RNA Scaffolds: Methods and Protocols [2 ed.] 1071614983, 9781071614983

Citation preview

Methods in Molecular Biology 2323

Luc Ponchon Editor

RNA Scaffolds Methods and Protocols Second Edition

METHODS

IN

MOLECULAR BIOLOGY

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

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

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

RNA Scaffolds Methods and Protocols Second Edition

Edited by

Luc Ponchon Faculté de Pharmacie, University of Paris, PARIS, France

Editor Luc Ponchon Faculte´ de Pharmacie University of Paris PARIS, France

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-0716-1498-3 ISBN 978-1-0716-1499-0 (eBook) https://doi.org/10.1007/978-1-0716-1499-0 © Springer Science+Business Media, LLC, part of Springer Nature 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature. The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.

Preface RNAs are at the center of numerous cellular phenomena and play very different roles in each. One of their roles is in particular that of organization center: the ribosome, RNA telomerase, and “Long Noncoding RNAs” are, among others, examples of RNA structures that recruit other molecules and organize biological processes. These RNAs possess structures allowing interactions with other molecules (proteins, ligands) and thus will potentialize molecular reactions. Advances in structural biology have permitted a definition of the rules with regard to the folding of RNA allowing us today to better understand exactly how they fold and interact. As opposed to DNA, RNAs are able to adopt very variable folds and therefore are able to adopt ligand-specific structures. Contrary to proteins, we are able to create structures composed of different “RNA” modules, each of which is able to keep its activity independent from the others. So, when they are stable, folded RNA can be used as a tool for biological, pharmacological, and/or molecular design studies. RNA presents the peculiarity, like Meccano, of being able to fold into structural domains, which can assemble and sometimes form supramolecular objects. We can isolate, modify, or create an RNA template de novo to make use of its recognition or enzymatic functions. From my point of view, an “RNA scaffold” is a synthetic or natural RNA whose structure, for example, allows one to optimize a reaction, to isolate a molecule, or to favor an interaction. Like Biobricks, the tools based on RNA scaffolds are an example of the emergence of synthetic biology. Indeed, they participate in the creation and construction of biological objects and systems for useful purposes. In this volume, we have tried to be as representative as possible of that which is done today. You will find detailed here processes and techniques that differ greatly from one to another. This book reviews recently developed techniques that use “RNA scaffolds” as molecular tools. These methods cover domains as various as molecular biology, cellular biology, nanotechnology, and structural biology. This book is composed of original chapters and updated chapters from the previous edition.

Contents In order to design a scaffold or understand its interaction with a ligand, it is sometimes necessary to possess certain structural data. A structure can be modeled from an RNA structure data bank. In the first chapter, Chen and his colleagues describe their prediction method, Vfold, that, from a primary sequence, allows one to determine a three-dimensional model of an RNA. This method is based on a pattern-based approach. Thus, they extract different patterns (such as hairpin loops, internal loops, pseudoknot loops, and three-way junctions) from the two-dimensional structure. Based on this data, the implemented algorithm calculates a three-dimensional model. In this chapter, they describe a hybrid method, which combines the motif template-based Vfold3D model and the loop template-based VfoldLA model, to predict RNA 3D structures. In Chapter 2, Sargueil and his colleague present a chemical footprinting procedure. Using small chemical reagents as a probe, this method allows the identification of the interaction site of a ligand with RNA but also RNA

v

vi

Preface

structural rearrangement upon ligand binding. Chemical footprinting could be a good alternative to RNAse footprinting. However, if one seeks to obtain an RNA structure at an atomic level (e.g., for the rationalized design of a scaffold), crystallography is the technique of choice. RNA crystallogenesis remains nonetheless complicated. Indeed, RNAs, as opposed to proteins, do not crystallize as easily and the X-ray diffraction is often weak. In Chapter 3, Ferre´ d’Amare´ and his colleagues describe techniques allowing an improvement in data acquisition and notably for large RNA (over 100 nucleotides) for which only very few structures exist. They propose a protocol allowing the increase in diffraction power of the crystals by combining two techniques: the first consisting of dehydrating the crystal, the second of substituting the divalent ions with strontium ions. Through the example of a gene regulatory tRNA-mRNA complex, they show us the importance of these two techniques on the quality of diffraction. Ke and his colleagues present a method to obtain RNA crystals. Indeed, RNA flexibility is one of the limiting factors to RNA crystallization. To overcome this difficulty, the RNA is embedded in a tRNA scaffold. The tRNA stabilizes the chimera and increases the conformation purity of the RNA target. Using a tRNAgly scaffold, they prove that this method can assist crystallization and phase determination. RNA scaffold can adopt different conformations. The molecular dynamics simulation is one of the tools to investigate these conformations. In Chapter 5, Pasquali and her colleague present a method to identify alternative RNA structures using a particular computational potential energy landscape framework. With the example of the 50 -hairpin of RNA 7SK they illustrate how the method can be applied to interpret experimental results and to obtain a detailed description of molecular properties. A limiting factor for RNA study (notably for long RNAs) is obtaining them in large quantities. Three classic methods exist in order to obtain the RNA: in vitro transcription, chemical synthesis, and cellular extraction. The following chapters describe the techniques permitting, as with proteins, the in vivo production of RNA. Chapter 6 describes a system of RNA-protein coproduction in the bacteria. This protocol is the logical progression of protocols already described in a previous book Recombinant and In Vitro RNA Synthesis in the same collection. In the aforementioned book, the authors describe the production of RNA protected from bacterial nuclease by a “tRNA” camouflage. The RNA is protected at its extremities by the tRNA chassis and accumulated in the bacteria. In this protocol, they propose to coproduce RNA–protein complexes and thus show the advantage of having a joint production of the two molecules. Fox and his colleagues offer a method also permitting the overproduction of RNA in bacteria in Chapter 7. They use a different chassis: the RNA ribosomal 5S. Indeed, like the tRNA, the RNA 5S possesses a fold that allows it to be accumulated in the bacteria. Their protocol thus goes into detail on the production of different RNAs and how to make the cleave. So as to liberate the RNA of interest from the 5S, they propose a cleavage via the DNAzyme. These short sequences of DNA hybridize themselves with the RNA to form a ribozyme-like structure and thus allow the RNA cleavage. In Chapter 8, Daros and his colleagues developed a third method to overexpress RNAs in E. coli. This method is adapted from a viroid system. The RNA of interest is flanked by domains of the viroid hammerhead ribozyme and is coproduced with a tRNA ligase resulting in circular RNAs. Because of the circular structure, the viroid scaffold facilitates the accumulation of RNAs in the bacteria but also the purification steps. Affinity or “pull-down” techniques allow the identification of molecular complexes. In these techniques, a protein linked to a matrix serves as bait. From the cellular extracts, one

Preface

vii

can isolate the linked molecules. Martinez-Salas and his colleagues propose a pull-down system in which RNA serves as bait in order to identify the IRES-binding proteins. To render the system specific and robust, their RNA bait is embedded in a tRNA scaffold and possesses an aptamer for the streptavidin in order to isolate IRES–protein complexes from extracts of eukaryotic cells. Fluorescence remains a technique of choice to perform cellular localization. Green fluorescent protein (GFP) and its numerous derivatives allow one to localize the proteins in cells or tissues. However, it remains difficult to specifically locate the RNA as no naturally fluorescent RNA has been identified to date. It is therefore necessary to use indirect techniques. The following chapters describe RNA scaffolds allowing one to make cellular localization of RNA or metabolites through fluorescence. In Chapter 10, Hammond and his colleagues propose to us a technique allowing the localization of a metabolite via an RNA scaffold, which is linked to a fluorescent chromophore. This RNA is a biosensor composed, on the one hand, of a “spinach aptamer” which links a fluorescent chromophore to DFHB1 and, on the other hand, of a specific riboswitch of the cyclic di-GMP. The idea being that the fixation of the fluorophore derives from the fixation of the di-GMP to its aptamer. In the next chapter, You and his colleagues describe the steps to rationally design, optimize, and apply fluorogenic RNA-based sensors. Through the example of the intracellular imaging of tetracycline in living E. coli cells, they show us how the recognition module induces a duplex formation of the transducer module, which further folds the fluorescence module, i.e., Broccoli, and activates the fluorescence of the fluorophore (DFHB1-1T). In Chapter 12, Winkler and his colleagues use a system composed of a riboswitch of the cyclic di-GMP upstream of a gene reporter. The metabolite-sensing riboswitch controls the expression of Yellow Fluorescent Protein and allows to determine the relative abundance of di-GMP in the cells by fluorescent microscopy. They present an example of their method apply in Bacillus subtilis. In the chapter titled “FRET Analysis of RNA-Protein Interactions Using Spinach Aptamers,” Hennig and his colleague present a method to analyze direct RNA-protein interaction using fluorescent light-up aptamers (FLAPs), and fluorescent proteins for the detection and quantification of a direct RNA-protein interaction. They describe the design and application of a homogenous assay to observe and quantify the interaction of the Pseudomonas aeruginosa bacteriophage coat protein 7 (PP7) with its cognate RNA sequence (pp7-RNA) using the Spinach-DFHBI aptamer as RNA fusion and the red fluorescent mCherry as protein fusion. RNA in vitro evolution or SELEX enables the artificial evolution and selection of RNA molecules that possess a desired property, such as binding affinity for a particular ligand or an activity such as that of an enzyme or catalyst. The first such selections involved isolation of various aptamers that bind to small molecules. The first catalytic RNAs produced by in vitro evolution were RNA ligases, catalytic RNAs that join two RNA fragments to produce a single adduct. Certain RNAs can see their activity regulated by a ligand link like a small molecule or another RNA in the case of aptazymes (allosteric ribozymes). The following two chapters (14 and 15) describe two methods allowing the identification of riboswitches from a random bank of aptazymes. These systems are logic gates; indeed, the aptazymes are composed of an aptamer RNA at a strategic position with regard to the self-cleaving ribozyme so that the structure of the ribozyme is stabilized or destabilized through the link of the ligand. These aptazymes are in the 50 position of an ARNm (coding for a reporter gene), which will or will

viii

Preface

not be degraded based on the efficiency of the riboswitch. Hartig and his colleagues have thus inserted random sequences of aptazymes in the 50 -UTR of the hRluc reporter gene on the plasmid psi-CHECK2 making the expression of the luciferase dependent on the ligand. As for Yokobayashi and his colleagues, they also work in the area of mammalian cells but use the reporter system EGFP. They describe a protocol allowing the screening, at a medium rate, around a hundred aptazymes directly into mammalian cells. Many RNAs, in particular, can assemble themselves in the form of nano-structures. We understand the determinants that govern the structure and the layout of these objects better and better. In the following chapter, a protocol allowing one to design and produce different RNA nano-structures is described. Saito and his colleagues show us how to design RNA origami. They thus manage to give their RNAs a triangular structure. The angles are produced by the interaction of the L7Ae protein with recognition sequences placed in the RNA. The final chapters are devoted to an RNA scaffold that serves as a genetic tool. “Gene silencing” can be induced by small noncoding antisense RNAs, which hybridize on an RNA messenger preventing the gene translation. Silencing RNAs (siRNAs) are thus small coding sequences allowing, in genetics, the switching off of a gene specifically and thereby understanding its function. In Chapter 17, Lee and his collaborators describe a method to us that allows the potentialization of the silencing activity of an siRNA. The siRNA is surrounded by two structures in a stem loop, which increase the half-life of the RNA and thus increase the silencing. This structure that they have named afsRNA (artificial small regulatory RNAs) could thus become a more efficient silencing tool. Yu and his colleagues choose another way to overexpress effective small RNAs using bioengineered RNA agents (BERAs) carrying warhead miRNAs, siRNAs, aptamers, or other forms of small RNAs. RNAs are produced in the cells because of an optimal hybrid tRNA/ pre-miRNA carrier. This system allows large-scale production of effective RNAs and their purification by fast protein liquid chromatography (FPLC) to a high degree of homogeneity. In Chapter 19, Li and his colleagues describe a bacteria-based cancer immunotherapy for cancer treatment. They have reported that bacteria can be engineered using synthetic biology technology to enable nonpathogenic bacteria to express gene silencer, invading transformed cells, escaping endosome and selective silence oncogenes in mammalian cells (termed transkingdom gene silencing or tkRNAi). In this chapter, they describe a novel transkingdom gene silencing vector capable of constitutively expressing long doublestranded RNAs enhancing target gene silencing. This approach can be adapted to multitarget gene silencing. In conclusion, I would like to thank all of the authors who have allowed this book to be published. Their work shows to what extent RNAs are fascinating and interest researchers in very different areas. Naturally, we are a long way from exploring their full potential and I hope that the reading of this book will inspire its readers, especially young researchers, to further study of this area. Paris, France

Luc Ponchon

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

1 Predicting RNA Scaffolds with a Hybrid Method of Vfold3D and VfoldLA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Xiaojun Xu and Shi-Jie Chen 2 RNA Footprinting Using Small Chemical Reagents . . . . . . . . . . . . . . . . . . . . . . . . . Gre´goire De Bisschop and Bruno Sargueil 3 Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jinwei Zhang and Adrian R. Ferre´-D’Amare´ 4 Using tRNA Scaffold to Assist RNA Crystallization . . . . . . . . . . . . . . . . . . . . . . . . . Changrui Lu, Rujie Cai, Jason C. Grigg, and Ailong Ke 5 RNA Modeling with the Computational Energy Landscape Framework . . . . . . . Konstantin Ro¨der and Samuela Pasquali 6 Coexpression and Copurification of RNA–Protein Complexes in Escherichia coli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Margot El Khouri, Marjorie Catala, Bili Seijo, Johana Chabal, Fre´de´ric Dardel, Carine Tisne´, and Luc Ponchon 7 In Vivo Production of Small Recombinant RNAs Embedded in 5S rRNA-Derived Protective Scaffold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Victor G. Stepanov and George E. Fox 8 Production of Circular Recombinant RNA in Escherichia coli Using Viroid Scaffolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jose´-Antonio Daro`s 9 Identification of RNA-Binding Proteins Associated to RNA Structural Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Javier Fernandez-Chamorro, Rosario Francisco-Velilla, Azman Embarc-Buh, and Encarnacion Martinez-Salas 10 Live Cell Imaging Using Riboswitch–Spinach tRNA Fusions as Metabolite-Sensing Fluorescent Biosensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sudeshna Manna, Colleen A. Kellenberger, Zachary F. Hallberg, and Ming C. Hammond 11 Rational Design of Allosteric Fluorogenic RNA Sensors for Cellular Imaging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Qikun Yu, Ru Zheng, Manojkumar Narayanan, and Mingxu You 12 Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell Imaging of Bacteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cordelia A. Weiss and Wade C. Winkler 13 FRET Analysis of RNA–Protein Interactions Using Spinach Aptamers. . . . . . . . . Laura Gerhard and Sven Hennig

ix

v xi

1 13

25 39 49

67

75

99

109

121

141

153 171

x

14

15 16

17

18

19

Contents

Engineering Aptazyme Switches for Conditional Gene Expression in Mammalian Cells Utilizing an In Vivo Screening Approach . . . . . . . . . . . . . . . . Charlotte Rehm, Benedikt Klauser, Monika Finke, and Jo¨rg S. Hartig Aptazyme-Based Riboswitches and Logic Gates in Mammalian Cells . . . . . . . . . . Yoko Nomura and Yohei Yokobayashi Folding RNA–Protein Complex into Designed Nanostructures. . . . . . . . . . . . . . . Tomonori Shibata, Yuki Suzuki, Hiroshi Sugiyama, Masayuki Endo, and Hirohide Saito An Effective Method for Specific Gene Silencing in Escherichia coli Using Artificial Small RNA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geunu Bak, Jee Soo Choi, Wonkyeong Kim, Shinae Suk, and Younghoon Lee Expression and Purification of tRNA/pre-miRNA-Based Recombinant Noncoding RNAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mei-Juan Tu, Halley K. Wright, Neelu Batra, and Ai-Ming Yu Synthetic Biology Medicine and Bacteria-Based Cancer Therapeutics. . . . . . . . . . Jaehyung Lee, Andrew C. Keates, and Chiang J. Li

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

199 213 221

233

249 267 281

Contributors GEUNU BAK • Department of Chemistry, KAIST, Daejeon, South Korea NEELU BATRA • Department of Biochemistry and Molecular Medicine, UC Davis School of Medicine, Sacramento, CA, USA RUJIE CAI • College of Chemistry, Chemical Engineering and Biotechnology, Donghua University, Shanghai, China MARJORIE CATALA • Expression ge´ne´tique microbienne, UMR CNRS 8261, Institut de biologie physico-chimique, Universite´ de Paris, Paris, France JOHANA CHABAL • EryPharm, Paris, France SHI-JIE CHEN • Department of Physics, Department of Biochemistry, and Institute for Data Science and Informatics, University of Missouri, Columbia, MO, USA JEE SOO CHOI • Department of Chemistry, KAIST, Daejeon, South Korea FRE´DE´RIC DARDEL • CiTCoM, UMR CNRS 8038, Universite´ de Paris, Paris, France JOSE´-ANTONIO DARO`S • Instituto de Biologı´a Molecular y Celular de Plantas (Consejo Superior de Investigaciones Cientı´ficas-Universitat Polite`cnica de Vale`ncia), Valencia, Spain GRE´GOIRE DE BISSCHOP • CiTCOM, Cibles The´rapeutiques et conception de me´dicaments, CNRS, Universite´ de Paris, Paris, France; Institut de Recherches Cliniques de Montre´al (IRCM), Montre´al, QC, Canada MARGOT EL KHOURI • Centre for Molecular Oncology, Barts Cancer Institute, Queen Mary University of London, London, UK AZMAN EMBARC-BUH • Centro de Biologı´a Molecular Severo Ochoa, CSIC-UAM, Madrid, Spain MASAYUKI ENDO • Institute for Integrated Cell-Material Sciences (WPI-iCeMS), Kyoto University, Kyoto, Japan JAVIER FERNANDEZ-CHAMORRO • Centro de Biologı´a Molecular Severo Ochoa, CSIC-UAM, Madrid, Spain; Garvan Institute of Medical Research, Darlinghurst, NSW, Australia ADRIAN R. FERRE´-D’AMARE´ • Biochemistry and Biophysics Center, National Heart, Lung and Blood Institute, Bethesda, MD, USA MONIKA FINKE • Department of Chemistry and Konstanz Research School Chemical Biology, University of Konstanz, Konstanz, Germany GEORGE E. FOX • Department of Biology and Biochemistry, University of Houston, Houston, TX, USA ROSARIO FRANCISCO-VELILLA • Centro de Biologı´a Molecular Severo Ochoa, CSIC-UAM, Madrid, Spain LAURA GERHARD • Division of Organic Chemistry, Department of Chemistry and Pharmaceutical Sciences, Vrije Universiteit Amsterdam, Amsterdam, Noord-Holland, The Netherlands JASON C. GRIGG • Department of Microbiology and Immunology, Life Sciences Institute, The University of British Columbia, Vancouver, BC, Canada ZACHARY F. HALLBERG • Department of Chemistry, University of California, Berkeley, Berkeley, CA, USA MING C. HAMMOND • Department of Chemistry, University of Utah, Salt Lake City, UT, USA; Henry Eyring Center for Cell and Genome Science, University of Utah, Salt Lake

xi

xii

Contributors

City, UT, USA; Department of Chemistry, University of California, Berkeley, Berkeley, CA, USA JO¨RG S. HARTIG • Department of Chemistry and Konstanz Research School Chemical Biology, University of Konstanz, Konstanz, Germany SVEN HENNIG • Division of Organic Chemistry, Department of Chemistry and Pharmaceutical Sciences, Vrije Universiteit Amsterdam, Amsterdam, Noord-Holland, The Netherlands AILONG KE • Department of Molecular Biology and Genetics, Cornell University, Ithaca, NY, USA ANDREW C. KEATES • Skip Ackerman Center for Molecular Therapeutics, Division of Gastroenterology and Hepatology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA COLLEEN A. KELLENBERGER • Department of Chemistry, University of California, Berkeley, Berkeley, CA, USA WONKYEONG KIM • Department of Chemistry, KAIST, Daejeon, South Korea BENEDIKT KLAUSER • Department of Chemistry and Konstanz Research School Chemical Biology, University of Konstanz, Konstanz, Germany JAEHYUNG LEE • Skip Ackerman Center for Molecular Therapeutics, Division of Gastroenterology and Hepatology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA YOUNGHOON LEE • Department of Chemistry, KAIST, Daejeon, South Korea CHIANG J. LI • Skip Ackerman Center for Molecular Therapeutics, Division of Gastroenterology and Hepatology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA CHANGRUI LU • College of Chemistry, Chemical Engineering and Biotechnology, Donghua University, Shanghai, China SUDESHNA MANNA • Department of Chemistry, University of Utah, Salt Lake City, UT, USA; Henry Eyring Center for Cell and Genome Science, University of Utah, Salt Lake City, UT, USA ENCARNACION MARTINEZ-SALAS • Centro de Biologı´a Molecular Severo Ochoa, CSIC-UAM, Madrid, Spain MANOJKUMAR NARAYANAN • Department of Chemistry, University of Massachusetts, Amherst, MA, USA YOKO NOMURA • Nucleic Acid Chemistry and Engineering Unit, Okinawa Institute of Science and Technology Graduate University, Onna, Okinawa, Japan SAMUELA PASQUALI • Laboratoire CiTCoM, CNRS UMR 8038, Universite´ de Paris, Paris, France LUC PONCHON • CiTCoM, UMR CNRS 8038, Universite´ de Paris, Paris, France CHARLOTTE REHM • Department of Chemistry and Konstanz Research School Chemical Biology, University of Konstanz, Konstanz, Germany KONSTANTIN RO¨DER • Yusuf Hamied Department of Chemistry, University of Chemistry, Cambridge, UK HIROHIDE SAITO • Center for iPS Cell Research and Application (CiRA), Kyoto University, Kyoto, Japan BRUNO SARGUEIL • CiTCOM, Cibles The´rapeutiques et conception de me´dicaments, CNRS, Universite´ de Paris, Paris, France BILI SEIJO • Faculte´ de biologie et me´decine Ludwig Lausanne Branch, Epalinges, Switzerland

Contributors

xiii

TOMONORI SHIBATA • Center for iPS Cell Research and Application (CiRA), Kyoto University, Kyoto, Japan VICTOR G. STEPANOV • Department of Biology and Biochemistry, University of Houston, Houston, TX, USA HIROSHI SUGIYAMA • Department of Chemistry, Graduate School of Science, Kyoto University, Kyoto, Japan SHINAE SUK • Department of Chemistry, KAIST, Daejeon, South Korea YUKI SUZUKI • Department of Chemistry, Graduate School of Science, Kyoto University, Kyoto, Japan CARINE TISNE´ • Expression ge´ne´tique microbienne, UMR CNRS 8261, Institut de biologie physico-chimique, Universite´ de Paris, Paris, France MEI-JUAN TU • Department of Biochemistry and Molecular Medicine, UC Davis School of Medicine, Sacramento, CA, USA CORDELIA A. WEISS • Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, MD, USA WADE C. WINKLER • Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, MD, USA HALLEY K. WRIGHT • Department of Biochemistry and Molecular Medicine, UC Davis School of Medicine, Sacramento, CA, USA XIAOJUN XU • Institute of Bioinformatics and Medical Engineering, Jiangsu University of Technology, Changzhou, Jiangsu, China YOHEI YOKOBAYASHI • Nucleic Acid Chemistry and Engineering Unit, Okinawa Institute of Science and Technology Graduate University, Onna, Okinawa, Japan MINGXU YOU • Department of Chemistry, University of Massachusetts, Amherst, MA, USA AI-MING YU • Department of Biochemistry and Molecular Medicine, UC Davis School of Medicine, Sacramento, CA, USA QIKUN YU • Department of Chemistry, University of Massachusetts, Amherst, MA, USA JINWEI ZHANG • Laboratory of Molecular Biology, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, USA RU ZHENG • Department of Chemistry, University of Massachusetts, Amherst, MA, USA

Chapter 1 Predicting RNA Scaffolds with a Hybrid Method of Vfold3D and VfoldLA Xiaojun Xu and Shi-Jie Chen Abstract The ever-increasing discoveries of noncoding RNA functions draw a strong demand for RNA structure determination from the sequence. In recently years, computational studies for RNA structures, at both the two-dimensional and the three-dimensional levels, led to several highly promising new developments. In this chapter, we describe a hybrid method, which combines the motif template-based Vfold3D model and the loop template-based VfoldLA model, to predict RNA 3D structures. The main emphasis is placed on the definition of motifs and loops, the treatment of no-template motifs, and the 3D structure assembly from templates of motifs and loops. For illustration, we use the ZIKV xrRNA1 as an example to show the template-based prediction of RNA 3D structures from the 2D structure. The web server for the hybrid model is freely accessible at http://rna.physics.missouri.edu/vfold3D2. Key words Structure prediction, Template-assembly, Secondary structural motifs, Single-stranded loops

1

Introduction To perform crucial cellular functions, RNA molecules fold up to form compact three-dimensional (3D) structures [1–5]. A comprehensive understanding of RNA structures, therefore, can not only provide the fundamental insights into the cellular functions of RNAs but also promote the development of structure-based bioengineering, such as precise gene regulation and editing, as well as effective drug design. However, experimental determination of RNA 3D structures is usually laborious and challenging, leading to the fast development of predictive computational modeling for RNA 3D structure predictions [6–24]. Structure predictions from sequence alone may achieve (near) atomic resolutions for most short RNAs, due to the relatively small degrees of freedom in the conformational space. Additional structural information, such as the base pairing pattern at two-dimensional (2D) level and tertiary contact maps at the 3D level, may be needed to further reduce the

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021

1

2

Xiaojun Xu and Shi-Jie Chen

conformational searching space and increase the prediction accuracy for large RNAs [19–33]. RNA 2D structure as defined by the base pairing pattern provides structural constraints for 3D structure folding. Different approaches have been developed to investigate the impact of RNA 2D structural constraint on 3D conformations [34–40]. For example, conformational analysis based on TOPRNA [38] and MC-sym [20] showed that RNA global conformation is largely defined by topological constraints of RNA secondary structure while the electrostatics, intra- and interloop, and other interactions select specific conformations from the accessible conformational ensemble. Recent investigation for RNA tertiary motifs containing the crosslinked base pairs indicates that, tertiary motif can further decrease the 3D conformational space and impose much stronger topological constraints than the secondary structural motifs, which lack cross-linked base pairs [40]. Taking the advantage of the strong topological constraints of secondary structural motifs, template assembly-based RNA 3D structure models, such as RNAcomposer [19], MC-sym [20], and Vfold3D [21, 22], select templates from known structures with homologous sequence information to predict 3D structures for large RNAs. For example, with the given RNA sequence and 2D structure, Vfold3D extracts the secondary structural motifs of helices, hairpin loops, internal/bulge loops, and multi-branched junctions, and predicts the 3D structures through the assembly of templates of non-helix motifs and A-form helices. With the existence of homologous structures for motifs, the knowledge-based methods for RNA 3D structure prediction can usually achieve higher accuracy and efficiency than the physics-based approaches. However, due to the limitation of current motif template database, the methods are severely limited by the low success rate of finding a proper template for a given motif. Recently, we introduced VfoldLA [23, 24], a hierarchical loop template-assembly method for RNA 3D structure prediction. Unlike the previous models which search for templates based on the whole motif or piece-wise fragments, VfoldLA searches for templates for single-stranded loops, and predicts 3D structures through the assembly of loop templates and A-form helices. It has a much higher success rate to find proper templates than the motifbased approaches, and has also a much higher computational efficiency than fragment-based approaches. In general, VfoldLA can predict a set of all-atom structures from any 2D structure. However, due to the large number of combinations for the different templates of the different loops, it can only handle RNAs containing small number of helices to avoid excessively long computational times.

Predicting RNA Scaffolds with a Hybrid Method of Vfold3D and VfoldLA

3

By combining Vfold3D and VfoldLA, in this chapter, we present a hybrid model for RNA 3D structure predictions with given 2D structures. The model takes the advantage of the conserved 3D structures for secondary structural motifs and the high success rate of finding templates for single-stranded loops. The predicted structures may serve as useful scaffolds for further structure refinement studies.

2

Algorithms

2.1 RNA Secondary Structural Motifs and Single-Stranded Loops

An RNA 2D structure contains helices of canonical base pairs and the single-stranded loops of unpaired nucleotides. According to the different loop–helix connection modes, we define four types of loops, as shown in Fig. 1a: “helix2” loop for a single strand that connects two helices; “hairpin” loop as a special “helix2” loop whose two ends connect to the same helix; “tail5” and “tail3” loops for segments of unpaired nucleotides at the 50 and 30 ends, respectively. The length of each loop L is the number of unpaired nucleotides. According to the different connection modes between helices, we define the secondary structural motifs of internal/bulge loops and multi-branched junctions, as shown in Fig. 1b (see Note 1). Motif size is defined by the length of each loop L within a motif. For example, the three-way junction has three helices connected by three loops of L1, L2 and L3. The size of a three-way junction is denoted by L1–L2–L3. Other motifs, such as pseudoknots and hairpin-hairpin kissing motifs with the cross-linked base pairs, are not included in this model. As shown by the red oval shapes in Fig. 1, we include the loopconnected terminal base pair(s) in the definition of loops and motifs to account for the loop–helix structural interference. The sequence of a loop/motif is formatted in the 50 !30 direction with W–W and N denoting the terminal base pair and the unpaired nucleotides, respectively. For example, W-50 W and W30 -W in the “helix2” loop are the base pairs connected to the 50 and 30 ends of the loop, respectively.

2.2 RNA Motif-based Template Library

The (3D structure) template library for secondary structural motifs was built from 4659 PDB structures (see Note 2), including RNA-involved complexes. It contains 3D templates for bulge loops, internal loops, and multibranched junctions. The motifbased template library can be built in four steps: 1. For a given RNA 3D structure, extract the A-form helices. From the base pairing pattern, the corresponding 2D structure is determined.

4

Xiaojun Xu and Shi-Jie Chen

Fig. 1 The definition of (a) single-stranded loops of hairpin, helix2, and tails, and (b) secondary structural motifs of internal/bulge loops and multibranched junctions. Double-stranded helices are represented by cylinders. The red oval shape in each helix denotes the terminal base pair attached to the loops (shown as red lines). Sequences of loops and motifs are in the format of W–W (terminal base pair) and N (unpaired nucleotide), with all loops in the 50 !30 directions

2. Identify all the secondary structural motifs for the given 3D structure, according to the linkage between helices. 3. Remove the redundant templates for those with root mean square deviation (RMSD) 1.5 Å for the same motif type, same size and identical sequence. 4. Collect all the nonredundant motif structures to construct a template library. Table 1 shows the statistics for the current template library for motifs.

Predicting RNA Scaffolds with a Hybrid Method of Vfold3D and VfoldLA

5

Table 1 RNA loop- and motif-based template libraries Loop name

Number of templates

Motif name

Number of templates

Hairpin loops

2562

Internal/bulge loops

4387

Helix2 loops

8612

Three-way junctions

1230

Tail5 loops

739

Four-way junctions

1168

Tail3 loops

1046

Five-way junctions

380

Six-way junctions

144

Seven-way junctions

217

2.3 RNA Loop-Based Template Library

The template library for single-stranded loops was built from 4659 PDB structures. It contains 3D templates for hairpin, helix2, tail5, and tail3 loops. The loop-based template library can be built in the following steps: 1. For a given RNA 3D structure, extract the A-form helices. 2. Identify all the single-stranded loops for the given 3D structure, according to the linkage between helices and loops. 3. Remove the redundant templates for those with RMSD 1.5 Å for the same loop type, same size and identical sequence. 4. Collect all the nonredundant loop structures to construct a template library. Table 1 shows the statistics for the current template library for loops.

2.4 Sequence Similarity-Based Score

3

Given a query loop/motif sequence, the model scores the sequence similarity according to the following rule: The score si for nucleotide position i is equal to 0 for perfect match, otherwise equal to 1 for purine/pyrimidine-type match, and 2 for no match. For example, sA!G ¼ 1 and sA!C ¼ 2 when nucleotide A is substituted with G and C, respectively. We rank the loop/motif templates according to the total score defined as the sum over all the nucleotides: Sloop/motif ¼ ∑isi. This scoring scheme considers the similarity of nucleotides, assuming that less changes in base substitution result in less changes in 3D structure.

Methods The Vfold3D and VfoldLA combined model predicts RNA 3D structure from a 2D structure by the structural assembly from the templates of motifs and loops. If the specific whole motif-based templates are available and can be successfully found, we predict the 3D structure from the motif-based templates. The motif-based

6

Xiaojun Xu and Shi-Jie Chen

Fig. 2 The workflow of the hybrid model

templates usually give more accurate predictions than the VfoldLA model alone. If a motif does not have available 3D templates in the library, the model breaks the whole motif into single-stranded loops and applies the VfoldLA strategy to predict the 3D structures. Such an approach would lead to a much broader applicability than the Vfold3D alone. As shown in the workflow in Fig. 2, the model works with the following steps.

