buffer DS. 6. Load the solution (0.5 mL, not more than 5 mg of ribosomes per gradient) on top of the 15 mL 10–40% sucrose gradient prepared with buffer DS, and centrifuge in SW32Ti rotor for 17 h at 40,000 ´ g at 4°C. 7. Collect the 70S ribosome fractions and precipitate by 2/3 volumes of ethanol. The ribosomes are then recovered by centrifugation for 30 min at 3,500 × g at 4°C. Resuspend the pellets in 1 mL of buffer CS0.1 and incubate for 30 min at 37°C for the dissociation of the ribosomal subunits. 8. Layer the solution (0.5 mL/per gradient, 1 mg) onto the top of the 15 mL 10–30% sucrose gradient prepared with buffer 3 and centrifuge in a SW32Ti rotor for 15 h at 40,000 × g at 4°C. 9. Collect and pool the 30S subunit fractions. Adjust MgCl2 concentration to 20 mM. Precipitate the 30S subunits with 2/3 volumes of ethanol for 2 h on ice, centrifuge for 30 min at 3,500 × g at 4°C, wash with ethanol 80%, and dry under vacuum. Dissolve the pellet in a minimal volume of buffer 1 and store in small aliquots at −80°C.

258

Fechter et al.

3.2.3. Preparation of the T. thermophilus Ribosomes

1. Add 1 mL of T. thermophilus HB8 stock culture to 1 L of the T. thermophilus medium and grow bacteria at 75°C overnight. Continue growth in 100 L fermentor and collect cells at OD600 = 1. 2. Wash cells (100 g) with 1 L of buffer AT, resuspend the pellet in 100 mL of the same buffer. Add DNaseI and PMSF to 1 U/mL and 1 μg/mL, respectively. 3. Disrupt cells in French Press and remove debris by centrifugation for 30 min at 30,000 × g at 4°C. 4. Layer 29 mL supernatant on top of 7 mL Cushion 1T and centrifuge in SW28 rotor at 100,000 × g for 20 h at 4°C. 5. Collect a 5-mL fraction from the bottom of the cushion and dilute it three times with buffer BT. Layer 29 mL ribosomes on 7 mL Cushion 2T and centrifuge in SW28 rotor at 100,000 × g for 40 h at 4°C. Collect a 4-mL fraction from the bottom of cushion and dialyse against buffer CT. 6. Load ribosomes (900 mg) on 200 mL Toyopearl Butyl 650S column equilibrated in buffer CT0, and CT1. Wash the column by 2 volumes of buffer CT0.8. Elute ribosomes by 900 mL reverse gradient of ammonium sulphate (from 80 to 40%) in buffers CT0/CT1 using the flow rate 6 mL/min and the fraction volume 12 mL. The 70S peak is collected as shown in Fig. 2 and ribosomes are dialysed against buffer BT.

Fig. 2. Elution profile of T. thermophilus ribosomes from Toyoperl Butyl 650S column.

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting

3.3. Toeprinting Assays 3.3.1. 5 ¢ End Labelling of Oligodeoxyribonucleotide

259

1. In order to prepare the radiolabelled DNA primer complementary to the 3¢ end of the mRNA, incubate 1 μL 100 μM oligodeoxyribonucleotide at 37°C for 1 h in the following mixture: 5 μL [γ-32P]ATP, 1 μL 10× T4 PNK buffer, 2 μL sterile water, 1 μL T4 PNK. 2. Add 10 μL DNA loading buffer to the labelling reaction and perform PAGE in a 12%/8 M urea gel to remove the excess of [γ-32P]ATP. 3. Cut the DNA band and elute DNA in DNA elution buffer at 4°C under continuous shaking for 12 h. Perform a phenol/ chloroform extraction and ethanol precipitation (see Subheading 3.3.8, step 4). 4. Wash the pellet with 70% ethanol, dry under vacuum and dissolved radioactive primer in sterile water at concentration ~200,000 cpm/μL.

3.3.2. mRNA-Primer Hybrid Formation and Renaturation

1. Mix 5 pmol of mRNA and 5 μL of labelled primer in Buffer TP1X− (final volume: 20 μL). Denature mRNA and primer for 1 min at 90°C and cool the mixture for 1 min on ice. 2. Adjust MgCl2 concentration to 10 mM with MgCl2100 solution (2.5 μL). Incubate the sample at 20°C for 20 min for formation of the mRNA/primer hybrid and mRNA renaturation.

3.3.3. Ribosome Activation

1. Reactivate both 70S ribosomes and 30S subunits stored at −80°C (see Note 5) by incubation at 37°C for 15 min (the storage buffer should contain 10 mM MgCl2). 2. After reactivation, dilute the 70S ribosomes 10 times in Buffer TP1X− to obtain final MgCl2 concentration of 1 mM. Incubate the ribosomes for 15 min at 37°C to facilitate subunit dissociation.

3.3.4. Formation of the Binary Complex

Mix the mRNA–primer mixture (1 μL, 0.25 pmol of mRNA) with 2 pmol of ribosomes or 30S subunits, and 2 μL Buffer TP15X- in the final volume of 10 μL. Adjust the MgCl2 concentration to 8 mM with MgCl2100 solution and form the mRNA–ribosome complex at 37°C for 15 min. Incubate free mRNA as a control.

3.3.5. Formation of the Ternary Complex

Add 0.5 μL (40 μM) of uncharged initiator tRNAfMet to the binary mRNA–ribosome complex. The stable ternary mRNA– tRNA–ribosome complex is then formed by incubation at 37°C for 5 min. Incubate free mRNA as a control.

3.3.6. Formation of Pre-initiation Complex Trapped by Repressor Protein

1. The E. coli S15 protein represses its own synthesis by stabilizing the complex between the rpsO mRNA and ribosome at the pre-initiation step. This regulation can be visualized by toeprinting (Fig. 1a). 2. Mix the mRNA–DNA mixture (1 μL, 0.25 pmol of mRNA) with 2 pmol of ribosomes or 30S subunits, 10 pmol of purified

260

Fechter et al.

S15 (see Note 6), 2 μL Buffer TP5X− in a final volume of 10 μL. Adjust MgCl2 concentration to 8 mM with MgCl2100 solution, and carry out formation of the mRNA–S15–ribosome complex at 37°C for 15 min. 3. To monitor the effect of ribosomal protein S15 on the formation of the mRNA–tRNA–ribosome complex, add 0.5 μL of uncharged initiator tRNAfMet to the pre-formed S15–mRNA– ribosome complex and incubate at 37°C for 5 min. As shown in Fig.1a, if tRNA is added after the repressor protein, the ternary mRNA–tRNA–ribosome complex cannot be formed. Conversely, if tRNA is added before the repressor protein, the stalled mRNA–S15–ribosome complex cannot be formed (8). 4. Incubation controls should be done under strictly identical conditions using the mRNA which is either free or bound to the repressor protein. 3.3.7. Effect of transActing Antisense RNA on the Formation of the Ribosomal Initiation Complex

1. Toeprinting assays can be used to visualize formation of the regulatory mRNA–antisense RNA complexes that cause inhibition of the mRNA–ribosome binding (Fig. 1b). 2. Refold the mRNA–DNA primer hybrid as described in Subheading 3.3.2. In parallel, denature and refold the regulatory RNA (here S. aureus RNAIII) as described in Subheading 3.3.2. Mix the mRNA–DNA mixture (1 μL, 0.25 pmol of mRNA) with 2 pmol of ribosomes, 1 μL of the regulatory RNA (perform several reactions with regulatory RNA concentration in the range of 20–250 nM), 2 μL of Buffer TP5X− in a final volume of 10 μL. Adjust MgCl2 concentration to 8 mM with MgCl2100 solution and carry out the binding reaction at 37°C for 15 min. 3. Add uncharged initiator tRNAfMet (0.5 μL, 40 μM) to the binary complex and incubate for 5 min at 37°C. 4. Incubation controls should be done as in Subheading 3.3.6, step 4.

3.3.8. Reverse Transcriptase Extension and Sequencing

1. Just prior reverse transcription, increase MgCl2 concentration in the samples to 15 mM with the MgCl2100 solution. This step is essential for stabilizing the complexes formed with the 70S ribosomes from T. thermophilus. 2. Add to all samples (incubation controls and reactions) 1.5 μL dNTP500 mix (50 μM final concentration), 1 μL Buffer TP5X+, 1.5 μL sterile water and 1 μL MMLV RT (3 U) or 1 μL AMV RT (2 U). The extension reactions (final volume 15 μL) are incubated for 15 min at 37°C. At high concentration, MMLV enzyme is less sensitive to secondary structures which might be important for the gene expression control (Fig. 1a).

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting

261

3. To help ethanol precipitation, add 1 μL total tRNA to the samples that do not contain tRNA or ribosome. 4. Carry out phenol extraction of all samples by addition 110 μL of 300 mM Na-acetate and 120 μL phenol/chloroform/isoamylic alcohol (24:24:1) mixture. Vortex the samples for at least 1 min at 20°C, and centrifuge at 15,000 × g for 3 min at 4°C. Carefully take the aqueous upper phase, and add 110 μL of chloroform/isoamylic alcohol mixture. Mix the samples vigorously for 1 min at 20°C, and centrifuge for 3 min at 15,000 × g at 4°C (see Note 7). 5. Take the upper phase and add 2.5 volumes of cold ethanol. Mix solutions and keep them in dry ice for 30 min. Centrifuge the samples at 4°C at 15,000 × g to precipitate the labelled DNA fragments. Wash the pellets twice with 70% cold ethanol, vacuum dry and dissolve in 10 μL of DNA loading buffer. 3.4. Gel Fractionation

1. The RT pauses are identified by running a sequencing reaction in parallel with toeprinting samples (see Chapter “Structural Probing of RNA Thermosensors”). The primer extension step of the sequencing reactions is performed with AMV reverse transcriptase (see Subheading 3.3.8, step 2) in the presence of one dideoxyribonucleotide ddXTP (2.5 μM), the corresponding deoxyribonucleotide dXTP (25 μM) and three other desoxyribonucleotides (100 μM). 2. After RT primer extension, heat all samples at 90°C for 3 min and centrifuge briefly. 3. Load 3 μL per well on a 8% polyacrylamide-(0.4% bis)/8 M urea slab gel and perform PAGE in 1× TBE. Store the remaining samples at −20°C. The migration time should be adjusted to the size of the analyzed fragments, assuming that during 8% PAGE, xylene cyanol, and bromophenol blue migrate as 81 nucleotide and 19 nucleotide long DNA, respectively. After PAGE, transfer the gel on a dispensable thin support (i.e. an old autoradiograph), and expose at −80°C with X-ray film and intensifying screen overnight.

4. Notes 1. For safety reason, manipulation with pathogenic bacteria such as S. aureus must be carried out in a Biosafety Level 2 laboratory (according to the country rule) that contains equipment required for cell growth, lysis, and centrifugation of bacteria.

262

Fechter et al.

2. For convenience, prepare a 2× stock solution of Buffer DE and mix it with 50% sucrose solution to prepare Buffer S1 and Buffer S2. 3. Ribosomes and ribosome subunits are flash frozen in liquid nitrogen and stored at −80°C. 4. Electrophoresis of ribosomes on 1% agarose gel allows the visualization of 23S rRNA (component of 50S subunits) and 16S rRNA (component of 30S subunits) which should appear as distinct bands. 5. No more than three cycles of flash freezing–defreezing should be performed for the 30S subunits. 6. E. coli S15 was purified according to ref. 8, stored at 4°C and reactivated for 15 min at 37°C in a buffer containing 20 mM MgCl2, 270 mM KCl, 50 mM Tris–HCl, pH 7.5, 3 mM DTT, 0.02 mg/mL bovine serum albumin. 7. After primer extension, phenol extraction followed by ethanol precipitation is not usually necessary. Therefore, 10 μL of DNA loading Buffer can be directly added to all samples. The assays are then heated for 3 min at 90°C and 6 μL are loaded per well on 8% gel. However, the migration of the labelled cDNA fragments might be sometimes perturbed (see Fig. 1b, left part). Phenol extraction helps to get a better resolution on the gel.

Acknowledgements We are grateful to Anne-Catherine Helfer for the toeprinting assays performed on E. coli rpsO mRNA. This work was supported by financial support from the Centre National de la Recherche Scientifique (UPR 9002 CNRS), from the University Louis Pasteur of Strasbourg, from the Ministère de la Recherche (ANR05-MIIME, ANR07-BLANC), from the European Community (FOSRAK, EC005120; BacRNA EC018618).

References 1. Serganov, A. and Patel, D. J. (2007). Ribozymes, riboswitches and beyond: regulation of gene expression without proteins. Nat. Rev. Genet. 8, 776–790. 2. Romby, P. and Springer, M. (2007). Translational control in prokaryotes. In Translational Control in Biology and Medicine (Hershey, J, Sonnenberg, N, Metthews, M, eds), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 807–832.

3. Narberhaus, F., Waldminghaus, T. and Chowdhury, S. (2006). RNA thermometers. FEMS Microbiol. Rev. 30, 3–16. 4 . Coppins , R. L. , Hall , K. B. and Groisman , E. A. (2007). The intricate world of riboswitches . Curr. Opin. Microbiol. 10 , 176 – 181 . 5. Romby, P. and Springer, M. (2003). Bacterial translational control at atomic resolution. Trends Genet. 19, 155–161.

Ribosomal Initiation Complexes Probed by Toeprinting and Effect of trans-Acting 6. Schlax, P. J. and Worhunsky, D. J. (2003). Translational repression mechanisms in prokaryotes. Mol. Microbiol. 48, 1157–1169. 7. Ehresmann, C., Ehresmann, B., Ennifar, E., Dumas, P., Garber, M., Mathy, N. et al. (2004). Molecular mimicry in translational regulation: the case of ribosomal protein S15. RNA Biol. 1, 66–73. 8. Marzi, S., Myasnikov, A. G., Serganov, A., Ehresmann, C., Romby, P., Yusupov, M. et al. (2007). Structured mRNAs regulate translation initiation by binding to the platform of the ribosome. Cell 130, 1019–1031. 9. Darfeuille, F., Unoson, C., Vogel, J. and Wagner, E. G. (2007). An antisense RNA inhibits translation by competing with standby ribosomes. Mol. Cell 26, 381–392. 10. Vogel, J. and Wagner, E. G. (2007). Target identification of small noncoding RNAs in bacteria. Curr. Opin. Microbiol. 10, 262–270. 11. Hartz, D., McPheeters, D. S., Traut, R. and Gold, L. (1988). Extension inhibition analysis of translation initiation complexes. Methods Enzymol. 164, 419–425. 12. Hartz, D., McPheeters, D. S. and Gold, L. (1989). Selection of the initiator tRNA by Escherichia coli initiation factors. Genes Dev. 3, 1899–1912. 13. Huttenhofer, A. and Noller, H. F. (1994). Footprinting mRNA–ribosome complexes with chemical probes. EMBO J. 13, 3892–3901. 14. Sacerdot, C., Caillet, J., Graffe, M., Eyermann, F., Ehresmann, B., Ehresmann, C. et al. (1998). The Escherichia coli threonyltRNA synthetase gene contains a split ribosomal binding site interrupted by a hairpin structure that is essential for autoregulation. Mol. Microbiol. 29, 1077–1090. 15. Yusupova, G. Z., Yusupov, M. M., Cate, J. H. and Noller, H. F. (2001). The path of messenger RNA through the ribosome. Cell 106, 233–241. 16. Jenner, L., Romby, P., Rees, B., SchulzeBriese, C., Springer, M., Ehresmann, C. et al. (2005). Translational operator of mRNA on the ribosome: how repressor proteins exclude ribosome binding. Science 308, 120–123. 17. Korostelev, A. , Trakhanov, S. , Asahara , H. , Laurberg, M. , Lancaster, L. and Noller, H. F. (2007). Interactions and dynamics of the Shine Dalgarno helix in the 70S ribosome. Proc. Natl Acad. Sci. U. S. A. 104, 16840– 16843 . 18. Hartz, D., McPheeters, D. S., Green, L. and Gold, L. (1991). Detection of Escherichia coli ribosome binding at translation initiation sites in the absence of tRNA. J. Mol. Biol. 218, 99–105.

263

19. Philippe, C., Eyermann, F., Benard, L., Portier, C., Ehresmann, B. and Ehresmann, C. (1993). Ribosomal protein S15 from Escherichia coli modulates its own translation by trapping the ribosome on the mRNA initiation loading site. Proc. Natl Acad. Sci. U. S. A. 90, 4394–4398. 20. Ringquist, S., MacDonald, M., Gibson, T. and Gold, L. (1993). Nature of the ribosomal mRNA track: analysis of ribosome-binding sites containing different sequences and secondary structures. Biochemistry 32, 10254–10262. 21. Qin, Y., Polacek, N., Vesper, O., Staub, E., Einfeldt, E., Wilson, D. N. et al. (2006). The highly conserved LepA is a ribosomal elongation factor that back-translocates the ribosome. Cell 127, 721–733. 22. Waldminghaus, T., Heidrich, N., Brantl, S. and Narberhaus, F. (2007). FourU: a novel type of RNA thermometer in Salmonella. Mol. Microbiol. 65, 413–424. 23. Brunel, C., Romby, P., Moine, H., Caillet, J., Grunberg-Manago, M., Springer, M. et al. (1993). Translational regulation of the Escherichia coli threonyl-tRNA synthetase gene: structural and functional importance of the thrS operator domains. Biochimie 75, 1167–7119. 24. Sharma, C. M., Darfeuille, F., Plantinga, T. H. and Vogel, J. (2007). A small RNA regulates multiple ABC transporter mRNAs by targeting C/A-rich elements inside and upstream of ribosome-binding sites. Genes Dev 21, 2804–2817. 25. Boisset, S., Geissmann, T., Huntzinger, E., Fechter, P., Bendridi, N., Possedko, M. et al. (2007). Staphylococcus aureus RNAIII coordinately represses the synthesis of virulence factors and the transcription regulator Rot by an antisense mechanism. Genes Dev. 21, 1353–1366. 26. Milligan, J. F. and Uhlenbeck, O. C. (1989). Synthesis of small RNAs using T7 RNA polymerase. Methods Enzymol. 180, 51–62. 27. Romaniuk, P. J., de Stevenson, I. L. and Wong, H. H. (1987). Defining the binding site of Xenopus transcription factor IIIA on 5S RNA using truncated and chimeric 5S RNA molecules. Nucleic Acids Res. 15, 2737–2755. 28. Jahn, M. J., Jahn, D., Kumar, A. M. and Soll, D. (1991). Mono Q chromatography permits recycling of DNA template and purification of RNA transcripts after T7 RNA polymerase reaction. Nucleic Acids Res. 19, 2786. 29. Marzi, S., Fechter, P., Chevalier, C., Romby, P. and Geissmann, T. RNA switches regulate initiation of translation in bacteria. Biol. Chem.389, 585–598. 30. Novick, R. P. and Jiang, D. (2003). The staphylococcal saeRS system coordinates environmental signals with agr quorum sensing. Microbiology 149, 2709–2717.

Chapter 19 Isolation and Characterization of the Heat Shock RNA 1 Ilya Shamovsky and Evgeny Nudler Summary The heat shock (HS) response is the major cellular defense mechanism against acute exposure to environmental stresses. The hallmark of the HS response, which is conserved in all eukaryotes, is the rapid and massive induction of expression of a set of cytoprotective genes. Most of the induction occurs at the level of transcription. The master regulator, heat shock transcription factor (HSF, or HSF1 in vertebrates), is responsible for the induction of HS gene transcription in response to elevated temperature. Under normal conditions HSF is present in the cell as an inactive monomer. During HS, HSF trimerizes and binds to a consensus sequence in the promoter of HS genes, stimulating their transcription by up to 200fold. We have shown that a large, noncoding RNA, HSR1, and the translation elongation factor eEF1A form a complex with HSF during HS and are required for its activation. Key words: Heat shock, HSF, HSR1, eEF-1A, Activation

1. Introduction The heat shock response is an evolutionarily ancient, conserved reaction of cells to unfavorable environmental conditions. During heat shock response major changes in the pattern of gene expression occur, including the rapid induction of a specific set of genes. Most of these genes encode heat shock proteins (HSPs), a large family of molecular chaperones, which includes several subfamilies grouped according with functional characteristics and molecular weight of the proteins. HSPs are essential for cell survival under normal conditions and are critical for cell survival during stress. Elevated expression of HSPs is a major cause of cancer cell resistance to therapies (1, 2). In some instances, HSPs were found

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_19 © Humana Press, a part of Springer Science + Business Media, LLC 2009

265

266

Shamovsky and Nudler

to be critical for malignant transformation due to their ability to stabilize mutant, metastable forms of oncoproteins, such as Ras or p53 (3). Expression of HSPs is regulated predominantly at the level of transcription by heat shock transcription factor (HSF) (4, 5). HSF is ubiquitously expressed as an inactive monomer. The activation of HSF during HS requires its trimerization, nuclear translocation, binding to a cognate DNA sequence in the promoter of HS genes, phosphorylation, and, finally, activation of transcription by rescue of transcriptionally engaged, arrested RNA polymerase II molecules in the 5¢ coding region of most of HSP genes (6, 7). We have shown that the first steps in the activation of HSF in response to HS require two positive regulators: a large, noncoding RNA, heat shock RNA 1 (HSR1), which provides thermosensing capabilities, and translation elongation factor eEF1A, which promotes trimerization of HSF (8). HSR1 is isolated by the affinity fractionation of a lysate of heat-shocked cells using a Sepharose column with covalently attached HSF. The functional characterization of HSR1 is performed in a reconstituted in vitro system using recombinant HSF1 and eEF1A isolated from HeLa cells. Since HSF spontaneously trimerizes at high concentrations, it is expressed and purified by procedures that keep its concentration low. Activation of HSF is determined by electrophoretic mobility shift assays (EMSAs).

2. Materials 2.1. Purification of Recombinant HSF

1. LB/Amp medium: Luria–Bertani (LB) broth supplemented with 50 mg/mL ampicillin. 2. Bacterial shaking incubator, 2-L flasks. 3. The TOP-10 E. coli cells transfected with GST-HSF1 plasmid. 4. Isopropyl b-D-1-thiogalactopyranoside (IPTG): 1 M solution. 5. Lysis buffer: 20 mM HEPES–Na, pH 7.9, 0.42 M NaCl, 0.5 mM ethylenediaminetetraacetic acid (EDTA), 0.5 mM dithiothreitol (DTT) (add just before use from 1 M stock), 10% (v/v) glycerol. 6. ATP-washing buffer: Lysis buffer containing 5 mM ATP and 25 mM MgCl2. 7. Thrombin cleavage buffer: Lysis buffer containing 0.0025% NP-40. 8. Glutathione-Sepharose 4B and thrombin (1 U/mL) (GE Healthcare). 9. Roche Complete Mini protease inhibitors cocktail (Roche, Basel, Switzerland).

Isolation and Characterization of the Heat Shock RNA

267

10. 0.1 M solution of PMSF in ethanol or isopropanol. 11. Glycine Running buffer (10×): 250 mM Tris–HCl, pH 8.3, 1.9 M glycine, 0.5% (w/v) SDS. 12. 0.375% Sodium dodecyl sulfate (SDS) polyacrylamide gels for electrophoresis (SDS–PAGE) using 1× Glycine Running buffer. The gels could be purchased or prepared according to Laemmli, with minor modifications (9). 13. Prestained molecular weight markers (Fermentas, Glen Burnie, MD). 14. Coomassie R-250 staining solution: 50% methanol, 10% acetic acid, 0.2% Coomassie R-250. Prepare Coomassie staining solution by first dissolving the dye powder in methanol (stir on a magnetic stirrer for at least 1 h), then add acetic acid and water. 15. Destaining solution: 10% methanol, 10% acetic acid. 16. Beckman Coulter Avanti J-25 centrifuge with JA-10 and JA-17 rotors or equivalent. 17. Sonicator (Branson 250 with microtip). 2.2. Preparation of HSF-Sepharose Column

1. Crosslinked Sepharose CL-4B (Sigma). 2. PD-10 disposable desalting columns (GE Healthcare). 3. Sodium m-periodate (Sigma). 4. Solid sodium borohydride (Sigma) and its solutions: 0.2 M, pH 9.0; 0.1 M, pH 8.0. 5. Disposable empty plastic columns (~12 mL total capacity) (Bio-Rad). 6. Ethanolamine solution: 0.1 M ethanolamine–HCl, pH 8.0, 0.5 M NaCl. 7. Bradford reagent (Bio-Rad).

2.3. Purification of eEF1A from S-100 Lysate

1. S-100 lysate (D. Reinberg laboratory, New York University). 2. Buffer A: 25 mM Tris–HCl, pH 7.5, 50 mM KCl, 1 mM DTT, 0.1 mM EDTA, 25% (v/v) glycerol. 3. Buffer B: 25 mM Tris–HCl, pH 7.9, 50 mM KCl, 1 mM DTT, 0.1 mM EDTA, 10% (v/v) glycerol. Buffer B650: buffer B with 650 mM KCl. 4. Diethylaminoethyl (DEAE) Sepharose column (~40 mL bed volume). 5. Phosphocellulose P11 column (~15 mL bed volume) prepared as described at http://mitchison.med.harvard.edu/protocols/ tubprep.html. 6. FPLC system (preferably) or peristaltic pump, UV-detector connected to a recorder, and fraction collector. 7. Centricons-10 cartridges (Amicon/Millipore, Billerica, MA).

268

Shamovsky and Nudler

2.4. Cell Culture and HSR Purification 2.4.1. HSR Purification

1. HeLa, BHK, or 3T3 cells. 2. Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen, Carlsbad, CA), supplemented with 10% fetal bovine serum (Gemini Bio-Tech), 2 mM glutamine, and antibiotic/ antimycotic cocktail (Invitrogen, Carlsbad, CA). 3. T-75 phenolic cap flasks and teflon cell scrapers (Corning). 4. Water bath large enough to accommodate 8–9 T-75 tissue culture flasks, set to 43°C. 5. 1× PBS: 8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4, 0.24 g KH2PO4 in 800 mL of water. Adjust the pH to 7.4 with HCl. Add water to 1 L. Sterilize by autoclave. 6. HEDGM buffer: 20 mM HEPES–Na, pH 7.9, 0.5 mM EDTA, 0.5 mM DTT, 1.5 mM MgCl2, 10% (v/v) glycerol, 0.0025% NP-40 (Igepal CA-630) (Sigma). 7. Dry ice/ethanol or liquid nitrogen bath for flash freezing. 8. Refrigerated centrifuge accepting Eppendorf tubes and capable of generating forces of at least 20,000 × g. 9. Tris–EDTA (TE) buffer: 10 mM Tris–HCl, pH 7.9, and 1 mM EDTA. 10. Reagents for proteinase K digestion: proteinase K, 10% (w/v) SDS, 0.5 M EDTA, pH 8.0. 11. Phenol/chloroform/isoamyl alcohol (25:24:1), pH adjusted to 8.0 with Tris–HCl. 12. Reagents for ethanol precipitation: 0.5 mg/mL glycogen, 100% ethanol, 3 M Na-acetate, pH.

2.4.2. RNA Denaturing Polyacrylamide Gel Electrophoresis

1. TBE buffer (10×): 0.89 M Tris–borate, pH 8.3, 5 mM EDTA. 2. 4% Acrylamide/N,N¢-bis-acrylamide (19:1) in 1× TBE/8 M urea. Store at 4°C, heat at 37°C to dissolve urea (see Note 1). 3. Ammonium persulfate: prepare 20% (w/v) solution in water. Store up to 4 weeks in the dark at +4°C. 4. N,N,N¢,N¢-Tetramethyl-ethylenediamine (TEMED). 5. 7 × 10 cm glass plates (Bio-Rad mini gel format), 1-mm thick comb and spacers. 6. RNA-loading buffer: 80% deionized formamide, 0.02% bromophenol blue, 0.02% xylene cyanol, 0.1% SDS in 1× TBE. 7. RNA silver staining solutions. Fixation solution: 7.5% acetic acid; silver impregnation solution: 0.15% (w/v) AgNO 3, 0.056% (v/v) formaldehyde; developer solution: 0.056% formaldehyde, 400 mg/L sodium thiosulfate in 3% Na2CO3.

Isolation and Characterization of the Heat Shock RNA

2.5. In Vitro Transcription of HSR1

269

1. pBluescript SK-vector with cloned HSR1 (pBSKM-HSR1). 2. 5× Transcription buffer (Promega). 3. 100 mM rNTP (Roche). Store at −20°C. 4. Pyrophosphatase, T7 RNA polymerase, RNase-free DNase I. Store at −20°C.

2.6. Electrophoretic Mobility Shift Assay

1. 40% (w/v) acrylamide/N,N¢-methylene-bis-acrylamide (19:1) solution, filtered through a 20-mm filter. Store at 4°C. 2. 20 × 20 cm glass plates, 0.75-mm thick comb and spacers. 3. TG buffer (10×): 250 mM Tris–HCl, pH 8.3, 1.9 M glycine, 0.5 mM EDTA. Store at room temperature. 4. Poly(dI–dC) solution (Sigma), nonspecific RNA carrier: dissolve in TE buffer at 1.25 mg/mL. Heat at 42°C for 15 min, aliquot, and store at −20°C. 5. EMSA binding buffer (10×): 200 mM HEPES–Na, pH 7.9, 1 M NaCl, 40 mM MgCl2, 2 mM DTT. Store in aliquots at −20°C. 6. HSE oligonucleotides: GCC TCG AAT GTT CGC GAA GTT TCG and its reverse complement (IDT DNA Technologies, Coralville, IA). 7. T4 polynucleotide kinase (PNK) supplied with 10× PNK reaction buffer (New England Biolabs, Ipswich, MA). 8. [g-32P]ATP, >6,000 Ci/mmol (MP Biomedicals, Solon, OH). 9. Micro Bio-Spin disposable columns (Bio-Rad, Hercules, CA). 10. PCR cycler (e.g., Eppendorf MasterCycler). 11. Loading dye solution: 0.02% (w/v) bromophenol blue, 25% glycerol. 12. UV measuring buffer: 6 M guanidine–HCl, 20 mM phosphate buffer, pH 6.5.

3. Methods Purified HSF trimerizes spontaneously at high concentration (10, 11). Therefore the conditions for HSF expression and purification are selected to minimize the spontaneous activation of HSF by keeping its concentration relatively low. Just before coupling to periodate-activated Sepharose the HSF is exchanged into high-pH borate buffer, which also helps to prevent spontaneous trimerization by keeping HSF molecules charged. The monomeric state of HSF to be bound to the activated Sepharose is essential for the successful capture of the eEF1A–HSR1 complex from the HS cell lysate.

270

Shamovsky and Nudler

Once HSR has been isolated and cloned (by a combination of 3¢ and 5¢ RACE or by another strategy of choice), it can be synthesized in large amounts by in vitro transcription by phage RNA polymerase. Of all commercially available RNA polymerases, T7 RNA polymerase provides the greatest yield. The yield tends to be higher when transcribing from a linearized plasmid template, but care should be taken to avoid the introduction of as fewer extra nucleotides from the plasmid into the resulting transcript as possible. The active trimeric form of HSF binds to the consensus heat shock element (HSE) sequence with high affinity. HSE is defined as an array of repeats of the sequence 5¢-nGAAn-3¢ in head-to-tail orientation. A minimum of three HSE units is required for the high affinity binding of HSF (12–14). We routinely use four-element arrays of the consensus HSE element as a probe in gel-shift experiments involving HSF activation. The in vitro activation is performed by incubating the RNA at the desired temperature, followed by the addition of the proteins (HSF and eEF1A) and the probe. The level of HSF activation is assayed by EMSA. The activation is conveniently performed in a PCR cycler, which allows for the fast and precise control of the incubation temperature. To screen antisense oligonucleotides, the oligonucleotide to be tested is premixed with HSR1 at 5- to 10-fold molar excess. 3.1. Purification of Recombinant HSF

1. Inoculate a single colony of TOP-10 E. coli cells transfected with GST-HSF1 plasmid from LB/Amp 1.2% agar plate into 3–5 mL of LB/Amp medium and grow overnight in the shaker at 37°C. 2. Dilute the overnight starter culture 1:100–1:200 into LB/ Amp medium and grow at 37°C in 2-L flasks until the OD600 reaches 0.6. 3. Remove the culture from the shaker and chill on ice or in the cold room for 10–15 min. In the meantime bring the temperature of the shaker to 25–28°C. 4. Add IPTG to a final concentration of 0.1 mM and return the culture to the shaker to allow GST-HSF1 synthesis to proceed for 4–6 h at 25–28°C. 5. Collect the cells by centrifugation at 5,000 × g for 15 min and discard the supernatant. At this point the cells may be frozen at −80°C for later use. 6. Resuspend the bacteria in 20 mL of the lysis buffer per 1 L of original culture. Add 1 tablet of Roche Complete Mini Protease Inhibitor Cocktail per 1 L of original culture. 7. Sonicate cells for 25 min with output control set to 6 and duty cycle of 60% with constant chilling in the ice-water bath (see Note 2). The lysate should become clearer and significantly darker in color.

Isolation and Characterization of the Heat Shock RNA

271

8. Transfer the lysate to the centrifuge tube (which fits Beckman JA-17 rotor) and centrifuge for 30 min at 35,000 × g at 4°C. During the centrifugation, prepare GSHSepharose as follows: resuspend the GSH-Sepharose by vigorous shaking of the bottle and transfer 1.33 mL of the suspension for every liter of original bacterial culture into a 15-mL screw-cap tube. 9. Wash twice with water and twice with the lysis buffer by resuspension/centrifugation at 500 × g. 10. Add 1 mL of the lysis buffer for every 1.33 mL of the initial GSH-Sepharose slurry. 11. Carefully remove the supernatant and place it in the new 50-mL screw-cap tube. Add washed GSH-Sepharose and incubate on the rotator for 30 min in the cold room. 12. Collect the beads by centrifugation at 500 × g, carefully remove and discard the supernatant, add 10 mL of the lysis buffer, mix and transfer the suspension to the new 15-mL screw-cap tube. 13. Wash the beads successively 3 times with 10 volumes of the lysis buffer by resuspension/centrifugation at 500 × g. 14. Resuspend the beads in 10 volumes of the ATP-washing buffer and incubate 15 min at room temperature. Repeat the wash two more times. 15. Wash the beads with 10 volumes of the thrombin cleavage buffer. Resuspend in 1 volume of the same buffer. 16. Add 50 U thrombin per each 1 mL of settled bead volume and incubate on the rotator overnight at room temperature. 17. Check the thrombin cleavage efficiency by measuring the protein concentration in the supernatant and/or analyzing both supernatant and beads suspension by SDS–PAGE using prestained molecular weight markers. After SDS–PAGE, stain the gel with Coomassie R-250 staining solution. Visualize the protein bands by shaking in destaining solution. 18. Stop the cleavage by adding 1/100 volume of PMSF. Centrifuge the beads, remove and save the supernatant, and wash the beads twice with 1–2 volume(s) of the thrombin cleavage buffer (these washes should contain a significant amount of the cleaved HSF1 as well). 3.2. Preparation of the HSF-Sepharose column

This procedure covalently couples HSF to the activated Sepharose via reaction of the e-amino groups of lysine residues in the protein with reactive aldehyde groups of the periodate-treated Sepharose. The mild oxidation of Sepharose by periodate results in conversion of 1,2-cis-diol groups to the corresponding dialdehydes. The aldehydes then react with the amino groups to yield Schiff’s bases, which are stabilized by mild reduction with sodium borohydrate.

272

Shamovsky and Nudler

3.2.1. Preparation of Periodate-Activated Sepharose CL-4B

1. Dispense ~1 mL Sepharose CL-4B resin (~1.33 mL 75% slurry in ethanol) into a 15-mL screw-cap tube. 2. Wash twice with 10 mL of water by centrifugation at 500 × g followed by resuspension. 3. Incubate for 2 h with 5 mL of 20 mg/mL sodium periodate at room temperature in the dark. 4. Wash twice with 10 mL of 0.2 M sodium borate, pH 9.0, by centrifugation followed by resuspension. Store at +4°C for up to 1 month.

3.2.2. Binding of HSF to Activated Sepharose Column

1. Equilibrate a PD10 desalting column with ~30 mL 0.2 M sodium borate, pH 9.0. 2. Apply all HSF1 (maximum 2 mL) prepared according to Subheading 3.1. 3. Wash the column repeatedly (10 times) with 1 mL of 0.2 M sodium borate, pH 9.0, collecting 1-mL fractions into Eppendorf tubes. 4. Assay the protein content of the fractions by dropping 5 mL from each fraction on a piece of Parafilm and mixing it with 20 mL Bradford reagent. Intense blue color indicates the presence of the protein. 5. Combine the fractions containing protein (usually fractions 3–5), save a ~50-mL aliquot and add the pooled fractions to the periodate-activated Sepharose. Incubate overnight on a rotator at room temperature. 6. Transfer the slurry into a disposable plastic column and wash with 10 volumes of 0.2 M sodium borate, pH 9.0, followed by 10 volumes of 0.1 M sodium borate, pH 8.0. 7. Dissolve preweighed sodium borohydride to yield a final concentration of 3 mg/mL in 0.1 M sodium borate, pH 8.0, and immediately add to the column. Plug the outlet of the column, resuspend the resin with a 1-mL pipetor, and incubate it for 5 min at room temperature. 8. Open the column outlet and wash the column with ~10 volumes of 0.1 M sodium borate until the pH of the eluate reaches 8 as measured by pH paper. If the flow stops due to the bubbles of hydrogen, resuspend the resin by pipetting it up and down. This should resume the flow by removing most bubbles from the solution. 9. Resuspend HSF1-Sepharose in ~5 volumes of ethanolamine solution and store at +4°C.

3.3. Purification of eEF-1A from S-100 Lysate

The S-100 lysate, which is the starting material for this procedure, can be obtained commercially or prepared according to the published method (15). This protocol is an adaptation of

Isolation and Characterization of the Heat Shock RNA

273

the original procedure described by Merrick (16). With minimal modifications, the procedure can be easily adapted to use tissue, for example rat liver, as the source of the protein. This protocol assumes that 50 mL of the S-100 lysate is used; the column and gradient volumes can be scaled up or down if greater or lesser amounts are used. 1. Load 50 mL of the S-100 lysate to a 40-mL DEAE-Sepharose column equilibrated with buffer A. Collect the flow-through, which contains eEF-1A. 2. Load the flow-through to a 15-mL phosphocellulose P11 column equilibrated with buffer B and wash the column with at least 3 volumes of the same buffer. 3. Apply a linear gradient of 50–650 mM KCl (10 column volumes) using buffers B and B650 and collect 3 mL fractions. Assay a 20-mL aliquot from each fraction by electrophoresis in a 10% SDS–PAGE and stain the gel with Coomassie. eEF-1A should be visible as a prominent band at ~50 kDa. 4. Pool the fractions containing eEF-1A, concentrate them using a Centricon or similar ultrafiltration device with a 10-kDa MWCO membrane to ~2 mL, and exchange into the storage buffer. The protein is stable at 4°C for at least a month or indefinitely at −80°C. 3.4. Isolation of HSR from Heat-Shocked Cells

This procedure has been used extensively with cells grown in monolayer, such as HeLa, BHK, or 3T3. However, it should be easily adaptable to the cells grown in suspension, if the ratio of cells to HSF-Sepharose is kept constant. A typical example of the isolation of HSR1 is shown in Fig. 1.

3.4.1. Isolation of HSR

1. Grown monolayer cells are in T-75 phenolic cap flasks in DMEM medium to ~75–90% confluency. 2. Tighten the caps of flasks to be heat shocked and submerge them in a water bath set to 43°C. Sealing the caps with Parafilm is not required when using phenolic caps that have been tightened properly. Allow heat shock to proceed for 45 min to 1 h. 3. Remove flasks from the water bath and pat them dry with paper towels to prevent water from entering the flasks. Aspirate the medium using a Pasteur pipet connected to a vacuum source. Rinse the cells once with ice-cold PBS (3 mL per T-75 flask) taking care not to dislodge any cells, and use a cell scraper to scrape the cells into 3 mL of fresh ice-cold PBS. 4. Collect the cells by centrifugation for 5 min at 500 × g. Remove PBS by aspiration. 5. Resuspend the cell pellet in 100 mL per T-75 flask of Lysis buffer. Incubate on ice for 10 min.

2

3

4

5

6

7

kDa

1 2 3 4 5 6

kDa

196 118

116

90

66.2

70 55

EF-1A 38

RNase A

1

DNase I

Shamovsky and Nudler

Input RNA

274

45

EF-1A

35 25 18.4

10% Laemmli SDS-PAGE

10% MOPS NuPAGE

BHK

HeLa

4% urea-TBE PAAG

Fig. 1. Isolation of HSF1-activating fraction and HSR1 from a lysate of heat-shocked HeLa (middle panel) and BHK (left panel) cells. Lane 1 (BHK and HeLa): lysate of heat-shocked cells; lane 2 (BHK): supernatant after incubation of the lysate with HSF1-Sepharose beads; lanes 3 (BHK) and 2 (HeLa): HSF1-Sepharose beads after incubation with a lysate of heatshocked cells and washes; lanes 4 (BHK) and 3 (HeLa): HSF1-Sepharose beads after three successive elutions at 43°C; lanes 5–7 (BHK) and 4–6 (HeLa): three successive elutions with buffer at 43°C. Right panel: silver-stained PAAG of RNA isolated from fractions 5–7 shown in left panel (BHK cells). Where indicated, RNA was treated with DNase I or RNase A prior to loading the gel.

6. Lyse the cells by three cycles of flash freezing in ethanol/ dry ice or liquid nitrogen/thawing at room temperature. Thawing can be performed at 37°C if the incubation time does not exceed the minimum required to thaw the cells completely. 7. Centrifuge the lysate for 15 min at 25,000 × g at 4°C. Discard the pellet. In the meantime prepare the HSF1Sepharose beads from Subheading 3.2 (15 mL per T-75 flask) by washing them once in lysis buffer containing 0.21 M NaCl. 8. Dilute the cell lysate 2-fold with HEDGM buffer (to give the final salt concentration of 0.21 M) and add it to the washed HSF-Sepharose. Incubate for a minimum of 4 h (or overnight) on a rotator at 4°C. 9. Wash the beads 3 × 15 min with 10 bed volumes of HEDGM buffer containing 0.21 M NaCl. 10. Resuspend the beads in an equal volume of HEDGM/0.21 M buffer and incubate 45 min at 43°C with occasional mixing. Centrifuge 1 min at maximum speed in a tabletop centrifuge, remove and save the supernatant, and add one bed volume of HEDGM/0.21 M buffer preheated to 43°C. Repeat two more times.

Isolation and Characterization of the Heat Shock RNA

275

11. Combine the resulting fractions saving a 15-mL aliquot for SDS–PAGE. Add to final concentrations: SDS to 0.25%, EDTA to 12.5 mM, and proteinase K to 0.2 mg/mL. Incubate at 50°C for 30 min. 12. Extract twice with phenol/chloroform/isoamyl alcohol. Precipitate the sample in the presence of 0.5 mg/mL glycogen by adding 3 volumes of 100% ethanol and 0.1 volume of 3 M sodium acetate. Place the sample at −20°C overnight or at −80°C for 30 min. Centrifuge and dissolve the RNA pellet in 20 mL TE buffer. 3.4.2. RNA Gel Electrophoresis and Silver Staining of the Gel

1. Prepare a gel assembly using 7 × 10 cm glass plates (Bio-Rad minigel format), 1-mm thick comb and spacers. Prepare a denaturing gel by mixing 10 mL 4% acrylamide/N,N¢bis-acrylamide (19:1) in 1× TBE/8 M urea, 0.025 mL APS, and 0.01 mL TEMED. Pour the gel onto the glass plates and let polymerize for at least 30 min. Preheat the gel for 30 min using 1× TBE as running buffer. 2. Mix 2 mL of the RNA with 3 mL of RNA-loading buffer, heat at 80°C for 5 min, load on the gel, and conduct electrophorese until the xylene cyanol (the second dye) has migrated to the end of the gel (see Note 3). 3. Soak the gel for 30 min in 7.5% acetic acid. Rinse 3× for 5 min in water. 4. Soak the gel for 30 min in the silver impregnation solution. Rinse briefly with copious amount of water. 5. Soak the gel in the development solution and watch for the bands to appear (usually, within 5–10 min). 6. Stop the development by pouring off the developer and soaking the gel in 7.5% acetic acid.

3.5. Synthesis of HSR1 by In Vitro Transcription

HSR1 is cloned in the pBluescript SK- vector (pBSKM-HSR1) in an orientation such that T7 RNA polymerase generates the sense transcript after digestion of the plasmid with SmaI and T3 RNA polymerase generates the antisense transcript after digestion of the plasmid with EcoRV. It is essential to cleave the plasmid with restriction enzymes that produce blunt ends, if possible, since this will minimize nonspecific transcription initiation on protruding single-stranded ends. Alternatively, a PCR product containing the T7 promoter sequence in one of the primers can be used as template to minimize the introduction of extra nucleotides in the RNA sequence. In vitro activation of HSF1 in the described system and its use to screen antisense HSR1 oligos is illustrated in Fig. 2. 1. Digest 20 mg pBSKM-HSR1 with the appropriate enzyme in a 50-mL reaction for 2–4 h.

T3

HeLa

BHK

T7

T3

− +

T7

− −

HSR1: EF-1A:

BHK

Shamovsky and Nudler HeLa

276

−

−

−

−

+

+

+

+

HSF-1

HSR1 (604 nt)

5’

3’ antisense DNA oligonucleotides

−F - 1

2

3

4

5

6

7

8

9 10 11 12 13 14 15

Fig. 2. In vitro reconstituted system for the activation of HSF1. Upper panel: recombinant mouse HSF1 (10 nM final) and eEF-1A (100 nM final, omitted as shown) were incubated in the absence of RNA, in the presence of HSR1 isolated from either BHK or HeLa cells, as indicated, in the presence of (25 nM final) in vitro transcribed sense (T3) or antisense (T7) mammalian HSR. Reactions were then incubated with the radiolabeled HSE oligo and separated on a native PAAG, as described in the text. The gel was dried and radioautographed. Lower panel: HSR1 antisense oligos screening. EMSA reactions were assembled as described earlier with the exception of the inclusion of antisense oligos at a 10-fold molar excess over HSR1.

2. Purify the digested plasmid or PCR fragment by agarose gel electrophoresis followed by phenol/chloroform extraction and ethanol precipitation. 3. Resuspend the digested plasmid or PCR fragment in ~20 mL TE. Prepare the transcription reaction by mixing (follow the order of reagent addition): water to 200 mL, 40 mL 5× transcription buffer, 10 mL 100 mM DTT, 30 mL 25 mM rNTP, template 50 mg/mL, pyrophosphatase to 1 U/mL, T7 RNA polymerase to 1,000–2,000 U/mL. 4. Incubate 4–6 h at 37°C. After 2 h incubation, another aliquot of RNA polymerase may be added.

Isolation and Characterization of the Heat Shock RNA

277

5. Add RNase-free DNase I (40–50 U) and incubate for 30 min at 37°C. 6. Precipitate the RNA by your method of choice (e.g., phenol/ chloroform extraction followed by ethanol or isopropanol precipitation, or LiCl precipitation, etc.). 3.6. In Vitro HSF Activation Assay

The typical range of concentrations in which HSF is activated in this system is 1–5 nM. Higher concentrations, especially in excess of 100 nM, tend to cause spontaneous trimerization and activation of HSF. Because of the fluctuations in the quality of the recombinant HSF preparations, it may be necessary to titrate each new preparation of HSF to determine the concentration at which it is not activated spontaneously. The final concentration of eEF1A in the assay is 100 nM. The concentration of both proteins is determined by UV absorption at 280 nm in UV measuring buffer (eHSF = 28,670 L/M/cm and eeEF1A = 45,755 L/M/cm). HSR1 is used in the assay at low (0.1 nM) concentration, because at higher concentrations the equilibrium concentration of the HSF-activating conformation of HSR1 is high enough to cause activation to occur independent of incubation temperature. However, for the antisense oligo screen, this “constitutive” activation is beneficial as it considerably simplifies the experimental setup. 1. Prepare the radioactively labeled HSE oligonucleotide by kinase reaction. In two separate tubes mix the following components: 27 mL water, 5 mL sense (or antisense) HSE oligonucleotide, 4 mL 10× PNK buffer, 3 mL [g-32P]ATP, and 1 mL PNK. Incubate at 37°C for 30 min. 2. Mix the content of the two tubes, incubate at 68°C for 10 min and let cool slowly to room temperature. 3. Purify the labeled HSE oligonucleotide by centrifugation through a Micro Bio-Spin 6 spin column. Store at −20°C for up to 2 weeks. 4. Prepare a 0.75-mm thick 4% PAAG/1× TG gel for electrophoresis by mixing 40 mL 4% PAAG/1× TG gel solution, 0.05 mL APS, and 0.015 mL TEMED. Polymerize for at least 30 min. 5. Combine 2.5 mL EMSA 10× mix, 0.75 mL poly(dI–dC), 0.25 mL 1 M DTT, and water according to the desired number of samples. The final volume of the reaction mixture after the addition of RNA and protein components should be 25 mL. Distribute into PCR tubes (0.2 mL) and equilibrate at 23°C in a PCR cycler. 6. Add RNA to the tubes and incubate at 23°C for 5 min. If performing an antisense oligo screen, add the oligonucleotide at 10-fold molar excess over HSR1. 7. Incubate at the HS temperature (43°C, or at 23°C if performing an antisense oligo screen) for 15 min.

278

Shamovsky and Nudler

8. Bring the temperature back to 23°C; add eEF-1A and HSF1 as quickly as possible and mix well by vigorous pipetting. Incubate at 23°C for 30 min. 9. Add 3 mL of 32P-HSE/loading dye mix and incubate for 20 min at 23°C. 10. Load reactions onto the gel and conduct electrophoreses in 1× TG at no more than 20 mA until the dye migrates to bottom of the gel (~3 h for a 20 × 20 cm gel). 11. Dry the gel and expose to X-ray film or phosphoimager screen. 12. The activation of HSF1 is observed as a diffuse band migrating at about 2 cm from the top of the gel. Normally, no or very little HSF1 activation should be seen in the samples without RNA or without heat treatment. The presence of HSF1 signal in the samples without RNA indicates that the concentration of HSF1 is too high and should be titrated down. The existence of HSF1 signal in the samples without heat treatment indicates that the concentration of RNA is too high and there is enough active conformation of HSR1 to cause constitutive HSF1 activation. In this case, RNA needs to be titrated down.

4. Notes 1. Avoid heating urea-containing solutions to high temperature, as urea will decompose. 2. The efficient cooling of the sample during sonication is essential. It is best achieved if the sonication is done in a metal vial (we routinely use tube adapters from an older Beckman ultracentrifuge) to ensure rapid and efficient dissipation of heat. 3. The volumes and incubation times are provided for a staining procedure for a 4% gel that was run using Bio-Rad mini electrophoresis apparatus. They should be adjusted accordingly for larger gels and/or higher concentration of acrylamide. Larger 4% gels may be difficult to handle.

Acknowledgment This work was supported by the NIH grant R01 GM069800 (E.N.).

Isolation and Characterization of the Heat Shock RNA

279

References 1 . Ciocca , D.R. and Calderwood , S.K. (2005) Heat shock proteins in cancer: diagnostic, prognostic, predictive, and treatment implications . Cell Stress Chaperones 10 , 86 – 103 . 2. Calderwood, S.K., Khaleque, M.A., Sawyer, D.B., and Ciocca, D.R. (2006) Heat shock proteins in cancer: chaperones of tumorigenesis. Trends Biochem. Sci. 31 , 164–172. 3. Whitesell, L. and Lindquist, S.L. (2005) Hsp90 and the chaperoning of cancer. Nat. Rev. Cancer 5 , 761–772. 4. Voellmy, R. (1996) Sensing stress and responding to stress. EXS 77 , 121–137. 5. Morimoto, R.I., Kroeger, P.E., and Cotto, J.J. (1996) The transcriptional regulation of heat shock genes: a plethora of heat shock factors and regulatory conditions. EXS 77 , 139–163. 6. Westwood, J.T. and Wu, C. (1993) Activation of Drosophila heat shock factor: conformational change associated with a monomerto-trimer transition. Mol. Cell. Biol. 13 , 3481–3486. 7 . Guettouche , T. , Boellmann , F. , Lane , W.S. , and Voellmy, R. (2005) Analysis of phosphorylation of human heat shock factor 1 in cells experiencing a stress . BMC Biochem . 6 , 4 . 8. Shamovsky, I., Ivannikov, M., Kandel, E.S., Gershon, D., and Nudler, E. (2006) RNAmediated response to heat shock in mammalian cells. Nature 440 , 556–560.

9. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage t4. Nature 227 , 680–685. 10. Zhong, M., Orosz, A. and Wu, C. (1998) Direct sensing of heat and oxidation by Drosophila heat shock transcription factor. Mol. Cell. 2 , 101–108. 11. Goodson, M.L. and Sarge, K.D. (1995) Heatinducible DNA binding of purified heat shock transcription factor 1. J. Biol. Chem. 270 , 2447–2450. 12. Orosz, A., Wisniewski, J., and Wu, C. (1996) Regulation of Drosophila heat shock factor trimerization: global sequence requirements and independence of nuclear localization. Mol. Cell. Biol. 16 , 7018–7030. 13. Wu, C. (1995) Heat shock transcription factors: structure and regulation. Annu. Rev. Cell. Dev. Biol. 11 , 441–469. 14. Abravaya, K., Phillips, B., and Morimoto, R.I. (1991) Heat shock-induced interactions of heat shock transcription factor and the human hsp70 promoter examined by in vivo footprinting. Mol. Cell. Biol. 11 , 586–592. 15. Dignam, J.D., Lebovitz, R.M., and Roeder, R.G. (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 11 , 1475–1489. 16. Carvalho, J.F., Carvalho, M.D., and Merrick, W.C. (1984) Purification of various forms of elongation factor 1 from rabbit reticulocytes. Arch. Biochem. Biophys. 234 , 591–602.

Chapter 20 Analysis of tRNA-Directed Transcription Antitermination in the T Box System In Vivo Tina M. Henkin Summary Regulation of gene expression in bacteria by cis-acting RNA elements can be investigated both in vivo and in vitro. Analyses in vivo can focus on changes in mRNA transcript levels or in protein production. Systems that are regulated at the level of premature termination of transcription are best analyzed by monitoring expression of a fusion to an easily assayable reporter gene construct or by direct measurement of the terminated and readthrough transcripts. These experimental approaches are described in the context of the Bacillus subtilis T box mechanism, which responds to uncharged tRNA as the effector, and are readily adaptable to other regulatory systems that respond to other signal molecules, and other experimental systems. Key words: Transcription attenuation, Antitermination, tRNA, Northern analysis, Reporter genes

1. Introduction The T box gene regulation mechanism represents a novel system in which binding of a specific uncharged tRNA to the nascent RNA transcript directs readthrough of a transcription termination signal located in the 5¢ region of the transcript, upstream of the start of the regulated coding sequence (1). The cis-acting regulatory region of the transcript is designated the “leader RNA” or 5¢-untranslated region (5¢-UTR). The T box mechanism is commonly used to regulate amino acid-related genes in Gram-positive bacteria, most prominently in the Firmicutes, and the regulated genes encode aminoacyl-tRNA synthetases, amino

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_20 © Humana Press, a part of Springer Science + Business Media, LLC 2009

281

282

Henkin

acid biosynthesis enzymes, and transporters (2). Multiple genes in a single organism can be regulated independently by the T box mechanism, with each transcriptional unit responding specifically to a tRNA that matches the amino acid specificity of the regulated genes. Genes regulated by the T box mechanism can be predicted by identification of a series of conserved sequence and structural elements in the leader region. These elements include a series of complex helical domains, designated Stem I, Stem II, the Stem IIA/B pseudoknot, Stem III, and an intrinsic transcriptional terminator, comprising a helix immediately followed by a run of uridines (1, 3). Sequences from the 5¢ side of the terminator helix can also pair with an upstream conserved sequence element to form a competing antiterminator helix. An internal bulge in Stem I (the Specifier Loop) contains a triplet sequence (the Specifier Sequence) that serves as a primary determinant for recognition of the appropriate tRNA effector, via codon–anticodon pairing (1). Four residues within a bulge of the antiterminator are responsible for a second pairing with the acceptor end of the tRNA; this second pairing occurs only with uncharged tRNA, allowing the system to respond specifically to uncharged tRNA (4). The T box mechanism was first characterized in depth in vivo, using the Bacillus subtilis tyrS gene, which encodes tyrosyltRNA synthetase, as a model (3–5). Transcript analysis and reporter gene fusions were used to demonstrate that tyrS regulation occurs at the level of premature transcription termination, and that decreased charging of tRNATyr is sufficient to induce readthrough of the tyrS leader region terminator. These studies also established the importance of a number of conserved sequence and structural elements in T box leader RNAs and in the tRNA effector molecule. Subsequent analyses were carried out in vitro, using the B. subtilis glyQS leader RNA as a model (6–8); glyQS represents a class of natural deletion variants that lack two large structural elements (Stem II and the Stem IIA/B pseudoknot) that are conserved in the majority of T box leader RNAs. Techniques important for analysis of the T box mechanism in vivo, in B. subtilis, will be described. These techniques can be adapted to other host organisms, and are also suitable for other classes of riboswitch RNAs that respond to d ifferent effectors, through small variations in the strain background and growth conditions tested. Regulatory RNAs from Gram-positive organisms other than B. subtilis are often functional when introduced into a B. subtilis host strain, allowing the use of genetic tools developed in this system.

Analysis of tRNA-Directed Transcription Antitermination in the T Box System In Vivo

283

2. Materials 2.1. Construction of Reporter Gene Fusions

1. Reporter gene fusions to lacZ are commonly used, and appropriate vectors are readily available. An example is plasmid pFG328 (9), which contains the E. coli lacZ coding sequence fused to a ribosome-binding site functional in Gram-positive bacteria, and a cat gene encoding chloramphenicol resistance for selection of B. subtilis isolates containing the fusion construct. Related plasmids are available from the Bacillus Genetic Stock Center (BGSC; http://www.bgsc.org). Plasmids that allow integration in single copy into the genome are preferable to multicopy plasmids. Plasmid pFG328 directs integration into an SPb prophage, which facilitates transfer to different strain backgrounds. Required strains are ZB307A (which is lysogenic for a variant of SPb that facilitates fusion construction) (10) and ZB449 (a strain lacking the prophage) (see Note 1). 2. Preferred growth media for B. subtilis is Tryptone Blood Agar Base (TBAB, 33 g/L) (Difco) for plates, and 2× YT broth (per liter): 16 g tryptone, 10 g yeast extract, 5 g NaCl. Chloramphenicol is added to growth medium at 5 mg/mL for selection of strains containing integrated fusions and at 0.1 mg/mL for induction of the chloramphenicol resistance gene. X-gal (5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside, Gold Biotechnologies) is added to plates at 40 mg/mL for detection of colonies producing b-galactosidase.

2.2. Transformation of B. subtilis

1. GMI medium (per liter): 2 g (NH4)2SO4, 14 g K2HPO4, 6 g KH2PO4, 1 g Na-citrate–H2O, 0.2 g MgSO4·7H2O, 0.5% glucose, 0.1% yeast extract, 0.05% casein hydrolysate, amino acids as required at 50 mg/mL; is used for preparation of competent cells of B. subtilis. 2. GMII medium is GMI with additional CaCl2 and MgCl2 (0.5 and 2.5 mM, respectively).

2.3. Growth Medium for Amino Acid Limitation

1. Minimal medium for amino acid limitation is Spizizen medium (11) (per liter): 2 g (NH4)2SO4, 14 g K2HPO4, 6 g KH2PO4, 1 g Na-citrate–H2O, 0.2 g MgSO4·7H2O, 0.5% glucose (see Note 2). 2. Amino acids are added as required at 50 mg/mL.

2.4. b-Galactosidase Assays

1. Z buffer (per liter): 5.5 g NaH2PO4–H2O, 16.1 g Na2HPO4·7H2O, 0.75 g KCl, 0.246 g MgSO4·7H2O, pH 7.0, to which b-mercaptoethanol (2.7 mL) is added immediately before use (12).