Predicting RNA Scaffolds with a Hybrid Method of Vfold3D and VfoldLA

7

1. Identify the structural components of double-stranded helices (see Note 3), secondary structural motifs, and single-stranded loops from the given 2D structure. For the motif types excluded in the motif library (such as pseudoknot, and hairpin-hairpin kissing), we break the whole motifs into the single-stranded loops. For example, a pseudoknot with two helices and three loops can be divided into three “helix2” loops. Same treatment can be applied to all the secondary structural motifs without templates. 2. Build the all-atom A-form helical structures. 3. Assemble the helix configurations in 3D space based on templates of all the motifs (with available templates) and “helix2” loops, according to the aforementioned sequence similaritybase scoring scheme. The optimal templates are selected with minimal scores. 4. Exclude helix configurations that involve steric clash or improper chain connectivity. To check steric clash between atoms in the assemble structures, we allow at most five atom pairs with a distance 2.0 Å. To check chain connectivity at the loop–helix interfaces, we allow at most a 5.0 Å heavy-atom RMSD of the helix terminal base pairs for each loop-closing (helix2) loops. 5. For each viable helix configuration, select the optimal templates for other (hairpin, tail5 and tail3) loops to build the whole structure. The optimal templates are chosen from the sequence similarity-based score. 6. Repeat step 5 until the assembled whole structure is clash-free (see Notes 4 and 5). The overall score is calculated by the sum over the scores of all the templates. 7. Terminate the assembly process when the number of the assembled 3D structures reaches a preset value or the computational time reaches a specific amount of time (see Note 6). 8. Cluster the overall score-ranked structures (see Note 7) and refine the structures of clusters. The assembled structures may contain structural clash and nonideal bond lengths, bond angles, or bond torsional angles, especially at the loop–helix interface regions. To avoid long computational time, the web server utilizes the IsRNA model [16], which employs a coarsegrained representation of RNA conformations and the knowledge-based interaction potentials to refine the all-atom structures assembled by the model (see Note 8). 9. The centroid structures of the top clusters are considered as the predicted structures for the given 2D structure (see Note 9). We use the ZIKV xrRNA1 [41] for the illustration of the 3D structure prediction (see Note 10). The native 2D structure

8

Xiaojun Xu and Shi-Jie Chen

Fig. 3 The prediction of the ZIKV xrRNA1. (a) Snapshot of the server input, which indicates the RNA sequence and 2D structure in dot-bracket format, as well as other job information, such as number of clusters, RMSD cutoff for clustering, job name and email address. The input 2D structure contains five helices, one three-way junction and several single-stranded loops, with a loop-kissing helix (H4 as shown in red). (b) Snapshot of the server output, which shows the detailed job information and a JSmol applet for visualization of predicted structures. The best prediction has the RMSD of 8.7 Å, compared with the experimentally determined structure

(Fig. 3a, see Note 11) contains five helices, one three-way junction, and several single-stranded loops with a loop-kissing helix (H4, shown in red). Because of the loop-kissing helix, Vfold3D gives no predictions. However, the hybrid model builds 3D structures of the 3-way junction and loops from the motif template and loop template libraries, respectively, and assemble A-form helical structures to generate all-atom structures. As shown in Fig. 3b, the RMSD between the (best) predicted structure and the experimental determined structure is 8.7 Å. This RMSD is smaller than the prediction from the VfoldLA model alone (RMSD ¼ 12.1 Å). The use of the Vfold3D strategy for the secondary structural motif structure prediction contributes to the improved accuracy.

Predicting RNA Scaffolds with a Hybrid Method of Vfold3D and VfoldLA

4

9

Notes 1. Other multiway junctions, such as four-way and five-way junctions, are not shown in Fig. 1b for clarity. 2. The list of the RNA PDB structures used for constructing the template library includes all the PDB entries released before February of 2020. 3. Each helix should contain at least two base pairs. A lone base pair is treated as an inter-loop interaction. 4. For each helix configuration, the model predicts one all-atom structure with the optimal score, since different loop conformations usually do not cause significant changes to the global fold. 5. If none of the loop templates can satisfy the clash-free criteria, the specific helix configuration would be removed. 6. Currently, the web server terminates predictions when the computational time exceeds 12 h or the number of assembled structures reaches 100, which may lead to no predictions. 7. As shown in Fig. 3a, the RMSD cutoff for clustering is set to 5.0 Å by default. The server users are free to select other values (10 Å) before submission. 8. The IsRNA can resolve most of the structural clash while keeping the overall 3D fold unchanged through small-RMSD conformational sampling. 9. As shown in Fig. 3a, the number of clusters is set to 5 by default. The server users are free to select other values (5) before submission. 10. The sequence of the ZIKV xrRNA1 is: 50 GGGUCAGGCCG GCGAAAGUCGCCACAGUUUGGGGAAAGCUGUGCA GCCUGUAACCCCCCCACGAAAGUGGG30 . 11. The native 2D structure is as follows: ....(((((((((....)))).((((((.. [[[[...))))))..)))))...]]]](((((....))))), which is extracted from the published paper [41] of the crystal structure (PDB id: 5tpy).

Acknowledgments This research was supported by NIH grants R01-GM063732, R01-GM117059, and R35-GM134919.

10

Xiaojun Xu and Shi-Jie Chen

References 1. Eddy SR (2001) Non-coding RNA genes and the modern RNA world. Nat Rev Genet 2:919–929 2. Sharp PA (2009) The centrality of RNA. Cell 136:577–580 3. Chappell J, Takahashi MK, Meyer S, Loughrey D, Watters KE, Lucks J (2013) The centrality of RNA for engineering gene expression. Biotechnol J 8:1379–1395 4. Mortimer SA, Kidwell MA, Doudna JA (2014) Insights into RNA structure and function from genome-wide studies. Nat Rev Genet 15:469–479 5. Cech TR, Steitz JA (2014) The noncoding RNA revolution—trashing old rules to forge new ones. Cell 157:77–94 6. Laing C, Schlick T (2011) Computational approaches to RNA structure prediction, analysis, and design. Curr Opin Struct Biol 21:306–318 7. Sim AY, Minary P, Levitt M (2012) Modeling nucleic acids. Curr Opin Struct Biol 22:273–278 8. Shapiro BA, Yingling YG, Kasprzak W, Bindewald E (2007) Bridging the gap in RNA structure prediction. Curr Opin Struct Biol 17:157–165 9. Sun L-Z, Zhang D, Chen S-J (2017) Theory and modeling of RNA structure and interactions with metal ions and small molecules. Annu Rev Biophys 46:227–246 10. Miao Z, Westhof E (2017) RNA structure: advances and assessment of 3D structure prediction. Annu Rev Biol 46:483–503 11. Strobel EJ, Yu AM, Lucks JB (2018) Highthroughput determination of RNA structures. Nat Rev Genet 19:615–634 12. Sharma S, Ding F, Dokholyan NY (2008) iFoldRNA: three-dimensional RNA structure prediction and folding. Bioinformatics 24:1951–1952 13. Xia Z, Gardner DP, Gutell RR, Ren P (2010) Coarse-grained model for simulation of RNA three dimensional structures. J Phys Chem B 114:13497–13506 14. Shi YZ, Wang FH, Wu YY, Tan Z-J (2014) A coarse-grained model with implicit salt for RNAs: predicting 3D structure, stability and salt effect. J Chem Phys 141:105102 15. Boniecki MJ, Lach G, Dawson WK, Tomala K, Lukasz P, Soltysinski T, Rother KM, Bujnicki JM (2016) SimRNA: a coarse-grained method for RNA folding simulations and 3D structure prediction. Nucleic Acids Res 44:e63

16. Zhang D, Chen S-J (2018) IsRNA: an iterative simulated reference state approach to modeling correlated interactions in RNA folding. J Chem Theory Comput 14:2230–2239 17. Das R, Baker D (2007) Automated de novo prediction of native-like RNA tertiary structures. Proc Natl Acad Sci U S A 104:14664–14669 18. Das R, Karanicolas J, Baker D (2010) Atomic accuracy in predicting and designing noncanonical RNA structure. Nat Methods 7:291–294 19. Popenda M, Szachniuk M, Antczak M, Purzycka KJ, Lukasiak P, Bartol N, Blazewicz J, Adamiak RW (2012) Automated 3D structure composition for large RNAs. Nucleic Acids Res 40:e112 20. Parisien M, Major F (2008) The MC-fold and MC-Sym pipeline infers RNA structure from sequence data. Nature 452:51–55 21. Cao S, Chen S-J (2011) Physics-based de novo prediction of RNA 3D structures. J Phys Chem B 115:4216–4226 22. Xu X, Zhao P, Chen S-J (2014) Vfold: a web server for RNA structure and folding thermodynamics prediction. PLoS One 9:e107504 23. Xu X, Chen S-J (2018) Hierarchical assembly of RNA three-dimensional structures based on loop templates. J Phys Chem B 122:5327–5335 24. Xu X, Zhao C, Chen S-J (2019) VfoldLA: a web server for loop assembly-based prediction of putative 3D RAN structures. J Struct Biol 207:235–240 25. Ponce-Salvatierra A, Astha, Merdas K, Nithin C, Ghosh P, Mukherjee S, Bujnicki JM (2019) Computational modeling of RNA 3D structure based on experimental data. Biosci Rep 39:BSR20181430 26. Das R, Kudaravalli M, Jonikas M, Laederach A, Fong R, Schwans JP, Baker D, Piccirilli JA, Altman RB, Herschlag D (2008) Structural inference of native and partially folded RNA by high-throughput contact mapping. Proc Natl Acad Sci U S A 105:4144–4149 27. Yang S, Parisien M, Major F, Roux B (2010) RNA structure determination using SAXS data. J Phys Chem B 114:10039–10048 28. Kladwang W, Vanlang CC, Cordero P, Das R (2011) A two-dimensional mutate-and-map strategy for non-coding RNA structure. Nat Chem 3:954–962 29. Ding F, Lavender CA, Weeks KM, Dokholyan NV (2012) Three-dimensional RNA structure

Predicting RNA Scaffolds with a Hybrid Method of Vfold3D and VfoldLA refinement by hydroxyl radical probing. Nat Methods 9:603–608 30. Falkner B, Schrder GF (2013) Cross-validation in cryo-EM-based structural modeling. Proc Natl Acad Sci U S A 110:8930–8935 31. Ramani V, Qiu R, Shendure J (2015) Highthroughput determination of RNA structure by proximity ligation. Nat Biotechnol 33:980–984 32. Nguyen TC, Cao X, Yu P, Xiao S, Lu J, Biase FH, Sridhar B, Huang N, Zhang K, Zhong S (2016) Mapping RNA-RNA interactome and RNA structure in vivo by MARIO. Nat Commun 7:12023 33. Cheng CY, Kladwang W, Yesselman JD, Das R (2017) RNA structure inference through chemical mapping after accidental or intentional mutations. Proc Natl Acad Sci U S A 114:9876–9881 34. Bailor MH, Sun X, Al-Hashimi HM (2010) Topology links RNA secondary structure with global conformation, dynamics, and adaptation. Science 327:202–206 35. Bailor MH, Mustoe AM, Brooks CL 3rd, Al-Hashimi HM (2011) Topological constraints: using RNA secondary structure to model 3D conformation, folding pathways, and dynamic adaptation. Curr Opin Struct Biol 21:296–305

11

36. Kim N, Laing C, Elmetwaly S, Jung S, Curuksu J, Schlick T (2014) Graph-based sampling for approximating global helical topologies of RNA. Proc Natl Acad Sci U S A 111:4079–4084 37. Sim AY, Levitt M (2011) Clustering to identify RNA conformations constrained by secondary structure. Proc Natl Acad Sci U S A 108:3590–3595 38. Mustoe AM, Al-Hashimi HM, Brooks CL 3rd. (2014) Coarse grained models reveal essential contributions of topological constraints to the conformational free energy of RNA bulges. J Phys Chem B 118:2615–2627 39. Mustoe AM, Bailor MH, Teixeira RM, Brooks CL 3rd, Al-Hashimi HM (2012) New insights into the fundamental role of topological constraints as a determinant of two-way junction conformation. Nucleic Acids Res 40:892–904 40. Xu X, Chen S-J (2020) Topological constraints of RNA pseudoknotted and loop-kissing motifs: applications to three-dimensional structure prediction. Nucleic Acids Res 48 (12):6503–6512 41. Akiyama BM, Laurence HM, Massey AR, Costantino DA, Xie X, Yang Y, Shi P, Nix JC, Beckham JD, Kieft JS (2016) Zika virus produces noncoding RNAs using a multi-pseudoknot structure that confounds a cellular exonuclease. Science 354:1148–1152

Chapter 2 RNA Footprinting Using Small Chemical Reagents Gre´goire De Bisschop and Bruno Sargueil Abstract RNA is a pivotal element of the cell which is most of the time found in complex with protein(s) in a cellular environment. RNA can adopt three-dimensional structures that may form specific binding sites not only for proteins but for all sorts of molecules. Since the early days of molecular biology, strategies to probe RNA structure have been developed. Such probes are small molecules or RNases that most of the time specifically react with single strand nucleotides. The precise reaction or cleavage site can be mapped by reverse transcription. It appears that nucleotides in close contact or in proximity of a ligand are no longer reactive to these probes. Carrying the RNA probing experiment in parallel in presence and absence of a ligand yield differences that are known as the ligand “footprint.” Such footprints allow for the identification of the precise site of the ligand interaction, but also reveals RNA structural rearrangement upon ligand binding. Here we provide an experimental and analytical workflow to carry RNA footprinting experiments. Key words RNA, Structure, Probing, Ribonucleoprotein, Structural rearrangement

1

Introduction The goal of this experiment is to map the site of interaction of (a) protein(s) on an RNA. In brief, it consists in probing the accessibility of the RNA nucleotides in presence and absence of the protein. The reactivity difference observed between the two conditions is considered to be the footprint of the protein on the RNA. Interestingly, such experiment also monitors the potential RNA structural rearrangement induced upon protein binding. Such experiment can be carried out in vitro using a purified protein, or crude extract, or even in vivo if it is possible to control the conditions in which the complex is formed. The below described protocol focuses on in vitro applications. The binding site of RNA interacting small molecules such as antibiotics [1], or conversely of huge complexes such as the ribosome [2] can also be footprinted owing to this technique.

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_2, © Springer Science+Business Media, LLC, part of Springer Nature 2021

13

14

Gre´goire De Bisschop and Bruno Sargueil

1.1 Footprinting with DMS or SHAPE Reagent

Footprinting experiments can be conducted with any of the different RNA structure probes [3, 4] including RNaseRNases and small molecules. Obviously RNaseRNases are more bulky than small molecules, precluding them to get too close of the protein bound. The footprint yielded by RNaseRNases are therefore often larger, and less precise than those obtained with small molecules. Similarly, RNaseRNases bulkiness preclude them from reaching the nucleotides buried within the structure or within the complex. This can be seen as an advantage because in some occasions small molecules may reach nucleotides within the RNA protein complex, yielding only a very small footprint region, sometime not very convincing. In addition, protein that binds only on double stranded region of the RNA may not yield any chemical footprinting because paired nucleotides are not reactive to such probes. An alternative is to use the double strand specific Cobra venom nuclease V1. Unfortunately, this nuclease is no longer commercially available at the moment. In summary we advise using chemical footprinting for which protocols are described below but, in some cases, enzymatic footprinting is an interesting alternative (see Subheading 3.4). A prerequisite for this experiment is to make sure that it is performed in conditions in which most of the RNA is complexed with the protein. Obviously to reach such state, the reaction mix must contain at least as many protein molecules as RNA molecules, but this may not be enough. As a rule of thumb, the protein concentration should be about tenfold the Kd of the equilibrium, if achievable. When using a purified protein Kd can be determined using different techniques, including electromobility gel-shift assays (EMSA) [5–7] or by filter binding assay [8–10]. When using crude cellular extracts, the situation is complicated because the concentration of the protein of interest is not known and other proteins that specifically or unspecifically bind the RNA may be present in the extract. However, strategies such as gel-shift assay can be attempted. If the system is not amenable to such experiments, different extract concentrations (0.01–1 μg/μl) of total protein added must be assayed [11]. From these preliminary experiments will be deduced the following parameters for the footprinting per se: l

Protein concentration (depending on the binding affinity (Kd)).

l

Cofactor (and concentration) requirement for the complex formation (e.g., ATP, nonhydrolyzable analogue of ATP).

l

Composition of concentration).

l

Incubation time.

the

binding

(reaction)

buffer

(salt

Critical parameters can be further optimized in a few parallel footprinting experiments, for example the protein concentration can be assayed at 5, 10, and 20-fold the Kd.

RNA Footprinting Using Small Chemical Reagents

15

Additionally, RNA folding must be as close to its native state as possible. RNA recovery protocol from cells or from in vitro transcription often include a denaturation step. In addition, RNA may also go through freeze and thaw cycles which also alter the native folding. In such cases, a denaturation–renaturation step helps to homogenize RNA folding. In case a nondenaturing protocol has been used to prepare the RNA, this step may be skipped. Numerous refolding protocols have been proposed, their relative relevance may depend on the RNA considered [12–15]. Herein, we propose a “classical” refolding protocol. The following protocols are designed for dimethyl sulfate (DMS), a reagent revealing adenosines and cytosines, and for SHAPE reagents that can be used for any nucleotides. For the latter we had more consistent results when using fast acting reagents such as the 1-methyl-7-nitroisatoic anhydride (1 M7) and benzoyl cyanide (BzCN). However, our protocol can be easily adapted to any probing reagents, such as cyclo-N0 -[2-(Nmethylmorpholino)ethyl] carbodiimide-p-toluenesulfonate (CMCT) that hits uridines and guanosines, or any RNase by modifying the reaction buffer and the incubation time when necessary [3]. In any case, the principle of the experiment consists in two reactions carried out in parallel in the exact same conditions, the first containing the RNA on its own, and the second with the RNA and its ligand in conditions were the complex is formed (see above). In addition, the two cognate control reactions excluding the probing reagent will be carried out. In summary 4 reactions will be carried out side by side. 1.2 RNA Retrieval and Sequencing

2

Once complex has been probed, the protein can be denatured and eliminated by phenol extraction to ensure that its presence does not influence the reverse transcription step. Although we advise to perform it, this step is optional. However, if you chose to do it, it must be performed on all four tubes. Nucleotides that have reacted with DMS or the SHAPE reagent are modified and induce premature reverse transcription stops. The reactivity is assessed for each nucleotide by the difference of intensity of the stop in the modification reaction and in the mock reaction. The nucleotide is mapped by comparison with the sequencing reaction.

Materials

2.1 Footprinting with DMS and SHAPE Reagent

1. Dry bath. 2. DMS buffer (10): 400 mM HEPES pH 7.5, 1 M KCl, 50 mM MgCl2. 3. DMS (diluted in ethanol 1:12).

16

Gre´goire De Bisschop and Bruno Sargueil

4. DMS stop reaction: 400 mM Tris pH 7.5. 5. Sodium hydroxide. 6. SHAPE buffer (10): 400 mM HEPES pH 7.5 (800 mM when using BzCN), 1 M KCl, 50 mM MgCl2. 7. SHAPE reagent 10 in anhydrous DMSO: 40 mM 1M7 or 400 mM BzCN. 8. Fume hood. 2.2 RNA Retrieval and Sequencing

1. Phenol GTC: 2.5 M guanidinium thiocyanate, 76 mM β-mercaptoethanol, 1% N-lauryl sarcosine, 60% phenol pH 7. 2. Glycogen 20 mg/ml or yeast tRNA 10 mg/ml. 3. Ammonium acetate 5 M. 4. 70% and 100% cold ethanol 5. Refrigerated centrifuge. 6. Dideoxynucleotide 10 mM, either ddTTP, ddATP, or ddCTP. 7. MMLV Reverse Transcriptase RNase H () 200 μ/μl. 8. MMLV RT buffer 5. 9. Deoxynucleotide mix dATP, dTTP, dCTP, and dITP, 10 mM each. 10. Two differently labeled fluorescent primers at 2 μM in water (e.g., D2-PA and D4-PA) (Sigma-Aldrich) if running the sequences on a CEQ8000 capillary sequencer (see Note 1). 11. Thermocycler. 12. RT stop buffer 5: 1.2 M sodium acetate pH 5, 40 mM EDTA, 4 mg/ml glycogen. 13. Sample loading solution (deionized formamide). 14. Capillary electrophoresis (ABI or Ceq8000 AB Sciex).

3

Methods

3.1 Footprinting with DMS and SHAPE Reagents

1. Prepare four tubes with 6 pmol of RNA in 35 μl of water— incubate for 5 min at 85  C. 2. Add 5 μl of prewarmed appropriate buffer (DMS or SHAPE) and let cool down for 10 min at room temperature, spin down briefly. Incubate for an additional 15 min at the probing temperature (30–37  C) (see Note 2). 3. Add 5 μl of the 10x protein in two tubes and 5 μl of the buffer in which the protein is contained in the two other tubes (see Notes 3 and 4). 4. Incubate for 5–10 min at a physiological relevant temperature (30–37  C in most cases).

RNA Footprinting Using Small Chemical Reagents

17

Probing with DMS: 5. Add 5 μl of DMS in two tubes (or 5 μl ethanol for the two mock reactions)—Incubate 5 min at 37  C. 6. Stop the reaction by addition of 400 mM of Tris pH 7.5 (final) and immediately place on ice. Probing with SHAPE Reagents: 7. Add 5 μl of SHAPE reagent 10 in two tubes (or 5 μl DMSO for the two mock reactions) (see Notes 5 and 6). 8. Incubate for five reagent half-lives at 37  C (30 min for NMIA, 2 min for 1M7 or few seconds for BzCN) after which the SHAPE compound is hydrolysed (no inactivation step required) (see Notes 7 and 8). 3.2 RNA Retrieval and Sequencing

Step 1a: RNA retrieval extraction (optional)

and

purification

by

phenol

1. To avoid losing too much of the material, the volume is brought to 500 μl adding 400 μl H2O and 50 μl ammonium acetate 5 M. 2. The reaction mix is extracted with an equal volume of Phenol or Phenol GTC. Vortex for 1 min. Centrifuge for 5 min. Carefully recover 450 μl the aqueous upper phase. Add an equal volume of chloroform. Vortex for 1 min centrifuge for 5 min. Carefully recover 400 μl the aqueous upper phase. Step 1b: RNA retrieval by ethanol precipitation 3. Add 1 μl of glycogen (20 mg/ml) or tRNA (10 mg/ml), 10% (v/v) of ammonium acetate 5 M (if step 1a was skipped), 250% (v/v) of 100% cold ethanol and incubate for at least 30 min at 20  C. 4. Centrifuge for 30 min at 16,000  g at 4  C. Discard the supernatant. 5. Wash with 400 μl of 70% cold ethanol. 6. Centrifuge for 5 min at 16,000  g at 40  C. Discard the supernatant—repeat this step. 7. Air-dry the pellet and resuspend it in 10 μl of water (see Note 9). Step 2: Reverse transcription 8. Transfer the resuspended RNAs to PCR tubes and prepare four PCR tubes with 6 pmol of RNA in 10 μl of water for the sequencing reactions. 9. Add 1 μl of DMSO to the samples (modified RNA, modified RNA–protein complex, mock reactions, and sequencing reactions), denature for 3 min at 95  C then flash cool on ice.

18

Gre´goire De Bisschop and Bruno Sargueil

10. Add 3 μl of 2 μM D2-labelled primer to the sequencing reaction and 3 μl of D4-labelled primer to the modified/mock RNA. Incubate for 5 min at 65  C then anneal for 10 min at 35  C and store on ice. 11. Add 4 μl of 5 RT buffer, 1 μl of the 10 mM deoxynucleotides mix, and 1 μl of MMLV reverse transcriptase. For the sequencing reactions, also add 1 μl of 10 mM of one dideoxynucleotide. Incubate for 2 min at 35  C, 30 min at 42  C and 5 min at 55  C then store on ice (see Note 10). 12. Add 5 μl of 5 RT stop buffer. 13. Combine one elongation from the modified RNA or the mock RNA with one sequencing reaction. 14. Add 10% (v/v) of ammonium acetate 5M and 250% (v/v) of 100% ethanol, incubate for at least 30 min at 20  C and precipitate the DNA as described above in Step 1b “Step 1b: RNA retrieval by ethanol precipitation” ). 15. Air-dry the pellet and resuspend in 40 μl of sample loading solution (see Notes 11 and 12). Step 3: Sequencing by capillary electrophoresis 16. Load side-by-side 20 μl of the elongated cDNA corresponding to the modified RNA and to the mock RNA. 17. Run the capillary electrophoresis. 18. Export the raw data as text files (see Note 13). 3.3

Data Analysis

1. Extract normalized reactivities and average them. Several software have been developed to treat cDNA traces to obtain reactivities [16–21]. The experimental procedure described here is compatible with QuSHAPE, but note that this set-up may need to be adapted in case if you wish to use a different software. A thorough workflow with detailed procedures to normalize and average reactivities is described elsewhere (Allouche et al. in press). Care should be taken when aligning the sequence to the (+) and () traces since even a single nucleotide frameshift in the base calling of the “free RNA” traces versus the “bound RNA” traces may lead to erroneous results. QuSHAPE provides a convenient “sequence alignment by reference” tool allowing a reference “project” (e.g., the free RNA condition) to serve as a template for the sequence alignment of another “project” (the bound RNA condition). We recommend performing footprint experiments in triplicate and average the reactivities. 2. Compare reactivities to infer the binding mode of a protein. The reactivity pattern obtained with the RNA only can be used to constrain computer assisted RNA secondary structure prediction using software or workflow such as RNAstructure [22],

RNA Footprinting Using Small Chemical Reagents

19

RSample [23], RNAfold [24], IPANEMAP [25]. This may ease the interpretation of the footprint, and yield a secondary structure protein binding site, but this is not an obligation to interpret a footprint. Independently of the secondary structure modelling, there are several ways of comparing the reactivity profiles obtained in the presence and absence of protein. We here propose a method which highlights the most significant differences. This heuristic approach can be adjusted in a caseby-case manner, with the aim of identifying nucleotides that fall into a different reactivity interval upon protein binding, for example, getting from “highly reactive” to “reactive.” An Excel file illustrating a simple yet powerful analysis chart is available on demand ([email protected]). Briefly, the reactivity difference Diff ¼ Rbound  Rfree and an bound R free j absolute Ratio ¼ jR Rbound þRfree are calculated, while all nucleotides with undetermined values (usually labeled with a negative value) are excluded from the analysis. Differences are considered significant when those variables exceed a threshold, typically 0.2 for the absolute difference and 0.2 for the ratio. The threshold ratio allows to mask differences between high values that would otherwise be overrepresented. Indeed, the measure of reactivity clearly gets noisier for high values. Conversely, the difference threshold removes weakly reactive nucleotides that exhibit only a small shift in reactivity yet with a high ratio. Finally, in case data contains multiple replicates, only the statistically significant differences are kept (t-test < 0.05). The provided Excel file contains two sheets; it is filled with an example dataset. Raw reactivities must be entered in the “Data” sheet along with the RNA sequence and its index. Nucleotides for which the absolute deviation to the mean reactivity is over 0.2 are highlighted, so that the consistency of the replicates can be quickly assessed. This threshold may be freely changed in the cell below the “threshold” cell. The differences, the ratios and the t-tests are automatically calculated in the “Analysis” sheet. The last column gives the footprint results according to the different thresholds described above (adjustable by the user). A histogram displaying the reactivity differences between the free and the bound RNA and highlights the footprint sites is provided. This allows the user to instantly visualize the effect of the threshold values modification. Differences between the bound-RNA reactivity profile and the RNA only reactivity profile are of three different types: (a) Significant decreases in reactivity can be interpreted as potential binding sites but keep in mind that they may also reflect a local stabilization of the structure upon protein binding or even a more extensive structural rearrangement.

20

Gre´goire De Bisschop and Bruno Sargueil

(b) Conversely, significant increases in reactivity can only be interpreted as sites that are destabilized in the complex, either because of a local structure alteration or because of a more global protein-induced structural rearrangement. (c) Finally, the majority of the nucleotides will likely have no reactivity change upon binding, consistent with the existence of specific interaction sites. In the opposite case, the absence of RNase from the protein sample should be checked. Of note double-strand specific binding protein may induce little or no reactivity change, especially if their mode of binding does not destabilize the targeted RNA helix. Lastly, the interpretation will benefit to account for the specificity of the chemical probe used. The reactivity profile obtained with base-specific probes such as DMS and CMCT is poorly sensitive to a protein that essentially binds to the phosphate – sugar backbone of the RNA. But conversely, footprints observed, if not due to a structural stabilization, can be attributed to direct contacts with the chemical group of the nucleotide targeted by the probe. Alternatively, SHAPE probes reactivity relies on the nucleotide flexibility and are therefore likely to be sensitive to any protein binding mode that restrain the degree of freedom of the RNA strand. Note that such analysis can be also be carried out with probing experiments coupled to NGS such as SHAPE-Map [26] or DMS-Map [27]. 3.4

Troubleshooting

1. No footprint observed (a) Check if any reactivity is observed in the no-protein condition by visually inspecting the raw traces. The signal decay of the (+) trace should be stronger than in the () trace. In case not, adjust the reaction conditions (add more probing reagent, increase the probing reagent concentration). Check if the probing reagent is still active (this is particularly sensitive in the case of SHAPE reagent which hydrolyzes with time). (b) The probing reaction was not carried out in good protein binding conditions. Check the binding with one of the techniques suggested in the introduction, adjust protein and/or salt concentration, incubation temperature. (c) The nucleotides involved in protein binding are not sensitive to the probe. Change the probing reagent including enzymes. In such case it may be particularly relevant to use the double strand RNase such as the Cobra Venom V1 enzyme to detect double strand RNA binding proteins.

RNA Footprinting Using Small Chemical Reagents

21

2. No signal is observed in the presence of the protein: (a) The protein (extract) sample is contaminated with an RNase. Check the sample. (b) RNA was carried out with the protein at the precipitation or phenol extraction step. Make sure to dissociate the complex by heating, adding some salts or detergent before one of these steps. 3. No signal observed in presence or absence of the protein or too many stops present even in the mock reaction: Check your buffers and reagents for RNase contamination.

4

Notes 1. Primer sequence should be chosen at least 50 to 75 nucleotides 30 (downstream) to the region of interest. 2. Reaction buffers mentioned above should be considered as a starting point and they should be adapted to each situation, notably monovalent and divalent cations nature and concentrations may be changed to optimize the specific binding of the RNA ligand. However, do not use Tris buffer with DMS because reacts with amine groups—and increase the buffer (any) concentration when using BzCN (SHAPE reagent) to compensate for the reagent acidification. If the protein volume added is significant, do not forget to take into account the mono and divalent salts brought by the protein sample. 3. If required the cofactor should also be added at this step. 4. If the protein is not available at tenfold the required concentration, the volume of protein added can be increased, in which case the volume in Subheading 3.1, step 1 should be decreased in consequence, and the buffer in Subheading 3.1, step 2 will also need to be adapted. If the volume in Subheading 3.1, step 1 goes under 20 μl we advise to pool the 4 tubes to limit evaporation. Remember that in such set up most of the buffer and salts will be brought by the protein sample. 5. In order to rapidly mix the solution, be careful to add the RNA solution onto the SHAPE reagent rather than the opposite. 6. All these reagents are genotoxic and/or skin and/or eye irritant. They should be handled with care under a fume hood wearing nitrile gloves and safety glasses. Waste tips and dispose soiled materials in a saturated sodium hydroxide solution, or a 10% acetic acid solution for DMS and CMCT respectively. SHAPE reagents get hydrolyzed relatively rapidly in water. 7. The solution will turn yellowish upon reaction with 1M7 or cloudy with BzCN.

22

Gre´goire De Bisschop and Bruno Sargueil

8. As the modification of one nucleotide may influence the rest of the structure, multiple modifications per RNA molecule should be avoided. As a rule of thumb, we consider that such conditions are met when the amount of the full-length product in at least 70% of that of that mock reaction. The modification conditions described above are pretty standard and work well for RNA between 200 and 500 nucleotides. They can be adjusted by modifying the incubation time and reagent concentration to get more or less modification (NB: concentration or time should be increased for shorter RNAs). This applies for DMS and CMCT, but not for SHAPE reagents which do not produce more than one modification per RNA under the above described conditions. 9. Probed RNA can be stored at 20  C for at least 6 months. 10. For the sequencing reaction pick a nucleotide which is well represented all along the RNA sequence. 11. Resuspension will take time at room temperature. It can be accelerated by incubating the samples for 10 min at 50  C. 12. Elongated cDNA can be stored at 20  C. 13. If you do not have access to a capillary sequencer, you may use radioactively labelled primers and run the samples on an appropriate sequencing P.A.G.E. Premature stops induced by the reagent will then be qualitatively assessed by comparison of the bands in the modification and the mock reaction. Note that the premature stop is one nucleotide 30 to the modified nucleotide.

Acknowledgments We wish to thank N Chamond, L Ponchon, and C Vasnier for sharing their protocols and tips, and for fruitful discussions, and M Pospiech for careful proofreading. References 1. von Ahsen U, Noller HF (1993) Footprinting the sites of interaction of antibiotics with catalytic group I intron RNA. Science 260:1500–1503 2. Angulo J, Ulryck N, Deforges J et al (2016) LOOP IIId of the HCV IRES is essential for the structural rearrangement of the 40SHCV IRES complex. Nucleic Acids Res 44:1309–1325 3. Brunel C, Romby P (2000) Probing RNA structure and RNA-ligand complexes with chemical probes. In: Methods in enzymology,

vol 318. Academic Press, Cambridge, Massachusetts, pp 3–21 4. Ehresmann C, Baudin F, Mougel M et al (1987) Probing the structure of RNAs in solution. Nucleic Acids Res 15:9109–9128 5. de Bisschop G, Ameur M, Ulryck N et al (2019) HIV-1 gRNA, a biological substrate, uncovers the potency of DDX3X biochemical activity. Biochimie 164:83–94 6. Ryder SP, Recht MI, Williamson JR (2008) Quantitative analysis of protein-RNA

RNA Footprinting Using Small Chemical Reagents interactions by gel mobility shift. Methods Mol Biol 488:99–115 7. Sargueil B, Pecchia DB, Burke JM (1995) An improved version of the hairpin ribozyme functions as a ribonucleoprotein complex. Biochemistry 34:7739–7748 8. Carey J, Cameron V, de Haseth PL et al (1983) Sequence-specific interaction of R17 coat protein with its ribonucleic acid binding site. Biochemistry 22:2601–2610 9. Chamond N, Deforges J, Ulryck N et al (2014) 40S recruitment in the absence of eIF4G/4A by EMCV IRES refines the model for translation initiation on the archetype of Type II IRESs. Nucleic Acids Res 42:10373–10384 10. Deforges J, de Breyne S, Ameur M et al (2017) Two ribosome recruitment sites direct multiple translation events within HIV1 gag open reading frame. Nucleic Acids Res 45:7382–7400 11. Vallejos M, Deforges J, Plank TD et al (2011) Activity of the human immunodeficiency virus type 1 cell cycle-dependent internal ribosomal entry site is modulated by IRES trans-acting factors. Nucleic Acids Res 39:6186–6200 12. Chillo´n I, Marcia M, Legiewicz M et al (2015) Chapter one - native purification and analysis of long RNAs. In: Woodson SA, Allain FHT (eds) Methods in enzymology, vol 558. Academic Press, Cambridge, Massachusetts, pp 3–37 13. Herschlag D (1995) RNA chaperones and the RNA folding problem. J Biol Chem 270:20871–20874 14. Jaeger L, Westhof E, Michel F (1993) Monitoring of the cooperative unfolding of the sunY group I intron of bacteriophage T4: the active form of the sunY ribozyme is stabilized by multiple interactions with 30 terminal intron components. J Mol Biol 234:331–346 15. Uhlenbeck OC (1995) Keeping RNA happy. RNA 1:4–6 16. Karabiber F, McGinnis JL, Favorov OV et al (2013) QuShape: rapid, accurate, and best practices quantification of nucleic acid probing information, resolved by capillary electrophoresis. RNA 19:63–73

23

17. Cantara WA, HatterschideJ WW et al (2017) RiboCAT: a new capillary electrophoresis data analysis tool for nucleic acid probing. RNA 23:240–249 18. Mitra S, Shcherbakova IV, Altman RB et al (2008) High-throughput single nucleotide structural mapping by capillary automated footprinting analysis. Nucleic Acids Res 36: e63–e63 19. Pang PS, Elazar M, Pham EA et al (2011) Simplified RNA secondary structure mapping by automation of SHAPE data analysis. Nucleic Acids Res 39:e151 20. Kim H, Cordero P, Das R et al (2013) HiTRACE-web: an online tool for robust analysis of highthroughput capillary electrophoresis. Nucleic Acids Res 41:W492–W498 21. Yoon S, Kim J, Hum J et al (2011) HiTRACE: high-throughput robust analysis for capillary electrophoresis. Bioinformatics 27:1798–1805 22. Deigan KE, Li TW, Mathews DH et al (2009) Accurate SHAPE-directed RNA structure determination. PNAS 106:97–102 23. Spasic A, Assmann SM, Bevilacqua PC et al (2018) Modeling RNA secondary structure folding ensembles using SHAPE mapping data. Nucleic Acids Res 46:314–323 24. Lorenz R, Luntzer D, Hofacker IL et al (2016) SHAPE directed RNA folding. Bioinformatics 32:145–147 25. Saaidi A, Allouche D, Regnier M, Sargueil B, Ponty Y (2020) IPANEMAP: integrative probing analysis of nucleic acids empowered by multiple accessibility profiles. Nucleic Acids Res 48:8276–8289 26. Siegfried NA, Busan S, Rice GM et al (2014) RNA motif discovery by SHAPE and mutational profiling (SHAPE-MaP). Nat Methods 11:959–965 27. Zubradt M, Gupta P, Persad S et al (2017) DMS-MaPseq for genome-wide or targeted RNA structure probing in vivo. Nat Methods 14:75–82

Chapter 3 Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment Jinwei Zhang and Adrian R. Ferre´-D’Amare´ Abstract The crystallization and structural determination of large RNAs and their complexes remain major bottlenecks in the mechanistic analysis of cellular and viral RNAs. Here, we describe a protocol that combines postcrystallization dehydration and ion replacement that dramatically improved the diffraction quality of crystals of a large gene-regulatory tRNA–mRNA complex. Through this method, the resolution limit of X-ray data extended from 8.5 to 3.2 Å, enabling structure determination. Although this protocol was developed for a particular RNA complex, the general importance of solvent and counterions in nucleic acid structure may render it generally useful for crystallographic analysis of other RNAs. Key words X-ray Crystallography, Crystal dehydration, Ion replacement, Riboswitch, T-box RNA, tRNA

1

Introduction The rapid exploration of the noncoding genome using highthroughput technologies is revealing critical roles for RNA structure in a wide range of cellular processes [1]. In addition, many DNA and RNA viruses utilize defined three-dimensional RNA folds to enable their life cycle and achieve infectivity [2–6]. Despite critical roles of structured RNAs in biology and their impact on human health, their functional elucidation is hampered by a paucity of available structural information [4, 7–11]. One hundred years after its invention, X-ray crystallography still provides the highestresolution structural information for macromolecules, especially for RNA domains. The rarity of diffraction-quality crystals of larger RNAs (longer than 100 nucleotides) remains a major roadblock that hinders their structure determination [4]. Compared to proteins, the polyanionic nature of the RNA backbone, reduced chemical diversity of the four nucleobases, paucity of long-range contacts, as well as inherent conformational flexibility, render it

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_3, © Springer Science+Business Media, LLC, part of Springer Nature 2021

25

26

Jinwei Zhang and Adrian R. Ferre´-D’Amare´

difficult for RNAs to form specific, stable crystal packing contacts [7]. Many RNA crystals only diffract X-rays to resolutions in the range of 5 to 8 Å, insufficient to provide biochemical insight (~3.5 Å or better is desirable). For protein crystals, many postcrystallization treatment strategies such as annealing, dehydration, and supplementing ligands have been developed [12–15]. For RNA, the marked sensitivity of crystals to hydration status, documented in the decades spanning publications on yeast tRNAPhe and the glmS riboswitch-ribozyme, hints that diffraction quality could also be significantly improved by postcrystallization treatments [16– 18]. In order to facilitate the development of general strategies and methods for crystallographic studies of larger RNAs, we detail and rationalize a protocol that enabled the crystallization and structure determination of a large tRNA–mRNA complex [19, 20]. Exploiting the general importance of RNA solvation and counterions in stabilizing compactly folded RNAs [21], this method concurrently dehydrates the RNA crystals and substitutes the divalent cations in them. This two-pronged approach drives quasi-rigid body movements of the RNA complexes in the crystal, causing them to achieve geometrically and energetically superior packing.

2

Materials 1. Oligonucleotides for PCR amplification. 2. Taq DNA polymerase, 5000 U/mL (New England Biolabs). 3. T7 RNA polymerase, 50,000 U/mL (New England Biolabs). 4. Diethylpyrocarbonate (DEPC)-treated water (see Note 1). 5. RNA Binding Buffer: 50 mM HEPES-KOH, pH 7.0, 100 mM KCl, 20 mM MgCl2, 5 mM tris (2-carboxyethyl) phosphine (TCEP). 6. 20 mM spermine solution, in DEPC-treated water, filtered through 0.2 μm filter. 7. Crystallization Solution: 50 mM Bis-Tris (HCl) pH 6.5, 0.3 M Li2SO4, 20 mM MgCl2, 20% (w/v) polyethylene glycol (PEG) 3350. 8. EasyXtal 15-Well Tool (Qiagen). 9. MicroSieves and MicroSaws (MiTeGen). 10. 90
angled MicroLoops or MicroMounts (MiTeGen) 11. Crystal Treatment Solutions: 50 mM Bis-Tris (HCl), pH 6.5, 100 mM KCl, 20–50 mM SrCl2 or 20–100 mM MgCl2, 40–45% PEG3350, 5 mM TCEP.

Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment

3

27

Methods

3.1 Design and Synthesis of T-Box RNA and tRNA for Crystallization

1. Initial biochemical and biophysical characterization of T-box RNA–tRNA complexes [22] was essential for design and engineering of crystallization constructs (Fig. 1). Glycine-specific glyQ/glyQS T-box constructs from 20 species were selected from a multiple sequence alignment, with preference given to thermophilic, extremophilic, and pathogenic organisms. The T-box and tRNA constructs were transcribed in vitro using T7 RNA Polymerase, purified by denaturing Urea-PAGE, electroeluted, washed once with 1 M KCl and extensively with DEPC-treated water, concentrated and stored at 4
C or 20
C before use [18, 23]. 2. T-box RNAs from a range of bacterial species were evaluated for their propensity to form monodisperse, stoichiometric complexes with tRNA using nondenaturing PAGE. T-boxes from a handful of bacterial species, such as the extremely halotolerant and alkaliphilic Oceanobacillus iheyensis eventually used in structural determination, exhibited robust tRNA binding, forming tRNA–mRNA complexes that migrated as relatively sharp bands on nondenaturing gels. 3. Full-length T-box RNAs that contain both the Stem I and the antiterminator domains (Fig. 1a) exhibited a tendency to form dimers, presumably due to the thermodynamic instability of the antiterminator [24]. Therefore, T-box RNAs were truncated at a series of lengths and their affinities toward tRNA and tendency to form monodisperse complexes evaluated using isothermal titration calorimetry (ITC) and nondenaturing gels. This analysis demonstrated that Stem I is the minimal T-box domain that is both necessary and sufficient for high-affinity, specific binding to tRNA (Fig. 1b, c) [19]. 4. To aid crystallization of RNA, several RNA-binding proteins have been successfully used as crystallization chaperones, such as the human spliceosomal U1A protein [25] and recombinant antibody fragments (Fabs) [26–28]. The Kink-turn (K-turn) is a widespread bistable RNA structural motif initially discovered on the ribosome that sharply kinks the RNA duplex backbone by 120
and is the landing platform to recruit several conserved proteins to accomplish a range of cellular functions [29– 32]. The T-box Stem I domain harbors a conserved, functionally important K-turn [30, 33, 34]. The crucial contribution of the K-turn to T-box architecture and function is further accentuated by recent structural and biochemical elucidations of a full-length T-box-tRNA complex and a novel class of translational T-box riboswitches [35, 36]. To stabilize the bistable K-turn structure, provide added opportunities for crystal

28

Jinwei Zhang and Adrian R. Ferre´-D’Amare´

Fig. 1 Sequences and secondary structures of full-length and truncated T-box riboswitch RNA used for crystallization. (a) Secondary structure and sequence conservation of a full-length B. subtilis glycineresponsive glyQS T-box riboswitch and its cognate tRNAGly. Circles with dark and light orange shades indicate highly conserved (>80%) and moderately conserved (50–80%) sequences, respectively. Salient structural features on both RNAs are boxed and annotated. Intermolecular T-box-tRNA base-pairing interactions are indicated by solid lines connecting the boxed sequences. (b) Secondary structure of Oceanobacillus iheyensis glyQ T-box Stem I domain used for cocrystallization. The italic, red sequences and red arrows denote engineered regions. (c) Secondary structure of engineered B. subtilis/O. iheyensis tRNAGly (identical sequences) used for cocrystallization. The original tRNA acceptor stem sequence is circularly permuted and capped with a stable GAAA tetraloop (red, italic sequences). The bidirectional arrow denotes the length variations to screen for optimal crystal contacts. (d) Representative crystals of T-box–Stem I–tRNA complexed with Methanococcus jannaschii L7Ae. All scale bars represent 200 μm. (e) Representative crystals of T-box–Stem I–tRNA complexed with B. subtilis YbxF. Note the differences in crystal morphology as dictated by the protein component in the complex

packing, and allow for phasing using selenomethionines, a panel of K-turn binding proteins were tested for their ability to support crystal growth and improve crystalline order. Interestingly, the choice of K-turn binding protein appreciably influenced the crystal morphology. While the presence of thermophilic Methanococcus jannaschii L7Ae protein [37] (and other species of L7Ae) yielded star-shaped nonsingle crystals (Fig. 1d), the addition of mesophilic B. subtilis YbxF [38] produced single, square-plate-shaped crystals (Fig. 1e). The

Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment

29

latter is much more amenable to diffraction data collection. In the absence of any K-turn-binding protein, only nondiffracting crystals of T-box-tRNA binary complexes were occasionally observed. 3.2 Crystallization of the T-Box Stem I– tRNA–YbxF Ternary Complex

1. Dilute concentrated tRNAGAAA (~ 1 mM; 24 g/L) to ~20 μM using DEPC-treated water to reduce intermolecular interaction and dimerization. 2. “Snap-cool” tRNAGAAA by incubating at 90
C for 3 min followed by rapid cooling to 4
C using a thermocyler (see Note 2). 3. Concentrate refolded tRNA to ~12 g/L (500 μM) using Amicon spin concentrators (10 kD MWCO; 0.5 mL). 4. Mix 200 μM each T-box Stem I RNA and snap-cooled tRNA in RNA Binding Buffer (Materials), incubate first at 50
C for 10 min and then at 37
C for 30 min. 5. Add one equivalent selenomethionyl B. subtilis YbxF to the RNA complex. 6. Add spermine to 2 mM. The mixture may become transiently cloudy. Mix gently with a pipette tip. 7. Heat to melt a stock of 2% low-melting-point agarose solution and allow it to cool to 37
C using a heat block to prevent it from solidifying. 8. Mix 1:1 the sample solution and Crystallization Solution and keep at 37
C. 9. Add 1/10 volumes of 2% low-melting-point agarose solution and gently mix by pipetting up and down. The presence of agarose fibers in crystal solvent channels has been shown to lend mechanical support to the crystals [39, 40]. The presence of 0.2% low-melting-point agarose effectively prevents the T-box cocrystals from cracking induced by the sudden change in osmolarity (Fig. 2a). In addition to providing mechanical support for the crystals, the agarose network also reduces convection and permits more uniform crystal growth into thicker dimensions. This beneficial effect on crystal habit and morphology has recently been observed again in the cocrystals of Nocardia farcinica ileS T-box-tRNA complex [36]. 10. Transfer the crystallization mixture onto cover slides and initiate crystallization experiments by hanging drop vapor diffusion.

3.3 Postcrystallization Treatments

1. Square-plate-shaped crystals of the T-box–tRNA–YbxF ternary complex start appearing as early as 1–2 days. Diffraction quality crystals tend to grow more slowly, reaching final dimensions of 300  300  50 μm3 over the course of 1–3 weeks (Fig. 2a).

30

Jinwei Zhang and Adrian R. Ferre´-D’Amare´

Fig. 2 Postcrystallization treatments dramatically improve diffraction quality of large RNA complexes. (a) Postcrystallization treatment procedures and effects on the crystal appearance. Due to the drastic changes in osmolarity (e.g., induced by a 20–40% change in PEG3350 concentration), pervasive crystal cracking and even disintegration occurs. Cracking is effectively prevented by the mechanical support from the agarose fibers in the solvent channels of the crystals. Note the agarose network that transferred together with the embedded crystals. All scale bars denote 200 μm. (b) Comparison of magnified portions of diffraction

Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment

31

These crystals have the symmetry of space group C2221, with unit cell dimensions of a ¼ 108.7 Å, b ¼ 108.8 Å, c ¼ 291.8 Å. As do many other macromolecular crystals with relatively long unit cell edges, the longest unit cell edge (291.8 Å) of these crystals is parallel to the shortest physical dimension of the crystals, that is, the edge that describes the thickness of the rectangular or rhombic plates. Thus, oscillation diffraction images that result from incident X-rays that traverse through the broad faces of the plates suffer from significant overlap of neighboring reflections. Such overlap is circumvented by the use of 90
bent crystal loops, which restrict the incident X-rays to only entering and exiting the crystals through their shortest physical “edges” but not their “faces” (Fig. 2a). 2. Depending on the final concentration of low-melting point agarose in the crystallization drop and temperature, the entire drop may exhibit consistencies ranging from fluid liquid, viscous liquid, jelly-like solid to robustly solid. Select appropriate tools to transfer crystals into ~200 μL Crystal Treatment Solutions in glass depression plates, that is, use conventional nylon loops to transfer individual crystals from nonviscous liquid drops, and use tools such as MicroSieves (MiTeGen) to transfer whole, solidified drops. The composition of Crystal Treatment Solutions will vary as it is based on both the RNA Binding Solution and the Crystallization Solution. A gradient of concentrations of the primary precipitant (20–50% PEG3350 in this example) is scouted to achieve a range of final solvent contents and the effect on diffraction quality is measured. Different concentrations of a panel of divalent cations in particular the alkaline earth metals (Mg2+, Ca2+, Sr2+, Ba2+) should be screened, both for supporting crystal growth and for postcrystallization treatment. In the case of the T-box complex crystals, crystal growth in Mg2+ combined with postcrystallization treatment in Sr2+ stood out as the optimal procedure, producing the best Bragg spots profiles required for de novo phasing using single-wavelength anomalous dispersion (SAD). 3. Seal each well of the depression plate using a glass cover slide and Vaseline. Incubate the crystals in Crystal Treatment Solution (Materials) for 16 h. For the crystals of the T-box ternary

ä Fig. 2 (continued) oscillation photographs of untreated (as-grown) crystals (left, PDB: 4TZP), partially treated crystals (middle panels and top right panel, PDB: 4TZV, 4TZW, and 4TZZ), and crystals that were subjected to full cation replacement and dehydration (lower right panel; PDB: 4LCK) to demonstrate the improvement in spot profile and order-to-order separation. Arrows indicate progressive additions of treatments. Diffraction limits are indicated below each panel. Postcrystallization treatment and resulting crystal properties are summarized in Table 1

32

Jinwei Zhang and Adrian R. Ferre´-D’Amare´

complex, shorter treatments (i.e., less than 4 h) generally do not produce the full effect of the treatment. 4. Carefully dissect the crystals out from their surrounding agarose network using MicroSaws (MiTeGen) and remove as much as agarose as possible (Fig. 2a). As the orientation of the crystals in the crystal loop is critical for reducing overlap during data collection, it is essential to trim nearly all agarose away from the crystal faces so that the plate-like crystals would be mounted parallel to the plane of the 90
bent loop due to surface tension. 5. Using a 90
bent loop such as the angled MicroLoops or MicroMounts (MiTeGen), pick up single, trimmed crystals and immediately plunge into liquid nitrogen for vitrification. As the Crystal Treatment Solution already contains at least 40% (w/v) polyethylene glycol (PEG) 3350, no additional cryoprotective agent is necessary. 3.4 Understanding the Basis of Treatment-Induced Improvement of Crystal Quality

1. Structure determination of as-grown, untreated crystals, and a number of crystals subjected to various combinations of postcrystallization treatments (Fig. 2b & Table 1) allowed the tracking of macromolecular movements in these crystals in response to the treatments received [19, 20]. 2. Structural alignment of untreated and optimally treated crystals revealed that the ternary complexes of T-box–tRNA–YbxF shift closer to each other in the crystal as quasi-rigid bodies (Fig. 3a; see Note 3), producing superior packing contacts such as three intimate base-stacking interactions between symmetry-related complexes (Fig. 3b) as well as a stable A-minor interaction between the engineered GAAA tetraloop on tRNA acceptor stem and the minor groove of the proximal region of T-box Stem I (Fig. 3c). 3. The unique preference for Sr2+ in postcrystallization treatments of T-box cocrystals may be rationalized by its specific association with the 30 cis-diols of neighboring symmetryrelated T-box RNAs (Fig. 3d), its frequent bidentate innersphere interactions with the Hoogsteen faces of purines (Fig. 3e), or its presence at bulges and junctions where phosphates cluster, or bridging across the narrow major groove. The ability of Sr2+ to bind RNA 30 termini and its flexible coordination geometry are properties that may allow it to improve crystalline packing of RNA [41].

Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment

33

Table 1 Select properties of crystals treated with varying degrees of ion replacement and dehydration Unit cell dimensions (Å)

VM (Å3/ Da)

VS (%)

C2221

108.7, 108.8, 291.8a

3.26

74.6

5.0

P43212

75.7, 75.7, 270.2a

2.93

71.7

20

4.7

P43212

75.3, 75.3, 268.9a

2.89

71.3

0

48

3.6

P21

70.6, 260.7, 70.7b

2.46

66.3

40

40

3.2

C2221

100.8, 109.7, 268.1a

2.81

70.4

PDB code

Li2SO4 (mM)

MgCl2 (mM)

SrCl2 (mM)

PEG 3350 (% w/v)

Resolution Space (Å) group

4TZP

300

20

0

20

8.5

4TZV 0

20

0

20

4TZW 0

0

50

4TZZ 0

100

4LCK 0

0

VM Matthews coefficient (Matthews, 1968) Vs Calculated solvent content a α ¼ β ¼ γ ¼ 90
b α ¼ γ ¼ 90
, β ¼ 92.8


4

Notes 1. DEPC is a toxic alkylating agent. It should be handled with appropriate personal protective equipment in a chemical fume hood. DEPC-treated water is nontoxic, because after mixing, the water–DEPC mixture is autoclaved. Heating in the presence of water converts DEPC into nontoxic carbon dioxide and ethanol. 2. tRNAGly and other tRNAs are known to form dimers in solution depending on conditions used for folding the RNA. To reduce dimerization, tRNAs are diluted in DEPC-treated water and “snap-cooled,” which favors tRNA folding while suppressing intermolecular association. 3. Note that the space group as well as the unit cell dimensions of the crystals have changed significantly in response to the postcrystallization treatments (Table 1).

Acknowledgments We thank the staff at beamlines 5.0.1 and 5.0.2 of the ALS and ID-24-C and ID-24-E of APS, in particular, K. Perry and K.R. Rajashankar of the Northeastern Collaborative Access Team (NE-CAT) of the APS for support in data collection and

34

Jinwei Zhang and Adrian R. Ferre´-D’Amare´

Fig. 3 Treatment-induced, in-crystal movements of T-box ternary complexes produce superior crystal contacts. (a) In-crystal redistribution of T-box ternary complexes as rigid bodies driven by dehydration and cation replacement. Overlay of T-box ternary complexes in untreated (as-grown) crystals (light blue, PDB: 4TZP) and fully dehydrated and cation-exchanged crystals (dark blue, PDB: 4LCK). The corresponding crystallographic unit cells are also shown, indicating close to ~10% compression along both a and c axes. The reference complexes in the center of the panel superimpose well (RMSD for 172 C10 < 1.4 Å), but the neighboring four complexes shift substantially closer as a result of the postcrystallization treatment (RMSDs range from 3 to 10 Å and 10 to 19 Å, for RNA C1’ and protein Cα, respectively). Red arrows denote directions of displacement (translation and rotation) of the four neighboring complexes. (b) Treatment-induced formation of an intimate crystal contact involving three symmetry-related T-box ternary complexes, shown in blue, green, and teal, respectively. Molecules from the untreated (PDB: 4TZP) and fully cation-replaced and dehydrated crystals (PDB: 4LCK) are overlaid and colored in pastel and solid colors, respectively. Parallel

Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment

35

processing; G. Piszczek (National Heart, Lung and Blood Institute, NHLBI), R. Levine, and D.-Y. Lee (NHLBI) for assistance with biophysical and mass spectrometric characterization; and N. Baird, T. Hamma, C. Jones, M. Lau, A. Roll-Mecak, and K. Warner for discussions. This work is partly based on research conducted at the ALS on the Berkeley Center for Structural Biology beamlines and at the APS on the NE-CAT beamlines (supported by National Institute of General Medical Sciences grant P41GM103403). Use of ALS and APS was supported by the US Department of Energy. This work was supported in part by the intramural programs of the National Heart, Lung and Blood Institute (NHLBI) and National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), and National Institutes of Health (NIH). References 1. Cech TR, Steitz JA (2014) The noncoding RNA revolution-trashing old rules to forge new ones. Cell 157:77–94 2. Watts JM, Dang KK, Gorelick RJ et al (2009) Architecture and secondary structure of an entire HIV-1 RNA genome. Nature 460:711–716 3. Fang X, Wang J, O’Carroll IP et al (2013) An unusual topological structure of the HIV-1 rev response element. Cell 155:594–605

4. Cantara WA, Olson ED, Musier-Forsyth K (2014) Progress and outlook in structural biology of large viral RNAs. Virus Res 193:24–38 5. Bou-Nader C, Gordon JM, Henderson FE et al (2019) The search for a PKR code-differential regulation of protein kinase R activity by diverse RNA and protein regulators. RNA 25:539–556 6. Hood IV, Gordon JM, Bou-Nader C et al (2019) Crystal structure of an adenovirus virus-associated RNA. Nat Commun 10:2871

ä Fig. 3 (continued) lines denote intermolecular stacking between nucleobases of symmetry-related complexes. Arrows indicate displacements between the untreated and fully treated states. The rear face of the interdigitated T-loops of Stem I distal region (opposite the face interacting with the tRNA elbow) form a prominent flat surface available for crystal packing [19]. Two patches of this flat surface (A39 and A60 respectively), upon full treatment, engage in direct stacking contact with the apical adenine of the GAAA tetraloop capping the tRNA acceptor stem (tA73) of a second complex (green), and with the terminal base pair of T-box Stem I (G1·C102) of a third complex (teal), respectively. Dotted triangle surrounds an intermolecular A-minor interaction (detail in c) present only in the fully treated crystals. tRNA residue numbers are preceded by ‘t’. (c) Detail of the intermolecular class-I A-minor interaction formed between the tetraloop of a tRNA and the minor groove of the proximal region of Stem I, colored as in (b). (d) Interfacial Sr2+ ions bridge symmetryrelated T-box RNA molecules by binding to their 30 termini. A well-defined Sr2+ ion (green sphere) is seen bound to the cis-diol of the T-box RNA 30 terminal nucleotide (C102, marine), and two nonbridging oxygen atoms of a symmetry-related T-box molecule (cyan), through inner-sphere coordination. Similar inner-sphere coordination between Sr2+ and two 30 cis-diol groups bridge two symmetry-related heptanucleotide derived from tRNAAla acceptor stem [42]. The Sr2+ bridging two symmetry-related T-box RNA thus may have contributed significantly to the improved crystal quality through Sr2+ soaking. (e) Pervasive Sr2+ binding to RNA nucleobases and backbone, such as a pair of well-defined Sr2+ ions next to T-box G43, one of which makes bidentate innersphere interactions with the Hoogsteen face of G43. The electron density shown in (d) and (e) is a portion of a composite simulated anneal-omit 2|Fo|  |Fc| synthesis contoured at 1.5 s.d. overlaid with the final refined model

36

Jinwei Zhang and Adrian R. Ferre´-D’Amare´

7. Zhang J, Ferre-D’Amare AR (2014) New molecular engineering approaches for crystallographic studies of large RNAs. Curr Opin Struct Biol 26:9–15 8. Jones CP, Ferre-D’Amare AR (2015) Recognition of the bacterial alarmone ZMP through long-distance association of two RNA subdomains. Nat Struct Mol Biol 22:679–685 9. Warner KD, Sjekloc´a L, Song W et al (2017) A homodimer interface without base pairs in an RNA mimic of red fluorescent protein. Nat Chem Biol 13:1195–1201 10. Stagno JR, Liu Y, Bhandari YR et al (2017) Structures of riboswitch RNA reaction states by mix-and-inject XFEL serial crystallography. Nature 541:242–246 11. Chen MC, Tippana R, Demeshkina NA et al (2018) Structural basis of G-quadruplex unfolding by the DEAH/RHA helicase DHX36. Nature 558:465–469 12. Heras B, Martin JL (2005) Post-crystallization treatments for improving diffraction quality of protein crystals. Acta Crystallogr D Biol Crystallogr 61:1173–1180 13. Russo Krauss I, Sica F, Mattia CA et al (2012) Increasing the X-ray diffraction power of protein crystals by dehydration: the case of bovine serum albumin and a survey of literature data. Int J Mol Sci 13:3782–3800 14. Deng X, Davidson WS, Thompson TB (2012) Improving the diffraction of apoA-IV crystals through extreme dehydration. Acta Crystallogr Sect F Struct Biol Cryst Commun 68:105–110 15. Awad W, Svensson Birkedal G, Thunnissen MM et al (2013) Improvements in the order, isotropy and electron density of glypican-1 crystals by controlled dehydration. Acta Crystallogr D Biol Crystallogr 69:2524–2533 16. Kim SH, Quigley G, Suddath FL et al (1973) Unit cell transormations in yeast phenylalanine transfer RNA crystals. J Mol Biol 75:429–432 17. Klein DJ, Ferre´-D’Amare´ AR (2009) Crystallization of the glmS ribozyme-riboswitch. Methods Mol Biol 540:129–139 18. Klein DJ, Ferre´-D’Amare´ AR (2006) Structural basis of glmS ribozyme activation by glucosamine-6-phosphate. Science 313:1752–1756 19. Zhang J, Ferre´-D’Amare´ AR (2013) Co-crystal structure of a T-box riboswitch stem I domain in complex with its cognate tRNA. Nature 500:363–366 20. Zhang J, Ferre´-D’Amare´ AR (2014) Dramatic improvement of crystals of large RNAs by cation replacement and dehydration. Structure 22:1363–1371

21. Draper DE (2004) A guide to ions and RNA structure. RNA 10:335–343 22. Grundy FJ, Henkin TM (1993) tRNA as a positive regulator of transcription antitermination in B. subtilis. Cell 74:475–482 23. Milligan JF, Uhlenbeck OC (1989) Synthesis of small RNAs using T7 RNA polymerase. Methods Enzymol 180:51–62 24. Zhang J, Ferre´-D’Amare´ AR (2014) Direct evaluation of tRNA Aminoacylation status by the T-box riboswitch using tRNA-mRNA stacking and steric readout. Mol Cell 55 (1):148–155 25. Ferre´-D’Amare´ AR (2010) Use of the spliceosomal protein U1A to facilitate crystallization and structure determination of complex RNAs. Methods 52:159–167 26. Shechner DM, Grant RA, Bagby SC et al (2009) Crystal structure of the catalytic core of an RNA-polymerase ribozyme. Science 326:1271–1275 27. Koldobskaya Y, Duguid EM, Shechner DM et al (2011) A portable RNA sequence whose recognition by a synthetic antibody facilitates structural determination. Nat Struct Mol Biol 18:100–106 28. Shelke SA, Shao Y, Laski A et al (2018) Structural basis for activation of fluorogenic dyes by an RNA aptamer lacking a G-quadruplex motif. Nat Commun 9:4542 29. Klein DJ, Schmeing TM, Moore PB et al (2001) The kink-turn: a new RNA secondary structure motif. EMBO J 20:4214–4221 30. Winkler WC, Grundy FJ, Murphy BA et al (2001) The GA motif: an RNA element common to bacterial antitermination systems, rRNA, and eukaryotic RNAs. RNA 7:1165–1172 31. Daldrop P, Lilley DM (2013) The plasticity of a structural motif in RNA: structural polymorphism of a kink turn as a function of its environment. RNA 19:357–364 32. Lilley DM (2012) The structure and folding of kink turns in RNA. Wiley Interdiscip Rev RNA 3:797–805 33. Zhang J, Ferre-D’Amare AR (2015) Structure and mechanism of the T-box riboswitches. Wiley Interdiscip Rev RNA 6:419–433 34. Suddala KC, Zhang J (2019) An evolving tale of two interacting RNAs-themes and variations of the T-box riboswitch mechanism. IUBMB Life 71:1167–1180 35. Li S, Su Z, Lehmann J et al (2019) Structural basis of amino acid surveillance by higher-order tRNA-mRNA interactions. Nat Struct Mol Biol 26:1094–1105

Improving RNA Crystal Diffraction Quality by Postcrystallization Treatment 36. Suddala KC, Zhang J (2019) High-affinity recognition of specific tRNAs by an mRNA anticodon-binding groove. Nat Struct Mol Biol 26:1114–1122 37. Hamma T, Ferre´-D’Amare´ AR (2004) Structure of protein L7Ae bound to a K-turn derived from an archaeal box H/ACA sRNA at 1.8 Å resolution. Structure 12:893–903 38. Baird NJ, Zhang J, Hamma T et al (2012) YbxF and YlxQ are bacterial homologs of L7Ae and bind K-turns but not K-loops. RNA 18:759–770 39. Biertu¨mpfel C, Basquin J, Suck D et al (2002) Crystallization of biological macromolecules

37

using agarose gel. Acta Crystallogr D Biol Crystallogr 58:1657–1659 40. Lorber B, Sauter C, The´obald-Dietrich A et al (2009) Crystal growth of proteins, nucleic acids, and viruses in gels. Prog Biophys Mol Biol 101:13–25 41. Hofer TS, Randolf BR, Rode BM (2006) Sr (II) in water: a labile hydrate with a highly mobile structure. J Phys Chem B 110:20409–20417 42. Mueller U, Schu¨bel H, Sprinzl M et al (1999) Crystal structure of acceptor stem of tRNA(ala) from Escherichia coli shows unique G.U wobble base pair at 1.16 a resolution. RNA 5:670–677

Chapter 4 Using tRNA Scaffold to Assist RNA Crystallization Changrui Lu, Rujie Cai, Jason C. Grigg, and Ailong Ke Abstract Recent studies have solidified RNA’s regulatory and catalytic roles in all life forms. Understanding such functions necessarily requires high-resolution understanding of the molecular structure of RNA. Whereas proteins tend to fold into a globular structure and gain most of the folding energy from tertiary interactions, RNAs behave the opposite. Their tertiary structure tends to be irregular and porous, and they gain the majority of their folding free energy from secondary structure formation. These properties lead to higher conformational dynamics in RNA structure. As a result, structure determination proves more difficult for RNA using X-ray crystallography and other structural biology tools. Despite the painstaking effort to obtain large quantities of chemically pure RNA molecules, many still fail to crystallize due to the presence of conformational impurity. To overcome the challenge, we developed a new method to crystallize the RNA of interest as a tRNA chimera. In most cases, tRNA fusion significantly increased the conformational purity of our RNA target, improved the success rate of obtaining RNA crystals, and made the subsequent structure determination process much easier. Here in this chapter we describe our protocol to design, stabilize, express, and purify an RNA target as a tRNA chimera. While this method continues a series of work utilizing well-behaving macromolecules/motifs as “crystallization tags” (Ke and Wolberger. Protein Sci 12:306–312, 2003; Ferre-D’Amare and Doudna. J Mol Biol 295:541–556, 2000; Koldobskaya et al . Nat Struct Mol Biol 18:100–106, 2011; Ferre-D’Amare et al. J Mol Biol 279:621–631, 1998), it was inspired by the work of Ponchon and Dardel to utilize tRNA scaffold to express, stabilize, and purify RNA of interest in vivo (Ponchon and Dardel. Nat Methods 4:571-576, 2007). The “tRNA scaffold,” where the target RNA is inserted into a normal tRNA, replacing the anticodon sequence, can effectively help the RNA fold, express in various sources and even assist crystallization and phase determination. This approach applies to any generic RNA whose 50 and 30 ends join and form a helix. Key words tRNA scaffold, RNA expression, Purification, Crystallization

1

Introduction RNA expression and purification has recently attracted attention across all biomedical fields as large amount of homogeneous RNA became critical in many RNA-related studies. However, RNA degradation, misfolding, and structural heterogeneity poses technical difficulties. In order to solve these problems, early approaches by Joachim Frank attach the target RNA onto the ribosome, replacing a loop on the ribosome surface [1]. Here we describe a similar

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_4, © Springer Science+Business Media, LLC, part of Springer Nature 2021

39

40

Changrui Lu et al.

approach, pioneered by Ponchon and Dardel in 2007, that replaces the tRNA anticodon sequence with a generic RNA sequence [2]. This “chimera” design enhances the target RNA stability without a cumbersome ribosome or any protein tags attached [3–9], enabling large-scale (recombinant) expression [10]. This chapter describes a protocol to obtain milligrams of target RNA–tRNA chimera construct through in vitro transcription. A similar approach using recombinant E. coli expression system uses exact same construct design principle and can be found in Ponchon and Dardel [2, 11].