284

Henkin

2. ONPG (o-nitrophenyl-b-D-galactoside, Sigma): at 4 mg/mL in Z buffer, stored at −20°C. 3. 1 M Na2CO3 (106 g/L). 4. Toluene. 2.5. RNA Extraction and FormaldehydeAgarose Gel Electrophoresis

1. RNA is extracted using the Fenozol kit (Active Motif, Carlsbad, CA). All solutions for RNA manipulation (including water used to make other solutions) are treated with diethylpyrocarbonate (DEPC), 1 mL/L; incubate with shaking at 37°C overnight followed by autoclaving. All chemicals are from Sigma Chemical Co. (St. Louis, MO, molecular biology grade), unless otherwise noted. 2. 10× MOPS gel buffer: dissolve 41.8 g MOPS in 700 mL DEPC-treated water; adjust pH to 7.0 with 1N NaOH; add 20 mL 1 M sodium acetate and 20 mL 0.5 M EDTA, pH 8.0; adjust volume to 1 L. 3. Gel solution: mix 1.2 g agarose with 72 mL water, microwave to dissolve, cool to 65°C, add 10 mL 10× MOPS buffer and 18 mL formaldehyde (37%). 4. RNA sample buffer: 500 mL deionized formamide (Ambion), 100 mL 10× MOPS buffer, 150 mL formaldehyde (37%, filtered through 0.45-mm filter), 250 mL water. 5. 20× SSC (per liter): 175.3 g NaCl, 88.2 g sodium citrate, adjust pH to 7.0 with 10N KOH. 6. Transfer buffer (per liter): 200 mL 20× SSC, 10 mL 1N NaOH, 790 mL water. 7. 50× Denhardt’s solution (per 100 mL): 1 g Ficoll 400, 1 g polyvinylpyrrolidone, 1 g bovine serum albumin; dissolve in water, adjust volume to 100 mL, store at −20°C. 8. Prehybridization/hybridization solution (per 100 mL): 50 mL deionized formamide, 25 mL 20× SSC, 10 mL 50× Denhardt’s solution, 10 mL 10% SDS (w/v), 5 mL water. Add denatured salmon sperm DNA (incubated at 95°C for 5 min followed by 5 min on ice) to 100 mg/mL final concentration immediately before use. Store at −20°C. 9. Wash buffers. Wash I: 2× SSC; Wash II: 2× SSC, 0.1% SDS; Wash III: 0.2× SSC, 0.1% SDS. 10. Sephadex G-50 column (GE Healthcare). 11. T7 RNA polymerase transcription kit (USB Corp.) supplied with 10× reaction buffer. 12. Individual nucleotide triphosphate solutions, 10 mM stocks. Store at −20°C. 13. [a-32P]-UTP, 3,000 Ci/mmol (GE Healthcare). 14. BrightStar Plus nylon membrane (Ambion).

Analysis of tRNA-Directed Transcription Antitermination in the T Box System In Vivo

3. Methods: Generation of Reporter Gene Fusions 3.1. Plasmid Constructions

3.2. Generation of Fusion Strains in B. subtilis 3.2.1. Transformation of B. subtilis

285

A DNA fragment containing the candidate promoter and leader region is fused to a promoterless reporter gene construct, so that expression of the reporter gene is dependent on transcription from the inserted fragment. For T box genes, this generally requires a 400-bp DNA fragment comprising the promoter and leader sequence, including the leader region terminator. The appropriate fragment is isolated by PCR, using oligonucleotides that introduce restriction endonuclease cleavage sites at the 5¢ and 3¢ ends of the insert fragment that match the sites available in the cloning site of the vector of choice. The resulting plasmid is propagated in a standard E. coli strain suitable for recombinant DNA technology. Isolates are checked for the appropriate insert and confirmed by DNA sequencing. 1. All growth is at 37°C. The recipient strain (for integration of a fusion into an SPb prophage, we use strain ZB307A (10), which is a lysogen of SPb c2del2::pSK10D6, a derivative of phage SPb-containing sequences that promote recombination with the reporter fusion plasmid) is grown on TBAB medium. Cells are inoculated into GMI medium (10 mL) and grown until early stationary growth phase, then are diluted (1:10) into GMII medium and grown for an addition 60 min. 2. Cells (1 mL) are mixed with plasmid DNA (1–5 mg), shaken for 30 min, then mixed with 2.5 mL 2× YT broth containing an inducing concentration of chloramphenicol (0.1 mg/mL) and incubated with shaking for 90 min to allow integration of the DNA and expression of the cat gene encoded on the plasmid. 3. Samples are plated on TBAB medium containing chloramphenicol (5 mg/mL) to select for colonies in which the fusion has integrated into the SPb prophage, and X-gal to detect lacZ expression, followed by overnight incubation.

3.2.2. Generation of Transducing Lysates

1. Transformants of strain ZB307A are pooled by rinsing the transformation plate with 2× YT medium containing chloramphenicol (5 mg/mL). The suspension is diluted to an A595 of 0.1, and the culture is grown with shaking until it reaches early logarithmic growth phase (A595 of 0.3). 2. The culture is incubated with gentle shaking at 52°C for 5.5 min to induce the prophage, followed by incubation at 37°C for 2 h. The resulting lysate is centrifuged (6,000 × g, 5 min) to remove unlysed cells, and the supernatant is filter sterilized (20-mm filters) and stored at 4°C.

286

Henkin

3.2.3. Purification of SPb Phage and Transduction of Recipient Strain

1. Phage-containing lacZ fusions are purified by passage through strain ZB449, which lacks an SPb prophage. Strain ZB449 is grown in 2× YT (containing 0.1 mg/mL chloramphenicol for induction of the phage-borne cat gene) until midlog phase (A595 of 0.5), and samples (1.5 mL) are mixed with varying amounts of the phage lysate (1, 10, and 100 mL). 2. The transduction mixtures are incubated with shaking for 90 min followed by plating of 0.1 mL samples onto TBABcontaining chloramphenicol (5 mg/mL) and X-gal. 3. Transductants from the transduction mixture generated from the lowest phage dosage that gives isolated colonies (i.e., the lowest multiplicity of infection, to prevent isolation of dilysogens) are purified by streaking for isolated colonies, and the resulting strains are grown to generate a secondary SPb lysate as described in Subheading 3.2.2. The secondary lysate is used to transduce the appropriate recipient strain (e.g., an auxotroph for the appropriate amino acid, to allow testing of fusion expression under amino acid limitation conditions), using a protocol identical to that used for transduction of strain ZB449.

3.3. Measurement of Reporter Gene Expression In Vivo

1. Cells (30 mL) are grown in Spizizen minimal medium containing all required amino acids until early logarithmic growth phase (A595 of 0.3).

3.3.1. Amino Acid Limitation

2. Cells are collected by centrifugation (6,000 × g, 10 min) and resuspended in Spizizen minimal medium lacking the appropriate amino acid. 3. The culture is then split, and the required amino acid is added at high concentration (50 mg/mL) to culture 1 (no starvation) and at low concentration (5 mg/mL) to culture 2 (starvation). 4. Growth is continued for 4–5 h, and samples (1 mL) are taken at 1-h intervals for measurement of b-galactosidase activity and growth (A595). Samples for b-galactosidase activity are collected in microfuge tubes (1.5 mL), and cells are pelleted by centrifugation in a microfuge (14,000 × g, 5 min) and stored at −70°C.

3.3.2. b-Galactosidase Activity Measurement

1. Cell pellets (from Subheading 3.3.1) are resuspended in 1 mL Z buffer. Toluene (10 μL) is added to each tube and tubes are vortexed for 35 s followed by incubation (10 min) in a fume hood with open lids to allow evaporation of the toluene. 2. Samples of permeabilized cells (0.4 mL) are mixed with Z buffer (0.6 μL) in 2.0-mL microfuge tubes. ONPG (0.2 mL) is added to start the reaction, and the time is recorded. Cleavage of ONPG results in release of a yellow cleavage product. As each reaction mixture reaches a light yellow color, the time

Analysis of tRNA-Directed Transcription Antitermination in the T Box System In Vivo

287

is recorded and the reaction is stopped by addition of 1 M Na2CO3 (0.5 mL) followed by vigorous vortexing. Reaction mixtures are pelleted in a microfuge to remove cells (14,000 × g, 5 min), and the yellow color is quantitated by spectrophotometric measurement at A420, using as a blank a control sample in which the permeabilized cells are replaced by additional Z buffer. b-Galactosidase activity (Miller units) (12) is calculated as (A420 × 1,000)/(A595 × 0.4 mL × min), where min represents the time of incubation of the reaction mixture, and A595 is the recorded growth measurement for that sample. Assays are carried out in duplicate, and independent growth experiments are carried out at least in duplicate to control for variability in growth conditions and amino acid pools. 3. The expected result if antitermination is occurring under the growth conditions tested is that the b-galactosidase activity will increase during amino acid limitation. A parallel experiment in which cells are limited for a noncognate amino acid (i.e., an amino acid different from that specified by the Specifier Sequence in the test leader RNA) serves as a useful control. Other controls include fusion constructs in which the terminator is mutated (by changes in the helix or the uridine run) or deleted; this should give high expression under both starved and unstarved conditions. Introduction of mutations in conserved residues in the 5¢ side of the antiterminator (e.g., the UGG residues that pair with the CCA end of the tRNA) is predicted to give low expression under all conditions. 3.4. Measurement of Terminated and Readthrough Transcripts In Vivo 3.4.1. Growth and RNA Isolation

1. Cells that are auxotrophic for the amino acid to which the test T box gene is predicted to respond are grown under conditions similar to those used for monitoring lacZ fusion expression (Subheading 3.3). The strain is initially grown to midlogarithmic growth phase (A595 = 0.3) in minimal medium containing all required amino acids. 2. Cells are pelleted by centrifugation and resuspended in minimal medium lacking the amino acid that will be tested. The culture is then split into two flasks, and the missing amino acid is added to high concentration (50 mg/mL, non-starvation conditions) or low concentration (5 mg/mL, starvation conditions) (see Note 3). 3. Cells samples (25 mL) are removed and cells are pelleted by centrifugation (6,000 × g, 5 min), resuspended in 1 mL total volume, transferred to a microfuge tube, pelleted at 14,000 × g, and the pellets are stored at −80°C. 4. RNA is extracted using the Fenozol kit and RNA concentration is determined by measuring A260 (A260 = 1 corresponds to 40 mg/mL RNA). An aliquot of RNA is run on an agarose/ formaldehyde gel (see Subheading 3.4.3) and stained with

288

Henkin

ethidium bromide to check for degradation by visualization of high molecular weight RNAs and ribosomal RNAs. 3.4.2. Synthesis of Radiolabeled RNA Probes

1. DNA templates for synthesis of radiolabeled probes are generated by PCR, using a primer containing a promoter for recognition by T7 RNA polymerase which directs transcription of an RNA transcript complementary to the target RNA region. PCR products should be approximately 250 nt in length, with similar representation of uridines within the transcribed region. For each gene, one template for probe synthesis corresponds to the leader region of the target gene, while the second template corresponds to a region within the 5¢ portion of the coding sequence. 2. RNA probes are synthesized by T7 RNA polymerase transcription using the buffer described by the manufacturer, in the presence of ribonucleoside triphosphates (ATP, GTP, and CTP at 500 mM, UTP at 12 mM, [a-32P]-UTP at 6.25 mM) and template DNA (20 mg/mL). The transcription reaction is incubated at 37°C for 2 h, followed by passage through a Sephadex G-50 column to remove unincorporated nucleotide triphosphates (see Note 4).

3.4.3. Northern Blotting Analysis

1. RNA is denatured by incubation at 65°C for 10 min in the RNA sample buffer containing formaldehyde/formamide. 2. RNAs are separated on 1.2% agarose-formaldehyde gels using 1× MOPS solution as a running buffer and transferred to a BrightStar Plus nylon membrane by downward transfer in transfer buffer, followed by UV crosslinking to the membrane for 5 min. 3. Membranes are prehybridized for 2 h at 68°C in the hybridization solution, followed by addition of the radioactive probe and hybridization for 14 h at 68°C. After hybridization, the membrane is washed twice for 15 min in 30 mL Wash I buffer at room temperature, twice for 30 min in 30 mL Wash II buffer at 68°C, and twice for 15 min in 30 mL Wash III buffer at 68°C. Hybridization and wash conditions may require adjustment depending on the G + C content of the target mRNA. Results can be visualized by PhosphorImager analysis or autoradiography. The leader region probe allows visualization of both the terminated and the readthrough transcript, while the coding region probe allows visualization only of the readthrough transcript.

3.4.4. Data Interpretation and Controls

The expected result if antitermination is occurring under the growth conditions tested is that the ratio between the readthrough and terminated products will increase in the course of the amino acid limitation (see Note 5). The readthrough transcript may

Analysis of tRNA-Directed Transcription Antitermination in the T Box System In Vivo

289

be difficult to visualize if the target transcription unit is a large operon, as large transcripts are subject to processing and degradation in bacteria. In this case, the decrease in the terminated product should still be apparent (see Note 6).

4. Notes 1. A variety of reporter gene fusion plasmids are available from the BGSC that allow integration directly into the chromosomal locus of the test gene (e.g., pMUTIN series) or into ectopic sites in the chromosome (e.g., into the amyE or thrC genes); see http://www.bgsc.org/Catalogs/Catpart4.pdf for available plasmids. The cat antibiotic resistance cassette is replaced by other resistance genes in some of these plasmids, requiring substitution of other antibiotics for chloramphenicol in the protocols. 2. Use single distilled water or tap water for preparation of minimal medium; do NOT use highly purified water as B. subtilis requires trace minerals, the absence of which results in poor growth in defined medium. 3. Some adjustment of the amino acid limitation conditions may be necessary to allow slow depletion of the amino acid supply, depending on the intracellular pool and utilization rate of the particular amino acid. Ideal conditions for expression of T box genes encoding aminoacyl-tRNA synthetases generally allow a decrease in growth rate under starvation conditions, without a total cessation of growth. 4. Radioactive materials are hazardous and require appropriate precautions. Chemiluminescent labeling protocols can be substituted, although they are generally less sensitive. 5. Alternative techniques can be employed for measurement of transcript levels, such as Real-time RT-PCR, in which reverse transcriptase is used to generate a cDNA copy of the transcript followed by amplification of the targeted region by the polymerase chain reaction. Separate amplification of the leader region and coding region of the transcript is necessary to determine the efficiency of terminator readthrough. Northern analysis allows a more direct comparison of the leader RNA and full-length transcripts, while Real-time RT-PCR is potentially more quantitative. 6. There is some evidence for processing of readthrough transcripts of certain T box genes in B. subtilis (13); the processing site is within the antiterminator domain and is predicted to result in an RNA product approximately 30 nt shorter than the terminated transcript.

290

Henkin

Acknowledgments The author would like to thank members of the lab who have contributed to the development of these technical approaches, particularly Frank Grundy, who was responsible for most of the initial work on the T box system; and Jerneja Tomsic, who standardized the protocols for RNA isolation and Northern analysis. This work was supported by NIH R01 GM47823.

References 1. Grundy, F. J., and Henkin, T. M. (1993). tRNA as a positive regulator of transcription antitermination in B. subtilis. Cell 74, 475–482. 2. Henkin, T. M., and Grundy, F. J. (2007). Sensing metabolic signals with nascent RNA transcripts: the T box and S box riboswitches as paradigms. Cold Spring Harb. Symp. Quant. Biol. 71, 231–237. 3. Rollins, S. M., Grundy, F. J., and Henkin, T. M. (1997). Analysis of cis-acting sequence and structural elements required for antitermination of the Bacillus subtilis tyrS gene. Mol. Microbiol. 25, 411–421. 4. Grundy, F. J., Rollins, S. M., and Henkin, T. M. (1994). Interaction between the acceptor end of tRNA and the T box stimulates antitermination in the Bacillus subtilis tyrS gene: a new role for the discriminator base. J. Bacteriol. 176, 4518–4526. 5. Henkin, T. M., Glass, B. L., and Grundy, F. J. (1992). Analysis of the Bacillus subtilis tyrS gene: conservation of a regulatory sequence in multiple tRNA synthetase genes. J. Bacteriol. 174, 6763–6770. 6. Grundy, F. J., Winkler, W. C., and Henkin, T. M. (2002). tRNA-mediated transcription antitermination in vitro: codon–anticodon pairing independent of the ribosome. Proc. Natl Acad. Sci. U. S. A. 99, 11121–11126.

7. Grundy, F. J., Yousef, M. R., and Henkin, T. M. (2005). Monitoring uncharged tRNA during transcription of the Bacillus subtilis glyQS gene. J. Mol. Biol. 346, 73–81. 8. Yousef, M. R., Grundy, F. J., and Henkin, T. M. (2005). Structural transitions induced by the interaction between tRNAGly and the Bacillus subtilis glyQS T box leader RNA. J. Mol. Biol. 349, 273–287. 9. Grundy, F. J., Waters, D. A., Allen, S. H. G., and Henkin, T. M. (1993). Regulation of the Bacillus subtilis acetate kinase gene by CcpA. J. Bacteriol. 175, 7348–7355. 10. Zuber, P., and Losick, R. (1987). Role of AbrB in SpoOA- and SpoOB-dependent utilization of a sporulation promoter in B. subtilis. J. Bacteriol. 169, 2223–2230. 11. Anagnostopoulos, C., and Spizizen, J. (1961). Requirements for transformation in Bacillus subtilis. J. Bacteriol. 81, 741–746. 12. Miller, J. (1972). Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 13. Condon, C., Putzer, H., and GrunbergManago, M. (1996). Processing of the leader mRNA plays a major role in the induction of thrS expression following threonine starvation in Bacillus subtilis. Proc. Natl Acad. Sci. U. S. A. 93, 6992–6997.

Chapter 21 In Vitro Selection of Conformational Probes for Riboswitches Günter Mayer and Michael Famulok Summary Riboswitches are non-coding RNA elements mainly located in the 5¢ untranslated regions (UTR) of bacterial genes. They bind to small metabolites and upon binding conformational changes occur that trigger the expression of a certain gene. Riboswitches have been identified that bind to amino acids, purines, and other small metabolites such as thiamine pyrophosphate. Riboswitches contain an aptamer domain which is necessary for interaction with the metabolite and a related expression domain which harbours structural and sequence information required for interference with gene expression. The binding of a metabolite to the aptamer domain induces structural rearrangements that are relayed to the expression domain, thereby interfering with gene expression. To investigate and determine domains of the riboswitches which undergo conformational changes upon metabolite binding we used a dynamic SELEX process and identified RNA aptamers that bind to the metabolite-free variant of the riboswitch but are released upon metabolite addition. By this means, and after determination of the binding region, domains which are necessary for proper function of a full-length riboswitch can be identified. Key words: SELEX, Aptamer, Riboswitch, Kissing complexes

1. Introduction Thiamine pyrophosphate (TPP)-binding riboswitches represent one of the most intensively studied riboswitch groups (1–7). They can be found in bacteria, regulating either transcription attenuation or translation initiation, and recently also an eukaryotic representative has been investigated and was determined to regulate mRNA splicing (8–10). Amongst the family of TPP riboswitches, the thiM riboswitch from E. coli was shown to regulate the expression of the enzyme hydroxyethylthiazol kinase by TPP-mediated inhibition Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_21 © Humana Press, a part of Springer Science + Business Media, LLC 2009

291

292

Mayer and Famulok

of translation initiation (1, 4). The hydroxyethylthiazol kinase is involved in the biosynthesis and transport of the coenzyme thiamine. The riboswitch exhibits strong TPP-binding activity, and the structure of the aptamer domain of the riboswitch in complex with TPP has recently been solved (11, 12). It is assumed that gene regulation depends on the accessibility of the Shine–Dalgarno (SD) sequence. In the absence of TPP the SD sequence is accessible for the ribosome, whereas TPP binding to the aptamer domain of the riboswitch induces conformational changes in the expression domain of the riboswitch that finally result in the sequestration of the SD sequence, rendering it inaccessible for the ribosome (1, 4, 6). Although riboswitches can be subdivided into an aptamer and an expression domain, both parts of the riboswitch are interconnected by a communication link (4). Upon metabolite binding this module facilitates the transduction of conformational changes from the aptamer- to the expression-domain. Elucidation of these communication links can be accomplished by structural analysis. However, crystallization of full-length riboswitches is difficult to be achieved and so are NMR studies of RNA molecules that bear more than 100 nucleotides in length. A possible alternative approach is offered by the in vitro selection of specific RNA molecules that distinguish between two different activation states of a riboswitch (13, 14). For this purpose we developed a specific selection protocol that allows the enrichment of RNA molecules that interact with the TPP-free variant of the thiM riboswitch and can be released in the presence of TPP (Fig. 1). The selected RNA molecules were shown to fold into defined short RNA hairpins which form so called kissing complexes with thiM riboswitch (4, 13).

Fig. 1. In vitro selection scheme for the enrichment of conformation-specific RNA aptamers that bind to the TPP-free thiM riboswitch of E. coli. ODN oligonucleotide. Reprinted with permission from Wiley-VCH.

In Vitro Selection of Conformational Probes for Riboswitches

293

One RNA hairpin was shown to bind tightly to the expression domain and mutational and functional analyses revealed that the binding site of the hairpin is important for the proper function of the thiM riboswitch in E. coli. This protocol describes the concept and in vitro selection of these specific riboswitch-binding RNA hairpin motifs and the analysis of their binding behaviour by surface plasmon resonance (SPR).

2. Materials 2.1. Oligodeoxynucleotides and RNA Molecules

1. Oligodeoxynucleotides were purchased from Metabion, Germany, in 0.04-mmol scale and HPLC grade. All oligodeoxynucleotides were dissolved in deionized water (ddH2O) (purified on a Millipore water purification system), and concentrations were determined using UV spectroscopy (l = 260 nm). List of oligodeoxynucleotides: (a) 5 ¢ -TCGTAATACGACTCACTATAGGAACCAAACGACTCG-3¢ (5¢tpp) (b) 5¢-TTGCGCTGGATCCAGCAGGTCGA-3¢ (3¢tpp) (c) 5¢-CGTGACTTCCCTACGCTGGCAT-3¢ (3¢tpp.91) (d) 5 ¢-TCGTAATACGACTCACTATAGACCACCAGGTCATTG-3¢ (5¢tpp.74) (e) 5¢-GGAACCAAACGACTCGGGGTGCCCTTCTGCGTGAAGGCTGAGAAATACCCGTATCACCTGATCTGGATAATGCCAGCGTAGGGAAGTCACGGACCACCAGGTCATTGCTTCTTCACGTTATGGCAGGAGCAAACTATGCAAGTCGACCTGCTGGATCCAGCGCAA-3¢ (tpp.RS) (f) 5¢-TGTACCTACGTCTGCAGTGAA-3¢ (N25.21) (g) 5¢-AATACGTAGACTGTCTCTCTCTCCC-3¢ (5¢oligo 1) 2. RNA library N25: 5¢-GGGAGAGAGAGACAGUCUACGUAUU-N25-UUCACUGCAGACGUAGUACA-3¢ was obtained by in vitro transcription from the corresponding dsDNA template. 3. Absolute ethanol.

2.2. Denaturing Polyacrylamide Gel Electrophoresis

1. Running buffer TBE: 89 mM Tris base, 89 mM boric acid, 2 mM EDTA (see Note 1). 2. RNA loading buffer: 50 mM EDTA of pH 8.0, 9 M urea. 3. 25% Acrylamide/bis solution (37.5:1) in D (this is a neurotoxin when unpolymerized and care should be taken not to receive exposure), Solution D: 8.3 M urea, Solution B: 10× TBE in 8.3 M urea.

294

Mayer and Famulok

4. N,N,N¢,N¢-Tetramethyl-ethylenediamine (TEMED) (Calbiochem, San Diego, CA) and ammonium peroxodisulphate (APS) (Merck, Darmstadt, Germany). 5. Gel electrophoresis chamber (30 cm × 30 cm) (Fisher Scientific, Pittsburgh, PA). 6. UV handlamp with l = 254 nm and silica gel plate with a fluorescing dye (Merck, Darmstadt, Germany). 2.3. Biotinylation of RNA

1. Solid and dried guanosine-5¢-thiophosphate (GMPS) (emp Biotech, Berlin, Germany) was dissolved in pure water and its concentration was determined by UV spectroscopy (l = 260 nm, e = 13,700 L/mol/cm). 2. Transcription buffer: 20 mM Tris–HCl, pH 7.9, 25 mM MgCl2, 5 mM DTT, 2.5 mM each NTP. 3. T7 RNA polymerase, 50 U/mL (Stratagene, La Jolla, CA). 4. Inorganic pyrophosphatase (Boehringer, Ingelheim, Germany). 5. A spatula tip of EZ-Link PEO-iodoacetyl biotin (Pierce, Rockford, IL) was freshly dissolved in dimethylformamide (DMF). 6. Reaction buffer: 10 mM Tris–HCl, pH 8.0, 50 mM EDTA. 7. RNasin, 40 U/mL (Promega, Madison, WI). 8. Phenol saturated with Tris–HCl, pH 8.0 (Roth, Karlsruhe, Germany). 9. Nucleotide triphosphate (NTP) mix, 25 mM each (Larova, Teltow, Germany). 10. NaOAc: 3 M sodium acetate adjusted to pH 5.4 by acetic acid. 11. NH4OAc: 6 M NH4OAc, pH 7.4. 12. G25 microspin columns (General Electric, Munich, Germany).

2.4. Magnetic Particle Preparation

1. Magnetic streptavidin-coated Dynabeads, 10 mg/mL (Invitrogen, Carlsbad, CA), were washed 5 times with coupling buffer prior to use. 2. Magnetic particle concentrator (Dynal, Oslo, Norway). 3. Head-over-tail shaker. 4. Coupling buffer: 50 mM Hepes-K, pH 7.5, 100 mM KCl, 0.5 M NaCl, 1 mM EDTA.

2.5. In Vitro Selection

1. Selection buffer: 50 mM Hepes-K of pH 7.5, 100 mM KCl, 5 mM MgCl2 2. Elution buffer: Coupling buffer supplemented with 5–50 mM thiamine pyrophosphate (TPP). 3. RT-PCR: One tube set up using superscript II reverse transcriptase (Invitrogen) and Taq DNA polymerase (Promega). For 100- m L reaction, mix 4 m L 5× First Strand buffer

In Vitro Selection of Conformational Probes for Riboswitches

295

(Invitrogen), 10 mL 10× PCR buffer (Promega), 2 mL 100 mM DTT, 1 mL 100 mM 5¢-primer, 1 mL 3¢-primer 100 mM, 6 mL 25 mM MgCl2, 3 mL 10 mM dNTP mix, 1 mL Superscript (200 U/mL) polymerase, 1 mL Taq polymerase (5 U/mL), template, and water to 100 mL. 4. Glycogen dissolved in water. 2.6. Surface Plasmon Resonance Analysis of RNA–RNA Interactions

1. SA sensor chips (Biacore, Uppsala, Sweden). 2. SA-coupling buffer: 50 mM Hepes-K, pH 7.5, 100 mM KCl, 0.5 M NaCl). 3. Running buffer: 50 mM Hepes-K of pH 7.5, 100 mM KCl, 5 mM MgCl2. 4. Regeneration buffer: 12.5 mM EDTA, pH 8.0.

3. Methods 3.1. Oligodeoxynucleotides and RNA Molecules

1. RNA molecules for the in vitro selection process, namely the 165 nucleotide containing thiM riboswitch and the RNA library N25, were prepared by in vitro transcription from dsDNA templates using standard T7 RNA polymerase-based transcription protocols (see Note 2). For 100-mL reaction, mix the DNA template, transcription buffer, 5 mL T7 RNA polymerase, 1 mL RNasin and inorganic pyrophosphatase (optional). 2. The dsDNA templates were performed by common PCR methods using the templates (tpp.RS) and primers (5¢tpp, 3¢tpp) from Subheading 2, whereas the 5¢-primer included the T7 promoter region (5¢-TAATACGACTCACTATA-3¢). 3. Purification of dsDNA molecules was achieved by phenol/ chloroform extraction, followed by precipitation with ethanol. After centrifugation the pellets were washed with 70% ethanol and dissolved in pure water.

3.2. Denaturing Polyacrylamide Gel Electrophoresis

1. The RNA molecules were purified using denaturing polyacrylamide gel electrophoresis. The glass plates were cleaned carefully with water and ethanol prior to use. 2. Prepare a 1.5-mm thick gel, 10% gel by mixing 28 mL of 25% acrylamide/bis solution, 35 mL solution D and 7 mL solution B, 28 mL TEMED, and 560 mL APS. After pouring the gel insert the comb and use clamps to fix the comb and let the gel polymerize for 30 min. 3. Assemble the electrophoresis chamber and fill the tanks with the running buffer TBE. Remove the comb carefully and rinse the wells using a syringe.

296

Mayer and Famulok

4. Let the gel pre-run about 30 min at 370 V. 5. Dissolve the RNA pellets after transcription work up (see Subheading 3.3) in the RNA loading buffer. Heat the samples to 80°C for 1 min and spin in a microcentrifuge to collect the entire sample at the bottom of the tube. Carefully load the well of the pre-run gel and let the gel run for 1.5 h at 370 V. 6. Disassemble the chamber and cover the gel with a thin foil. Place the gel on a fluorescing silica gel plate and use UV irradiation to visualize the corresponding RNA band. Cut the band out and do the work up according to the crush-andsoak method. Finally, the concentration of the purified RNA is measured by UV spectroscopy. 3.3. Biotinylation of RNA

1. Prepare a standard 100 mL in vitro transcription reaction with a 4-fold molar excess of GMPS (10 mM) over GTP (2.5 mM) (see Note 3). Incubate overnight at 37°C. 2. For work up of the GMPS-transcription reaction perform a phenol/chloroform extraction followed by ethanol precipitation. Add 100 mL phenol, vortex thoroughly, and spin for 3 min at 17,900 × g in a microcentrifuge. Take the supernatant and add 1 volume of chloroform. Vortex and centrifuge the mixture again. Recover the supernatant, add 30 mL NaOAc and 390 mL ethanol, incubate at −80°C for 10 min, centrifuge at 17,900 × g for 20 min, and remove the supernatant carefully. After washing the pellet with 70% ethanol, dissolve the pellet in 100 mL ddH2O. Remove excess of GMPS by filtration through G25 microspin columns two times. 3. The GMPS-RNA is then biotinylated by incubation with a 200fold excess of EZ-link PEO-iodoacetyl biotin in the reaction buffer for 2 h at 40°C. 4. After completion of the reaction, add 1 volume of 6 M NH4OAc of pH 7.4 and 3 volumes of ethanol and centrifuge for 30 min at 17,900 × g. Remove the supernatant, wash the pellet with 70% ethanol, and dissolve the dry pellet in the RNA loading buffer. 5. The biotinylated RNA can be purified by denaturing polyacrylamide gel electrophoresis as described in Subheading 3.2.

3.4. Magnetic Particle Preparation

1. Wash 50 mL of streptavidin-coated magnetic beads five times with coupling buffer using magnetic particle concentrator and re-suspend the beads in 250 mL of 2× coupling buffer. Add 10 pmol of biotinylated thiM riboswitch (250 mL in ddH2O) and incubate the solution in a head-over-tail shaker for 15 min. 2. Subsequently, the derivatized beads were washed five times with selection buffer and were finally re-suspended in 100 mL selection buffer. The beads were then directly used in the selection process.

In Vitro Selection of Conformational Probes for Riboswitches

3.5. In Vitro Selection

297

1. Prior to incubation with the thiM-bound streptavidin beads, the RNA library N25 should be incubated with the blocking oligonucleotides, which are complementary to the 5¢- and 3¢-constant primer-binding sites of the RNA library. Prepare a solution containing 50 pmol of the RNA library and 75 pmol of each oligodeoxynucleotide (5¢-oligo 1 and N25.21) in selection buffer (380 mL) without MgCl2. Heat the solution to 80°C for 3 min and let the solution cool down to room temperature for 15 min. Add 20 mL 0.1 M MgCl2 for the proper folding of the RNA library (see Note 4). 2. Combine the solution of the thiM-derivatized streptavidin beads (Subheading 3.4, step 2) with the RNA library solution (step 1 above) and incubate the reaction for 30 min at room temperature (see Note 5). 3. Remove all non-bound RNA molecules by washing 6 times with 100 mL selection buffer. 4. Elute the bound RNA molecules by the addition of 100 mL elution buffer containing TPP (see Note 6). Incubate for 15 min at room temperature, separate the beads using the magnetic particle concentrator, and remove the supernatant. 5. Precipitate the RNA from the supernatant by the addition of 1 mg glycogen, 10 mL 3 M NaOAc of pH 5.4, and 330 mL ethanol. Incubate at −80°C for 10 min and spin at 20,800 × g for 10 min. Remove the supernatant, wash the pellet with 70% ethanol, and dry. Dissolve the pellet in 50 mL ddH2O and use this solution as template for the RT-PCR. 6. RT-PCR was performed using superscript II reverse transcriptase and Taq DNA polymerase. Reverse transcription was carried out at 54°C for 10 min, followed by inactivation of the reverse transcriptase at 70°C for 15 min. PCR was performed using following settings: 94°C for 1 min, 60°C for 1 min, and 72°C for 90 s, 10–15 cycles (see Note 7). All amplified dsDNA products were analysed on 2.5% agarose gels and were subsequently purified by ethanol precipitation. The pellets were dissolved in 20 mL ddH2O, and 10-mL aliquot was used as template for the subsequent in vitro transcription yielding RNA for the next selection cycle. 7. After eight selection cycles, the RNA library was cloned and sequenced (Fig. 2). Monoclonal RNA aptamers can be analysed for binding to thiM RNA using SPR or electrophoretic mobility shift assays (EMSA).

3.6. Surface Plasmon Resonance Analysis of thiM–RNA Aptamer Interaction

1. The streptavidin-coated sensor chip was derivatized with 100 nM biotinylated thiM-RNA in SA-coupling buffer by consecutive injections of 5-mL aliquots of the biotinylated thiM RNA (flow rate: 5 mL/min) until a response of about 1,000 RU was reached (see Note 8).

298

Mayer and Famulok

2. Interaction of individual RNA aptamers with the thiM-derivatized surface was measured by injection of increasing concentrations of the aptamers in running buffer (Fig. 3) at flow rate 30 mL/min (see Note 9). 3. Regenerate the surfaces by two consecutive injections of the regeneration buffer for 30 s.

Fig. 2. DNA sequences of the selected RNA molecules. Only the initial random region is shown. Sequences can be grouped into two motifs, one unique sequence N25.3 and the motif I. Sequences that are complementary to the thiM riboswitch are highlighted in light and dark grey, respectively. Reprinted with permission from Wiley-VCH.

Fig. 3. Surface plasmon resonance analyses of aptamers. SPR analysis of N25.1, N25.4, and N25.3 aptamers. Biotinylated thiM RNA was coupled to the streptavidin surface of a CM5 sensor chip and aptamer RNA molecules were injected at indicated concentrations. Regeneration was achieved by 30 s injections with 12.5 mM EDTA, pH 8.0.

In Vitro Selection of Conformational Probes for Riboswitches

299

4. Notes 1. Avoid nuclease contamination in all buffers and solutions used. Prepare all buffers using DEPC-treated water. 2. Use gloves to avoid nuclease contamination. 3. Avoid oxidation of your GMPS transcriptions by O2 from the air since oxidation of mercaptane groups hampers the yield of the biotinylation reaction due to disulphide-bond formation of either the GMPS or the GMPS-labelled RNA. 4. Do not heat the RNA molecules in the presence of Mg2+ cations since this would favour autohydrolysis of the RNA. 5. The thiM-derivatized Dynabeads should be prepared freshly prior to each selection cycle. 6. The amount of TPP used to induce conformational changes of the thiM riboswitch and thus to remove bound RNA molecules can vary during the course of selection. Use higher concentrations (50 mM) in the first selection cycles and lower concentrations (5 mM) in the later cycles. 7. Avoid over-amplification of the DNA since this would negatively influence the outcome of the selection. 8. SPR analysis can be done either in the presence or absence of the blocking oligodeoxynucleotides. 9. Make sure that the running buffer used for SPR and the buffer used to dissolve the RNA are identical. Otherwise buffer changes result in worse resolution of the sensograms.

Acknowledgements We thank the Deutsche Froschungsgemeinschaft for financial support. Nicole Kuhn is acknowledged for excellent technical assistance. References 1. Winkler, W., Nahvi, A., and Breaker, R. R. (2002). Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression. Nature 419, 952–956. 2. Yamauchi, T., Miyoshi, D., Kubodera, T., Nishimura, A., Nakai, S., and Sugimoto, N. (2005). Roles of Mg2+ in TPP-dependent riboswitch. FEBS Lett. 579, 2583–2588. 3. Edwards, T. E., and Ferre-D’Amare, A. R. (2006). Crystal structures of the thi-box riboswitch bound to thiamine pyrophosphate

analogs reveal adaptive RNA-small molecule recognition. Structure 14, 1459–1468. 4. Rentmeister, A., Mayer, G., Kuhn, N., and Famulok, M. (2007). Conformational changes in the expression domain of the Escherichia coli thiM riboswitch. Nucleic Acids Res. 35, 3713–3722. 5. Miranda-Rios, J. (2007). The THI-box riboswitch, or how RNA binds thiamin pyrophosphate. Structure 15, 259–265. 6. Lang, K., Rieder, R., and Micura, R. (2007). Ligand-induced folding of the thiM TPP

300

7.

8.

9.

10.

Mayer and Famulok riboswitch investigated by a structure-based fluorescence spectroscopic approach. Nucleic Acids Res. 35, 5370–5378. Noeske, J., Richter, C., Stirnal, E., Schwalbe, H., and Wohnert, J. (2006). Phosphate-group recognition by the aptamer domain of the thiamine pyrophosphate sensing riboswitch. Chembiochem 7, 1451–1456. Tucker, B. J., and Breaker, R. R. (2005). Riboswitches as versatile gene control elements. Curr. Opin. Struct. Biol. 15, 342–348. Cheah, M. T., Wachter, A., Sudarsan, N., and Breaker, R. R. (2007). Control of alternative RNA splicing and gene expression by eukaryotic riboswitches. Nature 447, 497–500. Miranda-Rios, J., Navarro, M., and Soberon, M. (2001). A conserved RNA structure (thi box) is involved in regulation of thiamin bio-

11.

12.

13.

14.

synthetic gene expression in bacteria. Proc. Natl Acad. Sci. U. S. A. 98, 9736–9741. Thore, S., Leibundgut, M., and Ban, N. (2006). Structure of the eukaryotic thiamine pyrophosphate riboswitch with its regulatory ligand. Science 312, 1208–1211. Serganov, A., Polonskaia, A., Phan, A. T., Breaker, R. R., and Patel, D. J. (2006). Structural basis for gene regulation by a thiamine pyrophosphate-sensing riboswitch. Nature 441, 1167–1171. Mayer, G., Raddatz, M.S., Grunwald, J.D., and Famulok, M. (2007). Angrw. Chem. Int. Ed. Engl. 46, 557–560 Famulok, M., Hartig, J. S., and Mayer, G. (2007). Functional aptamers and aptazymes in biotechnology, diagnostics, and therapy. Chem. Rev. 107, 3715–3743.

Chapter 22 A Green Fluorescent Protein (GFP)-Based Plasmid System to Study Post-Transcriptional Control of Gene Expression In Vivo Johannes H. Urban and Jörg Vogel Summary Small non-coding RNAs (sRNAs) are an emerging class of regulators of bacterial gene expression, which mainly modulate the translation of trans-encoded mRNAs. Typically, these molecules are 50–200 nucleotides in size and do not contain expressed open reading frames (ORFs). In Escherichia coli, about 70 members of this group have been identified to date and further estimates assume hundreds of sRNAs per bacterial genome. Regulation of gene expression by sRNAs is predominantly mediated by physical sRNA/target mRNA interactions that are based on short and imperfect complementarity. Although the contribution of sRNAs to overall bacterial gene regulation is now being appreciated, the function of many sRNAs is still unknown and their targets await to be uncovered. We recently developed a modular two-plasmid system, based on the green fluorescent protein (GFP) as non-invasive reporter of gene expression, to rapidly monitor the regulatory potential of sRNA/target mRNA pairs under investigation in vivo. The specialized reporter plasmid series also provides a suitable platform to study the function of cis-encoded riboregulators such as natural riboswitches, thermosensors, or engineered aptamer-based regulatory switches. Key words: Green fluorescent protein, Post-transcriptional control, Small non-coding RNA, Riboregulator

1. Introduction Post-transcriptional regulation as mediated by sRNAs is often accomplished by complementary base pairing of an sRNA to the non-coding translation initiation region (TIR) of the targeted mRNA (1). Duplex formation is often supported by the bacterial RNA chaperone Hfq (reviewed in (2, 3)) and results in either translational repression or activation of the mRNA. Annealing of the Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_22 © Humana Press, a part of Springer Science + Business Media, LLC 2009

301

302

Urban and Vogel

sRNA to sequences that involve the ribosome-binding site (RBS) and/or the translational start codon of the mRNA typically leads to repression of protein synthesis by blocking ribosome entry and translation initiation. This mode of sRNA action seems to be most common in enterobacteria and was first described for the E. coli sRNA MicF (4) which represses translation of the ompF mRNA (coding for the major outer membrane protein F). Translational activation of mRNAs is achieved by sRNA-induced structural rearrangements of mRNA leader sequences. Some mRNA TIRs fold into selfinhibitory stem-loop structures, which mask the RBS, thereby rendering the mRNA translation incompetent. Such cis-repressed mRNAs, as demonstrated for the E. coli rpoS mRNA (encoding the stationary sigma factor sS), can be translationally activated by sRNA binding. Base pairing of the sRNA DsrA to nucleotides involved in maintenance of the self-inhibitory mRNA hairpin opens the stem within the rpoS leader and results in activation of translation by freeing the RBS and promoting ribosome entry (5, 6). Translational regulation by sRNAs is not limited to TIRs located in the 5¢ untranslated regions (UTR) of monocistronic mRNAs, but is also observed within intergenic regions of polycistronic messengers (7, 8). The approach described here is used in our laboratory to study sRNA-mediated translational control and to verify previously unknown sRNA targets in E. coli. The method is based on constitutive co-expression of sRNA and potential target mRNAs at high levels from two compatible plasmids within the same cell (9). Putative target mRNA regions, covering full-length 5¢ UTRs and the first coding residues, are translationally fused to gfp upon insertion into specialized reporter plasmids. These vectors allow, besides conventional cloning, cDNA cloning of primary mRNA transcripts (of genes for which the transcriptional start site is unknown) by an adapted 5¢ RACE strategy, and cloning of intraoperonic target regions derived from polycistronic mRNAs into a dual-reporter mini-operon. Target fusion regulation is rapidly assessed by colony fluorescence imaging, or quantitatively by fluorometric measurements upon growth in liquid culture, by flow cytometry or by immunoblotting.

2. Materials 2.1. Cloning Procedures

1. Competent E. coli Top10 and E. coli Top10F¢ (Invitrogen, Karlsruhe, Germany). Store at −80°C. 2. Reporter cloning plasmids (and control plasmids): pXG-0, pXG-1, pXG-10, pXG-20, pXG-30 (9). 3. RNA cloning plasmid (and control plasmid): pJU-334 (encoding the E. coli transcription factor agaR, which is removed upon sRNA cloning), pJV300 (10).

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

303

4. Media and antibiotics. LB-medium: 1% (w/v) tryptone, 0.5% (w/v) yeast extract, 1% (w/v) sodium chloride; SOCmedium: 2% (w/v) tryptone, 0.5% (w/v) yeast extract, 0.05% (w/v) NaCl, 20 mM glucose; chloramphenicol: 20 mg/mL in 70% (v/v) ethanol; ampicillin: 100 mg/mL in water. Store antibiotics at −20°C. 5. Restriction enzymes. BseRI, 4,000 U/mL (New England Biolabs, Ipswich, MA), provided with buffer 10× NE Buffer 2: 500 mM NaCl, 100 mM Tris–HCl, pH 7.9, 100 mM MgCl2, 10 mM dithiothreitol (DTT). BsgI, 3,000 U/mL (New England Biolabs), provided with buffer 10 NE Buffer 4: 500 mM K-acetate, 200 mM Tris-acetate, pH 7.9, 100 mM Mg–acetate, 10 mM DTT, and 400× SAM solution: 32 mM S-adenosylmethionine, 10% (v/v) ethanol, in 5 mM sulphuric acid. Mph1103I, 10,000 U/mL, (Fermentas, St.Leon-Rot, Germany), provided with buffer 10× TangoTM: 330 mM Tris-acetate, pH 7.9, 100 mM Mg-acetate, 660 mM K-acetate, 1 mg/mL (w/v) bovine serum albumin (BSA). DpnI, 10,000 U/mL (Fermentas). NheI, 10,000 U/ mL (Fermentas). XbaI, 10,000 U/mL (Fermentas). Store all enzymes and buffers at −20°C. 6. Shrimp alkaline phosphatase (SAP), 1,000 U/mL (Fermentas). 7. T4 DNA Ligase, 5,000 U/mL (Fermentas) provided with buffer 10× T4 DNA Ligation Buffer: 400 mM Tris–HCl, pH 7.8, 100 mM MgCl2, 100 mM DTT, 5 mM adenosine triphosphate (ATP). 8. T4 RNA Ligase, 20,000 U/mL (New England Biolabs), provided with buffer 10× T4 RNA Ligase Reaction Buffer: 500 mM Tris–HCl, pH 8.0, 100 mM MgCl2, 100 mM DTT, 10 mM ATP. 9. Tobacco Acid Pyrophosphatase (TAP), 10,000 U/mL (Epicentre Biotechnologies, Madison, WI), provided with TAP 10× Reaction Buffer: 0.5 M Na-acetate, 10 mM EDTA, 1% b-mercaptoethanol, 0.1% (v/v) Triton X-100. 10. RNasin RNase Inhibitor, 40,000 U/mL (Promega, Madison, WI). 11. DMSO. 12. Pfu DNA polymerase, 2,500 U/mL (Fermentas), provided with buffer 10× Pfu Buffer: 200 mM Tris–HCl, pH 8.0, 100 mM ammonium sulphate, 100 mM KCl, 0.1% (v/v) Triton X-100, 1 mg/mL (w/v) BSA, 20 mM magnesium sulphate. 13. PhusionTM High-Fidelity DNA polymerase, 2,000 U/mL (New England Biolabs), provided with 5× PhusionTM HF Buffer (composition not stated by the manufacturer; contains 7.5 mM MgCl2). 14. Taq DNA polymerase 5,000 U/mL (New England Biolabs), provided with 10× ThermoPol Reaction Buffer: 200 mM

304

Urban and Vogel

Tris–HCl, pH 8.8, 100 mM KCl, 100 mM sulphuric acid, 20 mM magnesium sulphate, 1% (v/v) Triton X-100. 15. Superscript III RT-PCR system (Invitrogen), components required here: Superscript III reverse transcriptase (RT), 200,000 U/mL; 5× first-strand buffer: 250 mM Tris–HCI, pH 8.3, 375 mM KCl, 15 mM MgCl2), 0.1 M DTT, 10 mM dNTP mix. 16. RNase H, 5,000 U/mL (New England Biolabs). 17. PCI mixture: water-saturated phenol/chloroform/isoamylic alcohol mixture, 25:24:1 (v/v/v). 18. Ethanol–Na–acetate mixture: ethanol containing 0.3 M Na–acetate (pH 5.7). 19. NucleoBond PC100 plasmid purification kit (MachereyNagel, Düren, Germany). 20. NucleoSpin Extract II DNA purification kit (Macherey-Nagel). 21. SV Total RNA Isolation System (Promega). 22. Phase Log Gel (PLG) heavy tubes (Eppendorf, Germany). 23. dNTPs: 10 mM solution (2.5 mM each dNTP). Store at −20°C. 24. Total RNA of E. coli strain MC4100. 25. Random hexamer oligonucleotides. 26. RNA oligonucleotide A4, GACGAGCACGAGGACACUGACAUGGAGGAGGGAGUAGAAA. 27. DNA oligonucleotides, all at 100 pmol/µL. PLlacoB, CGCACTGACCGAATTCATTAA; JVO-2164, AAACGAAATGAAACGAAAGTT; pZE-A, GTGCCACCTGACGTCTAAGA; pZE-CAT, TGGGATATATCAACGGTGGT; pZE-XbaI: TCGTTTTATTTGATGCCTCTAGA; JVO-0155, CCGTATGTAGCATCACCTTC. DNA oligonucleotides for MicC sRNA cloning: JVO0486, pGTTATATGCCTTTATTGTCACAGAT; JVO-0489, GTTTTTTCTAGACGATTAAATGCTCTGGATAAG. DNA oligonucleotides for ompC fusion cloning into pXG-10: JVO-0428, GTTTTTATGCATTGCCGACTGATTAATGAGG; JVO-0429 GTTTTTGCTAGCTGGGACCAGGAGG. DNA oligonucleotides for ompA cDNA fusion cloning into pXG-20: JVO-0367, ACTGACATGGAGGAGGGA; JVO-0433, GTTTTTGCTAGCGAAACCAGCCAGTG. DNA oligonucleotides for glmUS operon fusion cloning into pXG-30: JVO-1270, GTTTTATGCATCGTGTGCCGCAGACTCAG; JVO-1294, GTTTTGCTAGCGCCAACAATTCCAC.

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

2.2. Fluorescent Analysis 2.3. Immunoblotting

305

Fuji LAS-3000 image analyser or another suitable system. 1. 5× Protein Loading Buffer Pack (Fermentas). 2. Prestained Protein Marker, Broad Range (New England Biolabs). 3. Polyscreen PVDF Transfer Membrane (PerkinElmer, Waltham, MA). 4. Anti-GFP, mixture of two mouse monoclonal antibodies (Roche Applied Science, Indianapolis, IN). 5. Monoclonal anti-FLAG M2 antibody, from mouse (SigmaAldrich, Taufkirchen, Germany). 6. Anti-GroEL antibody (Sigma-Aldrich). 7. ECL anti-mouse IgG, peroxidase-linked whole antibody (GE Healthcare). 8. Western Lightning Chemiluminescence Reagent (PerkinElmer). 9. Roti-Free Western blot stripping solution (Roth, Karlsruhe, Germany). 10. TBST20: 20 mM Tris base, 150 mM NaCl, 0.1% Tween 20 (Roth).

3. Methods To monitor sRNA-mediated target-gfp fusion regulation, sRNA and reporter expression plasmids are co-maintained in E. coli Top10 (deficient in the recombinase recA), to reduce intermolecular recombination between the plasmids (Fig. 1a). Putative target sequences are translationally fused to gfp+ (11) and expressed from a constitutive PLtetO-1 promoter from a low-copy plasmid (pSC101* replicon, 3–4 copies/cell; cat resistance). To ensure efficient target regulation, an artificial surplus of sRNA expression as compared to the target fusion is created by expressing the sRNA from a high-copy plasmid (ColE1 replicon, ~70 copies/cell; bla resistance) under control of a constitutive PLlacO-1 promoter. Based upon the results of previous luciferase expression measurements (12), a tenfold higher transcription of the sRNA as compared to the gfp-fusion is expected. High-level, constitutive expression of both, the sRNA and the target-gfp fusion, largely uncouples the studied regulation from the chromosomal transcriptional network, thereby reducing the complexity of the sRNA/mRNA target pair regulation (9). To investigate mRNAs subjected to sRNA regulation or other types of control (riboswitches, thermosensors), suitable gfp

306

Urban and Vogel

Fig. 1. General approach to study sRNA-mediated translational control. (a) Putative sRNA target sequences are cloned as translational fusions to gfp in a low-copy vector (pSC101* replicon) that confers chloramphenicol resistance. The fusion is transcribed from a constitutive PLtetO-1 promoter. The sRNA is cloned in a high-copy vector (ColE1 replicon) under control of a constitutive PLlacO-1 promoter such that transcription will precisely start at the native +1 site of the sRNA. Upon co-transformation into E. coli, the effect of a given sRNA on a target fusion can be determined by monitoring GFP fluorescence of colonies grown on agar plates, of liquid cultures grown in standard laboratory flasks or in microtitre plates, or by flow cytometry. (b) Combinations of fusion and sRNA expression plasmids with control vectors are used to determine (i) the basal fluorescence of E. coli cells and how it is affected by sRNA overexpression, (ii) the general effect of plasmid-borne sRNA expression on the gfp gene, and (iii) the specific effect of an sRNA on a target fusion (cloned into one of the specialized reporter plasmids pXG-10, -20, or -30) of interest.

reporter gene fusions should include full-length 5¢ UTRs and ~20 coding residues of the gene of interest. These fusions most likely contain all regulatory mRNA elements and can circumvent overexpression of functional full-length protein genes (9). Transformation of E. coli cells harbouring a target-gfp fusion plasmid of interest with plasmids expressing either the cognate regulatory sRNA or a ~50 nt nonsense transcript (transcribed from the control plasmid pJV300) reveals the regulatory potential of the sRNA (Fig. 1b): fusions that exhibit a higher cellular fluorescence yield in the presence of an sRNA plasmid as compared to pJV300 are considered activated, whereas reduced fluorescence indicates target repression. To rule out regulatory effects of sRNA overexpression that do not act on the 5¢ fusion region of a putative target gene, the sRNA expression and pJV300 control plasmids are transformed into strains harbouring additional control plasmids. pXG-0 shares the backbone of all fusion plasmids but expresses a luciferase gene

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

307

(as non-fluorescent control) with an artificial 5¢ UTR containing a strong RBS (12) and reports a potential influence of an sRNA on cellular autofluorescence. In pXG-1, the luciferase gene of pXG-0 is replaced by a full-length gfp+ gene. Combination of pXG-1 with either an sRNA expression plasmid or pJV300 will reveal unintended regulation of fluorescence activity which might occur transcriptionally on the PLtetO-1 promoter, post-transcriptionally on the artificial 5¢ UTR or elsewhere in the gfp mRNA, or by altering the copy number of the reporter plasmid. 3.1. Cloning of sRNA Expression Plasmids

The sRNA gene of interest is cloned under control of the PLlacO-1 promoter such that transcription will precisely start at the native +1 site of the sRNA. This avoids addition of unwanted nucleotides to the 5¢ end of the sRNA transcript, which might interfere with sRNA stability or folding. The cloning protocol includes the PCR amplification of both the plasmid backbone and the sRNA gene and their direct ligation after restriction digestion to obtain the final expression plasmids. Knowledge of the sRNAs transcriptional start site and the approximate location of the transcription termination signal is required and can be obtained from databases (such as coliBASE: http://xbase.bham.ac.uk/colibase/; or EcoCyc: http://ecocyc.org/) for annotated sRNAs or by experimental determinations as described in ref. (13). 1. To obtain DNA for plasmid backbone preparation, a 50-µL PCR reaction is prepared by combining 1 µL (10–50 ng) pJU-334 template, 0.4 µL oligonucleotide PLlacoB, 0.4 µL oligonucleotide JVO-2164, 10 µL 5× PhusionTM HF Buffer, 1 µL dNTP mix (2.5 mM of each dATP, dCTP, dGTP and dTTP), 0.3 µL Phusion High-Fidelity DNA polymerase, and 36.9 µL water. The primers PLlacoB and JVO-2164 pair with the sense and antisense strands at −1 and +1 positions of the plasmid-borne PLlacO-1 promoter, respectively. The PCR reaction is heated at 98°C for 30 s, followed by 30 circles of the following steps: 98°C, 10 s; 58°C, 30 s; 72°C, 2 min 20 s. Incubation is continued at 72°C for 5 min and 5 µL of the reaction is analysed by 0.8% agarose gel electrophoresis for successful amplification of the ~3.1-kbp DNA fragment. 2. 1.5 µL DpnI is directly added to the remaining 45 µL of the PCR reaction, mixed by pipetting and incubated for 3 h at 37°C. DpnI destroys intact template plasmids by recognizing the methylated DNA without affecting the yield of the unmethylated PCR product. 3. The reaction is purified using the NucleoSpin Extract II DNA purification kit and DNA is eluted in 30 µL water. 4. The eluted PCR product is then digested in a 30-µL reaction containing 25 µL eluted DNA, 3 µL 10× TangoTM buffer, 2 µL XbaI for 6 h (or overnight) at 37°C. After the addition of

308

Urban and Vogel

1 µL SAP directly to the reaction, the incubation is continued for 1 h at 37°C. This digestion results in two DNA fragments of ~2.2 kbp and ~0.9 kbp, respectively. The ~2.2kbp fragment carries the PLlacO-1 promoter (from the −1 position), an ampicillin resistance cassette, a ColE1 replicon, and a strong rrnB terminator followed by the sticky end created by XbaI digestion. 5. The ~2.2-kbp DNA fragment is purified by preparative 0.8% agarose gel electrophoresis, and the DNA is eluted in 25 µL water using the NucleoSpin Extract II DNA purification kit. Approximately 0.5–1.0 µg of the ready-to-use plasmid backbone should be obtained at this step. 6. Primers are designed to amplify the sRNA gene of interest from a chromosomal DNA template. The sense primer pairs with the sRNA gene starting from the +1 transcriptional start nucleotide and is 5¢-phosphorylated for blunt end ligation to the −1 position of the PLlacO-1 promoter, located at the 3¢ end of the ‘backbone’ PCR product. The antisense primer pairs ~40 nt downstream of the terminator structure of the sRNA gene and carries a 5¢ extension with an XbaI restriction site and five additional nucleotides for optimal restriction digestion efficiency. 7. As an example, we provide a protocol of the PCR amplification of the E. coli micC sRNA gene in a 25-µL reaction containing 1 µL (~10–50 ng) chromosomal E. coli K12 template DNA, 0.2 µL oligonucleotide JVO-0486, 0.2 µL oligonucleotide JVO-0489, 2.5 µL 10× Pfu Buffer, 0.5 µL dNTP mix, 0.4 µL Pfu DNA polymerase, 20.2 µL water. The PCR protocol is following: 95°C/5 min; 30 circles: 95°C/45 s, 56°C/45 s, 72°C/30 s; 72°C/5 min. 5 µL of the reaction is analysed by 3% agarose gel electrophoresis for successful amplification of a ~160-bp DNA fragment. 8. The reaction is purified using the NucleoSpin Extract II DNA purification kit and the DNA is eluted in 15 µL water. 9. Eluted PCR products are digested in a 10-µL reaction containing 8 µL eluted DNA, 1 µL 10× TangoTM, and 1 µL XbaI for 3 h at 37°C. 10. The resulting ~150-bp DNA fragment is purified by preparative 3% agarose gel electrophoresis and eluted in 15 µL water using the NucleoSpin Extract II DNA purification kit. ~10 ng/µL concentration is expected. 11. Set up a small-scale ligation reaction (5 µL final volume) by mixing ~12 ng XbaI-digested backbone PCR product, ~5 ng XbaI-digested sRNA PCR product, 0.5 µL 10× T4 DNA Ligase Reaction Buffer, 0.5 µL 10× T4 DNA Ligase and incubate 1 h at room temperature. A control ligation where

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

309

the DNA insert is replaced by water should be included to check the religation rate of the vector. 12. Transform E. coli Top10F¢ cells with 2 µL of the reaction. Expect 50–200 colonies, whereas the number of religation events should be negligible. Note that in Top10F¢ cells the PLlacO-1 promoter is silent, due to the repression by a lacIq allele encoded on the F¢ episome. 13. Successful sRNA cloning is verified by colony PCR using primer pairs pZE-A/pZE-XbaI (~166 nt for flanking regions and additional 145 nt for the cloned micC sRNA fragment) or by sequencing of colony PCR products using either primer pZE-A or pZE-XbaI. Upon transformation of the plasmid pPL-MicC, described here, into lacI negative E. coli cells, the micC gene is constitutively expressed from the PLlacO-1 promoter and transcription of the 109 nt sRNA terminates at the predicted native Rho-independent terminator of the gene. Note that unintended read-through transcripts are terminated at the plasmid-borne rrnB terminator. 3.2. Target-gfp Fusion Cloning into pXG-10 (Standard Vector)

Target sequences are derived from either monocistronic genes or the first cistron of an operon, with known native transcriptional start site. These sequences are amplified using PCR and a chromosomal DNA template, and cloned into the standard fusion vector pXG-10 (Fig. 2). The sense primer used in the PCR adds an Mph1103I restriction site (ATGCAT) and pairs with the target gene beginning with the transcriptional start nucleotide (+1 site). The antisense primer is designed to pair with the N-terminal coding region of the gene (including ~20 coding residues) and to add an in-frame NheI restriction site (GCTAGC). The Mph1103I/NheI cloning of the amplified DNA fragment into pXG-10 leads to the expression of a transcript that includes the artificial AUGCAU sequence on its 5¢ end, the 5¢ UTR, and the gene encoding a chimeric gfp fusion protein. Generally, the gfp fusion cloning vectors are maintained in E. coli Top10 and stored as bacterial DMSO stocks at −80°C. 1. To prepare the vector for target fusion cloning, E. coli Top10 cells harbouring the pXG-10 plasmid are taken from the stock. 2. A single colony is inoculated into 4 mL LB containing 20 µg/mL chloramphenicol and grown overnight at 37°C with agitation. 3. Next day, the culture is diluted in 400 mL fresh LB medium and incubated overnight. 4. pXG-10 plasmid is then isolated from the culture using the NucleoBond PC100 plasmid purification kit. The whole culture volume can be purified on a single AX 100 column (note that this differs from the manufacturer’s recommendation; the

310

Urban and Vogel

Fig. 2. Putative target sequences are PCR amplified and cloned into specialized gfp fusion vectors. If the target sequence is derived from a monocistronic gene or the first gene of an operon, and its promoter is known (left panel), it is amplified with an upstream primer that binds at the +1 site of the target gene and adds an Mph1103 I site, and a downstream primer that binds in the N-terminal region of the target gene and adds an Nhe I site in frame with the target gene coding region. The resulting PCR product is inserted into the standard fusion vector, pXG-10, digested with Mph1103 I and Nhe I. If the promoter +1 site is unknown (middle panel ), the target sequence is amplified from cDNA established on total E. coli RNA that was ligated to a 5¢ linker oligonucleotide upon treatment with TAP. The amplified cDNA will carry a 5¢ BseR I site encoded in the linker sequence. Insertion of the BseR I/Nhe I-digested cDNA into Bsg I/Nhe I-digested RACE fusion plasmid, pXG-20, ensures that transcription of the fusion mRNA starts at the native +1 site of the target gene. Target sequences that are derived from within polycistronic mRNA regions are amplified and cloned into the operon fusion vector pXG-30 (right panel ). The upstream primer adds an Mph1103 I site in frame with the C-terminus of the upstream ORF; cloning into pXG-30 will create a C-terminal fusion to a short artificial reading frame composed of a FLAG-epitope and a truncated lacZ gene, thus mimicking operon composition.

column DNA-binding capacity far exceeds the total amount of low-copy plasmids). Use double volumes of washing buffer in each step mentioned in the manufacturer’s protocol. 5. Resuspend DNA in 80 µL water by pipetting and measure concentration. The typical plasmid yield should be ~10–15 µg. 6. 4 mg of the plasmid is then digested with 40 U NheI and 20 U Mph1103I in a 60-mL reaction containing 1× Tango buffer for 7 h or overnight at 37°C. 7. Following the restriction digest, 20 U SAP are added, the sample is mixed, and incubation continues for 3 h at 37°C. Note that the subsequent small-scale ligation requires full dephosphorylation of the vector.