2

Materials All reagents, solutions, supplies and lab equipment that contact the target RNA should be RNase-free. 1. Plasmid pUC19. 2. Restrictions enzymes EcoRI and SmaI. 3. PCR product cleanup kit. 4. Gel electrophoresis equipment. 5. RNase-free water. 6. Millipore centrifugation columns. 7. 1% (or any other appropriate percentage for the target construct) agarose gel: Mix 100 mL 0.5 TBE with 0.5 g agarose powder. Microwave for 30 s (or close to boiling) and gently swirl, removing all bubbles and particulates. Allow for slight cooling and add/mix 1.5μL 20 ethidium bromide stock, before pouring the mix into the casting tray and allow to cool to room temperature. 8. LB agar plate with ampicillin (or any other suitable antibiotic for the plasmid): Dissolve 15 g Bacto agar in 1 L LB medium and sterilize by autoclaving. Cool the medium to 50  C. Add 0.5 mL ampicillin (100 mg/mL). Gently mix and pour LB medium into plates under sterile conditions. 9. 1 M MgCl2: Dissolve 0.2 g MgCl2 in 1 mL water. 10. 100 mM NTPs. 11. 1 M Tris–HCl, pH 8.1. 12. 1 M DTT. 13. 1 M spermidine. 14. T7 RNA polymerase. 15. 1% Triton X-100: Mix 1μL 100% Triton X-100 with 99μL water. Store at 20  C.

Using tRNA Scaffold to Assist RNA Crystallization

41

16. 2 RNA loading buffer: 80% (w/v) deionized formamide and 10 mM EDTA (pH 8.0), add 0.2% (wt/vol) xylene cyanol FF and 0.2% (wt/vol) bromophenol blue. Store at 4  C. 17. 10 TBE buffer: Dissolve 108 g Tris, 55 g boric acid, and 7.5 g EDTA. Fill to 1 L with water. 18. 40% acrylamide solution (19:1): Weigh 266 g acrylamide and 14 g N,N0 - methylbisacrylamide in 600 mL of water. Adjust the volume to 700 mL with water and Sterilize the solution by 0.22μm filtration. Store in dark at 4  C. 19. 6% acrylamide–urea gel solution: Dissolve 480.48 g urea in water. Heating may contribute to dissolve the Urea. Then mix with 150 mL 40% acrylamide solution (19:1) and 100 mL 10 TBE. Make up to 1 L with water and filter through a 0.22μm filter. Store at 4  C. 20. Refolding buffer: 10 mM Na cacodylate pH 7 and 50 mM NaCl.

3

Methods

3.1 Design and Synthesize tRNA Scaffold Vector

The success of applying this method to an arbitrary RNA relies on producing the correct structure after fusing with a tRNA. In practice, the design principle includes two major aspects: (a) Retaining the core structure of both the tRNA scaffold and the target RNA; (b) Eliminating any unwanted interactions between the tRNA and target RNA. Since this method aims to target a variety of RNAs, we will provide a general guideline while individual cases can vary depending on the target RNA sequence (see Note 1). 1. The tRNA adopts a stable structure resistant to repeated folding/refolding and nuclease attack [12]. We chose the G. kaustophilus tRNAGly from an extreme thermophile as the platform scaffold to further strengthen these properties. The three dimensional structure of tRNAGly (PDB: 4MGM) demonstrates stability required for efficient expression in recombinant system [13]. 2. Amongst the three stems of the tRNA, TψC and D stem-loops remain unchanged to maintain the cloverleaf threedimensional structure, recognized by cellular factors, such as tRNA processing enzymes [14]. The anticodon stem (acceptor stem), where natural sequence variation occurs, is engineered for inserting the target RNA. This region also allows slight variations, such as base pair identities and helix lengths, to facilitate crystallization if required. However, any deviation from the wild type sequence should be followed by secondary

42

Changrui Lu et al.

Fig. 1 Schematic of tRNA scaffold construct containing plasmid. The tRNA scaffold is shown in blue. The anticodon stem is replaced by desired RNA

structure prediction in RNA fold [15] or mfold server [16] to avoid any alternative fold. (Fig. 1). For example, the stem requires a minimal length to prevent disruption to the tRNA. 3. The fusion point on the target tRNA will also affect its folding. For most RNAs, whose 50 and 30 join together forming a helix, no particular modification is required. In some rare cases, additional steps to find a safe fusion helix are required. Usually, we would use secondary prediction algorithms + conservation analysis to find a stable helix (without pseudoknot) that points outward from the core structure. Several constructs with varies helix lengths and location maybe required as this is a trial and error process. 4. If desired, a polylinker containing different restriction sites can replace the anticodon loop altogether. This allows the vector to accommodate different RNA projects. However, secondary structure checks using popular programs such as RNAfold or Mfold still need to be performed for every construct to avoid alternative conformations. We constructed the vector inside pUC19, preceded by the T7 RNA polymerase promoter [17]. Theoretically, any high copy number vector with an antibiotic marker would suffice. 3.2 tRNA–RNA Chimera Plasmid Construction

In order to avoid misfolding of tRNA–RNA chimera, the construct should satisfy the following two criteria: (1) The target RNA structure should have base paring 50 and 30 ends. (2) The final tRNA chimera should be tested in RNA fold [15] or mfold server [16]. Base identity modifications while retaining base complementarity on the scaffold may be required if misfoldings occur. In general, any RNA terminated by a stem could be inserted into the tRNA scaffold.

Using tRNA Scaffold to Assist RNA Crystallization

43

Fig. 2 Sequence inserted into pUC19 plasmid in this experiment. The rest of the pUC19 sequences are not shown

In our experiments, we constructed the entire tRNA–RNA chimera DNA sequence from scratch using overlapping PCR. Alternatively, one can construct a general purpose tRNA scaffold plasmid with polycloning site and insert various RNA targets (please see above). The DNA fragment corresponding to the target RNA (in this case pRNA domain II) and the tRNA scaffold is flanked by 50 AAAAGAATTCTAATACGACTCACTATAGCG GAAGTAGTTCAGTGGTAGAACAC and 30 CCCGGGTGGAG CGGAAGACGGGACTC . The fragment is amplified by PCR, double digested by EcoRI and SmaI, and ligated into cut pUC19, following standard molecular cloning techniques. The resulting plasmid is sequenced [10]. Figure 2 shows the sequence and secondary structure of the entire insert alongside pUC19 vector. 3.3 tRNA–RNA Sample Preparation

In this experiment, we opt to prepare the RNA samples via large scale PCR followed by in vitro run-off transcription. Similar results can be obtained by linearized plasmid run-off transcription. 1. Set up standard 10 0.1 mL PCR to amplify the required DNA template of tRNA–RNA chimeras (see Note 2). The primers choice generally covers the T7 site all the way to the 30 of the

44

Changrui Lu et al.

Table 1 A standard transcription reaction Component

Volume

100 mM NTPs

500μL of each

1 M Tris–HCl, pH 8.1

300μL

1 M MgCl2

250μL

1 M DTT

100μL

1% Triton X-100

100μL

1 M spermidine

20μL

PCR products

1 mL

T7 RNA polymerase

1 mL

Add water to total volume of

10 mL

acceptor stem. PCR product is then combined and isolated either by gel-extraction, ethanol precipitation or commercial PCR cleanup kit. The process can follow the typical PCR condition which has been described in Subheading 2. Then, the PCR products are purified by PCR cleanup kit (see Note 3). One drawback for this method is the random incorporation of terminal nucleotides, creating 30 overhangs. However, these single-stranded sequenced did not interfere with our downstream experiments. Alternatively, concerned users can attach the entire chimera sequence with a 30 self-cleaving ribozyme, leaving uniform ends (see Note 4). 2. Next, the purified PCR product is added to 10 mL (final) in vitro transcription mix with T7 RNA polymerase. A standard transcription reaction [8] is described in Table 1 (below). If additional reaction efficiency is required, concentrations of T7 RNA polymerase, Mg2+ and PCR products in transcription reaction can be optimized for individual RNA constructs. 3. Next we purify the transcription mix by denaturing gel electrophoresis [18]. Similar results can be obtained by electroelution or denaturing gel-filtration/ion-exchange chromatography [19, 20]. Mix 10 mL sample of in vitro transcribed tRNA– RNA chimeras with 10 mL 2  RNA loading buffer, then incubate at 65  C for 15 min. Separate and purify tRNA– RNA by denaturing polyacrylamide gel containing 8 M urea (50 W constant power with heat spreader over 2–5 h). The typical yield for a 10 mL reaction is 3–5 mg of target RNA (see Note 5). 4. The band corresponding to the target chimera RNA is then excised and crushed using a syringe. RNA is then eluted by swirling with RNase-free water overnight at 4  C (see Note 6).

Using tRNA Scaffold to Assist RNA Crystallization

45

5. The elution is then buffer exchanged and concentrated using a Millipore centrifugation column with appropriate MW cutoff (usually 2 that of the target RNA). This step eliminates urea, loose acrylamide, and other small molecule contaminants. The resulting RNA is ready for refolding or ligand combining. The sample can be flash-frozen and stored at 80  C (see Note 7). 6. we proceeded to refold the RNA in Refolding buffer, by heating up to 65  C for 10 min, followed by rapid cooling to 37  C using a PCR machine and the addition of 5 mM MgCl2 (see Note 8). 7. The pRNA domain we inserted into the scaffold does not crystallize by itself, probably due to structural heterogeneity. However, with the help of the scaffold stabilizing its structure, we obtained crystals in 5 mM cobaltHex, 12 mM spermine, 62 mM potassium chloride, 30 mM barium chloride, 40 mM PH 6 Na cacodylate, 10%MPD (see Note 9). Structure of the pRNA was solved by molecular replacement with the tRNA model and repeated round of building and refining. With the help of the tRNA scaffold, experimental phasing was not required (Fig. 3).

Fig. 3 Crystal structure of the tRNA–pRNA chimera. Pink: pRNA; blue: tRNA scaffold

46

4

Changrui Lu et al.

Notes 1. The success of the adapting this method to any arbitrary RNA relies on correctly predicting the resulting fusion’s secondary structures. Therefore, extreme care and multiple constructs significantly increase the chance of success. We strongly recommend running multiple secondary structure checks for every single construct or mutant. 2. We found that doing smaller fractions (50-100μL) yields best results. 3. Sometimes skipping the PCR cleanup step do not harm the transcription efficiency. However, we still recommend this step, at least during first tries. 4. An equally valid approach involves midi- or mega-prepping the target plasmid and linearizing with restriction enzymes. If the user found this step more suitable or familiar, please feel free to use it. 5. Several alternative methods are typically used here, each with its own advantages and drawbacks. Generally, the purification scheme should suit the target RNA (size or/and quantity) and its finally characterization. Users are recommended to use their own discretion. 6. Should the user have access to an electroeluter, the process can be accelerated. We also advise the users against proloned UV exposure on the target RNAs. 7. We recommend ample volumes of RNase-free water during this step to eliminate free acrylamide. Also, we recommend remixing the solution thoroughly every cycle as RNA tend to stick to the tube sides and the membrane. This often causes aggregation or percipitation. 8. We use a PCR machine for this step. Others have used water bath and ice buckets with equal success. 9. Making the fusion RNA does not guarantee crystallization. However multiple constructs with slight variations in nonconserved regions greatly increases the chance of success, provided that they still fold correctly.

Acknowledgments The work in Lu lab is supported by Shanghai Science and Technology Committee (19ZR1471100), Fundamental Research Funds for the Central Universities (19D210501). The work in the Ke lab is supported by NIH/NIGMS awards P41-GM103403 and S10-RR029205.

Using tRNA Scaffold to Assist RNA Crystallization

47

References 1. Slagter-J€ager JG, Allen GS, Smith D et al (2006) Visualization of a group II intron in the 23S rRNA of a stable ribosome. Proc Natl Acad Sci U S A 103:9838–9843 2. Ponchon L, Dardel F (2007) Recombinant RNA technology: the tRNA scaffold. Nat Methods 4:571–576 3. Ke A, Wolberger C (2003) Insights into binding cooperativity of MATa1/MATalpha2 from the crystal structure of a MATa1 homeodomain-maltose binding protein chimera. Protein Sci 12:306–312 4. Ferre-D’Amare AR, Doudna JA (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 5. Koldobskaya Y, Duguid EM, Shechner DM et al (2011) A portable RNA sequence whose recognition by a synthetic antibody facilitates structural determination. Nat Struct Mol Biol 18:100–106 6. Ferre-D’Amare AR, Zhou K, Doudna JA (1998) A general module for RNA crystallization. J Mol Biol 279:621–631 7. Ferre-D’Amare AR, Doudna JA (2001) Methods to crystallize RNA. Curr Protoc Nucleic Acid Chem; Chapter 7:Unit 7.6 8. Ke A, Doudna JA (2004) Crystallization of RNA and RNA-protein complexes. Methods 34:408–414 9. Zhang J, Ferre-D’Amare AR (2014) New molecular engineering approaches for crystallographic studies of large RNAs. Curr Opin Struct Biol 26:9–15 10. Cai R, Price IR, Ding F et al (2019) ATP/ADP modulates gp16-pRNA conformational change in the Phi29 DNA packaging motor. Nucleic Acids Res 47:9818–9828

11. Ponchon L, Beauvais G, Nonin-Lecomte S, Dardel F (2009) A generic protocol for the expression and purification of recombinant RNA in Escherichia coli using a tRNA scaffold. Nat Protoc 4:947–959 12. Engelke DR, Hopper AK (2006) Modified view of tRNA: stability amid sequence diversity. Mol Cell 21:144–145 13. Grigg JC, Ke A (2013) Structural determinants for geometry and information decoding of tRNA by T box leader RNA. Structure 21:2025–2032 14. Keiler KC, Waller PR, Sauer RT (1996) Role of a peptide tagging system in degradation of proteins synthesized from damaged messenger RNA. Science (New York, NY) 271:990–993 15. Gruber AR, Lorenz R, Bernhart SH et al (2008) The Vienna RNA websuite. Nucleic Acids Res 36:W70–W74 16. Zuker M (2003) Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 31:3406–3415 17. Grigg JC, Chen Y, Grundy FJ et al (2013) T box RNA decodes both the information content and geometry of tRNA to affect gene expression. Proc Natl Acad Sci U S A 110:7240–7245 18. Rio DC, Ares M Jr, Hannon GJ et al (2010) Polyacrylamide gel electrophoresis of RNA. Cold Spring Harbor Protocols 2010:pdb. prot5444 19. Easton LE, Shibata Y, Lukavsky PJ (2010) Rapid, nondenaturing RNA purification using weak anion-exchange fast performance liquid chromatography. RNA 16:647–653 20. Baronti L, Karlsson H, Marusˇicˇ M, Petzold K (2018) A guide to large-scale RNA sample preparation. Anal Bioanal Chem 410:3239–3252

Chapter 5 RNA Modeling with the Computational Energy Landscape Framework Konstantin Ro¨der and Samuela Pasquali Abstract The recent advances in computational abilities, such as the enormous speed-ups provided by GPU computing, allow for large scale computational studies of RNA molecules at an atomic level of detail. As RNA molecules are known to adopt multiple conformations with comparable energies, but different two-dimensional structures, all-atom models are necessary to better describe the structural ensembles for RNA molecules. This point is important because different conformations can exhibit different functions, and their regulation or mis-regulation is linked to a number of diseases. Problematically, the energy barriers between different conformational ensembles are high, resulting in long time scales for interensemble transitions. The computational potential energy landscape framework was designed to overcome this problem of broken ergodicity by use of geometry optimization. Here, we describe the algorithms used in the energy landscape explorations with the OPTIM and PATHSAMPLE programs, and how they are used in biomolecular simulations. We present a recent case study of the 50 -hairpin of RNA 7SK to illustrate how the method can be applied to interpret experimental results, and to obtain a detailed description of molecular properties. Key words Path sampling, Energy landscape, Alternative RNA structures

1

Introduction Over the last two decades, as multiple functions of RNA molecules have been unveiled, and more structures resolved at high resolution, it has become clear that single stranded RNA can assume multiple conformations. These conformations differ in their secondary structures as well as in their tertiary contacts, resulting in a range of overall architectures. Probably the most well-known examples of this behavior are riboswitches that can adopt alternative conformations in response to an external stimulus, such as the presence of a ligand [1]; while other molecules are known to adopt alternative structures in the same environment. For example, the 50 -hairpin of the RNA component of the ribonucleic

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_5, © Springer Science+Business Media, LLC, part of Springer Nature 2021

49

50

Konstantin Ro¨der and Samuela Pasquali

V

(a)

(b)

(c)

Fig. 1 Schematic illustration of the different topographies observed for potential energy landscapes in biomolecules. (a) A single funnel energy landscape often seen as the “standard” model for protein folding. (b) A multifunnel energy landscape exhibiting more than one accessible funnel, with a clearly identifiable global minimum—this topology is associated with multiple functions [14]. (c) This type of energy landscape is associated with intrinsically disordered proteins, exhibiting a highly frustrated system with a number of low-energy configurations separated by high-energy barriers

complex 7SK crystallized in two distinct conformations in several crystals and the two architectures have similar stability [2, 3]. Indeed, several studies, both experimental and computational, have highlighted the “rough” energy landscape of ssRNA molecules [4–8], consisting of several low-lying funnels. While proteins may exhibit an energy landscape with one main native basin (the folding funnel) [9, 10], the energy landscapes for some proteins [11–13] and for single-stranded RNA are characterized by several low-energy groups of minima of similar free energies separated by high barriers—the so-called multifunnel energy landscapes [14] (Fig. 1). Low-lying local minima correspond to metastable states, which can be very long-lived due to the high energy barriers surrounding them, which act as kinetic traps. These extremely long time scales of molecular motion associated with high energy barriers for refolding area common theme in biomolecular science. For example, G-quadruplexes, formed by both ssDNA and RNA, clearly exhibit this feature, with several alternative structures determined experimentally by NMR and crystallography, which can be stable for hours [15, 16]. These systems highlight the very complex energy landscapes possible even for molecules with a short sequence of less than 30 nucleotides [17, 18], which results from the high frustration [19] when both canonical and noncanonical base pairings can be formed, and therefore multiple possible secondary structures compete against each other. According to the structure–function paradigm, understanding the structure adopted by a molecule is essential to understand its function, and to conceive possible drugs for pharmaceutical RNA targets. The binding affinity for a ligand may change significantly based on the molecular conformations adopted by RNA, which can be observed for a protein partner, as seen in several binding essays for 7SK [3, 20],

RNA Modeling with the Computational Energy Landscape Framework

51

or for small-molecule drugs, which ideally target a unique, stable structure. We therefore need to understand the characteristics of different possible molecular structures, the context in which one structure might become dominant over others (for example based on temperature, pH or ligands), and the possible interconversions between alternative structures with respect to their relative stability. Common structure prediction methods are based on knowledge of the sequence and secondary structure prediction algorithms [21– 23], but are not necessarily accounting for the multifunnel nature of the energy landscape and the resulting dynamics of the molecule. Therefore, these tools provide only a snapshot view under some ideal “standard conditions” but are limited in giving the comprehensive picture needed to understand experiments. Physics-based models, on the other hand, have the ability to follow the dynamical behavior of the system and to properly include the environmental conditions specific to each experiment (temperature, pH, salt concentration, . . .). These models can help to interpret experimental data that, because of the plasticity of the molecule, may yield apparently contradicting information, and can provide a coherent view of the molecular structure, which is compatible with all observations. Hence, it is possible to formulate a working hypothesis of the relative stability of different configurations and their interconversion, which are only indirectly captured by experiments. As the energy landscape contains all the information needed to describe the thermodynamic, kinetic and structural properties of molecular systems, its exploration is the basis for a good computational description of a molecular system [24]. In principle, from molecular dynamics (MD) simulations it is possible to extract the landscape, but in practice, for the systems of the size of a full RNA (even a small one with a few dozen bases), this task is not feasible, as the simulations do not converge on any reasonable time scale, even with enhanced sampling methods and full parallelization. This issue mainly arises from the fact that transitions between different structural ensembles are rare events due to the high energy barriers that need to be overcome [25], and hence require a long simulation time. Instead, approaches that can sample independently of the actual time scales involved, or that target the rare transitions directly, are more effective. A useful description of the potential energy landscape is provided by the stable configurations a molecule can adopt (local minima), and the minimum energy paths connecting them (characterized by transition states1). Coarsegraining into a set of local minima and connecting transition states preserves the information contained in the landscape, and this set forms a graph representing the landscape faithfully. The transitions

1

Formally described as a Hessian-index 1 saddle point, that is, there is one unique imaginary normal mode frequency.)

52

Konstantin Ro¨der and Samuela Pasquali

from one configuration to another are described as a series of local minima with connecting transition states. Such a chain of local minima and intervening transition states is a discrete path, and the aim of the computational energy landscape framework [26] is to explore the energy landscape through sampling these discrete paths, hence the name Discrete Path Sampling (DPS) [27, 28]. In this contribution we discuss the set-up, algorithms and an example for the application of the computational potential energy landscape framework, with DPS as the key component, as implemented in the OPTIM and PATHSAMPLE software (see Note 1).

2

Algorithms The computational potential energy landscape framework consists of three key components. Firstly, suitable starting points for the exploration are required. These configurations may be obtained from experiment, such as X-ray, cryo-EM, or NMR structures (see Note 2). The low energy local minima are then added into a new database of stationary points and the sampling commences in PATHSAMPLE, with the transition state searches and connectivity determined by OPTIM. After sampling is completed, the postprocessing, including clustering of structures into ensembles, free energy calculations, and transition rate calculations, can be conducted, with most of these calculations available within PATHSAMPLE . In the following account, we will discuss the algorithms used in the second step, namely the search strategies and transition state location, and conclude with some discussion of the available energy models within the framework.

2.1 Selecting Pairs of Minima for Transition state Searches

The selection of local minima to connect is a crucial part of the energy landscape exploration. Whether a database is initiated from a small number of experimental structures or from the results of another simulation (see Note 2), we need to obtain an initial discrete path between them, before further sampling (see Subheading 2.3) is initiated. The algorithm used to identify the best pairs of minima to connect at this stage in PATHSAMPLE is also applied in OPTIM to run multiple, consecutive transition state searches between each pair of selected minima (see Note 3). The exact procedure is outlined below, and illustrated in Fig. 2. The initial path between two structures of interest, A and B, is found using a customized Dijkstra shortest path search [29]: 1. For each pair of local minima in the database, we define a weight for the transition between them. (a) This weight is set to zero, if the minima are connected by a transition state. (b) Otherwise, to the value a function (often exponential) of the distance between A and B.

RNA Modeling with the Computational Energy Landscape Framework d1

(a) A

53

B

E TS optimisation

(b) A

d2

T1

i

d3 j

E

E

TS optimisation

TS optimisation

(c)

B

A i

T1

j

B

Fig. 2 Left: An illustration of the search for a discrete path between two minima A and B, as described in Subheading 2.1, and the associated transition state searches with the DNEB algorithm, as detailed in Subheading 2.2. (a) The two minima A and B are separated by a distance d1, and there are no other minima or transition states in the database at this stage. A DNEB interpolation is conducted, with the optimized band shown, where the black dots are the discrete images of the interpolation. The highlighted image is a maximum, and therefore a transition state candidate. After further optimization, a transition state, T1 is located (green circle), which connects two new minima, i and j (blue circles). (b) A new analysis of the database finds that the shortest path contains two missing connections, namely between A and i, and j and B, with respective separations of d2 and d3. Subsequently, we attempt two DNEB interpolations, one for each gap in the discrete path. The first interpolation detects two transition state candidates, and the second one locates one. (c) After these three candidates are converged to transitions states, we have obtained a fully connected discrete path. Right: An illustration of the change to the interpolation bands as the DNEB algorithm progresses. The two minima to be connected are the blue regions of the surface (low potential energy). The linear interpolation as starting point for the band (small dashes) is shown, alongside two more optimized bands (dashed and solid line), where the solid line represents the final band, which is analyzed for transition state candidates. (Adopted from [30])

2. If a pair has been used in a search before, the weight is increased by a constant factor. This weighting scheme favors smaller gaps, and previously untried connections. 3. The discrete path between A and B with the smallest sum of weights is then deter-mined. This path contains pairs of minima that are unconnected, but have a low weight, unless we have already located a fully connected path between A and B. 4. For each of these unconnected pairs of adjacent minima on this path, a pathway search is initiated, based on locating transition states. 2.2 Transition State Searches

The characterisation of transition states within the computational potential energy framework is based on geometry optimization to find transition states. Details have been reviewed before, for example in [26]. Transition state candidates are located using a doubledended search—the doubly-nudged elastic band (DNEB) algorithm [31–33]. The algorithm relies on three steps:

54

Konstantin Ro¨der and Samuela Pasquali

1. An initial interpolation is created, using a set of discrete images (three-dimensional structures in a 3N-dimensional space for N atoms in the molecule) between the two endpoints (see Note 4). An artificial spring potential is added between every adjacent pair of images. The spring constants are set by the simulation input, and may be adjusted during the optimization. 2. The overall energy of the band is minimized with respect to selected components of the force acting on it to obtain a lower energy interpolation. The minimization is performed with respect to all interpolation images (i.e., 3 NP coordinates for P images for N atoms). 3. The local energy maxima in the resulting band are candidates for transition states, and they are subsequently refined using hybrid eigenvector-following [34, 35]. Subsequently, for every transition state the two connected local minima are found using energy minimization. As a result, a number of minima– transition state–minima triplets are located, and these are added to a database. Identical minima are identified using (a) similarity in energy and (b) the geometric distance after permutational alignment (see Note 5) [36]. The entire process is illustrated in Fig. 2. 2.3 Search Strategies

All search strategies to explore the potential energy landscape are chosen to provide a representative set of local minima and transition states, such that calculated observables are converged (i.e., unchanged with more sampling). To achieve this aim, a number of different strategies are available for further sampling. All of them select pairs of local minima for new transition state searches. 1. If particular areas of the energy landscape are under sampled, these regions can be targeted with enhanced local sampling [37]. Here, we select local minima close in energy or distance to the specified region, and connect them to one minimum in the region of interest. 2. Occasionally, the sampling can identify connections between low-energy minima and the rest of the landscape, but only via high-energy barriers, due to incomplete sampling. These local minima would act as artificial kinetic traps, and untrapping [38] removes these traps through further exploration. 3. Additionally, we may observe high energy barriers between different parts of the energy landscape, and the SHORTCUT BARRIER Scheme [38] aims to find alternative paths going around these high barriers. 4. The paths may be lengthened by irrelevant additional motions of side chains. Short-cutting in this context will cut the path length by removing these changes [38, 39]. Detailed descriptions of all these algorithms can be found elsewhere, and we

RNA Modeling with the Computational Energy Landscape Framework

55

note that the best combination of approaches depends on the system, and, to some extent, on the scientific question of interest. Once the sampling of the energy landscape is converged sufficiently, a number of analysis steps are possible. For example, PATHSAMPLE provides direct access to the calculation of heat capacities, (harmonic) free energies, and rate constants; in Subheading 3.3 more details of the possibilities are provided. 2.4 Visualization of the Energy Landscape

Once the minima and the transition states have been identified, an important step is the visual inspection of the energy landscape to see how the sampled states connect one another. This process allows us to identify energy basins. A major challenge in visualizing energy landscapes is their high dimensionality. The large number of degrees of freedom make a faithful representation difficult, and low dimensional projections are sometimes employed. Problematically, these projections will represent only a subset of the information gathered about the system, and projection errors may be introduced into the data. The representation using disconnectivity graphs [40, 41] gives a true presentation of the connectivity of the energy landscape. The graphs are tree graphs, where the vertical axis is the energy and the horizontal axis is arbitrary, but can be chosen to represent any property. A vertical line is associated with every local minimum, and lines merge when there is a discrete path between two minima at a given energy E (i.e., no part of the discrete path is higher in energy than E). This process leads to the hierarchy of the energy landscape being represented clearly, and funnels can be identified readily. The software package contains DisconnectionDPS, a tool to plot disconnectivity graphs. Plotting a disconnectivity graph throughout the sampling process can help in identifying under sampled regions and artifacts, and guide subsequent sampling strategies as described above (see Note 6). We illustrate a number of characteristic landscapes in Fig. 3, which highlight situations that commonly occur.

2.5 Available Potential Energy Models

A number of different potential energy models are available within the computational energy framework, in particular, for the simulation of RNA (in isolation or with other biomolecules) the software currently contains interfaces to AMBER (with and without GPU acceleration)2 and HiRE-RNA [42], as well as interfaces, for example, for GROMACS [43].

2.6 Analysis of Structural Ensembles

Once the energy landscape has been determined and visualized with disconnectivity graphs, one needs to analyze the structural content of the various basins that have been identified. The first

2

The use of AMBER requires an AMBER license.

56

Konstantin Ro¨der and Samuela Pasquali

(b)

C

B (a)

A

Fig. 3 An example disconnectivity graph illustrating features that suggest alternative search strategies. The graph was created as a synthetic example to highlight these pointers, and represents a typical disconnectivity graph after an initial path has been found between a minimum in funnel A and a minimum in funnel C. The entire landscape is only sampled sparsely, so more sampling is required in any case, though a number of features deserve special attention. Firstly, the disconnectivity graph exhibits three funnels, A, B, and C. However, due to the proximity, A and B may not be separate funnels, and local enhanced sampling can help to resolve this question. The highlighted region (a) exhibits a low-energy minimum with a high energy barrier to A. This feature is likely to be an artificial kinetic trap, and untrapping will remove such artefacts. Between A and C we encounter a single high energy barrier (b). While some systems exhibit such barriers, often they are associated with poor sampling, and shortcutting this barrier will lead to a more realistic pathway between the funnels

step consists of isolating the structures belonging to each basin, creating subsets that can be analyzed independently. This analysis is performed with the aid of disconnectivity graphs that allow us to set an energy cutoff to decide the boundaries of each basin. We then deploy the standard tools to analyze the structures of each basin. We generate a pseudotrajectory with all the structures in the set so that we can then use GROMACS [43] to perform most of the analysis. Typically we study the root mean square deviation (RMSD) of the set with respect to the structure of the minimum of the basin and the root mean square fluctuation for each

RNA Modeling with the Computational Energy Landscape Framework

57

nucleotide (RMSF), we look for different clusters, using a stringent cutoff since the structures are already very similar given they belong to the same basin, and we look at more specific features that are system dependent, such as the presence or absence of a particular base-pair, triplet, or bulge. This last part is achieved by writing analysis script dependent on the question investigated. A good tool for this purpose is python and the MDtraj package [44], which is designed to perform the analysis of MD trajectories and where one can easily define the properties (distances, angles, dihedrals) for which we want to assess the variability in the sample. Examples of the use of MDtraj are available in the online tutorial (http://mdtraj.org/1.9.3/examples/index.html). The last step, which is crucial for RNA molecules, but usually not integrated in the tools that have been mainly developed for proteins, is the analysis of the secondary structures characteristic of each minimum and each cluster. For this analysis one needs to use dedicated software, for example RNApdbee [45], which takes a PDB file as input and gives the 2D structure as output in various forms (e.g., dot brackets).

3

Method We describe here the typical flow for the exploration of the energy landscape of an RNA, using as an example a recently published case study on the 50 -hairpin of RNA 7SK [20], simulated using the AMBER ff99 [46] force field with the Barcelona α/γ backbone modification [47] and the χ modification tuned for RNA [48, 49]. We will focus on the native sequence, and all mutant sequences presented in the case study can be investigated in the same way. Example input files for all stages of the simulations, including additional commentary, are provided in the online resources for this article.

3.1

Starting Points

For the simulation, four experimental structures are available, three of which we used in the original study (PDB IDs: 5LYU [3] and 5IEM [50]). Step 1: Create the input files necessary for the potential model. This step requires us to fix the length of the fragment studied and, in some cases, the protonation states of the molecule under investigation (see Note 7). Here, we decided to study the upper end of the hairpin, nucleotides 37 to 70, rather than the full size of the reported crystal structures of 58 nucleotides. Generally, while always aiming to simulate the full system, if the degrees of freedom can be reduced, either by rigidification [51] of parts of the structure or considering only a certain part of it, it will significantly reduce computational costs. Of course, the simplification needs to be taken into account when analyzing the result, and requires validation (see the discussions in [20] for the justification for this system).