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

311

8. Load the mixture on 1% agarose gel and excise the 4.1-kbp band, corresponding to the pXG-10 vector backbone (lacZ fusion from the parental plasmid migrates at ~600 bp), from the gel. Purify DNA using the NucleoSpin Extract II kit and elute in 50 mL water. The eluted DNA should yield a concentration of ~20–50 ng/mL. 9. As an example, we provide a protocol of the PCR amplification of an E. coli ompC gene fusion fragment. Prepare a PCR reaction using oligonucleotides JVO-0428 and JVO-0429 as described in Subheading 3.1, step 7. Perform PCR using the following protocol: 95°C/5 min; 30 circles, 95°C/45 s, 58°C/45 s, 72°C/45 s; 72°C/5 min. The reaction should produce a 137-bp DNA fragment, containing the Mph1103I restriction site (from JVO-0428) directly upstream of the native ompC +1 site, the 81 nt 5¢ UTR of the gene and the first 12 amino-terminal coding residues followed by an in-frame NheI site (introduced by JVO-0429). 10. The PCR reaction is purified using the NucleoSpin Extract II kit and DNA is eluted in 15 mL water. 11. 12 mL of the PCR product is then digested with Mph1103I and NheI (7.5 U each) in a total reaction volume of 15 mL containing 1× Tango buffer for 3 h at 37°C. 12. The resulting ~120-bp DNA fragment is purified as described in Subheading 3.1, step 10. 13. Perform ligation and tranformation using Mph1103I/NheIdigested pXG-10 backbone and Mph1103I/NheI-digested ompC PCR product as described in Subheading 3.1, steps 11 and –12, but use E. coli Top10 as host strain for transformation. Note, that in E. coli Top10 cells the PLtetO-1 promoter is constitutively active, leading to expression of the established gfp fusion. 14. Successful fragment insertion is verified by colony PCR using primer pair pZE-CAT/JVO-0155 (418 bp for flanking regions, additional ~114 bp for ompC fusion inserts, and ~1,010 bp for religated pXG-10), by colony fluorescence (pXG-10 itself has poor fluorescence) or by sequencing the colony PCR products using either primer pZE-CAT or JVO-0155. 3.3. Target-gfp Fusion Cloning into pXG-20 (RACE Vector)

The RACE (rapid amplification of cDNA ends) vector pXG-20 was developed for fusion of a target gene with unknown transcriptional start (Fig. 2). RACE is frequently used to determine the 5¢ ends of transcripts and can be used to distinguish primary transcripts (originating from the +1 site of a promoter) from processed RNA species (14). In brief, total cellular RNA is split into two aliquots, and one of them is treated with TAP to convert the 5¢ triphosphates of primary transcripts to monophosphates.

312

Urban and Vogel

In a subsequent step, the 3¢ hydroxyl group of an RNA linker of known sequence is ligated to the 5¢ end of all RNA species in both samples. Following cDNA synthesis, a PCR with an RNA linker-specific sense and a gene-specific antisense primer is performed. Since TAP treatment provides 5¢ monophosphate substrates for the RNA ligase and results in a higher RNA-linker ligation efficiency, PCR products of 5¢ full-length transcripts are significantly enriched in the TAP-treated sample. To allow direct cloning of TAP-enriched PCR products as 5¢ full-length in-frame gfp fusions, a specialized RNA linker (A4) is used. This linker encodes a recognition sequence for the restriction enzyme BseRI, which becomes functional upon cDNA synthesis. BseRI cleaves the sense strand 10 nt downstream of the recognition sequence (which is the site of linker ligation) and leaves a 3¢AA overhang originating from the linker sequence in this strand. TAP-enriched PCR products, obtained with an RNA linker-specific sense primer and an antisense primer which contains an NheI site and pairs with the N-terminal coding region of the target gene, can be directly inserted into pXG-20. This vector contains BsgI and NheI restriction sites and a gfp gene lacking a start codon. BsgI cleaves the sense strand 14 nt upstream of its recognition sequence (which is the +1 site of PLtetO-1) and leaves a 3¢-TT overhang (which is the -1 and -2 position of the promoter) on the antisense strand. Ligation of a BseRI/NheI-digested RACE product to BsgI/NheItreated pXG-20 vector produces a gfp fusion that is transcribed precisely from the gene’s native +1 site. 1. The preparation of pXG-20 for cDNA fusion cloning follows the protocol described for pXG-10 preparation, but 4 mg of pXG-20 is digested with 40 U NheI and 20 U BsgI in a 60-mL reaction containing 1× NE Buffer 4 and 1× SAM for 7 h or overnight at 37°C. 2. To obtain an insert cDNA for reporter fusion cloning, DNA-free total bacterial RNA is required and should be ideally prepared under growth conditions which are known to express the gene of interest. If such conditions are unknown, it is advantageous to analyse RNA prepared at logarithmic and stationary phases in parallel. We routinely use Promega’s SV Total RNA Isolation System according to the manufacturer’s protocol or as described at www.ifr.ac.uk/safety/microarrays/protocols.html to prepare DNA-free total RNA from E. coli and other enterobacteria. 3. As example for cDNA target fusion cloning, the construction of an E. coli ompA (encoding the outer membrane protein A) fusion is outlined later. Other target fusion inserts can be easily obtained from the same cDNA preparation using an antisense oligonucleotide specific for the genes of interest. 4. To convert 5¢ triphosphates of primary transcripts to monophosphates, 12 mg total bacterial RNA is dissolved in 87.5 mL RNase-

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

313

free water. When 10 mL 10× TAP buffer and 0.5 mL RNase inhibitor have been added, the sample is mixed and split into two reactions (49 mL each). To one of the reactions 1 mL TAP enzyme is added and both samples are incubated for 30 min at 37°C. At the end of the incubation, 300 pmol of RNA oligonucleotide A4 is added to each reaction. 5. Buffer and enzyme are removed by PCI extraction: 50 mL RNase-free water is added to each sample, the reactions transferred to 0.5-mL PLG tubes and 100 mL of PCI mixture are added. After vigorous shaking for 30 s, the tubes are centrifuged at 13,000 × g for 15 min at room temperature. The aqueous phases (~100 mL) are transferred to fresh Eppendorf tubes. 6. The RNA is then precipitated upon addition of 300 mL ethanol–Na–acetate mixture for 1 h on ice and pelleted by centrifugation at 13,000 × g for 40 min at 4°C. The supernatants are carefully discarded using a pipette. The pellets are washed once by adding 200 mL 75% ethanol, the samples are centrifuged and the supernatants discarded as before. After air drying the pellets at room temperature, each RNA is dissolved in 13.5 mL water. 7. To ligate the A4 linker, the samples are denatured for 2 min at 90°C and subsequently chilled on ice for 5 min. After collecting the samples by point centrifugation, the tubes are set back on ice and 2 mL DMSO, 2 mL 10× T4 RNA Ligase Reaction Buffer, 2 mL T4 RNA Ligase, and 0.5 mL RNase inhibitor are added per tube. The samples are then incubated for 12 h at 17°C. 8. Following PCI extraction (the volume is adjusted to 100 mL by adding 80 mL water) and ethanol precipitation (Subheading 3.3, steps 5 and 6), the RNA pellets are air dried, dissolved in 10 mL water and the RNA concentration is measured. 9. For cDNA synthesis, 2 mg linker-ligated RNAs are mixed with 100 pmol random hexamer primers in a total volume of 10 mL adjusted with RNase-free water. The samples are transferred to 0.2-mL PCR tubes and placed in a PCR machine. The following steps can be programmed in a single routine, but it is important to add the required reagents and enzymes on time. 10. The samples are denatured for 5 min at 65°C and placed on ice. 9 mL of reverse transcriptase enzyme/buffer mix is added per tube. The mix is composed of components of the Superscript III RT-PCR system: 4 mL 5× first-strand buffer, 1 mL 10 mM dNTP mix, 2.5 mL RNase-free water, 1 mL 0.1 M DTT, and 0.5 mL RNase inhibitor. The samples are mixed and incubated for 10 min at 25°C. Then 1 mL Superscript III RT is added and the reaction is subsequently incubated for 15 min at each of the following temperatures: 41°C, 50°C, 55°C, 60°C. The

314

Urban and Vogel

RT enzyme is inactivated for 5 min at 85°C, the samples cooled down to 37°C, 1 mL RNase H is added, and the incubation proceeds at 37°C for 20° min. The ready-to-use cDNA is then kept on ice or stored at −20°C until PCR amplification. 11. For amplification of ompA gene fusion fragments, three PCR reactions (40 µL final volume) are performed using either 1 µL of the cDNA from untreated or TAP-treated samples, and 100 ng chromosomal E. coli K12 DNA (negative control) as template. The reactions also contain 0.2 µL of each oligonucleotide JVO-0367 and JVO-0433, 4 µL 10× ThermoPol Reaction Buffer, 1 µL 10 mM dNTP mix, 0.2 µL Taq DNA polymerase, and 33.4 µL water. The sense oligonucleotide JVO-0367 pairs with the sequence of the RNA linker (upstream of the encoded BseRI site), whereas JVO-0433 pairs with the antisense region of the ompA coding sequence starting from the 16th codon and carrying a NheI site extension in frame. The PCR protocol is the following: 95°C/10 min; 35 circles: 95°C/40 s, 57°C/40 s, 72°C/45 s; 72°C/10 min. 12. Half of the reactions are then analysed along with an appropriate DNA size marker by 3% agarose gel electrophoresis. The protocol described here typically amplifies four products specific for the ompA transcript (Fig. 3b). 13. The product that is enriched in the TAP-treated sample (band 2 in Fig. 3b for the ompA cDNA) represents fulllength transcripts arising from the native +1 site of the cognate promoter. TAP-enriched fragments are excised from the gel, the DNA purified and eluted in 12 µL water using the NucleoSpin Extract II DNA purification kit. 14. The cDNA is then digested in a 12-µL reaction containing 8.4 µL eluted DNA, 1.2 µL 10× NE Buffer 2, 1.2 µL 10× BSA, 0.6 µL BseRI, and 0.6 µL NheI for 3 h at 37°C. 15. The digestion is purified via preparative 3% agarose gel electrophoresis and eluted in 12 µL water using the NucleoSpin Extract II DNA purification kit. Note that in some cases the amount of TAP-specific DNA might not be sufficient for direct cloning and a 2nd PCR (using 1 mL of eluted DNA from step 13 as template) may be required to obtain more DNA for cloning. 16. A small-scale ligation reaction containing BsgI/NheI-digested pXG-20 backbone and BseRI/NheI-digested cDNA insert is set up as described in Subheading 3.1, step 11. 17. Transform E. coli Top10 cells with 2 µL of the reaction. Expect 10–50 colonies (without 2nd step PCR amplification of the cDNA), whereas the number of religation events should be negligible. 18. Verify the insertion by colony PCR using primer pair pZE-CAT/JVO-0155 (~1,400 bp for religated pXG-20),

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

315

Fig. 3. sRNA-mediated target-gfp fusion regulation monitored by in vivo colony fluorescence imaging and immunoblotting. (a) Colony fluorescence of E. coli strains harbouring pXG-1, pXG-0, or the ompC::gfp fusion plasmid in combination with either the mock control vector or the cognate pPL-MicC expression plasmid, respectively. Strains were grown overnight on LB-agar plates and bacterial colony images were taken in the fluorescence mode (upper panel) or visible light mode (lower panel). Pairing of the MicC sRNA to the ompC translation initiation region blocks ribosome entry(16) and results in inhibition of ompA::gfp synthesis. (b) cDNA cloning and sRNA-mediated regulation of an E. coli ompA::gfp fusion. Left panel, 3% agarose gel electrophoresis of RACE-amplified E. coli ompA cDNA for gfp-fusion cloning into pXG20. PCR was performed with linker A4- and ompA-specific primers. Direct cloning of the PCR product enriched in the TAP-treated sample (band 2), representing ompA primary transcripts, into pXG-20 results in 5¢ end full-length in-frame gfp-fusion cloning. A control PCR carried out using chromosomal E. coli DNA (sample: K12) and a DNA size marker were co-migrated on the gel. Right panel, colony fluorescence imaging of strains harbouring the ompA::gfp fusion in combination with either pJV300 (mock control) or a plasmid expressing the cognate MicA sRNA (pPL-MicA). Annealing of the MicA sRNA to sequences covering the ompA ribosome-binding site blocks ribosome entry (17) and inhibits ompA::gfp synthesis. (c) Schematic drawing of the FlacZ::glmU-glmS::gfp operon fusion (see also ref. (18)). The dual reporter fusion is expressed from a PLtetO promoter and contains an internal 234 bp fragment (covering the 161-bp non-coding IGR) of the dicistronic E. coli glmUS mRNA: in the first cistron, a translational fusion between the last 17 coding residues of GlmU (C-terminus) and a FLAG-epitope-tagged LacZ peptide is created, whereas the second cistron fuses the first 7 coding residues of GlmS (N-terminus) to GFP. (d) The GlmZ sRNA mediates the discoordinate expression of the glmUS::gfp operon. Whole-cell protein extracts of strains harbouring the glmUS::gfp operon fusion (shown in panel C) in combination with either pJV300 or a GlmZ sRNA expression plasmid (pPL-GlmZ) were subjected to Western blot analysis (left panel). GlmS::GFP or FLacZ::GlmU fusion proteins were detected with anti-GFP or anti-FLAG antibodies, respectively. Detection of the GroEL protein on the same membrane was used to confirm equal loading. The 5¢ UTR of glmS::gfp mRNA can fold into a self-inhibitory stem-loop that masks the glmS ribosome-binding site and represses glmS translation. Annealing of the GlmZ sRNA to the inhibitory hairpin is suggested to induce an alternative 5¢ UTR structure that allows ribosome entry and enhanced translation of GlmS::GFP without affecting FlacZ::GlmU synthesis (19). Colony fluorescence imaging (carried out as in panel A) of the strains used for Western blot analysis is shown in the right panel.

316

Urban and Vogel

by colony fluorescence (pXG-20 itself mediates no fluorescence) or by sequencing of colony PCR products using either primer pZE-CAT or JVO-0155. In case of the described ompA cDNA fusion cloning procedure, sequencing will confirm the cloning of full-length inserts starting from the ompA transcriptional +1 site up to the 16th residue of OmpA fused in frame with GFP. 3.4. Target-gfp Fusion Cloning into pXG-30 (Operon Vector)

Some sRNAs are known to target untranslated regions within polycistronic mRNAs to specifically regulate expression of a flanking cistron. Fusion cloning of intra-operon target sites is challenging for two main reasons. On the one hand, cloning of full-length fusions of the cistron of interest should be avoided, since intact upstream cistrons would be overexpressed and could cause pleiotropic effects in the cell. On the other hand, truncation of the fusion to an arbitrary 5¢ end in proximity to the cistron of interest might destabilize the fusion mRNA. The operon fusion vector pXG-30 enables the cloning of target sequences, located within an operon, to monitor the expression of the flanking genes by establishing dual reporter fusions (Fig. 2). pXG-30 mimics a dicistronic operon whose expression is driven by a PLtetO-1 promoter followed by a strong RBS. The first cistron is a 3× FLAG epitope-tagged short artificial ORF derived from the N-terminal coding region of the E. coli lacZ gene (FlacZ ¢). The second cistron is gfp. PCR-based directed cloning of an intraoperon target site fuses the C-terminal portion of the upstream target cistron with FlacZ ¢ and the N-terminal region of the downstream cistron with gfp. The sense and antisense primers used in the PCR introduce in-frame Mph1103I and NheI restriction sites for cloning into the cognate sites in pXG-30. In addition to monitoring gfp fusion expression by fluorescence imaging, expression of the FlacZ¢ and gfp cistrons is simultaneously studied by immunoblotting using anti-FLAG and anti-GFP antibodies, respectively. 1. The preparation of pXG-30 for dicistronic fusion cloning follows essentially the protocol described for pXG-10 preparation, since the same restriction sites are used to substitute the fusion module. In the parental pXG-30 plasmid, a 320-bp insert derived from the E. coli galETKM operon is inserted into the Mph1103I/NheI sites. This creates a C-terminal fusion of the last 58 coding residues of galT to FlacZ ¢ and an N-terminal fusion of the first 58 coding residues of galK to gfp. 2. As an example, we provide construction of a dual reporter fusion derived from the dicistronic E. coli glmUS operon (encoding essential functions in aminosugar metabolism). An internal 235 nt fragment of the glmUS operon is amplified by PCR using oligonucleotides JVO-1270 and JVO-1294 as described in Subheading 3.1, step 7. The sense primer JVO1270 pairs in the C-terminal coding region of glmU and adds an in-frame Mph1103I restriction site upstream of the last 17

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

317

coding residues, whereas the antisense primer JVO-1294 pairs in the N-terminal coding region of glmS and adds an in-frame NheI restriction site downstream of the first 7 coding residues of glmS. The PCR protocol is following: 95°C/5 min; 30 circles: 95°C/45 s, 57°C/45 s, 72°C/45 s; 72°C/5 min. 5 µL of the reaction is analysed by 3% agarose gel electrophoresis for successful amplification of a 255-bp DNA fragment. 3. Purification of the PCR fragment, digestion with Mph1103I/NheI, and cloning of the fragment into pXG-30 is carried out as described for cloning into pXG-10 (Subheading 3.2, steps 10–13). 4. Successful insertion is verified by colony PCR using primer pair pZE-CAT/JVO-0155 (control sizes: ~1.0 kb for the parental uncut pXG-30; ~680 bp for religated pXG-30 that has lost the parental insert), by change of colony fluorescence (pXG-30 shows high fluorescence), or by sequencing using primer JVO-0155. 3.5. Investigating Target-gfp Fusion Regulation

To assess the regulatory potential of sRNA expression on a translational target-gfp fusion under investigation, the sRNA and fusion plasmids are co-maintained in the same cell. The expression of both the sRNA (driven by the PLlacO-1 promoter) and the gfp fusion (driven by the PLtetO-1 promoter) can be tightly regulated in E. coli host strains that encode the LacI or TetR repressor proteins, respectively (12). Induced expression upon the addition of the appropriate inducer (IPTG or anhydroteracycline) might be desirable if investigation of short-term regulatory effects is required, or high transcription rate yields toxic RNA levels. For constitutive sRNA/target-gfp fusion expression, we routinely use E. coli TOP10 cells (negative for both, Lacl and TetR expression) which are recA- and thus better suited for two-plasmid systems (9).

3.5.1. Fluorescent Techniques

To combine the target-gfp fusion and sRNA expression plasmids within the same host cell, sRNA plasmids are transformed into competent cells of strains harbouring the gfp fusion plasmid of interest (see Note 1) and plated on LB-agar plates supplemented with ampicillin and chloramphenicol (100 mg/mL and 20 mg/ mL, respectively). In vivo whole-cell colony plate fluorescence imaging is a well-suited method to rapidly assess target fusion regulation (see Note 2). Herein, E. coli strains harbouring the target-gfp fusion of interest in combination with the sRNA expression plasmid or the pJV300 control are re-streaked next to each other onto a fresh LB-agar plate. For fluorescence comparison, care should be taken to obtain similar-sized single colonies of both strains upon overnight growth at 37°C. To visualize GFP-fusion protein accumulation in bacterial colonies, the LB-agar plate is photographed in an image analyser using a CCD camera after 2 s excitation at 460 nm with a 510-nm emission filter. An additional picture of

318

Urban and Vogel

the same plate in the visual light mode is taken to compare the colony morphology of the bacterial strains. Exemplary results for sRNA-mediated translational repression obtained with the earlier-established ompC::gfp and ompA::gfp fusions are shown in Fig. 3a, b, respectively. 3.5.2. Western Blotting Analysis

1. In addition to its function as fluorescent reporter for in vivo imaging, the GFP domain of the translational gene fusions represents a uniform tag that can be exploited in immunoblotting assays. Western blot analysis of E. coli whole-cell protein extracts using monoclonal antibodies against GFP enables to assess fusion protein accumulation quantitatively and with high sensitivity. An example result for sRNA-mediated translational activation obtained with the earlier-established dicistronic FlacZ::glmU-glmS::gfp operon fusion is shown in Fig. 3d. 2. We are generally using standard procedures for protein crude extract preparation and Western blot analysis. For a detailed protocol for Western blot detection of chimeric reporter fusion proteins see (9). The protein samples are typically loaded on the gel using 5× Protein Loading Buffer Pack and run along with Prestained Protein Marker. The anti-GFP, anti-FLAG, and anti-GroEL antibodies are used as 1:1,000; 1:1,000; and 1:20,000 dilutions in TBST20, respectively. Working solutions of anti-GFP and anti-FLAG antibodies can be stored at −20°C and reused up to 20 times without significant loss of detection properties. In contrast, the anti-GroEL solution should be prepared freshly, since the antibodies are directly conjugated to the horseradish peroxidase enzyme. The secondary antibody (anti-mouse IgG) is used as 1:5,000 dilution in TBST20. The antibodies are visualized using Western Lightning Chemiluminescence Reagent. Simultaneous detection of the proteins on the same membrane is achieved by cutting the PVDF membrane into pieces to enable parallel hybridization with the different antibody solutions, or by step-wise stripping and re-hybridization using Roti-Free Western blot stripping solution and standard protocols. The molecular weight of the GFP and FlacZ moieties (i.e. without fused peptides), and of GroEL protein is ~27 kDa, ~9 kDa, and ~60 kDa, respectively.

4. Notes 1. We routinely prepare competent E. coli strains harbouring gfp fusion plasmids by calcium chloride method. 2. Colony fluorescence imaging is a reliable and the fastest way to monitor target-gfp fusion regulation and ideally suited for

A Green Fluorescent Protein (GFP)-Based Plasmid System to Study

319

preliminary investigations. However, it should be noted that this method is not quantitative, since colony growth does not reflect a defined growth stage, due to heterogeneous gene expression within the different cell layers (15). Moreover, the fluorescence of chimeric gfp fusion proteins differs over a broad range (9) and a good fluorescence signal of the GFP-fusion under investigation is required for unambiguous visualization.

References 1. Vogel, J. and Wagner, E.G. (2007). Target identification of regulatory sRNAs in bacteria. Curr. Opin. Microbiol. 10, 262–270 2. Aiba, H. (2007). Mechanism of RNA silencing by Hfq-binding small RNAs. Curr. Opin. Microbiol. 10, 134–139 3. Valentin-Hansen, P., Eriksen, M. and Udesen, C. (2004). The bacterial Sm-like protein Hfq: a key player in RNA transactions. Mol. Microbiol. 51, 1525–1533 4. Mizuno, T., Chou, M.Y. and Inouye, M. (1984). A unique mechanism regulating gene expression: translational inhibition by a complementary RNA transcript (micRNA). Proc. Natl. Acad. Sci. U.S.A. 81, 1966–1970 5. Majdalani, N., Cunning, C., Sledjeski, D., Elliott, T. and Gottesman, S. (1998). DsrA RNA regulates translation of RpoS message by an anti-antisense mechanism, independent of its action as an antisilencer of transcription. Proc. Natl. Acad. Sci. U.S.A. 95, 12462–12467 6. Lease, R.A., Cusick, M.E. and Belfort, M. (1998). Riboregulation in Escherichia coli: DsrA RNA acts by RNA:RNA interactions at multiple loci. Proc. Natl. Acad. Sci. U.S.A. 95, 12456–12461 7. Møller, T., Franch, T., Udesen, C., Gerdes, K. and Valentin-Hansen, P. (2002). Spot 42 RNA mediates discoordinate expression of the E. coli galactose operon. Genes Dev. 16, 1696–1706 8. Massé, E. and Gottesman, S. (2002). A small RNA regulates the expression of genes involved in iron metabolism in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 99, 4620–4625 9. Urban, J.H. and Vogel, J. (2007). Translational control and target recognition by Escherichia coli small RNAs in vivo. Nucleic Acids Res. 35, 1018–1037 10. Sittka, A., Pfeiffer, V., Tedin, K. and Vogel, J. (2007). The RNA chaperone Hfq is essential for the virulence of Salmonella typhimurium. Mol. Microbiol. 63, 193–217

11. Scholz, O., Thiel, A., Hillen, W. and Niederweis, M. (2000). Quantitative analysis of gene expression with an improved green fluorescent protein. Eur. J. Biochem. 267, 1565–1570 12. Lutz, R. and Bujard, H. (1997). Independent and tight regulation of transcriptional units in Escherichia coli via the LacR/O, the TetR/O and AraC/I1-I2 regulatory elements. Nucleic Acids Res. 25, 1203–1210 13. Vogel, J., Bartels, V., Tang, T.H., Churakov, G., Slagter-Jager, J.G., Hüttenhofer, A. and Wagner, E.G. (2003). RNomics in Escherichia coli detects new sRNA species and indicates parallel transcriptional output in bacteria. Nucleic Acids Res. 31, 6435–6443 14. Bensing, B.A., Meyer, B.J. and Dunny, G.M. (1996). Sensitive detection of bacterial transcription initiation sites and differentiation from RNA processing sites in the pheromone-induced plasmid transfer system of Enterococcus faecalis. Proc. Natl. Acad. Sci. U.S.A. 93, 7794–7799 15. Shapiro, J.A. (1998). Thinking about bacterial populations as multicellular organisms. Annu. Rev. Microbiol. 52, 81–104 16. Chen, S., Zhang, A., Blyn, L.B. and Storz, G. (2004). MicC, a second small-RNA regulator of Omp protein expression in Escherichia coli. J. Bacteriol. 186, 6689–6697 17. Udekwu, K.I., Darfeuille, F., Vogel, J., Reimegard, J., Holmqvist, E. and Wagner, E.G. (2005). Hfq-dependent regulation of OmpA synthesis is mediated by an antisense RNA. Genes Dev. 19, 2355–2366 18. Urban, J.H., Papenfort, K., Thomsen, J., Schmitz, R.A. and Vogel, J. (2007). A conserved small RNA promotes discoordinate expression of the glmUS operon mRNA to activate GlmS synthesis. J. Mol. Biol. 373, 521–528 19. Urban, J.H. and Vogel, J. (2008). Two seemingly homologous noncoding RNAs act hierarchically to activate glmS mRNA translation. PLoS Biol. 6, e64

Chapter 23 High-Throughput Screens to Discover Synthetic Riboswitches Sean A. Lynch, Shana Topp, and Justin P. Gallivan Summary Synthetic riboswitches constructed from RNA aptamers provide a means to control bacterial gene expression using exogenous ligands. A common theme among riboswitches that function at the translational level is that the RNA aptamer interacts with the ribosome-binding site (RBS) of a gene via an intervening sequence known as an expression platform. Structural rearrangements of the expression platform convert ligand binding into a change in gene expression. While methods for selecting RNA aptamers that bind ligands are well established, few general methods have been reported for converting these aptamers into synthetic riboswitches with desirable properties. We have developed two such methods that not only provide the throughput of genetic selections, but also feature the quantitative nature of genetic screens. One method, based on cell motility, is operationally simple and requires only standard consumables; while the other, based on fluorescence-activated cell sorting (FACS), is particularly adept at identifying synthetic riboswitches that are highly dynamic and display very low levels of background expression in the absence of the ligand. Here we present detailed procedures for screening libraries for riboswitches using the two methods. Key words: Riboswitch, High-throughput screen, E. coli motility, FACS

1. Introduction Synthetic riboswitches are ligand-dependent genetic control systems that can be used to report on cellular metabolism, to construct synthetic gene circuits, or to reprogram cellular behavior. Because designing new synthetic riboswitches from existing RNA aptamers (1) remains challenging, genetic screens and selections represent powerful tools for discovering new synthetic riboswitches. Developing a screen or selection to discover new synthetic riboswitches presents a special challenge compared to traditional Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_23 © Humana Press, a part of Springer Science + Business Media, LLC 2009

321

322

Lynch, Topp, and Gallivan

methods employed in the directed evolution of enzymes because an assay for riboswitches must be able to quantitatively assess both the “on” and the “off” states of gene expression. Such counterselections, which select for one phenotype and against another, are powerful but can be difficult to implement. These challenges are further complicated if one wishes to exhaustively screen an entire library or to isolate only very rare events, such as those switches that display both high activation ratios and low background levels of gene expression. Because enzymatically screening large libraries (>100,000 members) can be cumbersome, time consuming, labor- and reagent-intensive, we have developed two selection methods that can be used independently, or in concert, to convert ligand-binding RNA aptamers into new synthetic riboswitches with favorable properties. Both methods are operationally simple and do not require the library members to be arrayed individually, yet rapidly yield synthetic riboswitches with dynamic activation ratios and low levels of background expression (2, 3). These methods may be used to discover riboswitches starting from a variety of RNA aptamers (1), and we expect that the resulting synthetic riboswitches will be generally useful for controlling bacterial gene expression. 1.1. Motility-Based Selection

Our laboratory previously developed a high-throughput robotic screen that was used to identify ligand-dependent riboswitches that displayed low background levels of gene expression in the absence of a ligand, and activated gene expression 36-fold in the presence of the ligand (4). We subsequently used these riboswitches to regulate expression of the E. coli cheZ gene to control bacterial motility in a ligand-dependent fashion (5) (see Note 1). Because the differences in cell motility at different ligand concentrations were easy to distinguish using only a ruler, we hypothesized that motility could be used as a reporter phenotype in a high-throughput selection to discover new synthetic riboswitches from large libraries. We envisioned that this method could equal, if not exceed, the throughput of our previously reported robotic screen (4) and could be performed at a fraction of the cost. The riboswitches identified using this assay have low background levels of gene expression without ligand and achieve 16- to 24-fold increases in gene expression in the presence of ligand (3).