58

Konstantin Ro¨der and Samuela Pasquali

Step 2: Obtain the input files for the actual simulations. The first set of input files is the input for the AMBER software package, namely a coordinates file in the rst format (named coords. inpcrd), a matching topology file (coords.prmtop, see Note 8), and a settings file (min.in). In the latter file, the selected mode is imin ¼ 1 with only one cycle of minimization (i.e., we only want to calculate the energy and force in AMBER), and the implicit solvent model, interaction cutoffs and ion concentration should be set appropriately (see Note 9). The second set of input files are specific to OPTIM, in particular the list of allowed permutations (perm.allow, see Note 10) and the parameter input (odata). Step 3: Minimize the endpoints to yield local minima with OPTIM. This step is necessary to obtain starting points for doubleended searches. Certain parts of the simulation are fixed from now, such as the convergence RMS force on the gradient, and whether vibrational frequencies are calculated and stored throughout (see Note 11). The minimizations produce a min.data.info output file, which contains all the information necessary about the local minima, and can be used to initiate a database in PATHSAMPLE . In the case of 7SK, this database only contains three minima and no transition states, and will serve as the starting point for landscape exploration. 3.2 Exploration of the Energy Landscape

At this stage, we aim to find a complete discrete path between the sets of local minima. Firstly, this connection allows us to describe the transition between the different configurations, although the initial path may be far from the actual, fastest transition. However, as a second advantage, once a discrete path is found, the amount of data we need to store and calculate, namely regarding the distance between different local minima, is drastically reduced, and subsequent sampling can proceed from the starting points that are already known. Step 4: Connect the starting configurations with PATHSAMPLE . For this system, as we have three starting configurations, there are two options to connect them. We could create three copies of the initial database, and in each of them sample to find a different discrete path (1 to 2, 1 to 3, and 2 to 3). Alternatively, we can use one database and connect them in subsequent searches. The first case requires a larger amount of available resources, but in return may give more stationary points, which can then be combined into a single database for all further sampling. While the second option produces less sampling at this stage, more sampling at later stages can compensate for this difference, and the smaller resource requirements made it our preferred option (see Note 12).

RNA Modeling with the Computational Energy Landscape Framework

59

Step 5: Extend the sampling with PATHSAMPLE. Once the initial paths are located, additional sampling is applied to converge the exploration. Enhanced local sampling can also be useful in the context of improving the sampling results from other strategies. For larger systems, the schemes described will usually identify a large number of new stationary points, but these might not be connected to the main set of stationary points yet. Local sampling after untrapping and shortcutting generally improves the connectivity, allowing the database to converge faster. Step 6: Test for convergence. As the sampling is strongly dependent on the system of interest and the scientific question(s) posed, there is no universal guide to convergence. However, there are a number of indicators that allow for some judgement of how well the energy landscape has been sampled. First, there are simple properties of the landscape: what is the ratio of minima to transition states and how large is the largest connected set? If we have many more minima than transition states and a small main set, then the sampling is probably insufficient. Similarly, a large number of low energy minima that are not connected to the main set and clearly under sampled regions in the disconnectivity graph are other indicators. Furthermore, convergence implies that the shape of the landscape, the number of funnels and their relative energies are unaffected by further sampling. Finally, the best judgement of convergence is comparison with experiment, that is, from properties like heat capacities, free energy differences, or rate constants. We can compare experimental and calculated values, and test whether the computed values are unchanged upon further sampling. Step 7: Extract relevant data from the potential energy landscape. Once we are satisfied the sampling is sufficient, the databases act as collections of stationary points for the particular systems, and we can extract structural information either straight from them, or export collections of minima (for example every minimum in a specific funnel, or in a specific transition) into MD trajectory files, allowing for postprocessing with widely used analysis tools. 3.3

Thermodynamics

Step 8: Compute free energies, heat capacities, and reaction rates in PATHSAMPLE. 1. The free energy is calculated for each stationary point individually, applying a harmonic approximation, and then combined using the superposition approach [52]. In addition, stationary points can be regrouped given an energy threshold [53]. This procedure allows us to smooth out smaller motions of side chains, and isolate the key steps along a rearrangement pathway. Importantly, the combination of these features is capable of identifying changes in the folding mechanism as a result of

60

Konstantin Ro¨der and Samuela Pasquali

raising or lowering the temperature. In the case of the 50 -hairpin in RNA 7SK, we identify three states, in agreement with experiment, and the regrouping allows us to compare the relative stability of the three structures and their occupancy. 2. It is also possible to calculate the heat capacity curve for a molecule. The observed features in the curve are directly related to changes in the occupation probability of different funnels [54], and the direct comparison between different mutants can be used to identify likely transitions between different funnels at biologically relevant temperatures [13]. 3. Finally, reaction rates can be calculated using the New Graph Transformation (NGT) approach [53, 55], a method that is computationally efficient and allows the computation of rate constants based on a master equation. These calculations require the definition of the “reactant” and “product” regions, but for clearly defined funnels, we see in general that one minimum in a funnel defines a region. The rate constants are particularly useful to track changes between mutated sequences, or between different fold mechanisms [56]. 3.4 Example Application

Figure 4 illustrates the results of these steps for the 50 -hairpin of RNA 7SK, together with the data that can be extracted from the energy landscape, which can be directly compared with experimental results. From the analysis of the potential energy landscape obtained starting our exploration from the four existing experimental structures, we were able to identify three main structural families, corresponding to three energy basins (Steps 1 to 7 applied to the wild type RNA structures). Through Step 8 we computed the free energy landscape and derived the thermal accessibility of the three families. Using basin-hopping global optimization [57– 59] for a peptide from the binding protein HEXIM docked to the RNA, followed by a fully atomistic MD run to achieve full relaxation, we identified possible stable RNA-protein complexes, for which we analyzed the details of the interactions. Mutating the sequence and generating new starting structures (Step 1 to 3) we explored the potential energy landscapes for several mutants, allowing comparison of the topographic variation of the mutant energy landscapes to the native sequence. This part of the study allowed us to correlate the effect of the mutations with the experimentally observed binding affinities for the HEXIM peptide, and to provide an explanation for the different affinities from the structural details of each mutant.

RNA Modeling with the Computational Energy Landscape Framework

61

ldi n

gm

ec h

anis

e n e r gy

Fre e

y N M R a n d X- r a

rma l

m

The

fro

WT

Fo

s accessi bility of ensemble

tant landscapes Mu erimental str exp uc t t en

n know es ur

HEXIM binding

s ure uct C o n x str ta ct m aps - Bound comple

Four d iffe r

embles - Alterations i n bi al ens r u t ndin truc s gp o t rop s e ert ng a ies Ch

e energy landscap

ral c tu ms u r t -S Landsca pe topography

es rti e p pr o

Fig. 4 Summary of the results for a study of the 50 -hairpin of RNA 7SK published in [20]. Based on four different experimentally described structures from NMR and X-ray studies [3, 50], an exploration of the energy landscape yields a description of the folding mechanisms between the different structures and properties of the structural ensembles they belong to. In addition with free energy calculations, it is possible to quantify occupation probabilities. From these descriptions, binding properties and the structures of bound complexes with protein targets can be described. Finally, repeating the exploration process for a number of sequence mutations allows us to interpret binding assays

4

Notes 1. The two programs are available under GPL license online: (a) OPTIM at http://www-wales.ch.cam.ac.uk/OPTIM/ (b) PATHSAMPLE at http://www-wales.ch.cam.ac.uk/PAT HSAMPLE /. Individual components may require

62

Konstantin Ro¨der and Samuela Pasquali

separate licenses (e.g., the AMBER interface requires an AMBER license). The websites also contains detailed descriptions of available keywords, various examples and some additional documentation. Furthermore, a wiki page is available here: https://wikis.ch.cam.ac.uk/rowalesdocs/wiki/index.php/MainPage. Furthermore, input files and additional commentary for the example in this publication is available (https://doi.org/10.5281/ zenodo.3757905). 2. For high-quality structures (no missing atoms, good resolution) experimental data can be used directly. In the absence of any data from experiment or with low-resolution data, low energy structures may be determined using the basin-hopping global optimization algorithm and GMIN [57–59], which is part of the software package. 3. The precise strategy for spreading the work between OPTIM and PATHSAMPLE depends on the available resources, as much as on the system studied. PATHSAMPLE is designed as a driver for OPTIM and the separate transition state searches it starts in OPTIM are truly parallel. As a driver, it is therefore best placed to decide on the pairs of minima to be connected, yet if for each pair there is only one DNEB search for transition states it is difficult to locate complete discrete paths. We generally run a maximum of between 5 and 20 DNEB cycles in OPTIM for each pair suggested from PATHSAMPLE, though in certain situations fewer, longer searches employing quasicontinuous interpolations (QCI, see Note 4) might be ideal to lower the distance between endpoints, which significantly increases the interpolation quality. 4. The creation of a good interpolation has been subject of much research, and we provide here only the most basic description. In many cases the DNEB algorithm presented here is sufficient, though a more recent addition, the quasi-continuous interpolation (QCI) is part of OPTIM, and should be considered the first choice interpolation for the future. 5. The alignment of structures is complicated by the allowed permutations between identical atoms (for example the hydrogen atoms in a methyl group, or the oxygens in the phosphate groups of nucleic acids). As a result, instead of a simple rotation, the alignment of structures needs to test for permutational isomers [60], which cannot be done deterministically. To reduce the computational costs it is assumed that minima with an energy difference larger than a certain threshold (for RNA we generally employ 104 kcal/mol with an convergence threshold for the L-BFGS minimization at least two orders of magnitude larger) are not identical. The structures are then

RNA Modeling with the Computational Energy Landscape Framework

63

aligned, testing for different permutations of identical atoms, and if the structures closely align (RMSD below for example 0.3 Å), they are considered identical. This threshold can be altered as well, depending on the requirements of the simulation. It should be noted that alignment takes place in OPTIM and PATHSAMPLE , and so consistency between the parameters is vital. 6. Disconnectivity graphs are a universal tool in visualizing the energy landscapes, and are often used in combination with coloring schemes based on order parameters to identify key properties of different ensembles and of the interensemble transitions. They can also be used to extract information from the database, such as the minima in a specific funnel, which can then be extracted and analyzed in the same way as an MD trajectory (see Subheading 2.6). 7. For this part of the work we prefer to use files in the pdb format, as it is easy to visualize the structures with standard software, and to manipulate the files, even with standard text editors. Any software package can be used in this step, as long as output files containing the sequence and three-dimensional structural information can be produced and are accepted by LEaP, or if the AMBER topology and coordinate files can be written directly. 8. The potential energy function must be symmetric, that is, the exchange of equivalent atoms must not change the energy of a configuration. Not all force fields are setup correctly with respect to this requirement, and additional steps are required. For AMBER, a python script is provided in the software package, which edits the topology file accordingly; for nucleic acids this editing only requires minor changes. 9. While there is a difference between implicit and explicit solvent, the implicit solvent models are, in our experience, sufficient, especially as the computational costs are much higher for explicit solvent. Instead of using explicit solvation and ions directly, we generally use implicit solvent, and afterward run MD simulations using the important structures located with DPS to confirm their stability in explicit solvent. 10. As explained in the Algorithms section, the structures have to be permutationally aligned. The perm.allow file contains a list of identical atoms to enable this procedure, and it is created from a pdb file using a python script provided in the software package. 11. Vibrational frequencies are needed to calculate rate constants and heat capacities, and therefore it is generally desirable to obtain them. However, for large systems, this process is expensive, and an alternative solution is to calculate them as part of

64

Konstantin Ro¨der and Samuela Pasquali

the post-processing. This option lowers the cost, as frequencies are only calculated once per stationary point, rather than for every encounter. 12. We note that the initial path search can be a time-consuming part of the sampling, as there are not many options to speed it up, other than by using more computing power, for example GPU acceleration. References 1. Quarta G, Sin K, Schlick T (2012) Dynamic energy landscapes of riboswitches help interpretconformational rearrangements and function. PLoS Comput Biol 8:e1002368 2. Martinez-Zapien D, Saliou JM, Han X et al (2015) Intermolecular recognition of the non-coding RNA 7SK and HEXIM protein in perspective. Biochimie 117:63–71 3. Martinez-Zapien D, Legrand P, McEwen AG et al (2017) The crystal structure of the 50 functional domain of the transcription riboregulator 7SK. Nucleic Acids Res 45 (6):3568–3579 4. Pan J, Woodson SA (1998) Folding intermediates of a self-splicing RNA: mispairing of the catalytic core. J Mol Biol 280:597–609 5. Chen SJ, Dill K (2000) RNA folding energy landscapes. Proc Natl Acad Sci U S A 97:646–651 6. Li PTX, Vieregg J, Tinoco I (2008) How RNA unfolds and refolds. Annu Rev Biochem 77:77–100 7. Solomatin SV, Greenfeld M, Chu S et al (2010) Multiple native states reveal persistent ruggedness of an RNA folding landscape. Nature 463:681–684 8. Schlatterer JC, Martin JS, Laederach A et al (2014) Mapping the kinetic barriers of a large RNA molecule’s folding landscape. PLoS One 9:e85041 9. Leopold PE, Montal M, Onuchic JN (1992) Protein folding funnels: a kinetic approach to the sequence-structure relationship. Proc Natl Acad Sci U S A 89(18):8721–8725 10. Bryngelson JD, Onuchic JN, Socci ND et al (1995) Funnels, pathways, and the energy landscape of protein folding: a synthesis. Proteins 21(3):167–195 11. Ferreiro DU, Hegler JA, Komives EA et al (2011) On the role of frustration in the energy landscapes of allosteric proteins. Proc Natl Acad Sci U S A 108(9):3499–3503 12. Kouza M, Hansmann UHE (2012) Folding simulations of the a and B domains of protein G. J Phys Chem B 116(23):6645–6653

13. Ro¨der K, Wales DJ (2017) Transforming the energy landscape of a coiled-coil peptide via point mutations. J Chem Theory Comput 13 (3):1468–1477 14. Ro¨der K, Wales DJ (2018) Evolved minimal frustration in multifunctional biomolecules. J Phys Chem B 122:10989–10995 15. Burge S, Parkinson G, Neidle S (2006) Quadruplex DNA: sequence, topology and structure. Nucleic Acids Res 34:5402–5415 16. Zhang AYQ, Balasubramanian S (2012) The kinetics and folding pathways of intramolecular G-quadruplex nucleic acids. J Am Chem Soc 134(46):19297–19308 17. Stadlbauer P, Mazzanti L, Cragnolini T et al (2016) Folding of human telomeric G-quadruplexes studied by coarse-grained and all atom simulations. J Chem Theory Comput 12:6077–6097 18. Cragnolini T, Chakraborty D, Sponer J et al (2017) Multifunctional energy landscape for a DNA G-quadruplex: an evolved molecular switch. J Chem Phys 147(15):152715 19. Bryngelson JD, Wolynes PG (1987) Spin glasses and the statistical mechanics of proteinfolding. Proc Natl Acad Sci U S A 84 (21):7524–7528 20. Ro¨der K, Stirnemann G, Dock-Bregeon AC et al (2020) Structural transitions in the RNA 7SK 50 hairpin and their effect on HEXIM binding. Nucleic Acids Res 48:373–389 21. Das R, Karanicolas J, Baker D (2010) Atomic accuracy in predicting and designing non canonical RNA structure. Nat Methods 7:291–294 22. Xu X, Zhao P, Chen SJ (2014) Vfold: a web server for RNA structure and folding thermodynamics prediction. PLoS One 9(9):107504 23. Popenda M, Szachniuk M, Antczak M et al (2012) Automated 3D structure composition for large RNAs. Nucleic Acids Res 40:e112 24. Wales DJ (2003) Energy landscapes. Cambridge University Press, Cambridge 25. Wales DJ, Salamon P (2014) Observation time scale, free-energy landscapes, and molecular

RNA Modeling with the Computational Energy Landscape Framework symmetry. Proc Natl Acad Sci U S A 111 (2):617–622 26. Joseph JA, Ro¨der K, Chakraborty D et al (2017) Exploring biomolecular energy landscapes. Chem Commun 53:6974–6988 27. Wales DJ (2002) Discrete path sampling. Mol Phys 100(20):3285–3305 28. Wales DJ (2004) Some further applications of discrete path sampling to cluster isomerization. Mol Phys 102(9–10):891–908 29. Carr JM, Trygubenko SA, Wales DJ (2005) Finding pathways between distant local minima. J Chem Phys 122(23):234903 30. Ro¨der K (2018) Energy landscaping - On the relationship between functionality and sequence mutations for multifunctional biomolecules PhD thesis University of Cambridge 31. Henkelman G, Uberuaga B, Jo´nsson H (2000) A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J Chem Phys 113(22):9901–9904 32. Henkelman G, Jo´nsson H (2000) Improved tangent estimate in the nudged elastic band method for finding minimum energy paths and saddle points. J Chem Phys 113 (22):9978–9985 33. Trygubenko SA, Wales DJ (2004) A doubly nudged elastic band method for finding transition states. J Chem Phys 120(5):2082–2094 34. Munro LJ, Wales DJ (1999) Defect migration in crystalline silicon. Phys Rev B 59 (6):3969–3980 35. Henkelman G, Jo´nsson H (1999) A dimer method for finding saddle points on high dimensional potential surfaces using only first derivatives. J Chem Phys 111(15):7010–7022 36. Griffiths M, Niblett SP, Wales DJ (2017) Optimal alignment of structures of finite and periodic systems. J Chem Theory Comput 13 (10):4914–4931 37. Ro¨der K, Wales DJ (2018) Energy landscapes for the aggregation of Aβ1742. J Am ChemSoc 140(11):4018–4027 38. Strodel B, Whittleston CS, Wales DJ (2007) Thermodynamics and kinetics of aggregation for the GNNQQNY peptide. J Am Chem Soc 129(51):16005–16014 39. Carr JM, Wales DJ (2005) Global optimization and folding pathways of selected alpha-helical proteins. J Chem Phys 123(23):234901 40. Becker OM, Karplus M (1998) The topology of multidimensional potential energy surfaces: theory and application to peptide structure and kinetics. J Chem Phys 106(4):1495–1517 41. Wales DJ, Miller MA, Walsh TR (1998) Archetypal energy landscapes. Nature 394 (6695):758–760

65

42. Crangolini T, Laurin Y, Derreumaux P et al (2015) Coarse-Grained HiRE-RNA model for ab initio RNA folding beyond simple molecules, including noncanonical and multiple base pairings. J Chem Theory Comput 14:3510–3522 43. Hess B, Kutzner C, Van Der Spoel D et al (2008) GROMACS 4: algorithms for highly efficient, load-balanced, and scalable molecular simulation. J Chem Theory Comput 4:435–447.16 44. McGibbon RT, Beauchamp KA, Harrigan MP et al (2015) MDTraj: a modern open library for the analysis of molecular dynamics trajectories. Biophys J 109(8):1528–1532 45. Antczak M, Zok T, Popenda M et al (2014) RNApdbee—a webserver to derive secondary structures from pdb files of knotted and unknotted RNAs. Nucleic Acids Res 42(W1): W368–W372 46. Wang J, Cieplak P, Kollman PA (2000) How well does a restrained electrostatic potential (RESP) model perform in calculating conformational energies of organic and biological molecules? J Comput Chem 21 (21):1049–1074 47. Pe´rez A, Marcha´n I, Svozil D et al (2007) Refinement of the AMBER force field for nucleic acids: improving the description of α/ γ conformers. Biophys J 92(11):3817–3829 48. Bana´s P, Hollas D, Zgarbova´ M, Jurecka P et al (2010) Performance of molecular mechanics force fields for RNA simulations: stability of UUCGand GNRA hairpins. J Chem Theory Comput 6(12):3836–3849 49. Zgarbova´ M, Otyepka M, Sponer J et al (2011) Refinement of the Cornell et al. Nucleic acids force field based on reference quantum chemical calculations of glycosidic torsion profiles. J Chem Theory Comput 7(9):2886–2902 50. Bourbigot S, Dock-Bregeon AC, Eberling P et al (2016) Solution structure of the 50 -terminal hairpin of the 7SK small nuclear RNA. RNA 22(12):1844–1858 51. Kusumaatmaja H, Whittleston CS, Wales DJ (2012) A local rigid body framework for global optimization of biomolecules. J Chem Theory Comput 8(12):5159–5165 52. Strodel B, Wales DJ (2008) Free energy surfaces from an extended harmonic superposition approach and kinetics for alanine dipeptide. Chem Phys Lett 466:105–115 53. Carr JM, Wales DJ (2008) Folding pathways and rates for the three-stranded β-sheet peptide Beta3s using discrete path sampling. J Phys Chem B 112:8760–8769

66

Konstantin Ro¨der and Samuela Pasquali

54. Wales DJ (2017) Decoding heat capacity features from the energy landscape. Phys Rev E 95:030105 55. Wales DJ (2009) Calculating rate constants and committor probabilities for transition networks by graph transformation. J Chem Phys 130:204111 56. Ro¨der K, Wales DJ (2018) Analysis of the Ub to Ub-CR transition in ubiquitin. Biochemistry 57(43):6180–6186 57. Li Z, Scheraga HA (1987) Monte Carlominimization approach to the multiple-minima problem in protein folding. Proc Natl Acad Sci U S A 84(19):6611–6615

58. Li Z, Scheraga HA (1988) Structure and freeenergy of complex thermodynamic systems. J Mol Struct 48:333–352 59. Wales DJ, Doye JPK (1997) Global optimization by basin-hopping and the lowest energy structures of Lennard-Jones clusters containing up to 110 atoms. J Chem Phys A 101 (28):5111–5116 60. Wales DJ, Carr JM (2012) Quasi-continuous interpolation scheme for pathways between distant configurations. J Chem Theory Comput 8(12):5020–5034

Chapter 6 Coexpression and Copurification of RNA–Protein Complexes in Escherichia coli Margot El Khouri, Marjorie Catala, Bili Seijo, Johana Chabal, Fre´de´ric Dardel, Carine Tisne´, and Luc Ponchon Abstract For structural, biochemical, or pharmacological studies, it is required to have pure RNA in large quantities. We previously devised a generic approach that allows for efficient in vivo expression of recombinant RNA in Escherichia coli. We have extended the “tRNA scaffold” method to RNA–protein coexpression in order to express and purify RNA by affinity in native condition. As a proof of concept, we present the expression and the purification of the AtRNA-mala in complex with the MS2 coat protein. Key words Expression plasmid, Recombinant RNA, RNA purification, tRNA scaffold , Protein expression

1

Introduction Recombinant transfer RNA (tRNA) has been successfully expressed in vivo in Escherichia coli [1–4]. This is possible because tRNAs are recognized and processed by a number of cellular enzymes. Furthermore, as a consequence of their three-dimensional structure, tRNAs are extremely stable to both heat-unfolding and nucleases [5]. This allows them to escape degradation and accumulate in E. coli when overexpressed from a recombinant plasmid. We took advantage of these specific features to express recombinant RNA as chimeras, by inserting them into the protective scaffold of a tRNA [6, 7]. We developed a new strategy to perform RNA–protein coexpression using E. coli. in particular when RNA are sensitive to RNase. This approach is derived from the armored RNA technology [8, 9]. The Armored tRNA (AtRNA) is composed of three parts: a tRNA scaffold, the MS2 operator hairpin and the RNA cloning site (Fig. 1a). Our system has more general applications, for instance an RNA of unknown structure could be inserted into the

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_6, © Springer Science+Business Media, LLC, part of Springer Nature 2021

67

68

Margot El Khouri et al.

Fig. 1 (a) schematic structure of standard AtRNA. The AtRNA is composed of three parts: tRNAscaffold, the MS2 operator hairpin, and the RNA cloning site. The insertion site was designed to maintain the hairpin structure. (b) Secondary structure of the AtRNA-mala. The malachite green aptamer is inserted into the AtRNA scaffold

tRNA scaffold. Concretely, we designed two plasmids for the His6MS2 coat protein expression and the AtRNA expression, respectively the pACYCT2-coat and the pBSKrna-AtRNA. The pACYCT2 and the pBSKrna are compatible with two different origin of replication and they contain, respectively, the ampicillin and chloramphenicol resistance genes, allowing selection of bacteria only cotransformed by the two plasmids. The induction of the protein expression is under the control of the lac operator and is thus triggered by the addition of IPTG. The His6-MS2 coat protein and the AtRNA are coproduced and copurified in one step by affinity on NiNTA. Here we present the expression and the purification of a chimeric RNA, in which the malachite green aptamer is fused to a tRNA scaffold with the MS2 operator hairpin instead of the anticodon loop (AtRNAmala). This chimeric RNA has been successfully purified via the high affinity interaction between the His-tagged MS2 coat protein and the MS2 operator hairpin. The malachite green aptamer remains functional under these production conditions and could be useful for in vivo fluorescence assay [10]. Our system allows RNA samples to be produced under native conditions without denaturation steps.

RNA-Protein Coexpression in E. coli

2

69

Materials Prepare all solutions using ultrapure water (prepared by purifying deionized water to attain a sensitivity of 15 MΩ cm at 25  C) and analytical grade reagents. Prepare and store all reagents at room temperature (unless indicated otherwise).

2.1 RNA–Protein Expression

1. RNA expression vector pBSKrna-AtRNAmala derived from the pBSTNAV plasmid (see Note 1). 2. Protein expression vector pACYCT2-coat derived from the pACYCduet plasmid (Novagen) (see Note 1). 3. Electrocompetent XL1-Blue E. coli cells (see Note 2). 4. 1 M Isopropyl β-D-1-thiogalactopyranoside (IPTG). Filtersterilize and store at 20  C. 5. 2 TY medium: Add 16 g tryptone, 10 g yeast extract and 5 g sodium chloride to 1 L H2O and sterilize by autoclaving. 6. 100 mg/mL ampicillin. Filter-sterilize and store at

20  C.

7. 20 mg/mL chloramphenicol. Filter-sterilize and store at 20  C. 8. Luria Broth (LB). 9. LB agar plates Add 12 g Bacto agar to 1 L LB medium before autoclaving. To prepare plates, allow medium to cool until flask or bottle can be held in hands without burning, then add 1 mL appropriate ampicillin stock(s), mix by gentle swirling and pour or pipet approximately 30 mL into each sterile petri dish (100 mm diameter). The final concentration of ampicillin should be 100 mg/L. 10. Electroporator and electroporation cuvettes. 11. Temperature-controlled shaking incubator. 12. 500-mL and 2-L culture flasks. 13. Apparatus for gel electrophoresis of RNA and protein. 14. 11% SDS-PAGE 15. Loading Buffer: 50% (v/v) glycerol, 0.2% (w/v) xylene cyanol, 1% SDS and 10 mM Tris–HCl pH 8.6. 16. Shrink wrap. 17. Hand-held UV lamp. 18. Coomassie staining solution: 0.2% Coomassie Brilliant Blue R250, 45% ethanol, 10% acetic acid. 19. Coomassie destaining solution: 20% ethanol, 10% acetic acid, 1% glycerol. 20. Water-saturated phenol.

70

Margot El Khouri et al.

21. RNA extraction buffer: 10 mM Mg-acetate, 10 mM MgCl2, pH 7.4. 2.2 RNA/Protein Purification

1. Mechanical device to disrupt E. coli cells (e.g., a sonicator, French press, or cell homogenizer). 2. Buffer A: 20 mM Tris–HCl and 300 mM NaCl, pH 8. 3. Buffer B: 20 mM Tris–HCl and 300 mM NaCl, 1 M Imidazole, pH 8. 4. AKTA FPLC chromatography system (or equivalent). 5. 5 mL Ni-NTA resin.

3 3.1

Methods Cell Growth

Carry out all procedures at room temperature unless otherwise specified. We recommend two steps: a pilot expression experiment followed by a large-scale cell growth and RNA–protein complex purification. 1. Cotransform 50 μL of electrocompetent XL1-Blue with 10–100 ng of pBSKrna-AtRNAmala vector and pACYCT2coat vector and spread 50 μL on an agar plate containing 100 μg/mL ampicillin and 20 μg/mL chloramphenicol. Incubate the plate overnight at 37  C. 2. Inoculate 100 mL LB medium containing 100 μg/mL ampicillin and 20 μg/mL chloramphenicol in a 500-mL bafflebottomed shake flask with a colony from the transformation; Shake overnight at 220 rpm and 37  C. 3. Add 25 mL of the saturated overnight culture to 1 L fresh LB medium containing 100 μg/mL ampicillin and 20 μg/mL chloramphenicol in a 2 L baffle-bottomed shake flask. 4. Shake the flasks at 250 rpm and 37  C until the cells reach mid-log phase OD600nm approximately 0.6. 5. Collect a sample for protein production analysis (see Note 3). 6. Add IPTG to a final concentration of 1 mM. Continue shaking for 4 h. 7. Collect a second sample for protein production analysis (see Note 3). 8. Pellet 5 mL culture by centrifugation for 10 min at 6000  g at 4  C. 9. Resuspend the pellet in 200 μL of RNA extraction Buffer. Add 200 μL of water-saturated phenol (see Note 4). Agitate gently for 20 min at room temperature in a polypropylene conical tube. Centrifuge 10 min at 10,000  g and collect the aqueous phase.

RNA-Protein Coexpression in E. coli

71

Fig. 2 Coexpression of AtRNA-mala/His6-MS2 coat protein in XL1-Blue in E. coli. Crude bacteria extracts (before and after IPTG induction /+) and crude RNA minipreps were separated on a 16% PAGE-SDS gel and visualized by Coomassie staining and UV shadowing. Ø (control): bacteria transformed by the vector with no insert. White boxes indicate the overexpressed RNA and protein. NiNTA elution: the RNA and the protein are eluted in the same fractions. The molecular weight (MW) of protein standards is given in kDa on the left. The black arrows indicate the AtRNA-mala and the protein

10. Add 0.1 volume of 5 M NaCl and 2 volumes of ethanol to the aqueous phase. Centrifuge 10 min at 10,000  g at 4  C in a polypropylene conical tube. Dissolve the pellet in 10 μL of water for electrophoresis analysis. 11. Analysis the expression of the RNA and the protein by electrophoresis on a 11% SDS-PAGE. Following electrophoresis, stained the gel with Coomassie staining solution, destained with Coomassie destaining solution. Place the gel on shrink wrap (see Note 5) and place on a white paper. Reveal the RNA by UV shadowing with the handheld UV lamp placed above the gel. The proteins appear as blue bands and RNA as gray bands (Fig. 2). 3.2 RNA–Protein Complex Purification

Perform all of the following procedures at 4  C. 1. Resuspend the pellet of the 1-L culture in Buffer A, using at least 10 mL/g wet cell paste. 2. Lyse the cell suspension using mechanical device to disrupt E. coli cells and centrifuge the disrupted cell suspension for 30 min at 15,000  g at 4  C. 3. Load onto a column of Ni-NTA resin equilibrated with buffer A (see Note 6). 4. Wash the column with buffer A supplemented with 20 mM imidazole until a stable baseline is reached.

72

Margot El Khouri et al.

5. Elute the RNA–protein complex with a linear gradient over 10 column volumes with buffer B. 6. Analyze the purification by electrophoresis on 11% SDS-PAGE as described in the previous paragraph (Fig. 2) (see Note 7).

4

Notes 1. The aptamer for the malachite was cloned between the EagI and SacII restriction sites of the pBSKrna-AtRNA plasmid and the resulting RNA was named AtRNA-mala (Fig. 1a). Sequence encoding for the MS2 coat protein (GenBank: AAA32260.1) was subcloned in the pACYCT2 between the NdeI and XhoI restriction sites. The vectors we described are derived from pBSTNAV and from the pACYCduet [9]. 2. As the lpp promoter on the expression vector is constitutive, there is no need for induction. RNA accumulates throughout the culture growth, up to the early stationary phase. Growing cells in rich medium is very important, as the expression of most tRNA processing enzymes is growth rate–dependent and thus tRNA chimera processing is also growth rate-dependent. For the same reason, it is important to use an “efficient” host strain, that is, not carrying mutations affecting growth rate. Most standard laboratory E. coli strains used for protein expression were found to be efficient in this respect (e.g., BL21 (DE3), DH5a, JM101, and other derivatives). 3. In order to observe the same background noise, centrifuge 100 μL of culture per 1 OD600nm before and after the induction. The pellets will be resuspended in 10 μL of water and 5 μL of Laemmli buffer. The samples have to be heated for 2 min at 95  C. 4. Phenol is toxic and corrosive, so wear gloves and handle under a fume hood. 5. You can use transparent film for wrapping food or any support UV transparent support that prevents the adsorption of the wet gel on the white paper. 6. You can choose between a batch purification using Ni-NTA Superflow onto a column with capped bottom outlet or a purification on chromatography workstations using Ni-NTA Superflow cartridges. 7. This purification step can be followed by a gel filtration step on a suitable chromatographic medium; typically, Superdex 75 can be used for small constructs (10–40 kDa) and Superdex 200 for larger constructs (40–150 kDa).

RNA-Protein Coexpression in E. coli

73

References 1. Masson JM, Miller JH (1986) Expression of synthetic tRNA genes under the control of a synthetic promoter. Gene 47:179–183 2. Meinnel T, Mechulam Y, Fayat G (1988) Fast purification of a functional elongator tRNAmet expressed from a synthetic gene in vivo. Nucleic Acids Res 16:8095–8096 3. Tisne´ C, Rigourd M, Marquet R, Ehresmann C, Dardel F (2000) NMR and biochemical characterization of recombinant human tRNA(Lys)3 expressed in Escherichia coli: identification of posttranscriptional nucleotide modifications required for efficient initiation of HIV-1 reverse transcription. RNA 6:1403–1412 4. Wallis NG, Dardel F, Blanquet S (1995) Heteronuclear NMR studies of the interactions of 15N-labeled methionine-specific transfer RNAs with methionyltRNA transformylase. Biochemistry 34:7668–7677 5. Engelke DR, Hopper AK (2006) Modified view of tRNA: stability amid sequence diversity. Mol Cell 21:144–145

6. Ponchon L, Dardel F (2007) Recombinant RNA technology: the tRNA scaffold. Nat Methods 4:571–576 7. Ponchon L, Beauvais G, Nonin-Lecomte S, Dardel F (2009) A generic protocol for the expression and purification of recombinant RNA in Escherichia coli using a tRNA scaffold. Nat Protoc 4:947–959 8. Pasloske BL, Walkerpeach CR, Obermoeller RD, Winkler M, DuBois DB (1998) Armored RNA technology for production of ribonuclease-resistant viral RNA controls and standards. J Clin Microbiol 36:3590–3594 9. Fang PY, Go´mez Ramos LM, Holguin SY et al (2017) Functional RNAs: combined assembly and packaging in VLPs. Nucleic Acids Res 45:3519–3527 10. Schwarz-Schilling M, Dupin A, Chizzolini F, Krishnan S, Mansy SS, Simmel FC (2018) Optimized assembly of a multifunctional RNA-protein nanostructure in a cell-free gene expression system. Nano Lett 18:2650–2657

Chapter 7 In Vivo Production of Small Recombinant RNAs Embedded in 5S rRNA-Derived Protective Scaffold Victor G. Stepanov and George E. Fox Abstract Preparative synthesis of RNA is a challenging task that is usually accomplished by either chemical or enzymatic polymerization of ribonucleotides in vitro. Herein, we describe an alternative approach in which RNAs of interest are expressed as a fusion with a 5S rRNA-derived scaffold. The scaffold provides protection against cellular ribonucleases resulting in cellular accumulations comparable to those of regular ribosomal RNAs. After isolation of the chimeric RNA from the cells, the scaffold can be removed, if necessary, by deoxyribozyme-catalyzed cleavage followed by preparative electrophoretic separation of the reaction products. The protocol is designed for sustained production of high quality RNA on the milligram scale. Key words Recombinant RNA, In vivo RNA production, 5S rRNA-derived scaffold

1

Introduction Recent advances in RNA technologies create a demand for costeffective production of RNA aptamers, antisense oligoribonucleotides, ribozymes, and other RNAs of therapeutic, environmental or industrial significance. As of now, the majority of small RNAs (70 nmoles of biotin-tagged ligands per 1 mL of wet beads. If the capacity is lower, the amount of the beads should be increased accordingly. 29. It is recommended to place the thermostat incubator near the centrifuge to minimize the time needed to transfer the tube from the incubator to the centrifuge. 30. In the case of strong residual interactions between the DNAzymes and the fragments of cleaved chimeric RNA, the number of washes should be doubled. 31. If necessary, the excised RNA can further be sequenced to confirm its identity. First, the RNA sample should be treated with T4 polynucleotide kinase to remove 30 -terminal cyclophosphate group and to phosphorylate the 50 end. Then, the RNA is ligated via its 30 end to a 50 -adenylated DNA adapter of defined 25-nucleotide sequence with dideoxyribonucleotide at the 30 end. The ligation reaction is catalyzed by T4 RNA ligase 2 truncated KQ (New England Biolabs, #M0373). The second adapter, an unmodified 25-mer oligoribonucleotide, is ligated to the 50 end of the RNA using T4 RNA ligase 1 (New England Biolabs, #M0437). Then, the tagged RNA is reversetranscribed using SuperScript IV Reverse Transcriptase (Invitrogen/ThermoFisher Scientific, #18090010) and the DNA primer complementary to the first adapter. The obtained DNA strand is amplified by PCR using an appropriate high-fidelity DNA polymerase, the DNA primer complementary to the first adapter, and the DNA primer matching the sequence of the second adapter. The dsDNA product is purified and sequenced bidirectionally by the Sanger method using the same primers as for the amplification reaction. The same approach can be used to prepare samples for RNA-seq analysis, but the adapter sequences should be compatible with the chosen sequencing instrument. RNA-seq analysis allows to not only validate the sequence of the produced RNA but also to reveal the presence of contaminating RNA species if any.

Recombinant RNA Production Using a 5S rRNA-Derived Scaffold

97

Acknowledgments This work was supported in part by NASA contract 80NSSC18K1139 under the Center for Origin of Life and NASA grant NNX14AK16G to GEF. References 1. Caruthers MH (2013) The chemical synthesis of DNA/RNA: our gift to science. J Biol Chem 288(2):1420–1427 2. Sousa R, Mukherjee S (2003) T7 RNA polymerase. Prog Nucleic Acid Res Mol Biol 73:1–41 3. Burroughs AM, Aravind L (2016) RNA damage in biological conflicts and the diversity of responding RNA repair systems. Nucleic Acids Res 44(18):8525–8555 4. Deutscher MP (2006) Degradation of RNA in bacteria: comparison of mRNA and stable RNA. Nucleic Acids Res 34(2):659–666 5. Yu AM, Batra N, Tu MJ, Sweeney C (2020) Novel approaches for efficient in vivo fermentation production of noncoding RNAs. Appl Microbiol Biotechnol 104:1927–1937 6. Pitulle C, Hedenstierna KO, Fox GE (1995) A novel approach for monitoring genetically engineered microorganisms by using artificial, stable RNAs. Appl Environ Microbiol 61:3661–3666 7. Ammons D, Rampersad J, Fox GE (1998) A genomically modified marker strain of Escherichia coli. Curr Microbiol 37:341–346 8. Tucker DL, Karouia F, Wang J et al (2005) Effect of an artificial RNA marker on gene expression in Escherichia coli. Appl Environ Microbiol 71:4156–4159 9. Ammons D, Rampersad J, Fox GE (1999) 5S rRNA gene deletions cause an unexpectedly high fitness loss in Escherichia coli. Nucleic Acids Res 27:637–642 10. Zhang X, Potty AS, Jackson GW et al (2009) Engineered 5S ribosomal RNAs displaying aptamers recognizing vascular endothelial growth factor and malachite green. J Mol Recognit 22:154–161 11. Pitulle C, D’Souza L, Fox GE (1997) A low molecular weight artificial RNA of unique size with multiple probe target regions. Syst Appl Microbiol 20:133–136 12. D’Souza LM, Larios-Sanz M, Setterquist RA et al (2003) Small RNA sequences are readily

stabilized by inclusion in a carrier rRNA. Biotechnol Prog 19:734–738 13. Liu Y, Stepanov VG, Strych U et al (2010) DNAzyme-mediated recovery of small recombinant RNAs from a 5S rRNA-derived chimera expressed in Escherichia coli. BMC Biotechnol 10:85 14. Cruz RP, Withers JB, Li Y (2004) Dinucleotide junction cleavage versatility of 8-17 deoxyribozyme. Chem Biol 11:57–67 15. Schlosser K, Gu J, Sule L et al (2008) Sequence-function relationships provide new insight into the cleavage site selectivity of the 8-17 RNA-cleaving deoxyribozyme. Nucleic Acids Res 36:1472–1481 16. Schlosser K, Gu J, Lam JC et al (2008) In vitro selection of small RNA-cleaving deoxyribozymes that cleave pyrimidine-pyrimidine junctions. Nucleic Acids Res 36:4768–4777 17. Lam JC, Kwan SO, Li Y (2011) Characterization of non-8-17 sequences uncovers structurally diverse RNA-cleaving deoxyribozymes. Mol BioSyst 7:2139–2146 18. Cairns MJ, King A, Sun LQ (2003) Optimisation of the 10-23 DNAzyme-substrate pairing interactions enhanced RNA cleavage activity at purine-cytosine target sites. Nucleic Acids Res 31:2883–2889 19. Borodina TA, Lehrach H, Soldatov AV (2003) DNA purification on homemade silica spincolumns. Anal Biochem 321:135–137 20. Brosius J (1984) Toxicity of an overproduced foreign gene product in Escherichia coli and its use in plasmid vectors for the selection of transcription terminators. Gene 27:161–172 21. Hartmann RK, Bindereif A, Scho¨n A et al (2005) Appendix: UV spectroscopy for the quantitation of RNA. In: Hartmann RK, Bindereif A, Scho¨n A, Westhof E (eds) Handbook of RNA biochemistry. Wiley-VCH Verlag GmbH & Co, KGaA, Weinheim, p 911 22. Cavaluzzi MJ, Borer PN (2004) Revised UV extinction coefficients for nucleoside-50 -monophosphates and unpaired DNA and RNA. Nucleic Acids Res 32:e13

Chapter 8 Production of Circular Recombinant RNA in Escherichia coli Using Viroid Scaffolds Jose´-Antonio Daro`s Abstract Viroids are small circular, noncoding, highly base-paired RNAs able to infect higher plants. Recently, it has been shown that viroids can be used as very stable scaffolds to produce recombinant RNA in Escherichia coli. Coexpression of an RNA precursor consisting of a viroid monomer, in which the RNA of interest is inserted, flanked by domains of the viroid hammerhead ribozyme, along with a host plant tRNA ligase, the enzyme that catalyzes viroid circularization in infected plants, allows for accumulation of large amounts of the chimeric viroid-RNA of interest in E. coli. Since viroids do not replicate in E. coli, high accumulation most probably results from viroid scaffold stability, resistance to exonucleases due to circularity, and accumulation as a ribonucleoprotein complex with tRNA ligase. Purification of the recombinant RNA from total E. coli RNA is also facilitated by the circular structure of the product. Key words Recombinant RNA, Circular RNA, Viroid scaffold, tRNA ligase, Escherichia coli

1

Introduction In parallel to what occurs with the production of recombinant proteins, the most cost-effective strategy to produce large amounts of recombinant RNAs is to take advantage of the native transcription systems of cells or entire organisms, which are then converted into molecular biofactories [1]. However, this approach is usually hampered by the intrinsic low half-life of RNA. Although RNAs of interest are efficiently transcribed, endogenous RNases will hydrolyze them in a rapid turnover. This limitation was satisfactorily circumvented by expressing the RNAs of interest embedded in the molecules of highly stable RNAs, such as tRNAs or rRNAs, which act as protective scaffolds in cells like Escherichia coli [2, 3]. Stability of the RNAs of interest expressed in bacteria was also shown to be improved by coexpression of a companion protein that forms a ribonucleoprotein complex [4], by circularization, which protects the RNA from exonuclease degradation [5], or by the use of particular bacterial species [6].

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_8, © Springer Science+Business Media, LLC, part of Springer Nature 2021

99

100

Jose´-Antonio Daro`s

Viroids are a unique class of pathogenic RNAs that infect higher plants. The more than thirty viroid species that are known to date exclusively consist of relatively small (246–430 nt), circular RNAs [7]. Remarkably, when inoculated in the right host, these circular RNAs, which do not code for any protein, are able to recruit the right plant enzymes and structures to mediate their replication, as well as their intracellular and intercellular traffic to invade the whole plant [8]. Amazingly, this occurs while the plant defensive systems, for example the host antiviral RNA silencing pathways, try to clear the pathogen infection [9]. In sum, the highly structured circular molecules of viroids, which consist of a series of double-stranded stretches separated by loops and bulges [10, 11], have been shaped by evolution to survive in the hostile environment of the host cell, and may represent an ideal scaffold to produce recombinant RNAs. In fact, in the course of a research aimed to understand how Eggplant latent viroid (ELVd) is able to recruit the chloroplastic isoform of eggplant tRNA ligase to mediate circularization of the viroid progeny [12], viroid suitability to perform as scaffolds to produce recombinant RNA in E. coli was discovered [13]. In addition to serve as a highly stable RNA scaffold, the viroid-derived system produces a circular RNA product that is resistant to exonucleases and also facilitates purification to homogeneity, based on this unique property [14]. The system likely produces a ribonucleoprotein complex between the chimeric viroid-RNA of interest molecule and the eggplant tRNA ligase that most probably contributes to increased accumulation in E. coli cells. In order to produce recombinant circular RNA in E. coli using a viroid scaffold, methods for plasmid construction, E. coli transformation, and recombinant RNA expression and purification, in the case of the ELVd-based system, has been recently detailed in this same series [15]. Here, I will focus on methods to enlarge the repertoire of scaffolds based on different viroids species, particularly those in the family Avsunviroidae (hammerhead viroids).

2

Materials Prepare all solutions using ultrapure water and analytical grade reagents, and follow all waste disposal regulations.

2.1 Bacterial Strains and Plasmids

1. Use the E. coli strains DH5α and BL21, or equivalents, for cloning and expression purposes, respectively. 2. Plasmid pULZ (Addgene plasmid # 169813) contains the E. coli murein lipoprotein promoter and the 5S rRNA (rrnC) terminator separated by a polylinker with two BsaI sites flanking the cDNA encoding the α peptide of β-galactosidase for blue–white cloning screening. It also contains a pUC

Production of Circular Recombinant RNA in Escherichia

101

Fig. 1 Schematic representations of plasmids (a) pULZ and (b) p15LZ to clone the precursor RNA and the plant tRNA ligase, respectively. Ori pUC and Ori p15A, replication origins pUC and p15A, respectively; AmpR and CamR, resistance to ampicillin and chloramphenicol selection markers, respectively; lpp, E. coli murein lipoprotein promoter; rrnC, E. coli 5S rRNA terminator; T7 50 UTR, T7 30 UTR, and T7 term, bacteriophage T7 50 untranslated region, 30 untranslated region, and terminator, respectively; LacZ, blue–white screening cassette that expresses the β-galactosidase α peptide

replication origin and an ampicillin resistance selection marker (Fig. 1a). This plasmid is intended to express the viroid scaffold with the RNA of interest inserted. 3. Plasmid p15LZ (Addgene plasmid # 169814), with a p15A replication origin and a chloramphenicol resistance selection marker, contains the E. coli murein lipoprotein promoter and the T7 bacteriophage terminator. They are separated by a polylinker with two BsaI sites that flank the cDNA of the of β-galactosidase α peptide (Fig. 1b). Derivatives of this plasmid, which will be compatible in E. coli with those of pULZ, will produce the tRNA ligases. 4. cDNA fragments for plasmid assembly can be obtained by gene synthesis of by amplification using standard reverse transcription (RT)-polymerase chain reaction (PCR) from RNA preparations from the appropriate plants. 5. Use the online NEBuilder assembly tool (http://nebuilder. neb.com/#!/) for the design and the NEBuilder HiFi DNA assembly master mix from New England Biolabs (see Note 1). 6. BsaI restriction enzyme. 7. Agarose gel electrophoresis system. 2.2

Culture Media

1. Luria-Bertani (LB) medium: 1% tryptone, 0.5% yeast extract, and 1% NaCl. Weigh 10 g of tryptone, 5 g of yeast extract, and 10 g of NaCl. Dissolve in water and bring it to 1 L. Transfer to

102

Jose´-Antonio Daro`s

a bottle and sterilize by autoclaving. For solid LB media, add 15 g of agar (1.5%) before autoclaving. 2. Super optimal broth with catabolite repression (SOC): 20 g/L tryptone, 5 g/L yeast extract, 0.5 g/L NaCl, 2.5 mM KCl, 10 mM MgCl2, 20 mM glucose, pH 7.0. Prepare 1 M MgCl2 stock solution by dissolving 20.3 g MgCl2 (·6H2O) in water. Bring to 100 mL and sterilize by autoclaving. Prepare 1 M KCl stock solution by dissolving 7.4 g KCl in water. Bring to 100 mL and sterilize by autoclaving. Prepare 1 M glucose by dissolving 18 g glucose in water. Bring to 100 mL and sterilize by filtration. Weigh 20 g of tryptone, 5 g of yeast extract, and 0.5 g of NaCl, and dissolve in water. Add 2.5 mL 1 M KCl and 10 mL 1 M MgCl2. Adjust the pH to 7.0 with NaOH and bring to 980 mL. Transfer to a bottle and sterilize by autoclaving. After cooling, add 20 mL of sterile 1 M glucose. 3. Terrific Broth (TB): 1.2% tryptone, 2.4% yeast extract, 0.4% glycerol, 17 mM KH2PO4, and 72 mM K2HPO4. Weigh 12 g of tryptone, 24 g of yeast extract, and 5.8 g of 87% glycerol (density 1.228 g/mL) and dissolve in 900 mL of water. Weigh 2.3 g of KH2PO4 and 16.4 g of K2HPO4 (·3H2O), and dissolve in 100 mL of water. Sterilize by autoclaving. Mix 9:1 parts of each solution before use. 4. 50 mg/mL ampicillin: weigh 1 g of ampicillin and dissolve in water. Bring to 20 mL. Sterilize by filtration and store at 20  C in aliquots. 5. 34 mg/mL chloramphenicol: weigh 340 mg of chloramphenicol and dissolve in ethanol. Bring to 10 mL final volume and store at 20  C. 6. 50 mg/mL 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal): weigh 0.25 g of X-gal and dissolve in 5 mL of N,N-dimethylformamide. Prepare 0.5 mL aliquots in tubes protected from light with aluminum foil and store at 20  C. 7. Electroporator, 1-mm cuvettes. 2.3 Purification and Analysis of Recombinant RNA

1. TE buffer: 10 mM Tris(hydroxymethyl)aminomethane (Tris), 1 mM ethylenediaminetetraacetic acid (EDTA), pH 8.0. Prepare 1 M Tris–HCl, pH 8.0. Weigh 60.6 g of Tris and dissolve in water. Adjust pH to 8.0 with HCl and bring to 500 mL. Autoclave and store at room temperature. Prepare 0.5 M EDTA, pH 8.0. Weigh 18.6 g of disodium EDTA (·2H2O) and add 75 mL of water. Adjust pH to 8.0 with 10 M NaOH while stirring the mix. Disodium EDTA will not dissolve until pH is close to 8. Bring the solution to 100 mL, sterilize by autoclaving and store at room temperature. Finally, mix 1 mL of 1 M Tris–HCl, pH 8.0, and 200μL of 0.5 M EDTA, pH 8.0, and bring to 100 mL with water.

Production of Circular Recombinant RNA in Escherichia

103

2. 1:1 (v/v) buffered phenol–chloroform, pH 8.0. Work in a hood when preparing and handling this reagent. Prepare water-saturated phenol mixing 400 mL of 90% phenol and 120 mL of water. Mix and transfer to an amber bottle. Store at 4  C. Mix 250 mL of water-saturated phenol and 50 mL of 1 M Tris–HCl, pH 8.0. Stir the mix for 5 min and settle to separate the phases. Remove most of the upper aqueous phase. Add 250 mL of chloroform. Stir the mix for 5 min and settle to separate the phases. Again remove most of the upper aqueous phase. Store at 4  C. 3. Chloroform.

3 3.1

Methods Plasmid Design

1. To produce in E. coli a recombinant RNA of interest embedded into a viroid scaffold, the corresponding construct must be coexpressed along with a plant tRNA ligase that will be involved in processing of the RNA precursor and accumulation of the final product. 2. Although any viroid may serve the purpose for a stable scaffold, the five viroids in the family Avsunviroidae at either plus or minus polarities are most suitable because: (a) they can be flanked by the endogenous hammerhead ribozymes that will self-cleave RNA precursor, (b) hammerhead ribozyme selfcleavage will produce the 50 -hydroxyl, 20 ,30 -cyclic phosphodiester termini that are directly recognized by the plant tRNA ligases, (c) each viroid in this family must be efficiently recognized by the cognate tRNA ligase from the corresponding host plant (see Note 2). 3. Select viroid sequence and polarity. GenBank reference sequences of the five currently known viroids that belong to the family Avsunviroidae are NC_001410.1, Avocado sunblotch viroid (ASBVd); NC_003636.1, Peach latent mosaic viroid (PLMVd); NC_003540.1, Chrysanthemum chlorotic mottle viroid (CChMVd); NC_039241, ELVd; and NC_028132.1, Apple hammerhead viroid (AHVd). Many natural sequence variants from these viroid species are available at GenBank or dedicated databases [16]. Deleted forms of viroid molecules may also be used [13]. 4. Select insertion site in the viroid molecule. Excellent results have been obtained inserting the RNA of interest between position U245 and U246 of plus polarity ELVd (NC_028132.1) [13]. This position corresponds to the terminal loop of a long hairpin. However, other insertion positions in ELVd or the other viroids may be equally useful.

104

Jose´-Antonio Daro`s

a

Ipp

rrnC

RNA of interest HH

5’ viroid

3’ viroid

HH

pULZ/BsaI

b

Plant tRNA ligase Ipp

T7

p15LZ/BsaI

Fig. 2 Schematic representation of the two plasmids that must be assembled to coexpress in E. coli (a) the precursor RNA that includes the RNA of interest and (b) the plant tRNA ligase. lpp, E. coli murein lipoprotein promoter; HH, hammerhead ribozyme; rrnC, E. coli 5S rRNA terminator; T7, bacteriophage T7 transcriptional terminator

5. Select the cDNA for the plant tRNA ligase. Eggplant (Solanum melongena) tRNA ligase (GenBank accession number JX025157.1) has been successfully used and a p15LZderivative to express this enzyme in E. coli is already available [13]. However, other plant tRNA ligases may be used, particularly with viroid scaffolds other than ELVd (see Note 3). Some examples are those from peach (Prunus persica; accession number XP_020421771.1) or apple (Malus domestica; accession number XP_028944329.1). 6. Finally, use the NEBuilder assembly tool to design the primers to amplify the PCR products that, in combination with digested vectors and synthesized DNAs, will be used for plasmid assembly. For the design follow the scheme in Fig. 2. 3.2 Plasmid Construction

1. Obtain the different DNA fragments by standard RT-PCR or PCR amplification techniques or by gene synthesis. The PCR products of interest must be purified after separation of the reaction products by electrophoresis in agarose gels. 2. Digest approximately 100 ng of plasmids pULZ and p15LZ with BsaI for 1 h at 37  C, separate the reaction products by electrophoresis in a 1% agarose gel and elute the linearized plasmid in which the polylinker was removed. 3. Set up the Gibson assembly reactions by mixing the different DNA fragments at molar ratios of 3 (each insert):1 (vector). Add one volume of NEBuilder HiFi DNA assembly master mix and incubate for 1 h at 50  C. Purify the reaction products using a silica gel column and use the eluate to electroporate E. coli DH5α. 4. Electroporate cells at 1500 V using 1-mm cuvettes. Quickly add 1 mL of SOC medium. Recover bacteria from the cuvette and incubate for 1 h at 37  C with shaking.

Production of Circular Recombinant RNA in Escherichia

105

5. Spread electroporated bacteria on LB plates containing 50μg/ mL ampicillin (pULZ derivatives) or 34μg/mL chloramphenicol (p15LZ derivatives), and X-gal (see Note 4). Grow overnight at 37  C. 6. Pick several white E. coli colonies and inoculate liquid LB media containing the appropriate antibiotic. Grow overnight at 37  C with shaking. 7. Miniprep the plasmids and analyze them by electrophoresis in an agarose gel. Use the empty pULZ y p15LZ as controls. Select plasmids with the expected electrophoretic mobility according to the inserted cDNA. Restriction analysis and sequencing may be required to finally select the right plasmids. 3.3 Recombinant RNA Production

1. Electroporate competent BL21 E. coli with the selected pULZderivative that will serve to express the viroid scaffold-RNA of interest precursor and the selected p15LZ-derivative that will mediate expression of the plant tRNA ligase (see Note 5). Repeat step 4 in Subheading 3.2. Spread electroporated bacteria in LB agar plates with 50μg/mL ampicillin and 34μg/mL chloramphenicol and incubate overnight at 37  C. 2. Pick colonies of cotransformed E. coli in 10 mL TB containing 50μg/mL ampicillin and 34μg/mL chloramphenicol, in 50 mL tubes (see Note 6). Incubate during 24 h at 37  C with vigorous shaking (225 revolutions per minute; rpm) (see Note 7).

3.4 Purification and Analysis of Recombinant RNA

1. Sediment E. coli cells by centrifugation at 10,000  g for 10 min. Discard supernatant. 2. Resuspend cells in some water by vortexing and completely fill the tube with water to extensively wash the cells. Sediment bacteria again by centrifuging in the same conditions as above. 3. Resuspend cells in 200μL TE buffer by vortexing. Bacteria can be frozen at 20  C at this point for further processing or purification continued. Add one volume (200μL) of 1:1 phenol–chloroform and vortex vigorously. Centrifuge for 10 min at 10,000  g to separate phases and recover the upper aqueous phase. 4. Add one volume (200μL) of chloroform. Vortex, separate phases by centrifugation as indicated above, and take the aqueous phase that will contain the total bacterial nucleic acids. 5. Analysis of the resulting recombinant RNAs can be performed by electrophoresis [15]. Methods for further purification of the recombinant RNA by chromatography or 2 dimensional electrophoresis have also been previously described in this series [15].

106

4

Jose´-Antonio Daro`s

Notes 1. Equivalent programs and reagents for Gibson assembly are also available from other companies. 2. Hybrid assemblies consisting of viroids belonging to the family Pospiviroidae flanked by hammerhead ribozymes from avsunviroids (family Avsunviroidae) may, in principle, also be possible. 3. Although use of the cognate tRNA ligase for each viroid scaffold is recommended, eggplant tRNA ligase efficiently circularizes precursors from different viroid species [17]. Thus, this observation suggests that different viroid scaffolds may be combined with ligases from different host or nonhost plants. 4. Before plating electroporated bacteria, spread 30μL of 50 mg/ mL X-gal on the plates for blue–white screening. 5. We obtain optimal results coelectroporating bacteria in a single step with approximately 100 ng of each plasmid. However, sequential transformation can be also performed. 6. Culture media and flask volumes can be varied according to the necessity. However, it is important to maintain a low media– flask volume ratio to ensure sufficient culture ventilation. Use baffled Erlenmeyer flasks for large scale preparations. 7. Although 24 h incubation may work in most instances, a timecourse analysis of recombinant RNA production is recommended, because RNA accumulation dynamics depend on many experimental factors such as culture media volume, flask volume and geometry, temperature, and agitation.

Acknowledgments This work was supported by grants BIO2017-83184-R and BIO2017-91865-EXP from the Ministerio de Ciencia e Innovacio´n (Spain), cofinanced by European Regional Development Fund (European Commission). References 1. Yu AM, Batra N, Tu MJ, Sweeney C (2020) Novel approaches for efficient in vivo fermentation production of noncoding RNAs. Appl Microbiol Biotechnol 104:1927–1937 2. Ponchon L, Dardel F (2007) Recombinant RNA technology: the tRNA scaffold. Nat Methods 4:571–576 3. Zhang X, Potty ASR, Jackson GW, Stepanov V, Tang A, Liu Y, Kourentzi K, Strych U, Fox GE,

Willson RC (2009) Engineered 5S ribosomal RNAs displaying aptamers recognizing vascular endothelial growth factor and malachite green. J Mol Recognit 22:154–161 4. Ponchon L, Catala M, Seijo B, El Khouri M, Dardel F, Nonin-Lecomte S, Tisne´ C (2013) Co-expression of RNA-protein complexes in Escherichia coli and applications to RNA biology. Nucleic Acids Res 41:e150

Production of Circular Recombinant RNA in Escherichia 5. Umekage S, Kikuchi Y (2009) In vitro and in vivo production and purification of circular RNA aptamer. J Biotechnol 139:265–272 6. Kikuchi Y, Umekage S (2018) Extracellular nucleic acids of the marine bacterium Rhodovulum sulfidophilum and recombinant RNA production technology using bacteria. FEMS Microbiol Lett 365:fnx268 7. Di Serio F, Flores R, Verhoeven JT, Li SF, Palla´s V, Randles JW, Sano T, Vidalakis G, Owens RA (2014) Current status of viroid taxonomy. Arch Virol 159:3467–3478 8. Daro`s JA (2016) Viroids: small noncoding infectious RNAs with the remakable ability of autonomous replication. Curr Res Top Plant Virol:295–322 9. Elena SF, Go´mez G, Daro`s JA (2009) Evolutionary constraints to viroid evolution. Viruses 1:241–254 10. Gigue`re T, Adkar-Purushothama CR, Perreault JP (2014) Comprehensive secondary structure elucidation of four genera of the family Pospiviroidae. PLoS One 9:e98655 11. Gigue`re T, Adkar-Purushothama CR, Bolduc F, Perreault JP (2014) Elucidation of the structures of all members of the Avsunviroidae family. Mol Plant Pathol 15:767–779

107

12. Cordero T, Ortola´ B, Daro`s JA (2018) Mutational analysis of eggplant latent viroid RNA circularization by the eggplant tRNA ligase in Escherichia coli. Front Microbiol 9:635 13. Daro`s JA, Aragone´s V, Cordero T (2018) A viroid-derived system to produce large amounts of recombinant RNA in Escherichia coli. Sci Rep 8:1904 14. Cordero T, Aragone´s V, Daro`s JA (2018) Large-scale production of recombinant RNAs on a circular scaffold using a viroid-derived system in Escherichia coli. J Vis Exp 2018: e58472 15. Ortola´ B, Daro`s J-A (2020) Production of recombinant RNA in Escherichia coli using eggplant latent viroid as a scaffold. Methods Mol Biol 2316: chapter 25 16. Rocheleau L, Pelchat M (2006) The subviral RNA database: a toolbox for viroids, the hepatitis delta virus and satellite RNAs research. BMC Microbiol 6:24 ´ , Molina-Serrano D, Flores R, 17. Nohales M-A Daro`s J-A (2012) Involvement of the chloroplastic isoform of tRNA ligase in the replication of viroids belonging to the family Avsunviroidae. J Virol 86:8269–8276

Chapter 9 Identification of RNA-Binding Proteins Associated to RNA Structural Elements Javier Fernandez-Chamorro, Rosario Francisco-Velilla, Azman Embarc-Buh, and Encarnacion Martinez-Salas Abstract RNA motifs guide the interaction with specific proteins leading to the assembly of ribonucleoprotein complexes that perform key functions in cellular processes. Internal ribosome entry site (IRES) elements are organized in structural domains that determine internal initiation of translation. In this chapter we describe a pull-down assay using streptavidin-aptamer tagged RNAs that combines RNA structuredependent protein isolation with proteomic analysis to identify novel interactors recognizing RNA structural domains. This approach takes advantage of tRNA-scaffold guided expression, allowing the identification of factors belonging to networks involved in RNA and protein metabolism. Key words RNA structure, RNA-binding proteins, Pull-down, Streptavidin aptamer, tRNA scaffold, IRES elements, Translation control

1

Introduction RNAs can fold into diverse structural motifs, which determine the capacity to interact with distinct type of RNA-binding proteins (RBPs), ultimately affecting their function. This property also affects mRNAs, and specifically, to regulatory elements located in the 50 and 30 noncoding regions. Translation control is a key step in gene expression regulation. In eukaryotic organisms, most mRNAs initiate translation using the so-called cap-dependent mechanism [1]. However, in response to different types of stresses, eukaryotic cells can reprogram the expression of specific genes allowing survival. Internal ribosome entry site (IRES) elements, which are mostly located at the 50 untranslated region, promote internal initiation of translation using cap-independent mechanisms [2]. Not surprisingly, IRES elements govern translation initiation in several positive strand RNA viruses under conditions of strong cap-dependent shutdown.

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_9, © Springer Science+Business Media, LLC, part of Springer Nature 2021

109

110

Javier Fernandez-Chamorro et al.

As in most regulatory RNAs, the structure of viral IRES elements plays a critical role for IRES function. Type II IRES elements, present in the genomic RNA of encephalomyocarditis (EMCV) and foot-and-mouth disease virus (FMDV), fold into modular domains [3]. In both IRES elements, the central domain is a self-folding cruciform structure that includes long-range interactions [4–6]. However, the implication of RNA structural features for the interaction with IRES transacting factors (ITAFs), presumably modulating their function, remains poorly studied. Beyond RNA structure, IRES activity depends on cellular proteins, specifically eukaryotic initiation factors (eIFs) and RBPs. Since the early reports on the discovery of ITAFs, the number of host factors identified as IRES-binding proteins has grown incessantly [7– 10]. Yet, there are differences in the factors required to assemble the translation initiation complex among distinct type of IRES elements. Thus, understanding how these elements perform their function demands the identification of ITAFs in the context of the folded (native) RNA. Given the relatively long length of picornavirus type II IRES, their RNA structure flexibility, as well as the diversity of cellular stresses that may influence IRES-dependent activity [11], it seems reasonable to propose that additional factors could interact, perhaps in a transient manner, with these noncoding regulatory regions, guiding stimulation or down-regulation of cap-independent translation. In this chapter, we describe a pulldown assay that combines the high affinity of streptavidin for streptavidin-aptamers with the capacity of the tRNA scaffold background to stabilize the secondary structure of RNA motifs (36–200 nucleotides in length) [12]. Streptavidin-aptamer tagged transcripts encoding individual structural IRES motifs in the tRNA scaffold context was recently used to perform a systematic characterization of IRES-bound factors followed by mass spectrometry analysis [13]. This pull-down approach allowed for the identification of proteins previously reported to interact with IRES elements [14], reinforcing the validity of the procedure. More importantly, we unambiguously identified host factors belonging to functional networks involved in a wide variety of RNA-dependent metabolism, including amino acid and nucleotide biosynthetic processes, ER-Golgi transport, innate immunity, besides RNA processing and translation. In this experimental setting, IRES transcripts also interacted with previously unknown factors as RNA-binding proteins, revealing the RNA-binding capacity of these proteins.

2

Materials 1. All solutions are prepared using RNase-free conditions in sterile glassware or disposable material.

Pull-Down Assay Using Streptavidin-Aptamer Tagged RNAs

111

2. RNA and protein purification, as well as RNA–protein association, is carried out in sterile polypropylene tubes. 3. HeLa cells are grown in sterile tissue culture disposable dishes. 4. Cellular extracts are prepared at 4  C, and frozen at 20  C until needed. 2.1 Pull-down and Cell Extract Preparation Materials

1. Streptavidin-coated magnetic beads Dynabeads (10 mg/ml) (ThermoFisher Scientific).

M-280

2. Magnetic stand. 3. Dounce homogenizer (1 ml). 4. Rotating wheel. 5. Tissue culture cells equipment. 6. Electrophoresis equipment.

2.2 RNA Chimera Prep

1. Escherichia coli cells transformed with a plasmid derived from vector pBSMrnaStrep expressing the RNA structural motif SL3a [12, 13]. 2. Luria-Bertani broth (LB) supplemented with 50 μg/ml Kanamycin. 3. 50 mM Tris–HCl pH 7.5. 4. Phenol (0.1 M Tris–HCl pH 7.5 equilibrated). 5. Ethanol absolute. 6. 1 M NaCl. 7. RNase-free water. 8. Acrylamide–bisacrylamide 8%, 8 M urea denaturing gels.