1.2. FACS-Based Selection

To further increase the throughput of our enzymatic screen (4), we chose to employ Fluorescence-Activated Cell Sorting (FACS) (6). Using a fluorescent reporter gene, FACS enables the analysis of large libraries (>1 × 108) of mutant bacteria to isolate those bacteria with the desired fluorescent phenotype. Here we describe a method employing FACS to identify and rapidly

High-Throughput Screens to Discover Synthetic Riboswitches

323

enrich the population of desirable riboswitch library members, i.e., those maintaining minimal levels of gene expression in the absence of a ligand while showing sizeable (>30-fold) increases in expression in the presence (2). This enrichment increases the likelihood of isolating highly dynamic riboswitches using a subsequent enzymatic screen, thereby avoiding the difficulties associated with enzymatically screening the entire library. The screen proceeds in two sorting steps. In the first, bacteria displaying low levels of fluorescence are collected from the library of potential riboswitches using FACS. The collected bacteria are then cultured overnight and subsequently grown in either the presence or absence of the ligand of interest. Analyses of these cultures should reveal a rare population of bacteria that display a ligand-dependent increase in fluorescence. These rare members can then be isolated and their riboswitches cloned upstream of a second reporter gene (lacZ) to rigorously quantify their behavior (7).

2. Materials 2.1. Library Construction and Motility-Based Selection

1. Luria-Bertani media (EMD Biosciences, Gibbstown, NJ). 2. Ampicillin (Fisher, Pittsburgh, PA), final concentration 50 µg/mL of media for all procedures. 3. Tryptone broth: 10 g/L Bacto Tryptone (Difco) and 5 g/L NaCl. 4. Tryptone motility agar, prepared as Tryptone broth with 2.5 g/L Bacto Agar (Difco). 5. Ligand of interest. 6. TOP10F’, electrocompetent Escherichia coli (Invitrogen, Carlsbad, CA). 7. JW1870, electrocompetent E. coli (DcheZ, Keio collection (8)). 8. Sterile Petri dishes, 85 mm (Fisher). 9. Bioassay tray, 241 mm × 241 mm (Nalgene, Rochester, NY). 10. QIAprep Spin Miniprep Kit (Qiagen, Valencia, CA).

2.2. Library Construction and FACS-Based Selection

1. PBS buffer: 177 mM NaCl, 2.7 mM KCl, 5.3 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4. 2. pDsRedExpress vector (Clontech, Mountain View, CA). 3. Becton Dickinson FACSVantage SE flow cytometer using an Innova70 spectrum laser tuned to 568 nm for excitation. Fluorescence was detected through a 630/22 bandpass filter.

324

Lynch, Topp, and Gallivan

2.3. Verification of Functional Riboswitches

1. Bioassay trays, 241 mm × 241 mm (Nalgene). 2. 96-well microtiter plates (Costar). 3. Lysis solution: use as 10:1 Pop Culture® (Novagen, Madison, WI):lysozyme (4 U/mL) mixture. 4. Z-buffer: 60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, 50 mM 2β-mercaptoethanol, pH 7.0. 5. Ortho-nitrophenyl-2-D-galactopyranoside (ONPG) (SigmaAldrich) solution: 4 mg/mL in 100 mM NaH2PO4, pH 7. 6. 1 M Na2CO3. 7. Chloroform. 8. 0.1% SDS. 9. X-Gal solution (US Biological, Swampscott, MA): 6.25 mg/ mL in dimethyl formamide (DMF). 10. Microplate reader (BioTek). 11. Multichannel pipettor.

3. Methods The following protocols assume that the user has identified an RNA aptamer (or series of aptamers) (1) that is (are) known to bind a nontoxic, cell-permeable ligand. The aptamer should be cloned 14–18 bases upstream of the start codon of a reporter gene (see Note 2), within a high-copy number plasmid that confers ampicillin resistance. Motility-based selections are performed with the cheZ (9) reporter gene, while FACS-based selections described here use the DsRedExpress reporter gene. In principle, any fluorescent reporter with absorption and emission characteristics that match the capabilities of the cell sorter can be used. It is advisable to use a constitutive promoter for both selection methods, although the optimal promoter strength depends upon which selection method is employed (see Notes 3 and 4). After performing selection experiments using either method, the library of enriched riboswitches can be subcloned into a lacZ vector to quantify their performance (7). Since cheZ, DsRedExpress, and lacZ readily accept N-terminal protein fusions, cloning steps may be simplified by introducing a unique restriction site within the context of a translational fusion at the N-terminus of either reporter gene. In addition to facilitating DNA manipulations, the short peptide fusion provides a second function: because riboswitch expression platforms can potentially interact with the aptamer or with ribonucleotides that encode for amino acids at the N-terminus of the reporter protein, incorporation of

High-Throughput Screens to Discover Synthetic Riboswitches

325

Fig. 1. A general strategy for constructing a riboswitch library using oligonucleotide-based cassette mutagenesis. Cassette A is generated using forward primer 1 and reverse primer 2, with primer 1 annealing 5¢ to the aptamer and reverse primer 2 containing the 10–12 base randomized region. The randomized region is flanked by a constant region complementary to the aptamer and a constant region that includes 4–6 constant bases and the start codon. Cassette B is created with forward primer 3 (the reverse complement of primer 2) and reverse primer 4, which anneals to a region located within the reporter gene. Following amplification and gel purification of each cassette, A and B can then be mixed together and assembled using outer primers 1 and 4 to yield PCR product C.

a short peptide fusion ensures that the sequence immediately 3¢ to the expression platform is constant throughout the selection and verification steps. It is best to allow the selection experiment to identify an optimal ribosome-binding site (RBS) for the expression platform; therefore, a fully randomized region of 10–12 bases is introduced between the aptamer and a region of 4–6 constant bases located immediately before the start codon. The sequence of these bases is not critical and an example is shown in Fig. 1. This constant region is important to provide suitable spacing between the start codon and the RBS that will be selected as part of the expression platform. Figure 1 illustrates a general strategy for constructing a riboswitch library using oligonucleotide-based cassette mutagenesis. The assembled riboswitch library (PCR product C) should then be digested with the appropriate restriction enzymes, gel purified, and cloned into the initial vector digested with the same enzymes. To prevent recircularization of the vector, the digested plasmid should be dephosphorylated with CIP prior to ligation. To achieve high transformation efficiency in E. coli, the ligation reaction (20 µL) should be precipitated at 4°C with butanol (10 volumes), washed with ethanol (15 volumes), and redissolved in sterile water (10 µL) to remove salts before introducing the DNA into freshly prepared, electrocompetent cells. 3.1. Motility-Based Selection

1. Transform electrocompetent TOP10F’ E. coli cells with the library ligation reaction and let cells recover for 1 h at 37°C with shaking (250 rpm). To determine the library size, plate 1 µL of recovered cells on a Petri dish (85 mm) with LB/ agar, supplemented with ampicillin. Grow overnight at 37°C. Plate the remaining cells on a large (241 mm × 241 mm) bioassay tray containing 300 mL of LB/agar, supplemented

326

Lynch, Topp, and Gallivan

with ampicillin. Grow at 37°C until colonies become visible (8–10 h). 2. Count the number of colonies on the small plate and then calculate the total library size. Scrape the large bioassay tray and suspend cells in 5 mL of LB supplemented with ampicillin. Incubate for ~2 h at 37°C with shaking (250 rpm). 3. Extract DNA from 3 mL of culture with the QIAprep Spin Miniprep Kit. 4. Transform electrocompetent JW1870 E. coli cells (8) with the plasmid library, and let cells recover for 1 h at 37°C with shaking (250 rpm). Plate the recovered cells on a large (241 mm × 241 mm) bioassay tray containing 300 mL of LB/agar, supplemented with ampicillin. Grow at 37°C until colonies become visible (8–10 h). 5. Scrape the bioassay tray and suspend cells in 5 mL of tryptone broth supplemented with ampicillin. Incubate for ~2 h at 37°C with shaking (250 rpm). 6. Use 100 µL of this culture to inoculate 5 mL of tryptone broth supplemented with ampicillin, and grow at 37°C with shaking for approximately 3 h to an OD600 between 0.5 and 0.7. Dilute the culture in tryptone broth to an OD600 of 0.2 (~200,000 cells/µL). 7. Spot 3 µL of diluted culture (~600,000 cells) at the center of duplicate motility agar plates supplemented with ampicillin and containing 0 mM or 1 mM ligand (see Note 5). Let cells dry in air for 10 min; then incubate plates (media-side down) at 30°C for 12–18 h. 8. Measure the diameter of the outermost ring for each motility plate to compare the migration distances of the cell population in the presence and absence of ligand. To select for library members for which gene expression is “off” in the absence of ligand (see Note 6), use a pipette tip to collect nonmotile cells from the center of the motility plate without the ligand, as shown in Fig. 2. Suspend the cells in 5 mL of tryptone broth supplemented with ampicillin, and grow the culture overnight at 37°C with shaking (250 rpm). 9. Repeat steps 6–8 to perform a second round of selection for library members that are “off” in the absence of ligand as shown in Fig. 2. 10. To identify library members for which gene expression is “on” in the presence of ligand, again repeat steps 6–8 with the exception that cells should be pipetted this time from the outside of the plate containing ligand, as shown in Fig. 2. Specifically, cells should be pipetted from just beyond the visible migration edge, and may also be collected by

High-Throughput Screens to Discover Synthetic Riboswitches

327

Fig. 2. Selection strategy to identify synthetic riboswitches using bacterial motility.

pipetting media at the outsides of the plate using a “sun ray” pattern. 11. Repeat steps 6 and 7, and then measure the diameter of the outermost ring for each motility plate to compare the migration distances of the cell population in the presence and absence of ligand. After the third round of selection, the enriched population should show a distinct increase in migration distance on the motility plate containing ligand compared to the plate lacking the ligand. 12. Using the QIAprep Spin Miniprep Kit, isolate plasmids from the overnight culture inoculated in step 10. These plasmids can now be used to subclone the enriched pool of riboswitches upstream of the lacZ reporter gene via the cloning strategy that was used to create the initial library, as shown in Fig. 1: Amplify the enriched pool of sequences with primers 1 and 4, digest the PCR product with the appropriate restriction enzymes, and ligate the gel-purified pool upstream of the lacZ reporter (see Subheading 3.3). 3.2. FACS-Based Selection

1. Transform TOP10F’ E. coli with the library ligation reaction. To determine the library size or number of transformed bacteria, library transformations should first be plated on large (241 mm × 241 mm) bioassay trays containing 300 mL of LB/agar, supplemented with ampicillin. Grow overnight at 37°C. 2. Record number of transformed bacteria. Scrape plates using 2 mL of liquid media. Use 500 µL cultivated bacteria to inoculate a 50 mL culture of LB supplemented with ampicillin. Incubate for 14 h at 37°C with shaking (250 rpm). 3. The following day, use 50 µL of the overnight culture to inoculate 5 mL of LB supplemented with ampicillin. 4. Incubate at 37°C with shaking for approximately 3 h to an OD600 between 0.3 and 0.5. 5. At this time, centrifuge 750 µL of culture at 5,000 × g for 10 min. Remove supernatant and resuspend the bacterial pellet in 1.5 mL of PBS buffer.

328

Lynch, Topp, and Gallivan

6. Immediately analyze bacteria on a Becton Dickinson FACSVantage SE flow cytometer using an Innova70 spectrum laser tuned to 568 nm for excitation. Fluorescence is detected using a 630/22 bandpass filter. 7. For initial negative selection, collect at least 2 × 105 clones displaying fluorescence levels within the first decade on the logarithmic scale as shown in Fig. 3. These bacteria should be sorted directly into 5 mL LB supplemented with ampicillin and then cultured overnight at 37°C. 8. The following day, use 50 µL of the overnight culture to inoculate 2 fresh 5 mL cultures of LB supplemented with ampicillin; one of which should contain the desired ligand at a concentration of 1 mM. If bacteria can tolerate higher concentrations of the ligand, a concentration greater than 1 mM may be used. 9. Repeat steps 4–6 for both cultures. 10. At this stage, the sorted culture grown without the ligand should display low levels of fluorescence following analysis with the flow cytometer (if a significant portion of the population remains fluorescent, an additional sort as described in step 7 should be performed). A comparison of the two cultures should reveal a small percentage (1–2%) of the bacteria grown with the ligand to be noticeably more fluorescent that those grown in the absence. For this positive selection step, a gate should now be set up so as to collect only the small fraction of clones grown with the ligand that display an increased fluorescence when compared to those grown without, as shown in Fig. 3. 11. Again, bacteria should be sorted directly into LB supplemented with ampicillin and then cultured overnight at 37°C. 12. The positively sorted culture should now be used to repeat step 8 followed by steps 4–6. 13. Analysis of these two cultures with the flow cytometer should reveal a highly dynamic, ligand-dependent shift in fluorescence, as shown in Fig. 3. By now, the population of riboswitches should be sufficiently enriched to the point where they can be readily screened enzymatically. If desired, a second positive selection can be performed and may, in some cases, further enrich the population. 14. Isolate plasmids from the overnight culture used in step 12. These plasmids can now be used to clone the enriched pool of riboswitches in front of a lacZ reporter gene. The cloning strategy should be similar to that used to create the initial library. PCR reactions can be set up using primers 1 and 4 to amplify the enriched pool of riboswitches. The resulting PCR product can then be easily digested with the appropriate

High-Throughput Screens to Discover Synthetic Riboswitches

329

Fig. 3. FACS histograms illustrating the selection strategy to enrich a population of synthetic riboswitches.

restriction enzymes, gel purified and ligated upstream of a lacZ reporter (see Subheading 3.3). 3.3. Enzymatic Screen for Functional Riboswitches

1. Transform TOP10F’ E. coli cells with the ligated library of enriched riboswitches that were cloned upstream of lacZ. For motility-based selections, transformations are plated on 85-mm Petri dishes containing 30 mL LB/agar supplemented with ampicillin. For blue-white screening of FACS-based selections, plate transformations to achieve a density of ~5,000 colonies/plate on large (241 mm × 241 mm) bioassay trays containing 300 mL of LB/agar, 4 mL of 6.25 mg/mL solution of X-Gal in dimethyl formamide, and ampicillin. 2. Grow for 14 h at 37°C. For FACS-based selection follow growth at 37°C by incubation at 4°C until blue color is readily visible. 3. For motility-based selections, pick 96 random colonies and inoculate in a 96-well microtiter plate containing 200 µL of LB supplemented with ampicillin. For FACS-based selections, pick the 96 “whitest” colonies from each bioassay tray to inoculate the 96-well plate. 4. Incubate overnight at 37°C with shaking (180 rpm). 5. The following day, use a multichannel pipettor to inoculate four 96-well plates of media (two sets of two) using 2 µL of the overnight culture. The first set of plates should contain 200 µL of LB supplemented with ampicillin. The second set of plates should contain 200 µL of LB supplemented with both ampicillin and your ligand (0.5 mM). 6. Incubate plates for approximately 2.5 h at 37°C with shaking (210 rpm) to an OD600 of 0.085–0.14 as determined by a microplate reader (0.3–0.5 with a 1-cm path length cuvette). Record OD600 for each well.

330

Lynch, Topp, and Gallivan

7. When the appropriate level of growth is reached, add 21 µL of a lysis solution and gently pipette the culture up and down. Allow cultures to lyse for 5 min at room temperature. 8. In a fresh plate, add 15 mL of lysed culture to 132 µL of Z-buffer. Incubate at 30°C for 10 min. 9. To each well, add 29 µL of ONPG solution and note the time. Allow ONPG to hydrolyze for approximately 20 min or until faint yellow color is observed. 10. Quench each reaction by 75 µL of Na2CO3. Record the time of hydrolysis and determine OD420 for each well. 11. Calculate Miller units for each well using the following equation: Miller units = OD420/(OD600 × hydrolysis time × relative volume of cell lysate). To identify functioning riboswitches, determine the ratio of Miller units observed for cultures grown in the presence of the ligand to those grown in the absence. Functioning riboswitches should show consistent results across the two plates while displaying normal growth rates and significant increases in β-galactosidase expression in the presence of the ligand (4). If these criteria are met, Miller unit ratios greater than 2 should indicate the presence of a functioning switch (see Note 7). To validate the function of identified switches, clones can be subcultured and assayed on a larger scale using a previously described protocol of Jain and Belasco (10).

4. Notes 1. The CheZ protein plays a critical role in E. coli chemotaxis by dephosphorylating the CheY-P protein, which binds to the flagellar motor and causes swimming cells to tumble (11). 2. The beginning of the aptamer should be positioned approximately 50 bases from the first transcribed base. Although the specific sequence of the short leader region is not generally critical, its presence renders the transcript less prone to degradation in the cell, thus improving its efficacy as a genetic switch. 3. Optimal levels of CheZ are necessary for E. coli cells to migrate on motility plates. If too little CheZ is present, the level of CheY-P will increase, and the cells will tumble incessantly and not migrate (12). If cells have excess CheZ, they will swim very smoothly and rarely tumble. Because cells that swim extremely

High-Throughput Screens to Discover Synthetic Riboswitches

331

smoothly can become embedded in the semisolid media, they cannot migrate (13). Thus it is critical to ensure that CheZ is not overexpressed in these assays. Because the strength of the promoter will ultimately dictate the maximum expression level of the cheZ gene, we suggest the use of two different promoters: the “weak” IS10 promoter (14) and the 60- to 100-fold stronger tac promoter (15). We anticipated that the motility selections would readily reveal which promoter provided the appropriate CheZ expression level. In our case, the weaker promoter provided ideal expression levels. It is useful to note, however, that riboswitches can be identified with a motilitybased selection even when a strong promoter is employed; riboswitches selected using a weak promoter tend to require strong RBS sequences, while riboswitches generated using a strong promoter generally use a very weak RBS sequence. Because these riboswitches function by interacting with the RBS, which is part of the randomized expression platform, it may be useful to perform a motility-based selection with several promoters to further enhance the diversity of riboswitches that may be identified using this motility-based method. 4. For a FACS-based enrichment, the construct should be under the control of a strong promoter such as tac (15), as this will provide the most detectable levels of gene expression. 5. When preparing motility plates, it is important that the agar remains evenly dissolved in the tryptone broth until the plates have been poured. If the agar is not fully dissolved, the uneven distribution will lead to some plates being too liquid and some plates being too solid. The stock media can be made at slightly higher concentration of agar (generally 0.35%) to permit the use of ligands that cannot be autoclaved. In this case, dissolve the ligand in tryptone broth, sterile filter the solution, and dilute the stock media with tryptone broth containing ligand; after dilution, the final concentration of agar in each plate should be 0.25%. The ligand concentration may be significantly higher or lower than 1 mM, depending upon the binding affinity of the aptamer and the intracellular concentration of the ligand in E. coli. To test whether a certain ligand concentration can be used for riboswitch selection using the motility method, it is important to determine whether the motility of wild-type E. coli is impacted by the ligand at that concentration. To test this factor, perform steps 6–8 of Subheading 3.1, with the exception that a single colony of wildtype bacteria should be used to inoculate the 5-mL culture described in step 6. If the migration distance is similar for wild-type cells grown with and without the ligand, then that ligand concentration can be used in a motility selection for riboswitches.

332

Lynch, Topp, and Gallivan

6. As explained in Notes 1 and 3, the use of a very strong promoter can complicate the selection because cells located at the center of the motility plate represent a mixed population of cells that have either low or high CheZ expression levels. 7. Careful inspection of the Miller unit data is recommended to ensure a high probability of identifying functional riboswitches (4). Low levels of gene expression in both the presence and absence of a ligand can give rise to large, misleading ratios. To account for this, it is useful to ignore those clones that display an OD420 less than 0.04 in the presence of the ligand. Also, inspect the data for irregularities in culture growth indicated by the OD600 as exceptionally high or low OD600 values can give rise to deceptive levels of calculated Miller units.

Acknowledgments Work in the Gallivan lab work is supported by the NIH (GM074070). J.P.G. is a Beckman Young Investigator, a Camille Dreyfus Teacher-Scholar, and an Alfred P. Sloan Research Fellow. S.A.L. is an NSF Problems and Research to Integrate Science and Mathematics (PRISM) Fellow. S.T. is a G.W. Woodruff Scholar and an ARCS Fellow. References 1. Tuerk, C., and Gold, L. (1990). Systematic evolution of ligands by exponential enrichment– RNA ligands to bacteriophage-T4 DNApolymerase. Science 249, 505–510 2. Lynch, S. A., and Gallivan, J. P. (2009). A flow cytometry-based screen for synthetic riboswitches. Nucl. Acids Res. 37, 184–192. 3. Topp, S., and Gallivan, J. P. (2007). Random walks to synthetic riboswitches – a highthroughput selection based on cell motility. ChemBioChem, 9, 210–213 4. Lynch, S. A., Desai, S. K., Sajja, H. K., and Gallivan, J. P. (2007). A high-throughput screen for synthetic riboswitches reveals mechanistic insights into their function. Chem. Biol. 14, 173–184 5. Topp, S., and Gallivan, J. P. (2007). Guiding bacteria with small molecules and RNA. J. Am. Chem. Soc. 129, 6807–6811 6. Link, A. J., Jeong, K. J., and Georgiou, G. (2007). Beyond toothpicks: new methods for isolating mutant bacteria. Nat. Rev. Microbiol. 5, 680–688

7 . Griffith , K. L. , and Wolf , R. E. (2002). Measuring b-galactosidase activity in bacteria: Cell growth, permeabilization, and enzyme assays in 96-well arrays. Biochem. Biophys. Res. Commun. 290, 397 8. Baba, T., Ara, T., Hasegawa, M., Takai, Y., Okumura, Y., Baba, M., Datsenko, K. A., Tomita, M., Wanner, B. L., and Mori, H. (2006). Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol. Sys. Biol. 2, 1–11 9. Mutoh, N., and Simon, M. I. (1986). Nucleotide-sequence corresponding to 5 chemotaxis genes in Escherichia coli. J. Bacteriol. 165, 161–166 10. Jain, C. , and Belasco, J. G. (2000). Genetic methodologies for detecting RNA–protein interactions, in Methods in Enzymology Vol. 318 (Celander, D. W. and Abelson, J. N., eds.) , Academic Press , New York , pp. 309–331 11. Zhao, R., Collins, E. J., Bourret, R. B., and Silversmith, R. E. (2002). Structure and catalytic

High-Throughput Screens to Discover Synthetic Riboswitches mechanism of the E. coli chemotaxis phosphatase CheZ. Nat. Struct. Biol. 9, 570–575 12. Wolfe, A. J., Conley, M. P., Kramer, T. J., and Berg, H. C. (1987). Reconstitution of signaling in bacterial chemotaxis. J. Bacteriol. 169, 1878–1885 13. Wolfe, A. J., and Berg, H. C. (1989). Migration of bacteria in semisolid agar. Proc. Natl. Acad. Sci. U.S.A. 86, 6973–6977

333

14. Jain, C., and Kleckner, N. (1993). Is-10 messenger-RNA stability and steady-state levels in Escherichia coli - Indirect effects of translation and role of rne function. Mol. Microbiol. 9, 233–247 15. Deboer, H. A., Comstock, L. J., and Vasser, M. (1983). The tac promoter – a functional hybrid derived from the trp and lac promoters. Proc. Natl. Acad. Sci. U.S.A. 80, 21–25

Chapter 24 A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage Laising Yen, Brent R. Stockwell, and Richard C. Mulligan Summary RNA cleavage is a catalytic reaction which defines many types of RNA processing events, including those of metabolite-sensing riboswitch, self-splicing introns, mRNA splicing, tRNA processing, polyA-cleavage, and various small ribozymes such as hairpin and hammerhead ribozyme. In this chapter, we describe a general methodology for developing a mammalian cell-based high-throughput screening assay useful for identifying small molecules capable of inhibiting RNA cleavage in mammalian cells. In the specific assay described, a plasmid DNA vector in which the expression of a luciferase reporter gene is controlled by hammerhead ribozyme cleavage was stably introduced into the human 293 cell line. Such a cell line enabled the rapid screening of chemical compound libraries and the identification of cell membrane-permeable inhibitory molecules capable of blocking ribozyme cleavage. The general strategy described later could in principle be adapted to identify small molecule inhibitors of many types of RNA cleavage reactions. Key words: RNA self-cleavage, Hammerhead ribozyme, Riboswitch, RNA–drug interaction, Highthroughput screening, Chemical libraries, Toyocamycin, Conditional gene regulation

1. Introduction The rules governing small molecule–RNA interactions in mammalian cells capable of affecting RNA cleavage reactions would appear to be necessarily more complex than those rules governing reactions in the test tube. In the cellular environment, many factors could theoretically limit interactions between small molecules and RNA that would lead to the inhibition of RNA self-cleavage. These factors include the complex secondary and tertiary structures of mRNA, the constant turnover and transient nature of mRNA, and the complex protein–RNA interactions that are unique to Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_24 © Humana Press, a part of Springer Science + Business Media, LLC 2009

335

336

Yen, Stockwell, and Mulligan

the mammalian cell. In addition, for a molecule to inhibit RNA cleavage in cells, it must be able to enter cells in sufficient quantity to effectively interact with cellular RNA in the appropriate cellular compartment. The toxicity of the molecule must also be limited, at least on the time-scale of the assay. While an in vitro biochemical screen may have the advantages of speed and ease, molecules identified in this manner may be ineffective when evaluated in vivo. Therefore, the choice between an in vitro biochemical screen versus a mammalian cell-based screen is ultimately a choice between simplicity versus applicability. In this chapter, we describe the necessary steps for developing a cell-based assay for screening small molecules capable of modulating RNA self-cleavage in vivo. The first essential component for such a cell-based assay is an RNA self-cleaving motif that functions efficiently in vivo. We have recently optimized a Schistosoma hammerhead ribozyme that is capable of extremely efficient self-cleavage in mammalian cells (1). This ribozyme, termed N79, was used to construct an RNA-only gene regulation system in which the spontaneous self-cleavage of ribozyme embedded in-cis leads to destruction of the mRNA and therefore a loss of reporter gene expression. Small molecules capable of inhibiting RNA self-cleavage should result in preservation of the intact mRNA and therefore “induce” gene expression (Fig. 1). Such an efficient RNA self-cleaving motif is pivotal to the

Fig. 1. Controlling gene expression via modulation of RNA self-cleavage. When cis-acting hammerhead ribozymes are embedded in the mRNA, self-cleavage leads to the destruction of the mRNA and results in the absence of gene expression. The administration of specific inhibitors of the ribozyme leads to the generation of intact mRNAs and results in protein expression.

A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage

337

development of a cell-based screen. It is strongly recommended that the readers optimize their RNA self-cleaving motif before tackling the high-throughput screening (HTS).

2. Materials 2.1. Cell Culture and Transfection

1. Regular mammalian cell culture medium for HEK 293 cells: Dulbecco’s Modified Eagle’s Medium (DMEM) with phenol red (Gibco/Invitrogen #11965–118, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS) (Hyclone, Logan, UT), 1× penicillin-streptomycin (Sigma), and 1 mM l-glutamine (Gibco/Invitrogen). Store DMEM at 4°C. Store aliquots of FBS, penicillin-streptomycin, and l-glutamine at –20°C. 2. Cell culture medium for luciferase assay: DMEM (Gibco/ Invitrogen # 21063–029) similar to the medium described earlier but without phenol red. The presence of phenol red reduces the signal strength of luciferase. 3. Solution of 0.25% trypsin/1 mM EDTA (Gibco/Invitrogen). Store aliquots at –20°C. 4. FuGENE6 (Roche, Basel, Switzerland) is used for plasmid and oligonucleotide DNA transfection. OptiMEM is from Gibco/ Invitrogen (#31985–070). Both reagents are stored at 4°C.

2.2. Puromycin Selection

1. Puromycin stock solution: 1 mg/mL (Sigma). Store aliquots at –20°C. Working solution (1 mg/mL) is prepared by diluting the stock solution 1,000-fold in culture medium.

2.3. Morpholino Oligo Transfection

1. Morpholino antisense oligos and the EPEI transfection agent are purchased from GeneTools (2) (www.gene-tools.com). Dissolve 300 nmol of prepaired duplex of morpholino/complementary DNA oligonucleotide in 0.6 mL water to make 0.5 mM stock oligonucleotide solution. Store oligonucleotide aliquots at –20°C. EPEI transfection agent (200 mM) is stored at 4°C.

2.4. Toyocamycin and 5F-uridine

1. Toyocamycin (1, 3, 4) (Berry & Associates, www.berryassoc.com) is first dissolved in 100% DMSO, and then diluted with PBS to a final concentration of 10 mM and 20% DMSO. 5F-uridine (4) (Sigma) is dissolved in PBS as 100 mM stock solution. Both stock solutions are filtered and stored at –80°C as 20 mL/ tube aliquots. The stock solutions are stable for up to 12 months at –80°C. Each aliquot is meant for one-time use since freeze/thaw cycle reduces the effectiveness. To make the cocktail solution, dilute toyocamycin 1:20,000 times and 5F-uridine 1:10,000 times in cell medium. This gives a working

338

Yen, Stockwell, and Mulligan

concentration of 0.5 mM toyocamycin and 10 mM 5F-uridine, respectively. Exercise caution since both 5F-uridine and toyocamycin are toxic. 2.5. Reagents for Luciferase Assay

1. Luciferin (Caliper LifeSciences/Xenogen, www.caliperls.com) is dissolved in water at 100 mM, or 1 g in 33 mL water. Store aliquots at –80°C protected from light. 2. ATP (Sigma): 0.1 M, or 1 g in 18.1 mL water. Store aliquots at –80°C. 3. Coenzyme A (Sigma): 50 mM, or 100 mg in 2.6 mL water. Store aliquots at –20°C. 4. Na2CO3 (Sigma): 0.25 M, or 2.65 g in 100 mL water. It is used for adjusting pH value of basic buffer. 5. DTT (DL-dithiothreitol) (Sigma): 1 M, or 1 g in 6.48 mL water. Store aliquots at –80°C. 6. Basic buffer: 25 mM Tricine, 0.5 mM EDTA, 0.54 mM Na-triphosphate, 16.3 mM Mg2SO4, 0.3% (v/v) Triton X-100. To make 500 mL of basic buffer, add 2.24 g of Tricine, 0.5 mL of 0.5 M EDTA, pH 8.0, 0.1 g of Na-triphosphate, 2 g of Mg2SO4·7 H2O, 1.5 mL of Triton X-100. Adjust pH to 7.8 using the solution of 0.25 M Na2CO3. Bring the final volume to 500 mL. Store at room temperature. 7. White opaque 96-well or 384-well cell culture microplates (Corning or Greiner) are used to seed cells and perform subsequent luciferase assay. White opaque microplates are preferred for luminescent applications since the white opaque background reflects light to increase the signal strength while crosstalk between wells remains low. However, if signal sensitivity is not limiting, black plates can be used with a similar signal-to-background ratio.

3. Methods 3.1. Generation of Reporter Cell Lines for Cell-Based Screening 3.1.1. Cotransfection of HEK293 Cells

1. For screening purposes, we chose to generate stable human HEK293-derived cell lines (see Note 1) which carry an integrated mammalian expression vector in which a luciferase reporter gene is placed under the control of the CMV promoter and two copies of the N79 ribozyme (Fig. 1). Stable cell lines are generated by cotransfection of HEK293 cells with the ribozyme-carrying plasmid and a separate vector encoding puromycin resistance gene at a 20:1 copy ratio. This 20:1 ratio ensures that cells receiving the puromycin resistant gene will also receive the ribozyme-carrying plasmid.

A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage

339

2. We typically cotransfect one million HEK293 cells at around 60% confluence in a 60-mm dish with 1 mg of ribozyme-carrying plasmid and 1/20th that of the plasmid-carrying puromycin resistance gene. A dish of control cells is transfected only with ribozyme-carrying plasmid without the plasmid-carrying puromycin resistant gene. The transfection mixture is prepared by combining 1 mg of plasmid DNA with 2.5 mL of FuGENE6 in 200 mL of OptiMEM and incubate at room temperature for 20 min. The mixture is then added to HEK293 cells in a 60-mm dish in the presence of serum. 3.1.2. Puromycin Selection

1. Three days after transfection, the cells are passaged at 1:10, 1:30, and 1:100 split, each with four dishes, and cultured in the presence of puromycin. We typically use puromycin at 1 mg/mL for HEK293 cells. The exact concentrations however should be determined empirically for each different cell line (see Note 2). Three to five days of puromycin treatment usually is sufficient to eliminate the nonresistant cells, as judged by the complete cell death in the control cells. 2. The surviving cell clones are allowed to grow into sizable colonies consisting of 50–100 cells per colony, and then are carefully picked up by pipette tips under microscopy and transferred to isolated wells for further expansion.

3.1.3. Treatment with Antisense Morpholino Oligonucleotides

Stable transfection produces cell clones different in the site of integration of the vector sequences and vector copy number. Thus, it is critical to characterize and validate the cell clones prior to starting a screen. We validate our cell clones by transfecting them with antisense morpholino oligo designed to block the ribozyme cleavage site in a sequence-specific manner to insure that the reporter is capable of being induced by virtue of inhibition of ribozyme self-cleavage (see Note 3). 1. Isolated cell clones are first seeded in a 24-well plate with a density of 120,000 cells per well. 2. Next day, prepare the morpholino transfection solution by adding 5.6 mL of 0.5 mM oligo stock and 5.6 mL of EPEI (see Note 4) to 188.8 mL of water. Vortex the solution immediately, and incubate at room temperature for 20 min. Add 1.8 mL of OptiMEM and vortex the solution. 3. We remove the culture medium and incubate the cells with 400 mL of the transfection mixture per well for 3 h, and then replace the mixture with fresh cell culture medium. 4. The cells are assayed for the induction of luciferase expression (see Subheading 3.1.5 for luciferase assay) 24–36 h post transfection. An example of cell clone validation is shown in Fig. 2 (see Note 5).

340

Yen, Stockwell, and Mulligan

Fig. 2. Characterization and validation of the cell clones by antisense morpholino oligo. (a) Bioluminescent images of stable cell clones transfected with an antisense morpholino oligo that blocks ribozyme cleavage. Stable cell clones differ significantly in their phenotypes in terms of background luciferase expression and inducible range. Plus sign “ + ” indicates morpholino oligo-treated; minus sign “-”untreated. (b) Quantitative luciferase activity measurements to determine background level and inducible range of stable cell clones. The numbers shown are averaged from triplicate experiments. Clones 7–9 exhibits a 20-fold induction range when treated with morpholino oligo, with signal levels within the linear range of the luminometer. These characteristics make clones 7–9 suitable for the use in the screen.

3.1.4. Characterization and Validation of the Cell Clones by Toyocamycin and 5F-Uridine

In addition to the use of antisense morpholino oligo, an alternative method for validating cell clones is to use toyocamycin or 5Furidine to inhibit RNA self-cleavage. These two potent compounds identified in our previous screening study (4) are incorporated into RNA through cellular RNA synthesis (5, 6). When incorporated at key positions of an RNA self-cleaving motif, they impair the self-cleavage reaction (7) (see Note 6). 1. Cell clones are seeded in a 24-well plate with a density of 120,000 cells per well. 2. In the next day, the culture medium is replaced with 500 mL per well of fresh medium containing 0.5 mM of toyocamycin and/or 10 mM of 5F-uridine. The cells are incubated for 24 h and then assayed for the induction of luciferase expression.

3.1.5. Luciferase Assay

Luciferase is the preferred reporter system for HTS for the following reasons. First, the signal is amplified by the enzymatic reaction of luciferase. Second, because there is no excitation beam involved in the assay, virtually no noise luminescence is emitted from the sample. This allows a highly sensitive assay with a superb signal-to-noise ratio. However, as an enzymatic reaction, luciferase assay is sensitive to temperature change, the presence of phenol red, and the time interval between the addition of luciferin substrate and the measurement (see Note 7). 1. Warm up all reagents to room temperature. To make 36 mL of assay buffer, add 144 mL 1 M DTT, 108 mL 0.1 M ATP, 252 mL 100 mM luciferin, and 360 mL 50 mM coenzyme A to 35 mL of basic buffer. For best results, the assay buffer should be made fresh prior to starting the assay (see Note 8).

A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage

341

2. Partially remove the cell medium from each well by aspiration. For experiments using the 24-well plate, we leave about 100 mL of medium in each well and add 400 mL of assay buffer per well. 3. Incubate the cells with assay buffer for 5 min to allow sufficient cell lysis (see Note 9). This is followed by photon output measurement with a luminometer. The intensity of luciferase signal decays slowly using the assay buffer described. 3.2. Optimization of the Cell-Based Screening Assay 3.2.1. Determination of DMSO Tolerance

Since many compound libraries are dissolved in DMSO in the range of ~10 mg/mL, the DMSO tolerance of the cell-based assay will determine the maximum screening concentration of the compound for the HTS. 1. Seed cells in the desired cell density and microplate format (either in 96- or in 384-well plates). 2. Allow cells to grow in DMSO-containing medium for the intended assay period (usually 2–3 days). 3. Determine the maximum DMSO concentration that the cells can tolerate in the assay period. For HEK293 cells, a final DMSO concentration of 0.1% is a good starting point.

3.2.2. Determination of Optimal Cell Plating Density

Cell growth rate is an important consideration to determine optimal initial seeding density. 1. Seed cells at different densities in the desired microplate format and allow cells to grow for the intended assay period (usually 2–3 days). DMSO must be used in these studies at the final concentration to be used in screening. 2. Choose an initial cell density that will not result in overgrowth by the end of the assay period. For HEK293 cells, an initial seeding density of 8,000 cells per well in 384-well plate is a good starting point.

3.2.3. Determination of Liquid-Handling Parameters for HTS

The following liquid-handling parameters are good starting points for HTS performed in 384-well format and can be scaled up to accommodate the 96-well format. 1. We first aliquot the compound libraries into daughter plates with a 1:50 dilution in DMEM. For screening, 8,000 cells are seeded in each well with 57 mL of culture medium (see Note 10), followed by the addition of 3 mL of diluted compound from the daughter plates (see Note 11). This yields a final DMSO concentration of 0.1% and final compound concentrations close to 10 mg/mL. 2. Cells are incubated with the compounds for 48 h. 3. Prior to the luciferase assay, cell media is partially removed by aspiration; approximately 15 mL of medium is left in each well (see Note 12). This is followed by the addition of 45 mL of luciferase assay buffer per well. The actual liquid-handling

342

Yen, Stockwell, and Mulligan

parameters should be empirically optimized for the particular robotic system and luminometer utilized (see Note 13). 3.2.4. Determination of the Z-Factor

Prior to starting an HTS, it is important to assess the general quality and signal-to-noise strength of the cell-based assay on a smaller scale. The Z-factor is a commonly used statistical tool to predict if useful data could be extracted from the assay if it is scaled up to millions of treated samples (8). 1. To calculate the Z-factor, perform the cell-based assay many times, with positive controls (such as morpholino oligotreated or toyocamycin-treated wells) and negative controls (untreated wells, or mock-treated with an inactive compound) on the same plates. 2. Compute the mean of all positive (mp) and negative (mn) controls, and the standard deviation of positive (SDp) and negative controls (SDn). Compute the Z-factor as follows: SSD = (SDp + SDn), R = |mp-mn|, Z-factor = 1-3(SSD)/R A perfect theoretical Z-factor equals 1, which means the assay has a large dynamic range with small standard deviations. An assay with a Z-factor between 0.5 and 1 is generally considered an excellent assay. A Z-factor less than 0.5 is marginal, and the assay may not yield useful data from the screen. In this case, it is essential to optimize assay parameters described in Subheadings 3.2.1 and 3.2.3 to improve the Z-factor prior to starting the HTS. In addition, Z-factors should be calculated for every plate of compounds screened, so that defective assay plates can be eliminated and then retested.

3.3. Plate Layout for HTS

1. Most high-throughput assays show a high amount of inherent variability and error. Therefore, it is recommended that all assays be performed in triplicate if possible. The simplest way to triplicate the entire screen is to generate three identical sets of assay plates instead of repeating three wells on the same plate. 2. Evaporation can cause the wells on the edge of the plate to dry out over a period of time of incubation. This is called “edge effect.” Edge effects can contribute to variability and should be avoided by excluding the use of wells on the edge of the plate. However, these edge wells should have the same medium inside the wells; otherwise, the outermost wells with medium become the new “edge” wells. 3. Control readings are essential to a well-designed assay. Each individual assay plate should have a few appropriate positive and negative controls. These controls can be used for identifying plate-to-plate variability, establishing assay background levels, and identifying potential fallouts during screening process. Most of the library plates are formatted with empty

A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage

343

Fig. 3. An example of a cell-based screen for identifying small molecule inhibitors of RNA self-cleavage. Cell-based HTS employing luciferase reporter usually is performed by a high-speed robotic system and luminometer. However, an imaging system can also be used for screens that require medium-throughput capacity. Cells were seeded in 96-well plate and each well was treated with a unique compound from the library for 2 days. The plate was then imaged using a Xenogen imager to visually reveal the levels of luciferase expression. As shown in the bioluminescent image, one of the treated wells has a marked increase in luciferase expression, indicating that the compound added to that particular well was able to inhibit ribozyme self-cleavage in cells. This compound was later identified as tubercidin (13), an adenosine analog.

wells to allow for controls. It is also a good practice to include one or two uniformly treated positive control plates to identify reading bias within the same plate due to luciferase signal decay over time or other unexpected factors. 4. An example of the small inhibitory molecules capable of modulating RNA self-cleavage in vivo which was identified by the cell-based assay described in this chapter is shown in Fig. 3. Although we used hammerhead ribozyme as the target for screen, the experimental principle for identifying small inhibitory molecules for other types of RNA self-cleavage in vivo remains the same. Several sources of commonly used small molecule libraries are listed in Note 14, and several web resources for general guidelines of High-Throughput Screening are noted in Note 15. Readers are encouraged to explore those web sites for more information prior to initiating the HTS.

4. Notes 1. Choosing the right kind of cell lines is critical for the success of cell-based high-throughput screening. Issues to consider include the efficiency of transfection, promoter and gene

344

Yen, Stockwell, and Mulligan

expression level, cell growth rate, origin of the cell type, property of cell adhesion, and the existing antibiotic resistance. We chose HEK293 cell because of its human origin, the ease of transfection, and because of its relative slow growth rate that would allow 48 h of incubation with small molecules without being overly grown. We did not choose HEK293T cells because this cell line is already resistant to neomycin and is not suitable for screening similar aminoglycosides that were reported to inhibit hammerhead ribozyme cleavage in vitro. We employed the CMV promoter that is known to generate strong expression in HEK293 cells. When other cell types are used, a suitable promoter should be chosen. 2. To determine the puromycin concentrations for specific host cell lines, treat the untransfected host cells with a series of diluted puromycin (0, 0.25, 0.5, 1, 1.5, 2 mg/mL) for a period of 5 days. At the end of the treatment, use cell death as the gauge to determine the effective concentration. For example, if only part of cells dies on the plate at 0.5 mg/mL and all cells die at 1 mg/mL, then a concentration of 1 mg/mL is suitable for the selection and maintenance of stably transfected clones. 3. To identify specific cell clones capable of responding to putative inhibitory molecules, the cells should be tested for both their basal levels of luciferase expression and for the extent of induction of luciferase expression when RNA self-cleavage is inhibited. One way to inhibit RNA cleavage without having an a priori small inhibitory molecule in hand is to use antisense oligos to block the RNA cleavage in a sequence-specific manner. We have tested a variety of antisense oligos with different chemical structures. We found that transfection with peptide nucleic acid (PNA) (9), locked nucleic acid (LNA)10), and “grip” nucleic acid (11) had no measurable effects on self-cleavage activity. Transfection with phosphorothiolate, 2¢-O-methyl, and phosphorothiolate 2¢-O-methyl-derived RNAs (12) led to modest inhibition of self-cleavage (fivefold to tenfold induction)(1). Transfection of a morpholino oligo (2), however, led to a strong inhibition of ribozyme self-cleavage, as revealed by a dramatic increase in reporter gene expression (100- to 2,000-fold)(1). We have also tested different morpholino oligos designed to target different regions of the ribozyme structure and found that the best oligos are the ones that block the ribozyme cleavage location. Online computational help for designing morpholino oligo sequences is available on GeneTools web site (www. gene-tools.com). 4. GeneTools offers two different ways of transfecting morpholino oligos: EPEI and Endo-porter. Although EPEI transfection reagent is more toxic to cells than Endo-porter, the transfection mediated through EPEI seems to be more effective in our

A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage

345

hands. This however may simply reflect authors’ familiarity with the EPEI method. Readers should consult with GeneTools for updated suggestions. 5. We transfected many cell clones using EPEI with an antisense morpholino oligo (Morph-3) known to block the cleavage of N79 ribozyme(4). Several cell clones tested in this way expressed low basal levels of luciferase, yet were readily induced to express luciferase after administration of oligo (Fig. 2). One of these cell lines, termed “7–9,” showed a 20-fold induction of luciferase expression after oligo administration, and was chosen for high-throughput screening that led to the identification of toyocamycin, 5F-uridine, and other molecules capable of inhibiting RNA self-cleavage in mammalian cells(4). 6. Toyocamycin is an adenosine analog, and 5F-uridine is a uridine analog. Incorporation of these compounds in cellular RNAs will replace A and U bases in the RNA sequences, respectively. The effectiveness of the compounds in inhibiting a particular RNA self-cleavage depends on whether A or U bases are critical to the structure or function of the RNA self-cleaving motif. Because of this, toyocamycin and 5F-uridine treatment may be used as tools to probe the structure-function relationships of an RNA self-cleaving motif in vivo. Since these molecules replace different bases in RNA, the combined administration of toyocamycin and 5F-uridine will lead to a better inhibition of RNA self-cleavage and an improved induction of gene expression, relative to the use of toyocamycin alone(4). For applications in mammalian cells with short experimental periods, a cocktail of 0.5 mM of toyocamycin and 10 mM of 5F-uridine was found effective. If toxicity becomes a concern, the concentrations can be scaled down in proportion according to the toxicity effect on the cells. 7. To increase the signal strength and the reproducibility of the assay, the following steps are recommended. (1) Equilibrate all reaction reagents to room temperature prior to use and perform the assay at room temperature, which is close to the temperature optimum of luciferase activity. (2) Include negative controls (untreated or mock-treated wells) and positive controls (morpholino oligo-treated or toyocamycin-treated wells) on each microplate. The values of these controls can be used to normalize signal variations between the plates. (3) Grow cells and perform assays in DMEM medium without the phenol red. The presence of phenol red reduces the signal intensity. (4) Remove any bubbles in the wells that may interfere with the measurement. (5) Always work in the linear range of the luciferase assay, and read the signals at a fixed time point within the plateau period.

346

Yen, Stockwell, and Mulligan

8. Promega offers several luciferase assay kits for different purposes. These kits are good for small-scale experiments, but the cost for large-scale experiments using these kits can be too high. 9. If cells are to be recovered, one can measure luciferase activity using luciferin alone, without other assay buffer components. To measure luciferase activity without cell lysis, we routinely add 20 mL of 100 mM sterile luciferin to each well in 96-well plate without removing the cell medium. The luciferase signal intensity obtained in this assay is less reliable and usually peaks and fades relatively quickly (within 5 min) without a steady plateau. After the measurement, replace luciferincontaining medium with fresh cell culture medium and let cells to recover. 10. Although puromycin is used in our routine maintenance of the cell clones, it is removed from cell medium during the HTS. 11. One problem of seeding cells in 384-well plate is that air bubbles are frequently trapped in the bottom of the wells thus preventing cells to settle. The problem is usually solved by a quick spin after seeding the cells. 12. It is not recommended to remove all medium from the treated wells since this can lead to a loss of loosely attached cells and can possibly decrease assay reproducibility. 13. In general, assay volumes in 384-well plates range from 5 mL to 100 mL. At the low end of this range, inaccuracies in pipetting can cause signal variation, while there is a risk of spillage and crosscontamination at the high end. 14. A few popular sources of small molecule libraries include: LOPAC, ~1,280 compounds from Sigma (www.sigmaaldrich. com); Microsource Discovery’s ~2,000 compound collection (www.msdiscovery.com); NCI’s 1,990 compound diversity set (dtp.nci.nih.gov); Prestwick chemical set, ~1,100 compounds (www.prestwickchemical.fr/index.php?pa = 1); Specs world diversity set, 10,000 compounds (www.specs. net); Chembridge Diverset, 52,000 compounds (www.chembridge.com/chembridge/compound.html). 15. A few excellent web resources for general guidelines of High-Throughput Screening: NIH Chemical Genomics Center (NCGC) (http://www.ncgc.nih. gov/); Pittsburgh Molecular Libraries Screening Center (PMLSC) (http:// pmlsc.pitt.edu/content.asp?id = 1050); Harvard Medical School-Longwood screening facility (ICCB) (http://iccb. med.harvard.edu/ ); NIH Molecular Libraries Small Molecule Repository (MLSMR) (http://mlsmr.glpg.com/ MLSMR_HomePage/); Information of small molecules (Pubchem) (http://pubchem.ncbi.nlm.nih.gov/).

A Mammalian Cell-Based Assay for Screening Inhibitors of RNA Cleavage

347

Acknowledgments The authors would like to thank Kourosh Salehi-Ashtiani for his critical reading of the manuscript. L. Yen is supported by the Duncan Scholar Award, ACS-IRG-93–034–12, and the start-up fund from the Department of Pathology, Baylor College of Medicine. R.C. Mulligan is supported by R01 GM075127 and R01 GM79187. B.R. Stockwell is supported by A. Beckman Young Investigator Award from the Arnold and Mabel Beckman Foundation.

References 1. Yen, L., Svendsen, J., Lee, J. S., et al. (2004). Exogenous control of mammalian gene expression through modulation of RNA selfcleavage. Nature. 431, 471–476 2. Morcos, P. A. (2001). Achieving efficient delivery of morpholino oligos in cultured cells. Genesis. 30, 94–102 3. Aszalos, A., Lemanski, P., Robison, R., Davis, S. & Berk, B. (1966). Identification of antibiotic 1037 as toyocamycin. J. Antibiot. (Tokyo). 19, 285 4. Yen, L., Magnier, M., Weissleder, R., Stockwell, B. R. & Mulligan, R. C. (2006). Identification of inhibitors of ribozyme self-cleavage in mammalian cells via highthroughput screening of chemical libraries. RNA. 12, 797–806 5. Suhadolnik, R. J., Uematsu, T. & Uematsu, H. (1967). Toyocamycin: phosphorylation and incorporation into RNA and DNA and the biochemical properties of the triphosphate. Biochim. Biophys. Acta. 149, 41–49 6. Glazer, R. I. & Lloyd, L. S. (1982). Association of cell lethality with incorporation of 5-fluorouracil and 5-fluorouridine into nuclear RNA in human colon carcinoma cells in culture. Mol Pharmacol. 21, 468–473 7. Hammann, C., Norman, D. G. & Lilley, D. M. (2001). Dissection of the ion-induced folding of the hammerhead ribozyme using

8.

9.

10.

11.

12.

13.

19F NMR. Proc Natl Acad Sci U.S.A. 98, 5503–5508 Zhang, J. H., Chung, T. D. & Oldenburg, K. R. (1999). A simple statistical parameter for use in evaluation and validation of high throughput screening assays. J. Biomol. Screen. 4, 67–73 Nielsen, P. E., Egholm, M., Berg, R. H. & Buchardt, O. (1991). Sequence-selective recognition of DNA by strand displacement with a thymine-substituted polyamide. Science. 254, 1497–1500 Wahlestedt, C., Salmi, P., Good, L., et al. (2000). Potent and nontoxic antisense oligonucleotides containing locked nucleic acids. Proc Natl Acad Sci U.S.A. 97, 5633–5638 Efimov, V. A., Choob, M. V., Buryakova, A. A., Kalinkina, A. L. & Chakhmakhcheva, O. G. (1998). Synthesis and evaluation of some properties of chimeric oligomers containing PNA and phosphono-PNA residues. Nucleic Acids Res. 26, 566–575 Braasch, D. A. & Corey, D. R. (2002). Novel antisense and peptide nucleic acid strategies for controlling gene expression. Biochemistry. 41, 4503–4510 Acs, G., Reich, E. & Mori, M. (1964). Biological and biochemical properties of the analogue antibiotic tubercidin. Proc. Natl. Acad. Sci. U.S.A. 52, 493–501

Chapter 25 In Vitro Selection of glmS Ribozymes Kristian H. Link and Ronald R. Breaker Summary Riboswitches modulate gene expression in eubacteria and eukaryotes in response to changing concentrations of small molecule metabolites. In most examples studied to date, riboswitches achieve both metabolite sensing and gene control functions without the obligate involvement of protein factors. These findings validate the hypothesis that RNA molecules could be engineered to function as designer gene control elements that sense and respond to different ligands. We believe that reverse engineering natural riboswitches could provide an intellectual foundation for those who wish to build synthetic riboswitches. Also, natural riboswitches might serve as starting points for efforts to change ligand specificity or gene control function through mutation and selection in vitro. In this chapter, we describe how in vitro selection can be used to create variant glmS ribozymes. Additionally, we discuss how these techniques can be extended to other riboswitch classes. Key words: Allosteric ribozyme, Aptamer, Hammerhead ribozyme, In vitro selection, Riboswitch

1. Introduction Riboswitches are metabolite-binding RNA elements that are commonly located within the noncoding regions of mRNAs where they regulate the expression of downstream genes via directly binding to small molecule metabolites (1–3). Riboswitch classes are distinguished by their natural aptamer domains (4) which are responsible for selectively sensing a variety of natural ligands, including coenzymes, amino acids, and nucleobases and their derivatives. Each riboswitch aptamer typically is associated with an expression platform (1, 5) that transduces a ligand-binding event into a change in gene expression. In many instances, natural riboswitches function as allosteric RNAs because ligand

Alexander Serganov (ed.), Riboswitches, Methods in Molecular Biology, vol. 540 DOI: 10.1007/978-1-59745-558-9_25 © Humana Press, a part of Springer Science + Business Media, LLC 2009

349

350

Link and Breaker

binding at one site alters folding of the mRNA at distal sites to control gene expression. One riboswitch class that has several remarkable characteristics compared to most riboswitches is the glmS ribozyme motif (6, 7) (Fig. 1a). To date, all glmS ribozymes have been discovered in bacterial mRNAs that encode L-glutamine:fructose 6-phosphate amidotransferase. This biosynthetic enzyme produces glucosamine-6-phosphate (GlcN6P), which is a

Fig. 1. Consensus and Bacillus cereus glmS ribozyme sequences. (a) Consensus sequence and secondary structure model for glmS ribozymes. Nucleotide numbers conform to those for the B. cereus RNA (Fig. 1b). Boxed nucleotide is absent from the B. cereus sequence, and open circles represent the two extra nucleotides present in B. cereus and Bacillus anthracis ribozymes (Fig. 1b). Nucleotides in black circles are conserved in at least 97% of natural glmS ribozyme representatives. N represents any nucleotide identity, where some stretches are variable in length. Nucleotides denoted Y are either C or U, and nucleotides denoted R are either G or A. Arrowhead identifies the site of ribozyme cleavage. (b) The sequence and secondary structure model for the glmS ribozyme from B. cereus based on comparative sequence analysis (6,7) with revisions based on additional bioinformatics (unpublished data) and X-ray structural data (11,12). Nucleotides 61 and 62 (circled) are found in strains of B. cereus and B. anthracis, but are typically not present in glmS ribozymes from other bacteria (6, 9). Nucleotides depicted in gray were mutagenized to a degeneracy level of 0.09 (14). Mutations were not introduced at nucleotide 52 due to restrictions of the method used to generate the DNA templates for transcription of the G0 RNA population. Arrowhead identifies the site of ribozyme cleavage and the boxed nucleotides identify the minimal functional core of the ribozyme (6). Images reproduced from reference (10) with permission.

In Vitro Selection of glmS Ribozymes

351

precursor compound used in the production of bacterial cell walls. GlcN6P is selectively recognized by glmS ribozymes from Bacillus subtilis (6) and Bacillus cereus (8, 9), and GlcN6P binding results in efficient self-cleavage of glmS ribozymes. Since ligand binding induces ribozyme self-cleavage, glmS ribozymes are ideally suited for use with in vitro selection methods that rely on the separation of functional from inactive RNAs via size purification. We have used the methods described below to assess whether glmS ribozyme variants could be found that either changed or broadened their ligand specificity, which may yield ribozymes that trigger gene control using analogs of GlcN6P (10). Numerous variant ribozymes were isolated that retained function with the original ligand. However none were identified that had altered ligand specificity. Most likely, this was due to the fact that GlcN6P is not a simple allosteric effector that induces cleavage by changing the folding pattern of the RNA. Rather, GlcN6P serves as a cofactor used at the active site of the ribozyme that may protonate the 5¢-oxyanion leaving group during RNA cleavage (11–13). Regardless, methods similar to those described here may be useful to select for true allosteric self-cleaving ribozymes that can be used for gene regulation applications.

2. Materials 2.1. Construction of the Starting Pool

1. Appropriate chemically synthesized oligonucleotides that encode the mutagenized population (see Note 1). Oligonucleotide 1: 5¢-TAATACGACTCACTATAGGATTGTAAATTATAGAAGCGCCAGAACTACAAGTAGTGTAGTTGACGAGGTGGGGTTTATCGAGATTTCGGCGGATGGCTCCCGGTTG. Oligonucleotide 2: 5¢-CCGTGCCTCTTCTCTCATCACACTTTCACCTTTGTCCACTAAGTCAGCTTAATGATTTAAGTAAAAGCTTGCGGTTGTGATGTACAACCGGGAGCCATCCGCCG. Underlined nucleotides represent position that were synthesized with a degeneracy of 0.09 per position (14) 2. 5× reverse transcription buffer: 250 mM Tris–HCl (pH 8.3 at 23°C), 375 mM KCl, 15 mM MgCl2. 3. 100 mM dithiothreitol (DTT). 4. 10× deoxynucleoside 5¢-triphosphate (dNTP) mix: 2 mM each of deoxyguanosine 5¢-triphosphate (dGTP), deoxyadenosine 5¢-triphosphate (dATP), deoxythymidine 5¢-triphosphate (dTTP) and deoxycytidine 5¢-triphosphate (dCTP) (see Note 2).

352

Link and Breaker

5. Reverse transcriptase (e.g., SuperScript II, Invitrogen). 6. 10× transcription buffer: 500 mM HEPES (pH 7.5 at 23°C), 150 mM MgCl2, 20 mM spermidine and 50 mM DTT (see Note 3). 7. 3 M sodium acetate, pH 5.2. 8. 100% ethanol. 9. 5× nucleoside 5¢-triphosphate (NTP) mix: 12.5 mM each of guanosine 5¢-triphosphate (GTP), adenosine 5¢-triphosphate (ATP), uridine 5¢-triphosphate (UTP) and cytidine 5¢-triphosphate (CTP). 10. T7 RNA polymerase (T7 RNAP). 11. [a-32P] UTP, 3,000 Ci/mmole (PerkinElmer), and appropriate equipment for handling 32P. 12. Crush/soak buffer: 10 mM HEPES (pH 7.5 at 23°C) 200 mM NaCl, and 1 mM ethylenediaminetetraacetate (EDTA). 13. 2× denaturing gel-loading buffer: 32% (w/v) sucrose, 0.16% (w/v) sodium dodecyl sulfate, 0.08% (w/v) bromophenol blue, 0.08% (w/v) xylene cyanol, 7.25 M urea, 144 mM Tris–HCl, 144 mM boric acid, and 200 mM EDTA. 14. Reagents and apparatus for denaturing 8 M urea polyacrylamide gel electrophoresis (PAGE). 2.2. Positive Selection

1. Appropriate chemically synthesized oligonucleotides for the amplification of recovered sequences (see Note 1). Oligonucleotide 3: 5¢-TAATACGACTCACTATAGGATTGTAAATTATAGAAGCGCCAGAACTACAAG and Oligonucleotide 4: 5 ¢ -CCGTGCCTCTTCTCTC. 2. 2× glmS assay buffer: 100 mM HEPES (pH 7.5 at 23°C), 20 mM MgCl2 and 400 mM KCl. 3. 10× effector solution (a single compound or a mixture chosen by the experimenter). 4. X-ray film and reagents for developing the X-ray film. 5. PhosphorImager cassettes and PhosphorImager (e.g., Storm PhosphorImager, GE HealthCare).

2.3. Reverse Transcription and PCR Amplification

1. Reverse transcriptase (e.g., SuperScript II, Invitrogen). 2. 10× PCR Buffer: 100 mM Tris–HCl (pH 8.3 at 23°C), 15 mM MgCl2, 400 mM KCl, and 0.1% gelatin. 3. Taq DNA polymerase. 4. Reagents and apparatus for agarose gel electrophoresis.