2.3

Cell Extract Prep

1. HeLa cells (or mammalian cells of interest). 2. Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% calf fetal serum. 3. Phosphate buffered saline (PBS) pH 7.4. 4. Hypotonic buffer: 10 mM HEPES–KOH pH 7.4, 10 mM KAc, 1.5 mM MgAc, 2.5 mM DTT. 5. Bradford reagent. 6. Glycerol.

2.4 RNA–Protein Pull Down

1. Solution A: 50 mM NaCl, 0.1 M NaOH. 2. Solution B: 10 mM NaCl. 3. Binding & Washing buffer: 5 mM Tris–HCl pH 7.5, 1 M NaCl, 0.5 mM EDTA. 4. Binding buffer: 100 mM HEPES pH 7.4, 200 mM NaCl, 6 mM MgCl2.

112

Javier Fernandez-Chamorro et al.

5. 10 mg/ml BSA (RNase-free grade). 6. 20 ng/μl yeast tRNA. 7. 0.1 M DTT. 8. 3 Laemmli buffer. 9. 10% SDS-PAGE. 10. PANTHER software.

3

Methods

3.1 RNA Chimera Purification

1. Grow Escherichia coli cells transformed with a plasmid derived from vector pBSMrnaStrep expressing the RNA structural motif SL3a [12, 13] in LB (50 ml) supplemented with kanamycin (50μg/ml) overnight at 37  C. 2. Recover cells spinning for 10 min 4000  g at 4  C, and resuspend the pellet in 1.5 ml of 50 mM Tris–HCl pH 7.5. 3. Add 1 ml phenol (1:1 vol/vol), vortexing 20–30 s at room temperature. 4. Spin 10 min 10,000  g at room temperature to separate phases. Collect the aqueous phase in a clean tube. 5. Add 2 volumes of ice-cold absolute ethanol and 0.1 volume of 1 M NaCl to precipitate RNA. Store at 20  C for 30 min. 6. Collect the RNA spinning 15 min 10,000  g at 4  C. Dry the pellet at room temperature for 5 min. 7. Resuspend the pellet in RNase-free water (200μl). 8. Load a small aliquot of RNA (10μl), in parallel to a control RNA, in denaturing 8% acrylamide–bisacrylamide, 8 M Urea; run the gel 35 min at 300 V (see Note 1). 9. To visualize the transcript, stain the gel with EtBr, SYBR Green, or UV shadowing. RNA (full-length size) should appear as a single band after binding to streptavidin-coated beads.

3.2 Preparation of S10 Cell Lysates

1. Grow HeLa cells (or the mammalian cell of interest) in 10-cm dishes (5–10 dishes) to 100% confluence in DMEM supplemented with 10% calf fetal serum, in a humidified incubator 5% CO2 at 37  C (see Note 2). 2. Wash twice cell monolayers with ice-cold PBS. 3. Scrape cells in 1 ml ice-cold PBS/dish and collect them by centrifugation (15 ml polystyrene sterile tubes) 2000  g 5 min at 4  C. 4. Resuspend the cell pellet in 1 volume of ice-cold hypotonic buffer.

Pull-Down Assay Using Streptavidin-Aptamer Tagged RNAs

113

5. Lyse cells by 30 strokes in 1 ml glass Dounce homogenizer on ice. 6. Transfer the lysate to a clean polystyrene tube. 7. Remove cell debris by centrifugation at 5000  g for 5 min at 4  C. 8. Transfer the clear lysate to a clean polypropylene tube. 9. Spin down the clear lysate at 10,000  g for 5 min at 4  C. 10. Transfer the supernatant to a clean polypropylene tube, and adjust the supernatant to 3% glycerol. 11. Measure total protein concentration by the Bradford assay. 3.3 Pull-down Assay Using Streptavidin-Aptamer Tagged RNA

As a general rule, it is recommended to use 1 mg (100μl) of streptavidin-coated magnetic beads (Dynabeads M-280, 10 mg/ ml) for 20 pmol of RNA. This ratio should be optimized for the RNA of interest (see Note 3).

3.3.1 Binding of the RNA Chimera to Streptavidin-Coated Magnetic Beads

1. Wash streptavidin-coated magnetic beads with 500μl Binding & Washing buffer in a polypropylene 1.5 ml tube for 5 min at room temperature in a rotating wheel. 2. Recover the beads in the tube wall using a magnetic stand for 2 min. 3. Wash twice the beads with 600μl solution A, recovering the pellet on the tube wall standing on the magnet 2 min. 4. Wash once the beads with solution B; then, resuspend the beads in 100μl solution B. 5. Add RNA chimera (20 pmol) to prewashed streptavidin-coated magnetic beads in 100μl of solution B, 100μl binding buffer (see Note 4) and RNase-free water in a final volume of 500μl. Incubate 30 min in a rotating wheel at room temperature (see Fig. 1 for an overview of the protocol). 6. Collect beads-RNA in the tube wall standing on the magnet 2 min. 7. Remove the supernatant. Wash beads-RNA three times with 0.5 ml binding buffer to eliminate unbound RNA, recovering each time the pellet on the tube wall standing on the magnet 2 min. 8. Resuspend the beads-RNA pellet in 10μl binding buffer (5). At this step, the RNA-bound to beads can be visualized following elution in binding buffer, heating at 92  C 2 min, and fractionation in denaturing acrylamide–urea gels (see Note 5). RNA should appear as a single band of full-length size (Fig. 2).

114

Javier Fernandez-Chamorro et al.

Fig. 1 Overview of the pull-down procedure. Schematic representation of a representative example RNA chimera (SL3a inserted into tRNA scaffold); proteins in total cell lysate are depicted in black (specific RBPs) and gray (unspecific-associated factors); for simplicity, only main steps are indicated

3.3.2 Block RNA-Bound Streptavidin Coated Magnetic Beads

This optional step may reduce unspecific binding (see Note 6). 1. Add 25μl BSA (10 mg/ml) in RNase-free water to 250μl (final volume), incubating 15 min in a rotating wheel at room temperature to decrease unspecific protein binding.

Pull-Down Assay Using Streptavidin-Aptamer Tagged RNAs

115

Fig. 2 Analysis of RNA capture by streptavidin-coated beads. Images of denaturing acrylamide gel loaded with control RNA (tRNA scaffold) and SL3a chimera in denaturing gels (8% acrylamide 7 M urea). Red arrows point to the RNA obtained by streptavidin purification (+); ribosomal RNA (rRNA and 5S RNA) are detected only in the input sample

2. Collect beads-RNA in the tube wall standing on the magnet 2 min and remove the supernatant. 3. Wash beads three times with binding buffer. 3.3.3 Binding of Proteins to RNA Chimera

1. Resuspend the beads-RNA pellet (optionally blocked with BSA) in 10μl binding buffer 5. Then, add 100μg cytoplasmic soluble proteins (about 6–12μl of S10 HeLa cell extract), 2 nM yeast tRNA, 1 mM DTT, and RNase-free water in 50μl (final volume). 2. At time 0 take aliquots (1%) as Input samples for subsequent comparisons. 3. Incubate the mixture of beads-RNA with soluble proteins 30 min in a rotating wheel, at room temperature. 4. Collect bead–RNA–protein complexes in the tube wall standing on the magnet 2 min and remove the supernatant. 5. Wash the bead–RNA–protein complexes three times with five volumes of binding buffer incubating 5 min at room temperature. 6. Add 25μl Laemmli buffer (3), and heat the samples 5 min at 95  C.

116

Javier Fernandez-Chamorro et al.

7. Analyze the eluted proteins from beads by electrophoresis in 10% SDS-PAGE using an aliquot (about 10% of the volume). 8. Silver stain the gel to compare the pattern obtained with the RNA chimera of interest with the control RNA (see Note 7). 3.4 Mass Spectrometry Identification of RNA-Eluted Factors

1. Send the remaining sample (about 90%) for mass spectrometry analysis (LC/MS-MS). It is important to analyze at least two independent biological replicates (see Note 7). 2. Take into consideration factors identified in both replicates with more than 2 unique peptides (FDR 10%) and control RNA subtraction

118

Javier Fernandez-Chamorro et al.

Fig. 5 Binding of SYNCRIP to SL3a RNA chimera. (a) Band-shift conducted with increasing concentration of purified His-SYNCRIP and a fixed amount of labeled transcript SL3a. To purify SYNCRIP, the coding region was transferred from pGEXKG-SYNCRIP [18] to pET28a(+) using standard procedures. The graph represents the adjusted curve curves obtained from the quantifications of the retarded complex relative to the free probe (mean  SD) from three independent assays. (b) SYNCRIP stimulates IRES-dependent translation. Cells were transfected with the indicated siRNAs, and 24 h later with a plasmid expressing luciferase in IRES-dependent manner. The level of SYNCRIP was determined by western blot in comparison to siRNAcontrol 48 h posttransfection. The effect of SYNCRIP silencing on protein synthesis was calculated as the % of Luciferase activity/μg of protein relative to the control siRNA. Asterisks denote statistically significant differences (P < 0.005) between cells treated with the siRNAcontrol and siSYNCRIP RNA

8. Experimental validation of the proteins identified is highly recommended. As representative example, we show the interaction of SYNCRIP (hnRNPQ) with the RNA chimera (Fig. 5).

Acknowledgments We thank L. Ponchon for reagents, J. Ramajo for technical assistance, and A. Escos for comments on the protocol. This work was supported by MINECO (grant BFU2017-84492-R), Comunidad de Madrid (B2017/BMD-3770) and an Institutional grant from Fundacio´n Ramo´n Areces. References 1. Hinnebusch AG (2017) Structural insights into the mechanism of scanning and start codon recognition in eukaryotic translation initiation. Trends Biochem Sci 42:589–611

2. Lozano G, Francisco-Velilla R, Martinez-Salas E (2018) Deconstructing internal ribosome entry site elements: an update of structural motifs and functional divergences. Open Biol 8(11):180155

Pull-Down Assay Using Streptavidin-Aptamer Tagged RNAs 3. Lozano G, Martinez-Salas E (2015) Structural insights into viral IRES-dependent translation mechanisms. Curr Opin Virol 12:113–120 4. Lozano G, Fernandez N, Martinez-Salas E (2016) Modeling three-dimensional structural motifs of viral IRES. J Mol Biol 428:767–776 5. Fernandez-Miragall O, Ramos R, Ramajo J, Martinez-Salas E (2006) Evidence of reciprocal tertiary interactions between conserved motifs involved in organizing RNA structure essential for internal initiation of translation. RNA 12:223–234 6. Jung S, Schlick T (2013) Candidate RNA structures for domain 3 of the foot-andmouth-disease virus internal ribosome entry site. Nucleic Acids Res 41:1483–1495 7. Yu Y, Abaeva IS, Marintchev A, Pestova TV, Hellen CU (2011) Common conformational changes induced in type 2 picornavirus IRESs by cognate trans-acting factors. Nucleic Acids Res 39:4851–4865 8. Sweeney TR, Abaeva IS, Pestova TV, Hellen CU (2014) The mechanism of translation initiation on type 1 picornavirus IRESs. EMBO J 33:76–92 9. Pineiro D, Fernandez N, Ramajo J, MartinezSalas E (2013) Gemin5 promotes IRES interaction and translation control through its C-terminal region. Nucleic Acids Res 41:1017–1028 10. Lee KM, Chen CJ, Shih SR (2017) Regulation mechanisms of viral IRES-driven translation. Trends Microbiol 25:546–561 11. Lozano G, Francisco-Velilla R, Martinez-Salas E (2018) Ribosome-dependent conformational flexibility changes and RNA dynamics of IRES domains revealed by differential SHAPE. Sci Rep 8:5545

119

12. Ponchon L, Beauvais G, Nonin-Lecomte S, Dardel F (2009) A generic protocol for the expression and purification of recombinant RNA in Escherichia coli using a tRNA scaffold. Nat Protoc 4:947–959 13. Fernandez-Chamorro J, Francisco-Velilla R, Ramajo J, Martinez-Salas E (2019) Rab1b and ARF5 are novel RNA-binding proteins involved in FMDV IRES-driven RNA localization. Life Sci Alliance 2(1):e201800131 14. Martinez-Salas E, Francisco-Velilla R, Fernandez-Chamorro J, Lozano G, DiazToledano R (2015) Picornavirus IRES elements: RNA structure and host protein interactions. Virus Res 206:62–73 15. Mi H, Huang X, Muruganujan A, Tang H, Mills C, Kang D, Thomas PD (2017) PANTHER version 11: expanded annotation data from gene ontology and Reactome pathways, and data analysis tool enhancements. Nucleic Acids Res 45:D183–D189 16. Francisco-Velilla R, Fernandez-Chamorro J, Lozano G, Diaz-Toledano R, Martinez-Salas E (2015) RNA-protein interaction methods to study viral IRES elements. Methods 91:3–12 17. Lopez de Quinto S, Martinez-Salas E (2000) Interaction of the eIF4G initiation factor with the aphthovirus IRES is essential for internal translation initiation in vivo. RNA 6:1380–1392 18. Francisco-Velilla R, Fernandez-Chamorro J, Ramajo J, Martinez-Salas E (2016) The RNA-binding protein Gemin5 binds directly to the ribosome and regulates global translation. Nucleic Acids Res 44:8335–8351

Chapter 10 Live Cell Imaging Using Riboswitch–Spinach tRNA Fusions as Metabolite-Sensing Fluorescent Biosensors Sudeshna Manna, Colleen A. Kellenberger, Zachary F. Hallberg, and Ming C. Hammond Abstract The development of fluorescent biosensors is motivated by the desire to monitor cellular metabolite levels in real time. Most genetically encodable fluorescent biosensors are based on receptor proteins fused to fluorescent protein domains. More recently, small molecule–binding riboswitches have been adapted for use as fluorescent biosensors through fusion to the in vitro selected Spinach aptamer, which binds a profluorescent, cell-permeable small molecule mimic of the GFP chromophore, DFHBI. Here we describe methods to prepare and analyze riboswitch–Spinach tRNA fusions for ligand-dependent activation of fluorescence in vivo. Example procedures describe the use of the Vc2-Spinach tRNA biosensor to monitor perturbations in cellular levels of cyclic di-GMP using either fluorescence microscopy or flow cytometry. In this updated chapter, we have added procedures on using biosensors in flow cytometry to detect exogenously added compounds. The relative ease of cloning and imaging of these biosensors, as well as their modular nature, should make this method appealing to other researchers interested in utilizing riboswitchbased biosensors for metabolite sensing. Key words Fluorescence imaging, Fluorescence microscopy, Flow cytometry, RNA biosensor, Cyclic dinucleotide, Cyclic di-GMP, Vc2 riboswitch, Spinach aptamer

1

Introduction Fluorescent biosensors provide a measurable output that changes in real time in response to binding of a specific small molecule, making them ideal tools for quantifying and tracking cellular signals in vivo. Most biosensors employed in live cell imaging utilize Fo¨rster resonance energy transfer (FRET) between two fluorescent proteins fused to a receptor protein, which undergo conformational changes in response to the target analytes [1]. However, many small molecule metabolites and signals exist for which no corresponding protein receptor is known. Furthermore, engineering a biosensor from a natural protein receptor that produces a large change in signal remains difficult.

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_10, © Springer Science+Business Media, LLC, part of Springer Nature 2021

121

122

Sudeshna Manna et al.

Riboswitches are natural RNA-based receptors for metabolites and signaling molecules that bind cognate ligands with high selectivity and affinity, making them strong candidates as biosensor scaffolds [2]. Previously, reporter systems have been developed using riboswitches to control the expression of a downstream reporter gene such as β-galactosidase or GFP [3]. These systems have the advantage of providing signal amplification, but the readout of metabolite concentrations is indirect and not responsive in real time. The recent development of the Spinach aptamer [4] has enabled the generation of RNA-based fluorescent biosensors that function in vivo. Spinach is an in vitro selected RNA aptamer that binds to 3,5-difluoro-4-hydroxybenzylidene imidazolinone (DFHBI), a profluorescent small molecule analog of the GFP chromophore. The Spinach aptamer forms a G-quadruplex motif structure that binds DFHBI and shields the chromophore from solvent- and rotation-mediated forms of energetic decay, causing a 1000-fold increase in fluorescence [5, 6]. Furthermore, it was shown that DFHBI binding and fluorescence activation by Spinach can be modulated by controlling the formation of the P2 stem with a small molecule-binding aptamer fused to this stem [7]. Following this strategy, we used a natural riboswitch aptamer to develop a first-generation fluorescent biosensor for cyclic-di-GMP [8], which is a key second messenger in bacteria that regulates processes including biofilm formation, cell cycle progression, and virulence [9]. A series of second-generation biosensors for cyclic di-GMP was then developed using the improved Spinach2 aptamer [10] and a phylogenetic screening strategy [11]. Recently, we showed that a second-generation biosensor was able to monitor the dynamic rise of cyclic di-GMP levels in Escherichia coli upon zinc depletion [12]. Expression plasmids encoding these biosensors, the Spinach2 tRNA control, and cyclic di-GMP-related enzymes have been deposited in Addgene (search Ming Hammond Lab Plasmids). Here we describe the application of an RNA-based biosensor for live cell imaging of the bacterial second messenger cyclic di-GMP. We demonstrate two methods, fluorescence microscopy and flow cytometry, to analyze changes in cyclic di-GMP levels in E. coli upon coexpression of the biosensor and diguanylate cyclase enzymes that synthesize cyclic di-GMP in the cells. In this update to our original 2015 chapter [13], we have added procedures on using biosensors in flow cytometry to detect exogenously added compounds [14]. These procedures are expected to be generalizable for other Spinach-based biosensors developed for different cellular metabolites (Table 1 provides a comprehensive list of in vivo RNA-based biosensors) [7, 8, 11, 15–25]. For additional information, please refer to our recent review on the development of RNA-based biosensors for imaging small molecules and RNAs in vivo [26].

Live Cell Imaging Using RNA-Based Biosensors

123

Table 1 Summary of RNA-based biosensors for sensing of small molecules Fluorogenic aptamer used

Target molecules

Dye used

Year/ Reference

ATP, theophylline, FMN

Malachite green-binding aptamer

Malachite green

2004 [14]

ADP, GTP

Blue fluorescent RNA (BFR)

Bisbenzimidazole chromophore

2010 [15]

S-adenosyl-L-methionine (SAM), ADP, GTP, guanine, adenosine

Spinach

DFHBI

2012 [7]

Cyclic di-GMP, cyclic GMP-AMP (30 ,30 -cGAMP)

Spinach

DFHBI

2013 [8]

Cyclic GMP-AMP (30 ,30 -cGAMP)

Spinach

DFHBI

2015 [16]

Cyclic di-AMP

Spinach2

DFHBI

2015 [17]

Thiamine pyrophosphate (TPP), SAM, guanine, adenine

Spinach

DFHBI

2015 [18]

S-adenosyl-L-homocysteine (SAH)

cpSpinach2

DFHBI

2016 [19]

Mammalian cyclic GMP-AMP (20 ,30 -cGAMP)

Spinach2

DFHBI

2016 [20]

Cyclic di-GMP

Spinach2

DFHBI

2016 [11]

5-hydroxy-L-tryptophan (5-HTP), 3,4-dihydroxy-L-phenylalanine (L-DOPA)

Broccoli

DFHBI-1T

2017 [21]

SAM

Spinach and mango

DFHBI and thiazole orange 3

2018 [22]

Theophylline, cyclic-di GMP, cyclic-di-AMP

Broccoli

DFHBI-1T

2019 [23]

Tetracycline

DFHBI-1T and Broccoli and DNB sulforhodamine (dinitroaniline-binding B-dinitroaniline aptamer)

2020 [24]

Guanidine

Spinach2

2021 [14]

2

DFHBI-1T

Materials

2.1 Equipment and Supplies

1. Micropipettor. 2. Vortex mixer. 3. Microcentrifuge. 4. Millipore water filter with a BioPak unit. 5. Micropipettor tips.

124

Sudeshna Manna et al.

6. 1.5 mL microcentrifuge tubes. 7. Sterile filter units. 8. PCR thermocycler. 9. 14 mL culture tubes and petri dishes. 10. VWR 96-well deep well plates (2.2 mL/well). 11. Incubator shaker set to 37  C. 12. 96 well round bottom plates (330 μL/well). 13. VWR VistaVision Cover Glasses No. 1 ½. 14. VWR VistaVision Microscope Slides. 15. Zeiss 200 M AxioVert microscope (Zeiss, Jena, Germany) equipped with a mercury light source X-Cite 120 Series (Exfo Life Science Divisions, Ontario, Canada), a 63/1.4 PlanApochromat oil DIC objective lens and a 1.6 tube lens. For monitoring fluorescence, a GFP filter set with an excitation 470/40 BP, FT 495 beamsplitter, and emission 525/50 BP was used (Filter Set 38 HE). 16. ImageJ data analysis software. 17. FlowJo software for flowcytometry data analysis. 18. 5 mL Falcon polypropylene tubes with cell-strainer cap. 19. BD Influx v7 cell sorter with BD FACS Software (Version 1.0.0.650). 2.2

Reagents

1. Phusion DNA polymerase with 5 HF and 5 GC buffers (NEB). 2. 10 dNTPs (2 mM each ATP, CTP, GTP, TTP). 3. 1 TAE buffer: 40 mM Tris, 20 mM acetic acid, 1 mM EDTA, pH 8.4. 4. 1% agarose solution: 1 g agarose in 100 mL 1 TAE buffer. 5. 2-log DNA ladder (NEB). 6. Gel extraction or PCR cleanup kits. 7. Restriction enzymes: BglII and XhoI with NEBuffer 3.1 or EagI-HF and SacII with NEB CutSmart buffer along with NdeI and XhoI with NEB CutSmart buffer. 8. Calf intestinal alkaline phosphatase (Fisher). 9. T4 DNA ligase with 10 ligase buffer (NEB). 10. Luria Broth (LB, Fisher Scientific). 11. E. coli TOP10 chemically competent cells (Life Technologies). 12. LB/agar (EMD Millipore). 13. Carbenicillin: 50 mg/mL stock concentration, filtered through a 0.2 μm nitrocellulose filter.

Live Cell Imaging Using RNA-Based Biosensors

125

14. Kanamycin: 50 mg/mL stock concentration, filtered through a 0.2 μm nitrocellulose filter. 15. pET31b(+) plasmid. 16. pET24a or pCOLADuet-1 plasmid. 17. 40 μM stock concentration of the following DNA oligonucleotide primers (BOLD ¼ T7 promoter; UNDERLINED ¼ tRNA scaffold; CAPS ¼ Spinach sequence; BOLD UND ERLINED ¼ T7 terminator; ITALICS ¼ restriction enzyme recognition site; lower case ¼ other): (a) 50 -cgatcccgcgaaatTAATACGACTCACTATAGGGGCC CGGATAGCTCAGTCGGTAGAGCAGCGGCCGGAC GCGACTGAATGAAATGGTGAAGGACGGG (b) 50 -AGAGGCCCCAAGGGGTTATGCTATGGCGCC CGAACAGGGACTTGAACCCTGGACCCGCGGCCG GACGCGACTAGTTACGGAGCTCACACTCTACTC (c) 50 -cagtcaAGATCT cgatcccgcgaaatTAATACGACTCAC TATAGGG (d) 50 -catcagCTCGAG CAAAAAACCCCTCAAGACCCG TTTAGAGGCCCCA AGGGGTTATGCTA (e) 50 -gatcCGGCCGGACGCGACTGAATGAAATGGTG. (f) 50 -catgCCGCGGCCGGACGCGACTAGTTACGGAGC TC. (g) 50 -cttgCATATGcacaaccctcatgagagcaag. (h) 50 -catgCTCGAGtcagcccgccggggc. 18. E. coli BL21 (DE3) Star chemically competent cells (Life Technologies). 19. Borate buffer: 2.1 g/L boric acid, 9 g/L sodium tetraborate decahydrate; pH to 8.5. Filter through a 0.2 μm nitrocellulose filter and store at ambient temperature. 20. 1 M HCl solution. 21. 4 Poly-D-Lysine solution: 2 mg/mL poly-D-lysine hydrobromide (Sigma, CAS: 27964-9904). Filter through a 0.2 μm nitrocellulose filter and store at 20  C. 22. Ethanol. 23. M9 minimal media, pH 7.0: 48 mM Na2HPO4, 22 mM KH2PO4, 8.6 mM NaCl, 20 mM NH4Cl, 5 mM MgSO4, 0.4% glucose, 100 μM CaCl2. 24. Non inducing (NI) media: 1 mM MgSO4, 0.5% glucose, 1 NPS (25 mM NH4SO4, 50 mM KH2PO4, 50 mM Na2HPO4, pH 6.75) [27].

126

Sudeshna Manna et al.

25. Auto induction (AI) media (200 mL): 186 mL ZY media (10 g tryptone and 5 g yeast extract in 925 mL water), 1 mM MgSO4,1 5052 (0.5% glycerol, 0.05% glucose, 0.2% α-lactose), 1 NPS (25 mM NH4SO4, 50 mM KH2PO4, 50 mM Na2HPO4, pH 6.75) [27]. 26. DFHBI (3,5-difluoro-4-hydroxybenzylidene imidazolinone): stock concentration of 20 mM DFHBI in DMSO, stored in 50 μL aliquots at 20  C. 27. DFHBI-1T((Z)-4-(3,5-difluoro-4-hydroxybenzylidene)-2methyl-1-(2,2,2-trifluoroethyl)-1H-imidazol-5(4H)-one): stock concentration of 50 mM DFHBI-1T in DMSO stored in 50 μL aliquots at 20  C [28]. 28. 1 PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4 (Fisher).

3

Method

3.1 Construction and Transformation of Biosensor and Enzyme Constructs

In these experiments, two separate plasmids are used to separately express the biosensor or the enzyme (see Note 1, Addgene ID #79158 through 79167). The Vc2-Spinach tRNA biosensor is cloned into pET31b and alleles of the WspR diguanylate cyclase from Pseudomonas fluorescens [29] are cloned into pET24a or pCOLADuet-1. Previous reports have shown that plasmids containing the same origin of replication (e.g., pET vectors) and different antibiotic resistances are incompatible for coexpression as both are not stably maintained in the cell due to competition for the same replication factors [30]. On the contrary, recent reports have shown that the degree of incompatibility depends on the expression system and that plasmids with the same origin can stably persist for several overnight growth cycles [31]. We have found that both incompatible plasmids (pET31b and pET24a) as well as compatible plasmids (pET31b and pCOLADuet-1) function in these experiments to detect altered levels of cdiG, but it is advised to use compatible expression vectors for highly sensitive experiments in order to minimize plasmid loss during bacterial growth. Though the expression of both the biosensor and enzyme are driven by IPTG-induction in the following experiments, use of the RNA-based biosensor is highly adaptable for many experimental systems. For instance, various promoters can be cloned directly upstream of the Vc2-Spinach tRNA construct to achieve either constitutive or inducible biosensor expression in E. coli or other bacterial species. Various enzyme alleles, gene homologs, or mutant enzyme libraries can be coexpressed to assess in vivo enzymatic activity. Furthermore, the biosensor can be expressed alone to study cyclic dinucleotide levels in different strains, in bacterial transposon libraries, or in response to chemical inputs.

Live Cell Imaging Using RNA-Based Biosensors

127

Fig. 1 Design of RNA-based fluorescent biosensor and enzyme constructs. (a) Model of Vc2-Spinach tRNA construct bound to cdiG and DFHBI. The tRNA scaffold serves to stabilize the biosensor fold in vivo. (b) Diagram of biosensor and enzyme expression plasmid constructs 3.1.1 Cloning of Biosensor Expression Vector

For stable expression in vivo, the riboswitch-Spinach fusion, the construct used for in vitro experiments, is inserted within a tRNALys3 scaffold to reduce susceptibility to RNases and to ensure homogeneous end processing in cells (Fig. 1a) [32]. In vitro analysis has confirmed that placement of the Vc2-Spinach construct within the tRNA scaffold has little to no effect on the ligand binding affinity [8]. For initiation and termination of transcription, respectively, the Vc2-Spinach tRNA construct is flanked by T7 promoter and T7 terminator sequences. The existing T7 promoter sequence in the expression plasmid pET31b is removed during the restriction digest to ensure homogenous expression of the RNA from a single promoter site (see Note 2). Two procedures for constructing the biosensor expression plasmid are described below. In the first case, the Vc2-Spinach tRNA insert is generated through PCR and cloned into the original pET31b vector (see Note 3). Alternatively, the Vc2-Spinach biosensor can be cloned into a pET31b plasmid already containing the tRNA scaffold (Fig. 1b). The second strategy is recommended when the user already has a biosensor expression plasmid and is interested in swapping out the specific biosensor construct for another. Shown below is the sequence of the Vc2-Spinach tRNA biosensor construct (italics ¼ restriction enzyme recognition site; BOLD ¼ T7 promoter; UNDERLINED ¼ tRNA scaffold; CAPS ¼ Spinach sequence; lowercase ¼ Vc2 sequence; BOLD ITALICS ¼ T7 terminator): (a) agatctCGATCCCGCGAAATTAATACGACTCACTATAGG GGCCCGGATAGCTCAGTCGGTAGAGCAGCGGCCGGA CGCGACTGAATGAAATGGTGAAGGACGGGTCCAcacgca

128

Sudeshna Manna et al.

cagggcaaaccattcgaaagagtgggacgcaaagcctccggcctaaaccagaagacat ggtaggtagcggggttaccgatgTTGTTGAGTAGAGTGTGAGCTC CGTAACTAGTCGCGTC CGGCCGCGG GTCCAGGGTTC AAGTCCCTGTTCGGGCGCCA TAGCATAACCCCTTGG GGCCTCTAAACGGGTCTTGAGGGGTTTTTTGctcgag. Procedure Starting from Original pET31b Vector Step 1a—Generation of Vc2-Spinach tRNA Construct (Estimated Time: 2 h)

1. To prepare the tRNA-scaffold biosensor insert, combine the following in a PCR tube: (a) 10 μL 5 Phusion HF buffer. (b) 5 μL 10 dNTPs. (c) 0.8 μL 40 μM primer a. (d) 0.8 μL 40 μM primer b. (e) 1 μL Vc2-Spinach DNA template (10–50 ng) (see Note 4). (f) 31.4 μL ddH2O. (g) 1 μL Phusion DNA polymerase. 2. Amplify the construct using the following standard thermocycler protocol: initial denaturation 98  C for 1 min; 35 cycles of 98  C for 5–10 s, 66  C for 20 s, 72  C for 15 s, final extension 72  C for 5 min. 3. Analyze the product on a 1% agarose gel (expected size: 290 bp). Purify the product via commercial gel extraction or PCR cleanup kits following the manufacturer’s protocol and elute the product in ddH2O or the provided elution buffer.

Step 2a—Generation of Vc2-Spinach tRNA Construct with T7 Promoter and Restriction Sites (Estimated Time: 2 h).

1. Complete construction of the biosensor construct containing restriction enzyme recognition sites by combining the following in a PCR tube: (a) 10 μL 5 Phusion HF buffer. (b) 5 μL 10 dNTPs. (c) 0.8 μL 40 μM primer c. (d) 0.8 μL 40 μM primer d. (e) 1 μL Vc2-Spinach-tRNA DNA template (10–50 ng). (f) 31.4 μL ddH2O. (g) 1 μL Phusion DNA polymerase. 2. Amplify the construct using the same protocol as in Subheading “Step 1a—Generation of Vc2-Spinach tRNA Construct (Estimated Time: 2 h)”, step 2 above, but with an annealing temperature of 68  C. 3. Analyze and purify the product (expected size: 338 bp) as in Subheading “Step 1a—Generation of Vc2-Spinach tRNA Construct (Estimated Time: 2 h)”, step 3 above.

Live Cell Imaging Using RNA-Based Biosensors Step 3a—Cloning of Biosensor Insert into Expression Vector (Estimated Time: 2 Days)

129

1. To insert the Vc2-Spinach tRNA construct into the expression plasmid pET31b, perform a double digest on both the insert and the vector using BglII and XhoI restriction enzymes. Prepare 50 μL digestion reactions by combining either 43 μL of the complete Vc2-Spinach-tRNA construct or pET31b plasmid with 5 μL NEBuffer 3.1, 1 μL BglII, and 1 μL XhoI. 2. Incubate the reaction at 37  C for 2 h. 3. To reduce plasmid religation, dephosphorylate the digested pET31b vector by adding 1 μL calf intestinal alkaline phosphatase to the digested reaction and incubate at 55  C for 30 min. 4. Check the digested and dephosphorylated products on a 1% agarose gel to ensure the integrity of the products and to check that the plasmid was linearized. Purify the products via commercial PCR cleanup kit following the manufacturer’s protocol and elute the DNA in ddH2O or the provided elution buffer. 5. Ligate the digested products by combining the following in a PCR tube: 1 μL 10 T4 DNA ligase buffer, 25 ng digested/ dephosphorylated plasmid, 20 ng digested insert, ddH2O to 9.5 μL, and 0.5 μL T4 DNA ligase. Incubate the reaction for 10 min at room temperature or at 16  C overnight. 6. Deactivate the ligase by heating it to 65  C for 10 min. 7. Use 1–5 μL of the ligation reaction to transform 50 μL E. coli TOP10 chemically competent cells following the manufacturer’s protocol. Plate the cells on LB/Carb (50 μg/mL Carb) plates and incubate 12–16 h at 37  C. 8. Check transformed colonies for clones with the correct sequence by streaking single colonies on a LB/Carb master plate. Inoculate overnight LB/Carb cultures from the master plate, then isolate the plasmids using a commercial kit and submit the plasmids for sequencing. Use only a plasmid containing the desired sequence for future experiments.