In Vitro Selection of glmS Ribozymes

353

3. Methods 3.1. General Selection Strategy

In vitro selection was initiated (10) by using a population of variant ribozymes that carry 122 mutagenized nucleotides of a total of 142 nucleotides that comprise the glmS ribozyme from B. cereus (Figs. 1b and 2). The initial population for generation zero (G0) was generated by reverse transcription extension of two synthetic DNAs synthesized with a degeneracy of 0.09 per mutagenized position (14). Transcription of approximately 200 pmol (1.2 × 1014 molecules) of the resulting DNAs yielded a population of RNAs that provide complete coverage of all 6-error mutants. Full-length precursor RNAs were purified by denaturing

Fig. 2. Outline of the basic glmS selection scheme. (a) Two synthetic oligonucleotides are extended using SSII RT to generate double-stranded DNAs encoding the mutagenized glmS ribozymes. (b) The double-stranded DNAs encoding the mutagenized glmS ribozymes are transcribed using T7 RNA polymerase. (c) The full-length RNAs are purified, and the RNAs are subjected to positive selection in the presence of the effectors. (d) The 3¢ cleavage products carrying the randomized ribozyme domains are gel purified and reverse transcribed. (e) The cDNA is then amplified in a PCR reaction that generates double-stranded DNAs encoding the selected glmS ribozyme variants.

354

Link and Breaker

6% PAGE, extracted from the gel, concentrated by precipitation with ethanol, and resuspended in selection buffer. Self-cleavage reactions were initiated by the addition of the effector compounds, and the cleavage reaction incubations were terminated by adding an equal volume of 2× denaturing gel-loading buffer. The mutagenized 3¢ cleavage products were purified by denaturing 6% PAGE, and full-length RNAs were generated by reverse transcription and polymerase chain reaction (RT–PCR). The PCR products from this G0 round of in vitro selection were used to initiate the next round (G1) of selective amplification by repeating the process as described. 3.2. Construction of the Starting Pool

1. Use standard solid-phase methods to chemically synthesize the appropriate DNA oligomers using a defined mixture of phosphoramidites for the randomized region (see Note 1). 2. To prepare the double-stranded DNA template for transcription, incubate 200 pmol each of the two oligomers that encode the G0 population in 400 mL 1× reverse transcription buffer supplemented with 10 mM DTT, 200 mM of the four dNTPs, and 8 U/mL reverse transcriptase. The mixture may be heated to 95°C for 1 min and cooled to room temperature to allow the oligonucleotides to anneal before adding the reverse transcriptase (see Note 4). 3. Incubate at 37–42°C for 2 h. 4. Run a small aliquot out on a 2% agarose gel to confirm that double-stranded DNA was synthesized and is the correct size. 5. Precipitate the double-stranded DNA by adding 0.1 volumes of 3 M sodium acetate and 2.5 volumes of cold ethanol. Incubate the mixture at −20°C for 20 min and pellet by centrifugation at 10,000 × g for 20 min. 6. Resuspend the DNA to yield a 200-mL transcription reaction using 1× transcription buffer supplemented with 2.5 mM of the four NTPs and 20 mCi of [a-32P] UTP. Add 40 U/mL T7 RNA polymerase. 7. Incubate at 37°C for 1–2 h (see Note 5). 8. To concentrate the RNA, precipitate with ethanol and centrifuge as in step 5. 9. Resuspend the RNA pellet in sterile deionized water and add an equal volume of 2× denaturing gel-loading buffer. 10. Separate full-length precursor RNAs from 3¢ cleavage products, premature transcription products, and unincorporated [a-32P] UTP by denaturing 6% PAGE (see Note 6). 11. Visualize the RNA product bands by UV shadowing, excise the gel band containing the full-length RNAs, and cut the excised band into ~1-mm cubes.

In Vitro Selection of glmS Ribozymes

355

12. Elute the RNA from the diced gel slices overnight at 4°C using crush-soak buffer. 13. Transfer the solution containing the eluted RNAs into a fresh tube. 14. Concentrate the sample by precipitation with ethanol and centrifuge as in step 5. 15. Resuspend the RNA in sterile deionized water. 16. Determine the concentration of RNA by measuring the extent of incorporation of [a-32P] UTP (see Note 7). 3.3. Positive Selection

1. Incubate half of the RNAs (minimum of 200 pmol) in 400 mL of 1× glmS assay buffer. This reaction contains no added effectors and serves as the negative control reaction used to monitor the extent of cleavage without effector. Incubate the remaining RNAs (minimum 200 pmol) in the same buffer supplemented with 200 micromolar effectors. This reaction is the positive selection reaction (see Notes 8 and 9). Both reactions are incubated for 30 min at 37°C (see Note 10). 2. To concentrate the RNAs from the reactions, precipitate with ethanol and centrifuge as in Subheading 3.2 , step 5 . 3. Resuspend the RNAs in sterile deionized water and combine with an equal volume of 2× denaturing gel-loading buffer. 4. Separate cleaved RNAs from full-length RNAs by denaturing 6% PAGE. A cleavage marker (e.g., wild-type RNAs allowed to self-cleave to 50% completion) should be run in parallel. Leave at least one lane between the control lane (no effectors), cleavage marker, and positive selection lanes to avoid crosscontamination (see Note 11). 5. Image the resulting gel using a phosphorimager and calculate the fraction of the RNAs that are cleaved (see Note 12). 6. Use an autoradiogram of the gel to precisely locate the 3¢ cleavage products from the positive selection lane, excise the gel band containing the cleaved RNAs, and cut the excised band into ~1-mm cubes (see Note 13). 7. Elute and concentrate the RNAs as in Subheading 3.2, steps 12–15.

3.4. Reverse Transcription and PCR Amplification

1. Add 50 pmol of the 3¢ primer (primer 4) to the resuspended RNAs (unless it has been added as a carrier in Subheading 3.3, step 7; see Note 14) 2. Reverse transcribe the recovered RNAs in 50 mL of 1× reverse transcription buffer supplemented with 10 mM DTT, 200 mM

356

Link and Breaker

of the four dNTPs, and 8 U/mL reverse transcriptase. The mixture may be heated to 95°C for 1 min and cooled to room temperature to allow the oligonucleotides to anneal before adding the reverse transcriptase. 3. Incubate at 37–42°C for 2 h. 4. Prepare PCR mixture by adding 5–10 mL of the reverse transcription reaction and 20–50 pmol each of primers 3 and 4 to a 100-mL PCR reaction containing 1× PCR buffer, 200 mM each of the four dNTPs, and 1 U/mL of Taq DNA polymerase (see Notes 15 and 16). 5. Amplify full-length double-stranded DNAs using the following PCR cycling parameters: segment 1: 94°C for 3 min; segment 2: 94°C for 30 s, 50°C for 30 s, and 72°C for 30 s; segment 3: 72°C for 5 min (see Note 17). 6. Analyze a small aliquot of the PCR reaction on a 2% agarose gel to check the amount and size of the PCR products (see Note 18). 7. Use the amplified DNAs as templates to transcribe the RNA pool for the next round of selection, as in Subheading 3.2, step 6. Save a portion of the PCR product. 3.5. Evaluating the Progression of the Selection

Rounds of selection should be performed by iterations of the protocol described earlier with reduced amounts of RNA (~10 pmol) in the positive selection and negative control reactions. Once the amount of 3¢ cleavage product within the positive selection reaction (effector-dependent RNAs) is at elevated levels relative to the negative control reaction (see Note 19), the selective pressure on the population of RNAs can be increased. This can be achieved by reducing the duration of the positive selection reaction (and subsequently the negative control reaction) or by reducing the concentration of effectors (see Note 20). In response to this pressure, the rate of self-cleavage of the positive selection reaction should start to increase because the population is being enriched for efficient effector-dependent self-cleaving RNAs. Although effector-independent self-cleaving RNAs could also survive the positive selection reaction, these RNAs should be selected against because they should also cleave during transcription. When satisfactory effector-dependent activity is exhibited by the population, DNAs encoding representative members of the population can be cloned and sequenced (see Note 21). The individual RNAs from these DNA clones can be prepared and assayed to determine the performance characteristics of these molecules. Important functional parameters include fold activation in the presence of the appropriate effector, apparent KD of the RNA for its effector, rate constant for RNA cleavage, and the extent of RNA cleavage.

In Vitro Selection of glmS Ribozymes

3.6. Evolving Non SelfCleaving Riboswitches for Altered Specificity

357

The methods described herein can be adapted for use with various ribozyme and aptamer constructs. For some applications it might be beneficial to employ in vitro selection to modify the function of an aptamer domain of a riboswitch that does not undergo self-cleavage. One way to achieve this is to employ allosteric selection techniques (15–18). This can be accomplished by fusing a population of mutagenized riboswitch aptamer domains to a hammerhead ribozyme (or another self-cleaving ribozyme type) using a communication module (Fig. 3) (see Note 22). These RNAs can then be enriched using these techniques for effector-dependent activation of hammerhead self-cleavage. The resulting engineered riboswitch aptamer domains can be subsequently isolated by disintegration of the aptamer domain from the corresponding allosteric ribozymes (18). The methods described here have proven to be useful for isolating glmS ribozyme variants. However, there are numerous strategies in which effector-independent RNAs can survive an allosteric selection. It is worth noting that these effectorindependent RNAs can even dominate the population of RNAs despite the constraints imposed by the selection. Therefore, it is critical to implement an appropriate selection strategy to remove these molecules from the population of RNAs. One common way for effector-independent RNAs to survive a selection is to exhibit slow observed rate constants for selfcleavage. While the total number of effector-independent RNAs

Fig. 3. Construct design for the in vitro selection of allosteric ribozymes. A mutagenized riboswitch aptamer can be grafted onto stem II of the hammerhead ribozyme via a communication module to create a population of RNAs. Allosteric ribozymes that are controlled by ligand binding to the aptamer undergo self-cleavage at the site indicated by an arrowhead.

358

Link and Breaker

will be drastically reduced during the in vitro transcription step (see Note 4), some RNAs will inevitably survive intact. These effector-independent molecules then self-cleave during the positive selection step. Since these effector-independent molecules can be more prevalent in the initial population compared to effector-dependent RNAs, they can dominate a population of RNAs. One way to remove these RNAs from the population is to employ more aggressive negative and positive selection steps (see Note 23). The most common way to determine if effector-independent RNAs have overtaken a population is to observe an increase in the amount of RNAs cleaving during the transcription and/or the negative selection step over several rounds of selection. Another common way an effector-independent RNA can survive an allosteric selection is to partition itself between active and inactive conformation during the negative selection step. The fraction of these molecules that fold into the active conformation can be removed during the negative selection step. However, the fraction of the molecules that do not fold correctly during the negative selection will survive if they correctly fold and cleave during the positive selection step. Altering the incubation times of the negative and positive selection steps will not remove these molecules because these RNAs will always portion into active and inactive conformation. The best way to remove these molecules from the population is to iteratively perform denaturing and refolding treatments during the negative selection reaction. For example, periodic heat spikes (65°C for 1 min) at regular intervals (10–30 min) followed by incubation at room temperature (10–30 min) over a 2 h period can help reduce the selective advantage of these misfolding variants. Even thermal denaturation will lose effectiveness as rounds of selection proceed because some RNAs will emerge that resist denaturation at high temperature. The most effective way to remove these molecules is to perform iterations of negative selection followed by chemical denaturation. Detailed protocols of allosteric selection techniques including these negative selection methods have been previously published (17).

4. Notes 1. For each selection a total of 4 oligonucleotides should be synthesized. Two of the oligomers should encode the population of mutagenized RNAs. The 5¢ oligomer (primer 1) should encode a T7 RNA polymerase promoter and overlap the 3¢ oligomer (primer 2) by 15–20 nucleotides. Two oligomers should be synthesized for use in RT–PCR. The 5¢

In Vitro Selection of glmS Ribozymes

359

oligomer (primer 3) should include approximately 15 nucleotides downstream of the cleave site. The 3¢ oligomer (primer 4) should include nucleotides complementary to the last 15–20 nucleotides of the RNA construct. The DNAs should be purified by denaturing PAGE or high-performance liquid chromatography (HPLC) to ensure length homogeneity. 2. The optimum concentration of dNTPs in a reaction containing SSII RT is 0.5 mM. However, we have observed that SSII RT is capable of producing full-length products in the presence of 0.2 mM dNTPS. This allows the same dNTP mix to be used in reactions containing SSII RT and Taq DNA Polymerase. 3. All buffers that glmS ribozymes are exposed to before or during the in vitro selection reaction should not contain tris(hydroxymethyl)aminomethane (Tris) because this buffer is known to activate the ribozyme (8, 9). 4. The decision to begin the selection with 200 pmol of variant RNAs balances the need for a highly diverse sequence population with the ability to efficiently perform laboratory experiments. Increasing the diversity of the starting population is always advantageous because more sequence space is being searched. However, we do not recommend scaling up the initial population to greater than 2 nmol because populations larger than this are not easily manipulated. Conversely, we do not recommend reducing the initial population to amounts lower than 20 pmol because manipulation does not become easier, but the decreased sequence diversity of the G0 population could impede the selection. 5. A white precipitate may form during the course of in vitro transcription, resulting from the formation of an insoluble complex between Mg2+ and inorganic pyrophosphate. Typically, the formation of the white precipitate indicates that a large amount of RNA has been produced. Inorganic pyrophosphatase can be added to the in vitro transcription reaction at a final concentration of 0.1 U/mL to prevent precipitation. 6. In vitro transcription steps carried out in the absence of ligands and antisense oligonucleotides can serve as a negative selection step, wherein RNAs that cleave in the absence of ligands are removed from the population during gel purification of the full-length RNAs that remain. 7. Typically, in vitro transcription reactions produce approximately 10–20 copies of RNA for each DNA. However, we have determined that in vitro transcription reactions of several glmS ribozyme constructs produce only 1–2 copies of each sequence. Therefore, to insure that there is no loss of

360

Link and Breaker

sequence diversity during in vitro transcription, the researcher should confirm that the amount of RNA produced is greater than the amount of template DNA. Sometimes the number of copies of RNAs can be increased by slightly changing the length of the DNA template or by adding 0.1 U/mL inorganic pyrophosphatase to the transcription reaction. 8. The order of addition of the components of the positive selection is critical. First, the deionized water and the assay buffer are combined. Then RNA is added to the reaction, and the effectors are added last. This order of addition allows the RNAs to prefold in the presence of Mg2+ before the effectors are added to the reaction. 9. The concentration of possible effectors in the positive selection reaction was chosen based on the known properties of the parental RNA. It was known that the apparent KD of GlcN6P and glucosamine-6-sulfate (two known glmS ribozyme effectors) are approximately 200 mM and 20 mM, respectively, assuming that the reduced activity of glucosamine-6-sulfate is entirely due to its poorer affinity (6). Furthermore, nonspecific interactions between RNA and some effectors can be observed when the concentration of the effector is above 1 mM. Therefore, we chose 200 mM as a starting point for all the effector candidates. 10. The incubation time for the positive selection reaction must be chosen to favor effector-dependent self-cleaving RNAs relative to RNAs that display increased rate constants for self-cleavage in the absence of effector. Typically, during the early rounds of selection the incubation time is long (30 min) to favor the recovery of the maximal number of effector-dependent self-cleaving RNAs. During the later rounds of selection, the duration of the incubation time is drastically reduced (~1 min) over several rounds of selection. This change favors the isolation of the most efficient effector-dependent ribozymes and should further reduce the number of RNAs selected that cleave without requiring effector binding. 11. It is possible for the sizes of the RNAs undergoing selectiveamplification to change as the rounds of in vitro selection progress. Therefore, it is often useful to include a marker on the gel to insure that the size of the full-length and 3¢ cleavage products are the correct lengths. 12. The fraction RNAs cleaved in the positive selection reaction should be approximately the same as the negative control during the initial rounds of selection, because inactive variants and variants with wild-type-like activity dominate the population. If the fraction of RNAs cleaved in the positive

In Vitro Selection of glmS Ribozymes

361

selection reactions is reduced compared to the negative control reaction, it could be an indication that one of the effectors within the cocktail is inhibiting self-cleavage. Conversely, if there is significantly more cleavage observed in the positive selection reaction compared to the negative control, before effector-dependent RNAs have been enriched, one or more of the effectors could be nonspecifically enhancing enzyme function. In either case, such an effector will interfere with the isolation of the desired RNAs and should be removed from the effector cocktail and the selection restarted. 13. The amount of ribozyme cleavage product present in early RNA populations may be sufficient to allow detection on an autoradiogram after a brief exposure. However, if an overnight exposure is required, the film cassette should be stored at 4°C to reduce the rate of diffusion of the RNAs within the gel. If an overnight exposure does not permit detection of the 3¢ cleavage product, the size marker can be used to infer the location of these products. 14. Since the presence of efficient effector-dependent ribozymes in the RNA populations of the early selection rounds, it is important to maximize the recovery of the 3¢ cleavage product from the positive selection reaction. Sometimes the amount of 3¢ cleavage product formed during the positive selection is too low for detection, and there is a high risk of losing some (or all) of the enriched molecules during the ethanol precipitation step. To improve the yield of RNAs recovered during the early rounds of selection it is recommended that additional material be added as a “carrier” to aid precipitation by ethanol. The preferred carrier is the 3¢ PCR primer (primer 4; 50 pmol) because it is subsequently required in the reverse transcription step. However, other molecules such as glycogen (20 mg) or linear acrylamide (5 mg) can be used as well. 15. During the early rounds of selection, it is recommended that most (or all) of the cDNA be amplified. This will increase the probability that even the rarest of the active molecules will be amplified. 16. To avoid the amplification of undesired nonspecific PCR products it is suggested that the concentration of KCl in the PCR be optimized. Typically, this can be accomplished by using PCR to amplify a small amount of template DNAs in the presence of a range of KCl (10–50 mM). We have found that a final KCl concentration of 40 mM reduces or eliminates the amplification of undesired PCR products. 17. The number of cycles required for the complete incorporation of all primers in a PCR reaction is dependent on the amount

362

Link and Breaker

of template DNA. Therefore, it is critical to stop the PCR amplification of the cDNA before undesired higher molecular weight products are formed. The optimal way to determine the number of cycles for complete incorporation of the PCR primers is to analyze aliquots of the PCR reaction withdrawn from a scout PCR as the number of cycles increase. 18. Typically, smaller undesired PCR products are transcribed more efficiently than the larger specific products that encode the desired RNAs. If smaller PCR products are present, it is recommended that the desired PCR products be purified by cutting the appropriate agarose band and recovering the DNA by using crush/soak elution, electroelution or by using a Qiagene gel extraction kit. 19. If a cocktail of effectors has been used in the positive selection reaction, it is likely that this response is resulting from a single effector or a subset of the effectors, and it is highly unlikely that it is resulting from all the effectors. Assays should be performed on the RNAs with each individual effector to determine which effectors are responsible for promoting the observed increase in activity. In subsequent rounds of selection, the population should be split into two pools. To optimize ribozyme activity for the most active effector, one pool should be subjected to selection only with the effector responsible for the increased activity. The second pool can be subjected to further rounds of selection with a mixture of the remaining effectors. This process can be repeated each time a dominant effector-dependent class of RNAs is identified. 20. The duration of the positive selection reaction (and subsequently the negative control) should be gradually decreased over several rounds of selection. If the duration of the positive selection reaction (and subsequently the negative control) is reduced too much, there will be a loss of selective advantage for effector-dependent activity and the round of selection should be repeated using a longer incubation time. This loss of effector-dependent activity might not manifest itself until the subsequent round of selection. If this occurs, both rounds of selection should be repeated. Alternatively, several incubation times can be used for the positive selection reaction (and subsequently the negative control). This will allow for several populations of RNAs to be isolated after each round of selection, decreasing the rounds of selection that need to be repeated. Furthermore, if three or more incubation times are used, the observed rate constant for the positive selection reaction (and subsequently the negative control) can be determined. 21. It is very convenient to clone the DNAs encoding the RNAs using a TOPO TA cloning kit (Invitrogen). This method

In Vitro Selection of glmS Ribozymes

363

does not require restriction enzyme sites at the 5¢ and 3¢ ends of the PCR product. 22. Caution has to be used whenever a pre-existing communication module is employed to generate an allosteric ribozyme because communication modules are not always functional when placed in new contexts. Therefore, the observed rate constant of the parent allosteric ribozyme (containing the wild-type sequence of the natural aptamer domain) should be determined in the presence and absence of the small molecule metabolite. If the ribozyme does not sufficiently activate in the presence of the small molecule metabolite, a selection should be performed to isolate a communication module that provides sufficient switching characteristics (19). 23. One way to employ more aggressive negative selection is to conduct a conventional negative selection step prior to the positive selection step. During this negative selection step, the population of RNAs is incubated in selection buffer in the absence of effector. The full-length RNAs are then purified and subjected to the positive selection step using a shorter incubation time to favor the isolation of effector-dependent RNAs. A detailed discussion of negative selection strategies is presented elsewhere (17).

Acknowledgments We would like to thank Adam Roth and Michelle Meyer for critically reading the manuscript. In vitro selection research in the Breaker laboratory is supported by NIH and by the Howard Hughes Medical Institute.

References 1. Mandal, M. and Breaker, R. R. (2004). Gene regulation by riboswitches. Nat. Rev. Mol. Cell Biol. 5, 451–463 2. Winkler, W. C. and Breaker, R. R. (2005). Regulation of bacterial gene expression by riboswitches. Annu. Rev. Microbiol. 59, 487–517 3. Coppins, R. L., Hall, K. B., and Groisman, E.A. (2007). The intricate world of riboswitches. Curr. Opin. Microbiol. 10, 176–181 4. Breaker, R. R. (2006). Riboswitches and the RNA world, in The RNA World (Gesteland,

R. F., Cech, T. R. and Atkins, J. F., eds.), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, NY, pp.89–107 5. Winkler, W. C., Cohen-Chalamish, S. and Breaker, R. R. (2002). An mRNA structure that controls gene expression by binding FMN. Proc. Natl. Acad. Sci. U.S.A. 99, 15908–15913 6. Winkler, W. C., Nahvi, A., Roth, A., Collins, J. A., and Breaker, R. R. (2004). Control of gene expression by a natural metaboliteresponsive ribozyme. Nature 428, 281–286

364

Link and Breaker

7. Barrick, J. E., Corbino, K. A., Winkler, W. C., Nahvi, A., Mandal, M., Collins, J., Lee, M., Roth, A., Sudarsan, N., Jona, I., Wickiser, J. K. and Breaker, R. R. (2004). New RNA motifs suggest an expanded scope for riboswitches in bacterial genetic control. Proc. Nat. Acad. Sci. U.S.A. 101, 6421–6426 8. McCarthy, T. J., Plog, M. A., Floy, S. A., Jansen, J. A., Soukup, J. K. and Soukup, G. A. (2005). Ligand requirements for glmS ribozyme selfcleavage. Chem. Biol. 12, 1221–1226 9. Roth, A., Nahvi, A., Lee, M., Jona, I., and Breaker, R. R. (2006). Characteristics of the glmS ribozyme suggest only structural roles for divalent metal ions. RNA 12, 607–619 10. Link, K. H., Guo, L. and Breaker, R. R. (2006). Examination of the structural and functional versatility of glmS ribozymes by using in vitro selection. Nucleic Acids Res. 34, 4968–4975 11. Klein, D. J., and Ferre-D’Amare, A. R. (2006). Structural basis of glmS ribozyme activation by glucosamine-6-phosphate. Science 313, 1752–1756 12. Cochrane, J. C., Lipchock, S. V., and Strobel, S. A. (2007). Structural investigation of the glmS ribozyme bound to its catalytic cofactor. Chem. Biol. 14, 97–105

13. Cochrane, J. C., Strobel, S. A. (2008). Catalytic strategies of self-cleaving ribozymes. Acc. Chem. Res. 41, 1027–1035 14. Breaker, R. R. and Joyce, G. F. (1994). Inventing and improving ribozyme function – rational design versus iterative selection methods. Trends Biotechnol. 12, 268–275 15. Koizumi, M. Soukup, G. A., Kerr, J. N., and Breaker, R. R. (1999). Allosteric selection of ribozymes that respond to the second messenger cGMP and cAMP. Nat. Struct. Biol. 6, 1062–1071 16. Soukup, G. A., Emilsson, G. A. and Breaker, R. R. (2000). Altering molecular recognition of RNA aptamers by allosteric selection. J. Mol. Biol. 298, 623–632 17. Roth, A., and Breaker R. R. (2004). Selection in vitro of allosteric ribozymes. Methods Mol. Biol. 252, 145–164 18. Soukup, G. A., DeRose, E. C., Koizumi, M., and Breaker, R. R. (2001). Generating new ligand-binding RNAs by affinity maturation and disintegration of allosteric ribozymes. RNA 7, 524–536 19. Soukup, G. A. and Breaker, R. R. (1999) Engineering precision RNA molecular switches. Proc. Natl. Acad. Sci. U.S.A. 30, 3584–3589

INDEX A Adenine riboswitch ............................................. 15, 65–67 2-aminopurine ........................................................... 17, 25 fluorescence ......................................................... 26, 49 modification .............................................................. 15 pair............................................................................. 26 2AP. See 2-aminopurine A-riboswitch, See adenine riboswitch Arthrobacter aurescens ...................................................... 3–4

FMN ......................................................................... 49, 57 extinction coefficient ................................................. 57 fluorescence ............................................................... 49 riboswitch .......................................................49, 53, 55 FRET ........................................................................ 65–66 single-molecule (sm).................................66–67, 73–74 5F-uridine ......................................................337, 340, 345

G

Bacillus cereus.............................................90, 196, 351, 353 Bacillus subtilis.................................... 40, 42–43, 53, 55, 93, 98, 101, 161–162, 212, 281–283, 289, 351 Biotin............................................ 45, 50, 67, 294, 296, 299 bovine serum albumin.....................................68, 73, 75 RNA ...........................................................69, 296, 297

β-galactosidase ............................... 235, 283, 286–287, 330 Glucosamine-6-phosphate (GlcN6P) .................... 129–130, 193, 196, 199, 350 Glycine aptamer ............................................................ 142 Gram-negative bacteria ........................................... 42, 255 Gram-positive bacteria ....................................42, 129, 208, 212, 255, 281 Green fluorescent protein (GFP) .......................... 235, 301 Guanine riboswitch ............................................25, 98, 101

C

H

Caged ligand ..................................................161, 164, 168 Calorimetry ..................................................................... 97 Chaperone ......................................................211, 265, 301 Chemotaxis.................................................................... 330 CMCT ...................................................237–238, 240, 243 Covariance model .......................................1–2, 4–6, 10–12

Heat shock ............................................................ 233, 265 element .................................................................... 270 Hydrodynamic methods .................................................. 77 Hypoxanthine ................................................................ 163 photolabile ............................................................... 166

B

K D Diaminopurine (DAP) ...................................... 97–98, 101 extinction coefficient ............................................... 106 DMS .............................................. 234, 237–238, 240, 243

E Elongation complexes (EC) .................................40, 45, 49 Elongation factor ............................................249, 265–266

F FACS......................................................321–324, 329, 331 Fluorescent labels emission ............................................................... 71–72 excitation ................................................................... 72 extinction coefficient ................................................. 71 fluorescein.......................................................67, 69, 71 indocarbocyanine-3 (Cy3) ........................67, 69, 71, 74 indocarbocyanine-5 (Cy5) ........................67, 69, 74, 75

Kethoxal ......................... 174, 234, 237–238, 240, 243–244 Kinetics....................40, 53, 55, 74, 169, 181, 220, 235–236 K-turn ............................................................................. 11

L Laser .................... 74, 78, 161, 164–165, 168, 170, 323, 328 Lead ........................174, 215–216, 218–220, 224, 229, 234 Ligase .............................................................................. 20 T4 DNA ....................................... 17, 20, 122, 303, 308 T4 RNA .......................................... 17, 20, 28, 71, 218, 223, 303, 312–313 Listeria monocytogenes ..................................................... 234 Luciferase ............................... 305–306, 335, 337–340, 346

M M-box ....................................................................... 93–94 Mfold .............................................................119, 235–236 Molecular dynamics....................................................... 164

365

IBOSWITCHES 366 RIndex

N NAIM ................................................................... 174, 193 NAIS ..................................................................... 193, 196 NMR spectroscopy .................................124, 141, 161–162 Nuclease S1 ............................ 218, 220–221, 224, 229–230

R Rfam database ................................................... 2–3, 11–12 Ribozyme ............................. 2, 36, 123–124, 131, 229, 335 glmS ...........................129–131, 133, 137, 193, 196, 349 group I ..................................................................... 235 group II.................................................................... 235 hairpin ..................................................................... 335 hammerhead ..................... 119, 137, 335–336, 344, 357 HDV ....................................................................... 137 crystals ..................................................................... 136 Rifampicin ................................................................. 42, 47 RNase III........................................................210–211, 224 RNase A ............................98, 110, 112, 218, 220, 224, 228 RNase H.....................................................42, 48, 304, 314 RNase P......................................................................... 235 RNase T1.......................................................130, 133, 135, 174, 176, 181, 218, 220, 224–226, 228 RNase T2 .......................................................218, 220, 224 RNase V1 ........................174, 216, 218, 220, 224, 229, 230 Roadblock.................................................................. 46–47 ROSE ............................................................................ 233 Ruminococcus gnavus ........................................................ 10

SELEX .......................................................................... 291 Serratia proteamaculans....................................................... 3 SHAPE ..........................................................173, 174, 178 Shine-Dalgarno sequence ................. 66, 208, 249, 253, 292 Sigma factor .......................................................... 233, 302 Specifier Sequence ..................................212–213, 282, 287 sRNA ..................................... 209, 215–216, 248–249, 301 Staphylococcus aureus ............................... 218, 221, 247, 249, 255, 257, 261 Stokes radii ...................................................................... 81 Synchronized transcription.............................................. 56

T T box .............................................. 212, 281–282, 285, 289 Thermoanaerobacter tengcongensis .............101, 129–130, 133 Thermodynamics............................................. 53, 235–236 Thermosensor......................... 209, 216, 233, 249, 301, 305 Thermus thermophilus ...................... 247, 249, 255, 258, 260 Toeprinting ............................................................ 247–248 Toyocamycin ...........................................337, 340, 342, 345 TPP riboswitch ..........................................3–4, 48, 93, 291 Transcription termination..........................2, 39–40, 53–55, 208–209, 281–282, 307 Translation initiation .........................................39, 66, 116, 208–212, 234, 247, 255, 291, 302 region................................................208, 233, 248, 301 site. .......................................................................... 235

V

S

Vibrio cholerae ......................................................... 141–142

SAM riboswitch .................................................25, 31, 101 SAXS ..................................................................... 141–142 Sedimentation ......................................................81, 83, 94 coefficient .................................................81, 86, 88, 94 equilibrium ................................................................ 80 velocity....................................................................... 80

W “Walking” .................................................................. 40, 45

X X-ray crystallography...................... 115, 117, 129, 174, 193