Alternative Procedure Starting from Existing Biosensor Construct Step 1b—Generation of Biosensor-Spinach Insert

1. Amplify the Vc2-Spinach insert with EagI and SacII restriction sites by combining the following in a PCR tube: (a) 10 μL 5 Phusion HF buffer. (b) 5 μL 10 dNTPs. (c) 0.8 μL 40 μM primer e. (d) 0.8 μL 40 μM primer f. (e) 1 μL Vc2-Spinach-tRNA DNA template (10–50 ng). (f) 31.4 μL ddH2O. (g) 1 μL Phusion DNA polymerase. 2. Amplify the construct using the following thermocycler protocol: Initial denaturation 98  C for 1 min; 35 cycles of 98  C

130

Sudeshna Manna et al.

5–10 s, 66  C for 20 s, 72  C for 15 s, final extension 72  C for 5 min. 3. Check the product on a 1% agarose gel (expected size: 175 bp) and purify the product via commercial PCR cleanup or gel extraction kits following the manufacturer’s protocol and elute the DNA in ddH2O or the provided elution buffer. Step 2b—Cloning of Biosensor Insert into Expression Vector (Estimated time: 1 Day)

1. Digest the pET31b-Spinach-containing plasmid and the Vc2-Spinach insert by combining either 43 μL of the Vc2-Spinach construct or pET31b-Spinach plasmid with 5 μL NEB CutSmart buffer, and 1 μL SacII. Incubate the reaction at 37  C for 1 h, then add 1 μL EagI-HF and digest at 37  C for another hour. It is necessary to do a sequential digest since first cutting with SacII eliminates the second EagI restriction enzyme recognition site. 2. Follow Subheading “Step 3a—Cloning of Biosensor Insert into Expression Vector (Estimated Time: 2 days)”, steps 3– 8 above to ligate, transform, and sequence to confirm the pET31b Vc2-Spinach-tRNA biosensor construct.

3.1.2 Construction of Enzyme Expression Vector Step 4—Generation of WspR Diguanylate Cyclase Insert (Estimated Time: 3 h)

1. Amplify the WspR diguanylate cyclase insert by combining the following in a PCR tube: (a) 10 μL 5 Phusion GC buffer. (b) 5 μL 10 dNTPs. (c) 0.8 μL 40 μM primer g. (d) 0.8 μL 40 μM primer h. (e) 1 μL WspR DNA template (10–50 ng). (f) 3 μL DMSO. (g) 28.4 μL ddH2O. (h) 1 μL Phusion DNA polymerase. 2. Amplify the construct using the following thermocycler protocol: Initial denaturation 98  C for 1 min; 35 cycles of 98  C 5–10 s, 68  C for 20 s, 72  C for 45 s, final extension 72  C for 10 min (see Note 5). 3. Check the amplified product on a 1% agarose gel (expected size: 1061 bp) and purify the product by commercial PCR cleanup or gel extraction kits following the manufacturer’s protocol and elute the DNA in ddH2O or the provided elution buffer.

Step 5—Cloning of WspR Diguanylate Cyclase into Expression Vector (Estimated Time: 2 Days)

1. Digest the amplified WspR insert along with pET24a or pCOLADuet-1 plasmid by combining the following: 43 μL of either the WspR insert or the plasmid with 5 μL NEB CutSmart buffer, 1 μL NdeI, and 1 μL XhoI. Incubate the reaction at 37  C for 2 h.

Live Cell Imaging Using RNA-Based Biosensors

131

2. Follow Subheading “Step 3a—Cloning of Biosensor Insert into Expression Vector (Estimated Time: 2 days)”, steps 3– 8 in Subheading 3.1.1, but with cultures grown in LB/Kan (50 μg/mL) in order to obtain sequence-confirmed pET24a or pCOLADuet-1 WspR plasmid. 3.1.3 Transformation of Expression Vectors for Live Cell Imaging

1. To prepare cells coexpressing the biosensor and enzyme, transform BL21 (DE3) Star E. coli cells with ~60 ng each of the pET31b and pET24a constructs following the manufacturer’s protocol.

Step 6a—Generation of Strains Containing Both Biosensor and Enzyme Constructs (Estimated Time: 1 Day)

2. Plate the cells on LB/Carb/Kan (Carb: 50 μg/mL, Kan: 50 μg/mL) plates and incubate at 37  C for 12–16 h. Colonies should contain both plasmids and the plates can be stored at 4  C for ~3 weeks.

Step 6b—Alternative Procedure for Generation of Strains Containing Biosensor Only (Estimated Time: 1 Day)

1. To prepare cells expressing the biosensor, transform BL21 (DE3) Star E. coli cells with ~60 ng of the pET31b construct following the manufacturer’s protocol.

3.2 Live Cell Imaging of the RNA-Based Biosensor

In general, cellular imaging with Spinach-based biosensors requires that several conditions be met, including that DFHBI can diffuse into the cell and that the RNA biosensor is expressed at sufficient levels for visualization. The permeability of DFHBI depends upon the composition of the cell membrane or cell wall and thus may fluctuate between different bacterial species and strains or under different growth conditions. Similarly, RNA expression levels may vary between different conditions or stressors on the cell (e.g., coexpression of a heterologous enzyme). Thus, to ensure that observed fluorescence changes are due to changes in metabolite or signaling molecule levels rather than changes in intracellular DFHBI and RNA concentrations, it is recommended that the Spinach aptamer alone be used as a control (Addgene ID: #79783). If cellular fluorescence of the Spinach-tRNA construct with DFHBI and RNA expression levels remain constant under the experimental conditions, then differences in biosensor fluorescence under these conditions should provide an accurate readout of changing metabolite levels. Two distinctly advantageous techniques for live cell imaging of the fluorescent biosensor include fluorescence microscopy and flow cytometry. For the former, microscopy allows for direct visualization of cell morphology, for tracking cellular fluorescence over time, and for monitoring spatial dynamics of fluorescence (Fig. 2a). Generally, any fluorescence microscope with power to

2. Plate the cells on LB/Carb (50 μg/mL) plates and incubate at 37  C for 12–16 h. Colonies should contain the plasmid and the plates can be stored at 4  C for ~3 weeks.

132

Sudeshna Manna et al.

a

WT Vc2-Spinach DIC

M1 Vc2-Spinach

Fluorescence

DIC

Fluorescence

b

40

p < 0.001 p < 0.001

Mean Fluorescence Intensity

D70E WspR

WT WspR

G249A WspR

35

G249A WspR

30

WT WspR

20

Vc2

p < 0.001 p < 0.001

15 10 5 0

DIC

D70E WspR

25

WT Vc2Spinach

M1 Vc2Spinach

Vc2

Fluorescence

D70E WspR

Fig. 2 The Vc2-Spinach biosensor detects cdiG through fluorescence microscopy of live E. coli. (a) Differential interference contrast (DIC) and fluorescence microscopy images are shown of representative E. coli expressing variants of the Vc2-Spinach biosensor and the WspR cdiG synthase. Scale bars represent 10μm. (b) Quantitation of mean fluorescence intensity from fluorescence microscopy experiments. Error bars represent the SEM for at least 50 quantified cells. Indicated p values were calculated using a student’s t-test

a

b 80 Empty vector

80

WT WspR 60 40 20

Mean Fluorescence Intensity

Normalized Cell Count

100

70 60 Empty vector

50

WT WspR

40 30 20 10 0

0 100

101 102 103 Fluorescence

104

WT Vc2Spinach

Fig. 3 The Vc2-Spinach biosensor detects cdiG through flow cytometry of live E. coli. (a) Flow cytometry data of cells coexpressing the WT Vc2-Spinach biosensor along with empty vector or WT WspR. (b) Quantitation of mean fluorescence intensity from flow cytometry experiments. Error bars represent the standard deviation of independent biological replicates of at least 35,000 cells

resolve individual bacterial cells can be used, but analysis of results with the currently available software is time-consuming. On the contrary, flow cytometry or fluorescence activated cell sorting (FACS) allow for rapid and high-throughput analysis of thousands of cells (Fig. 3a). These flow-based methods provide a snapshot of

Live Cell Imaging Using RNA-Based Biosensors

133

bacterial characteristics of a large population at a specific time, and the ability to sort cells exhibiting differential fluorescent outputs. However, flow techniques are not suited for tracking the behavior of individual cells over time. Nevertheless, both experimental techniques have proven robust and reproducible using RNA-based biosensors. 3.2.1 Preparation of Poly-D-Lysine Coverslips for Fluorescence Microscopy

Step 7—Acid Rinse of Coverslips (Estimated Time: 11 h)

In performing live cell fluorescence microscopy experiments, the coverslips used for adhering the cells are first prepared and can be stored at room temperature for several months. Cells will adhere to the polylysine-coated coverslips through electrostatic interactions between the positive charge of the lysine residues on the coverslips and the negative charge of the bacterial cell. Specially designed plates and other polylysine-coated materials can alternatively be ordered from commercial vendors such as MatTek. 1. To coat the coverslips with a negative charge, soak an adequate number of coverslips in 1 M HCl and heat the solution to ~50  C for 1 h. Let the solution cool to ambient temperature for 2–10 h with occasional mixing. 2. Carefully remove the coverslips from the acid solution, then rinse the coverslips twice with ddH2O and once with 100% ethanol. Let the coverslips dry at ambient temperature for at least 1 h. Take care to neutralize the acid solution and to follow proper waste disposal guidelines.

Step 8—Poly-D-Lysine Rinse of Coverslips (Estimated Time: 11 h)

1. To coat the coverslips with poly-D-lysine, dilute 100 μL 4 poly-D-Lysine solution with 300 μL ddH2O. Let the coverslips soak in this solution in a petri dish for 1–10 h on a rocker at ambient temperature. 2. Rinse the coverslips carefully twice with ddH2O water and once with 100% ethanol. Remove the coverslips from the petri dish and dry them on filter paper for at least 1 h at ambient temperature. Coverslips can be stored at room temperature for ~3 months.

3.2.2 Fluorescent Microscopy Experiments

1. Inoculate a 3 mL LB/Carb/Kan culture with a single colony of the transformed BL21 star cells. Incubate the culture at 37  C in an incubator/shaker for 12–16 h or overnight.

Step 9—Bacterial Growth and Induction Conditions (Estimated Time: 16–20 h)

2. In the morning, inoculate a fresh 3 mL LB/Carb/Kan culture with 200 μL of overnight culture. Grow the cells in a 37  C incubator shaker and monitor growth by measuring the OD600. 3. Once the cells have reached an OD600 0.4–0.6 (approx. 2 h), induce the expression of the biosensor and plasmid with

134

Sudeshna Manna et al.

isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final concentration of 1 mM. Continue growing the cells in a 37  C incubator shaker for 2.5 h. Step 10—Harvesting Cells and Preparation of Slides (Estimated Time: 3.5 h)

1. To collect cells and remove LB, pellet 100 μL cells at 4500 rcf for 3 min at ambient temperature. Remove the supernatant and resuspend the cell pellet in 500 μL M9 minimal media, pH 7.0. Wash the cells of LB by centrifuging again, removing the supernatant, and resuspending in another 200 μL M9 minimal media, pH 7.0 (see Note 6). 2. Adhere the suspended cells to the prepared poly-D-lysine coverslips by pipetting the cells onto each coverslip and allow the cells to attach for ~30 min at 37  C. 3. To remove any unadhered cells, gently rinse the coverslips with 3 mL M9 minimal and aspirate off the solution. 4. Add 200 μL of 200 μM DFHBI in M9 minimal media, pH 7.0 to the cells and incubate the cells at 37  C for ~1–1.5 h to allow DFHBI to permeate the membrane and to bind the biosensor RNA. Keep the cells in the dark from this point on as DFHBI is light-sensitive. 5. Remove excess liquid by aspiration and place the coverslip on top of a microscope slide. Seal the edges of the coverslip with nail polish to prevent drying of the cells.

Step 11—Imaging Cells Using Fluorescence Microscopy (Estimated Time: 1 h)

1. To image cells via microscopy, first place the brightest expected sample on the microscope platform and focus on the cells using bright field imaging. Determine the fluorescence exposure to be used by maximizing the exposure time to increase fluorescence signal without overexposing any cells (see Note 7). Note this fluorescence exposure setting and revert to bright field exposure while fluorescence imaging is not active. 2. Image all samples by first taking a snapshot of cells in the bright field, then switch to green fluorescence filters using the previously determined exposure time and take a snapshot of the same field visualized by fluorescence. For each sample, repeat this procedure for at least three different viewing sections, where at least a total of 50 individual cells can be visualized. Follow this procedure for all samples of interest and save all images as files that can be opened in ImageJ (e.g., tiff files).

Step 12—Analysis of Fluorescence Microscopy Data (Estimated Time: 3 h)

1. To normalize all microscope images to the brightest sample of the set, open the fluorescent tif file in ImageJ of the brightest sample tested. In the adjustment category of the image menu, select Brightness/Contrast and set it to Auto. Note the maximum displayed value of the 8-bit image as this will be used to set the maximum brightness of all images.

Live Cell Imaging Using RNA-Based Biosensors

135

2. To determine the background fluorescence of each sample, which varies based upon the amount of DFHBI left on the coverslip, outline at least four areas of the image that contain no cells and use the measurement tool to determine the median fluorescence for each area. The average median fluorescence of the four areas (typically  2 units within the minimum fluorescence set by the Auto function) is the background fluorescence of the image. 3. Adjust each image individually by setting the minimum displayed value to the background fluorescence found in step 2, and the maximum displayed value to the maximum brightness found in step 1. Click ok and then apply within the Brightness and Contrast tool bar. Complete this process for all images to be analyzed. 4. To determine the mean fluorescence of each cell, individual cells will be manually outlined in the bright field image and then overlaid on the corresponding fluorescence image. Open the bright field image in ImageJ then manually outline each cell using the heart-shaped freehand selection tool. Hold down the Shift key while outlining to keep all previously outlined cells selected and periodically save the outlined cells as regions of interest (ROIs). If at any point an error is made during this process, click on the background of the image to clear the mistake, then select the last object in the ROI manager to redo that step. 5. Once all cells in the bright field image have been outlined, delete all ROI entries except for the last object containing all outlined cells. Open the fluorescence image, then select the object in the ROI manager to have the outlines appear on this image. Some adjustment may be necessary if the camera shifted between taking snapshots of the bright field and fluorescence images. If so, simply move the ROI selections using the arrow keys. Save the image as a separate file by overlaying the ROI selections from the ROI manager. 6. Obtain the median fluorescence for each outlined cell by selecting to Split the combined object in the ROI manager. This separates each outlined object, and the fluorescence statistics can be obtained by highlighting the individual cells in the ROI manager and selecting Measure. 7. Record the obtained cell statistics and also measure the background fluorescence of four areas of the image that have no cells. Determine the mean fluorescence intensity of each cell by subtracting this background from the mean fluorescence of each cell. For each sample tested, determine the fluorescence

136

Sudeshna Manna et al.

of at least 50 cells to calculate the average mean fluorescence intensity for this sample. Assess the statistical significance of the differences in sample fluorescence using a student’s t test (Fig. 2b). 3.2.3 Flow Cytometry Experiments

Biosensor and enzyme expression can be induced with either IPTG or lactose in autoinduction (AI) media. We have generally observed more consistent cell growth and expression between biological replicates in AI media following inoculation in noninducing (NI) media. Therefore, it is recommended to use the AI induction protocol for experiments involving biosensor response to exogenous analytes, which usually result in less fluorescence changes than enzyme overexpression experiments.

Step 13a—Growth of Cells with IPTG Induction for Flow Cytometry (Estimated: 16–18 h)

1. To prepare cells for flow cytometry, follow Subheading “Step 10—Harvesting Cells and Preparation of Slides (Estimated Time: 3.5 h)” as discussed in Subheading 3.2.2 (see Note 8).

Step 14a—Preparation of Cells for Flow Cytometry (Estimated Time: 1 h)

1. To collect cells and remove LB, pellet 250 μL cells at 4500 rcf for 3 min at ambient temperature. Remove the supernatant and resuspend the pellet in 500 μL 1 PBS (see Notes 6 and 9). 2. Add 1–5 μL of the suspended cells to 250 μL of the 25 μM DFHBI in 1 PBS. Assuming that OD600 ¼ 1 corresponds with approximately 1  109 cells/mL, this should give a final concentration of between 3000 and 15,000 cells/μL. 3. Filter the suspension through the cell-strainer caps into polypropylene tubes for analysis.

Step 13b—Alternative Procedure for Growth of Cells with Autoinduction Media for Flow Cytometry (Estimated: 42 h)

Step 14b—Alternate Procedure for Treatment of Cells with Exogenous Analytes for Flow Cytometry (Estimated Time: 25 min)

1. Inoculate a 0.5 mL NI media/Carb (50 μg/mL) culture with a single colony of the transformed BL21 star cells. Incubate the culture at 37  C in an incubator/shaker for 24 h until OD600 reaches 3.0–4.0 (see Note 10). 2. In the morning, inoculate a fresh 3 mL AI media/Carb (50 μg/mL) culture with 30 μL of overnight NI culture. Grow the cells in a 37  C incubator shaker for 16–18 h until the OD600 reached to 3.0–4.0. 1. To preincubate cells with dye, add 2 μL of overnight AI culture cells to 98 μL of 1 PBS containing 50 μM DFHBI-1T in each well of a 96 well plate (330 μL/well). Mix with a micropipette and incubate for 10 min to allow for equilibration. Typically, between 3 and 4 replicates are prepared for each condition and analyzed.

Live Cell Imaging Using RNA-Based Biosensors

137

2. Add 2 μL of metabolite/analyte stock solution (concentration is 50 desired final concentration) to the cell solutions. Incubate for 15 min to measure the endpoint response. 3. Prepare control samples by adding 2 μL of water instead to the cell solutions and incubate for the same amount of time (15 min) (see Note 11). Step 15—Analysis of Cells by Flow Cytometry (Estimated Time: 1 h)

1. For future analysis of cells via flow cytometry, establish the appropriate gating settings using both positive and negative fluorescence controls. To do this, first establish the forward scatter (FSC) and side scatter (SSC) regions corresponding to cells by first analyzing DFHBI-PBS solution alone, followed by a solution containing cells in DFHBI-PBS solution. 2. To optimize the voltage gain settings, pass the negative control—either cells expressing no dinucleotide cyclase or a nonfunctional biosensor—through the analyzer, and compare to the positive control with cells coexpressing the Vc2-Spinach biosensor with WT WspR. 3. Analyze all samples by collecting data for at least 10,000 cells within the gating window (see Notes 12 and 13).

Step 16—Analysis of Flow Cytometry Data (Estimated Time: 1 h)

1. Using FlowJo software, generate a gate from the FSC/SSC dot plot to analyze data points solely corresponding to cells and limiting the amount of debris analyzed. Apply this same gate to all samples tested. 2. From this gate, use the FlowJo statistical analysis to determine the mean fluorescence intensity of each sample. Calculate the standard deviation from the mean fluorescence intensity of at least two independent biological replicates (see Fig. 3b).

4

Notes 1. To use the biosensor to analyze the effect of an enzyme on metabolite levels, it is advisable to compare cells harboring both biosensor and enzyme versus cells harboring both biosensor and inactive enzyme (e.g., WspR G249A, Addgene ID 79163). The latter control eliminates variation in expression caused by cells grown under different antibiotic conditions, different plasmid loads, and different expression loads. 2. The protocol describes the cloning of the pET31b Vc2-Spinach tRNA biosensor expression vector. Similar cloning approaches can be applied to construct other riboswitchSpinach tRNA biosensors expression vectors.

138

Sudeshna Manna et al.

3. As described in Subheading 3.1, other pET vectors may be used, but it is recommended to use compatible plasmids if coexpressing the biosensor with an enzyme. 4. The Vc2-Spinach DNA template can either be ordered as a DNA oligo from a commercial vendor or can be generated through overlapping PCR [8]. 5. Amplification of genes with high GC-content such as WspR can be difficult. If necessary, 1.5–3 μL of DMSO can be added to the PCR amplification. Alternatively, consider using a different enzyme gene that would suit the experiment. 6. One of the most important parameters with both microscopy and flow cytometry analyses is to ensure that the DFHBI solution is the same concentration across all samples. We generally make the 1 PBS or M9 Minimal Media solution containing DFHBI fresh for each set of experiments. 7. If no fluorescence is observed, yet the biosensor has been confirmed to function in vitro, examine the effect of the tRNA scaffold on fluorescence in vitro. Alternatively, test whether the Spinach-tRNA construct alone functions in vivo to assess if the problem is due to RNA expression, DFHBI permeability, or the microscopy protocol. If Spinach expression can be visualized, consider that ligand concentration might be too low for fluorescence activation or other effects of ligand on fluorescence visualization. If Spinach expression cannot be visualized, check that the RNA is being expressed via Northern blot or RT-PCR analysis. 8. Flow cytometry experiments can also be carried out by growing and analyzing cells in 96-well plate format. In a 1000 μL deepwell plate, inoculate 200 μL of LB/Carb/Kan media with 5 μL of overnight culture and incubate the cells in a 37  C shaker for ~1.5 h. Induce the cells with 1 mM IPTG, then harvest the cells after ~3 h incubation by centrifuging the entire plate at 4500 rcf for 3 min at ambient temperature. Decant the supernatant and resuspend the cells in 500 μL 1 PBS. Add 1 μL resuspended cells to 150 μL of the DFHBI-PBS solution in a 96 well plate and read using Attune Flow Cytometer with autosampler. 9. We have found that both M9 minimal media, pH 7.0 and 1 PBS are suitable for preparing cells for flow cytometry. Due to the longer incubation for adhering cells in the microscopy procedure though, we recommend using M9 minimal media for microscope experiments. 10. Initial inoculation in NI media from a single colony helps to achieve more consistent cell growth and biosensor expression between biological replicates compared to direct inoculation in AI media. Cells grown in NI media retain the plasmid and can

Live Cell Imaging Using RNA-Based Biosensors

139

be stored in 4  C for up to a week to be used for the next inoculation in AI media. We can use either culture tubes or 96-well plates for cell growth and induction depending on the number of samples. However, for experiments involving biosensor response to exogenous analytes, which result in lesser net change in cellular fluorescence compared to enzyme overexpression, we have found growth in culture tubes to give more consistent and reliable results. 11. Another advised control is to analyze cells expressing the Spinach2-tRNA construct (Addgene ID 79783) in the presence and absence of the metabolite of interest. This control will help determine whether changes in cellular fluorescence are due to specific interactions between the analyte and the biosensor versus nonspecific interactions with the Spinach2 aptamer. 12. Cell counts for flow cytometry analysis should be 10,000 at minimum, but preferably 30,000 or more to get better statistical parameters. 13. During the flow cytometry sample preparation and experiments, it is advisable to cover the plates or tubes containing samples with aluminum foil to protect DFHBI/DFHBI-1T fluorophore from photobleaching.

Acknowledgments The work on which this updated chapter is based was supported by Office of Naval Research grant N000141712638 (to M.C.H.). References 1. Newman RH, Fosbrink MD, Zhang J (2011) Genetically Encodable fluorescent biosensors for tracking signaling dynamics in living cells. Chem Rev 111(5):3614–3666 2. Serganov A, Nudler E (2013) A decade of riboswitches. Cell 152(1-2):17–24 3. Gao X, Dong X, Subramanian S, Matthews PM, Cooper CA, Kearns DB, Dann CE 3rd (2014) Engineering of Bacillus subtilis strains to allow rapid characterization of heterologous diguanylate cyclases and phosphodiesterases. Appl Environl Microbiol 80(19):6167–6174 4. Paige JS, Wu KY, Jaffrey SR (2011) RNA mimics of green fluorescent protein. Science 333(6042):642–646 5. Huang H, Suslov NB, Li N-S, Shelke SA, Evans ME, Koldobskaya Y, Rice PA, Piccirilli JA (2014) A G-quadruplex-containing RNA activates fluorescence in a GFP-like fluorophore. Nat Chem Biol 10(8):686–691

6. Warner KD, Chen MC, Song W, Strack RL, Thorn A, Jaffrey SR, Ferre´-D’Amare´ AR (2014) Structural basis for activity of highly efficient RNA mimics of green fluorescent protein. Nat Struct Mol Biol 21(8):658–663 7. Paige JS, Nguyen-Duc T, Song W, Jaffrey SR (2012) Fluorescence imaging of cellular metabolites with RNA. Science 335 (6073):1194–1194 8. Kellenberger CA, Wilson SC, Sales-Lee J, Hammond MC (2013) RNA-based fluorescent biosensors for live cell imaging of second messengers cyclic di-GMP and cyclic AMP-GMP. J Am Chem Soc 135(13):4906–4909 9. Ro¨mling U, Galperin MY, Gomelsky M (2013) Cyclic di-GMP: the first 25 years of a universal bacterial second messenger. Microbiol Mol Biol Rev 77(1):1–52 10. Strack RL, Disney MD, Jaffrey SR (2013) A superfolding Spinach2 reveals the dynamic

140

Sudeshna Manna et al.

nature of trinucleotide repeat–containing RNA. Nat Methods 10(12):1219–1224 11. Wang XC, Wilson SC, Hammond MC (2016) Next-generation RNA-based fluorescent biosensors enable anaerobic detection of cyclic di-GMP. Nucleic Acids Res 44(17):e139 12. Yeo J, Dippel AB, Wang XC, Hammond MC (2018) In vivo biochemistry: single-cell dynamics of cyclic Di-GMP in Escherichia coli in response to zinc overload. Biochemistry 57 (1):108–116 13. Kellenberger CA, Hallberg ZF, Hammond MC (2015) Live cell imaging using riboswitch-spinach tRNA fusions as metabolite-sensing fluorescent biosensors. In: Ponchon L (ed) RNA scaffolds: methods and protocols. Springer New York, New York, NY, pp 87–103 14. Manna S, Truong J, Hammond MC (2021) Guanidine biosensors enable comparison of cellular turn-on kinetics of riboswitch-based biosensor and reporter. ACS Synth Biol 10(3):566–578 15. Stojanovic MN, Kolpashchikov DM (2004) Modular aptameric sensors. J Am Chem Soc 126(30):9266–9270 16. Furutani C, Shinomiya K, Aoyama Y, Yamada K, Sando S (2010) Modular blue fluorescent RNA sensors for label-free detection of target molecules. Mol BioSyst 6 (9):1569–1571 17. Kellenberger CA, Wilson SC, Hickey SF, Gonzalez TL, Su Y, Hallberg ZF, Brewer TF, Iavarone AT, Carlson HK, Hsieh Y-F, Hammond MC (2015) GEMM-I riboswitches fromGeobacter sense the bacterial second messenger cyclic AMP-GMP. Proc Natl Acad Sci U S A 112(17):5383 18. Kellenberger CA, Chen C, Whiteley AT, Portnoy DA, Hammond MC (2015) RNA-based fluorescent biosensors for live cell imaging of second messenger cyclic di-AMP. J Am Chem Soc 137(20):6432–6435 19. You M, Litke JL, Jaffrey SR (2015) Imaging metabolite dynamics in living cells using a spinach-based riboswitch. Proc Natl Acad Sci U S A 112(21):E2756 20. Su Y, Hickey SF, Keyser SGL, Hammond MC (2016) In vitro and in vivo enzyme activity screening via RNA-based fluorescent biosensors for S-Adenosyl-l-homocysteine (SAH). J Am Chem Soc 138(22):7040–7047

21. Bose D, Su Y, Marcus A, Raulet DH, Hammond MC (2016) An RNA-based fluorescent biosensor for high-throughput analysis of the cGAS-cGAMP-STING pathway. Cell Chem Biol 23(12):1539–1549 22. Porter EB, Polaski JT, Morck MM, Batey RT (2017) Recurrent RNA motifs as scaffolds for genetically encodable small-molecule biosensors. Nat Chem Biol 13(3):295–301 23. Jepsen MDE, Sparvath SM, Nielsen TB, Langvad AH, Grossi G, Gothelf KV, Andersen ES (2018) Development of a genetically encodable FRET system using fluorescent RNA aptamers. Nat Commun 9(1):18 24. You M, Litke JL, Wu R, Jaffrey SR (2019) Detection of low-abundance metabolites in live cells using an RNA integrator. Cell Chem Biol 26(4):471–481 25. Wu R, Karunanayake Mudiyanselage APKK, Ren K, Sun Z, Tian Q, Zhao B, Bagheri Y, Lutati D, Keshri P, You M (2020) Ratiometric Fluorogenic RNA-based sensors for imaging live-cell dynamics of small molecules. ACS Appl. Bio Mater 5:2633–2642. https://doi. org/10.1021/acsabm.9b01237 26. Su Y, Hammond MC (2020) RNA-based fluorescent biosensors for live cell imaging of small molecules and RNAs. Curr Opin Biotec 63:157–166 27. Studier FW (2005) Protein production by auto-induction in high-density shaking cultures. Protein Expres Purif 41(1):207–234 28. Song W, Strack RL, Svensen N, Jaffrey SR (2014) Plug-and-play fluorophores extend the spectral properties of spinach. J Am Chem Soc 136(4):1198–1201 29. Malone JG, Williams R, Christen M, Jenal U, Spiers AJ, Rainey PB (2007) The structure–function relationship of WspR, a Pseudomonas fluorescens response regulator with a GGDEF output domain. Microbiology 153 (4):980–994 30. Novick RP (1987) Plasmid incompatibility. Microbiol Rev 51(4):381–395 31. Velappan N, Sblattero D, Chasteen L, Pavlik P, Bradbury ARM (2007) Plasmid incompatibility: more compatible than previously thought? Protein Eng Des Sel 20(7):309–313 32. Ponchon L, Dardel F (2007) Recombinant RNA technology: the tRNA scaffold. Nat Methods 4(7):571–576

Chapter 11 Rational Design of Allosteric Fluorogenic RNA Sensors for Cellular Imaging Qikun Yu, Ru Zheng, Manojkumar Narayanan, and Mingxu You Abstract Fluorescence-based tools are invaluable in studying cellular functions. Traditional small molecule or protein-based fluorescent sensors have been widely used for the cellular imaging, but the choice of targets is still limited. Recently, fluorogenic RNA-based sensors gained lots of attention. This novel sensor system can function as a general platform for various cellular targets. Here, we describe the steps to rationally design, optimize, and apply fluorogenic RNA-based sensors, using the intracellular imaging of tetracycline in living E. coli cells as an example. Key words Fluorogenic RNA, Aptamers, Genetically encoded, Allosteric RNA sensor, Antibiotics, Intracellular imaging

1

Introduction By allowing live-cell imaging of targets, fluorescent probes help us better understand the basics of cell biology. Large numbers of fluorescent protein-based sensors have been developed for this purpose [1–4]. However, the wide applications of these genetically encoded sensors are influenced by the limited choice of targetbinding protein domains and their low signal-to-noise ratio [5]. Recently, the discovery and development of fluorogenic RNA aptamers provided another option for engineering genetically encoded fluorescent sensors. Aptamers are single-stranded oligonucleotides that can specifically bind to their targets with high affinity. Fluorogenic RNA aptamers, such as Spinach [6] and Broccoli [7], can bind to the fluorophores like 3,5-difluoro-4-hydroxybenzylidene-1-trifluoroethyl-imidazolinone (DFHBI-1T) and activate their fluorescence signal. We and others have developed several Spinach and Broccoli-based allosteric sensors for the intracellular imaging of proteins, RNAs, signaling molecules, metabolites, and metal ions [8–14]. Here, using Broccoli as an example, we will illustrate how to rationally develop genetically encoded

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_11, © Springer Science+Business Media, LLC, part of Springer Nature 2021

141

142

Qikun Yu et al.

fluorogenic RNA sensors for imaging in living cells. Several other fluorogenic RNA aptamers have been recently reported [15–19], while the design principle of the corresponding allosteric sensors can be quite similar. A typical fluorogenic RNA-based allosteric sensor consists of a fluorescence module, a transducer module and a recognition module (Fig. 1). In the presence of targets, their binding with the recognition module induce a duplex formation of the transducer module, which further folds the fluorescence module, that is, Broccoli, and activates the fluorescence of DFHBI-1T. In the rest of this chapter, we will discuss the detailed procedures on how to design and optimize these sensors for both in vitro and intracellular studies.

2

Materials Prepare all solutions using ultrapure water (prepared by Milli-Q system with a sensitivity of 18.2 MΩ cm at 25
C) and molecular biology grade chemicals. Prepare all reagents at room temperature and store at 4
C (unless indicated otherwise).

2.1 In Vitro RNA Preparation and Characterization

1. EDTA solution: Add 93.1 g of disodium EDTA·2H2O to 400 mL of H2O. Stir and adjust the pH to 8.0 with NaOH (see Note 1). Make up to 500 mL with water. 2. Tris–HCl solution: Dissolve 60.6 g of Tris base in 400 mL of H2O, stir and adjust the pH to 8.0 by adding concentrated HCl. 3. TE buffer (1): Mix 2 mL of above-made 0.5 M EDTA (pH 8.0) and 10 mL of 1 M Tris–HCl (pH 8.0) with 988 mL of water. 4. High-fidelity DNA polymerase (2 U/μL). Store under 20
C. 5. High-fidelity DNA polymerase Buffer. Store under 20
C. 6. 10 mM dNTP mix. Store under 20
C. 7. Forward and reverse primers for PCR. Store under 20
C. 8. Commercial PCR cleanup kit. Store at room temperature. 9. HiScribe™ T7 high yield RNA synthesis kit (New England Biolabs, Ipswich, MA, USA). Store under 20
C. 10. DNase I (RNase-free) enzyme. Store under 20
C. 11. Sephadex G-25 column: Mix Sephadex G-25 medium beads with 1x TE buffer (see Note 2). Use empty column (remove filter from commercial protein purification columns) and put siliconized glass wool (see Note 3) at the bottom of the column. Pack the column by adding the beads and TE buffer

Fluorogenic RNA Sensors

143

Fig. 1 Design and working principle of allosteric fluorogenic RNA sensors. (a) The sensor comprises a fluorescence module (green), a transducer module (gray), and a recognition module (blue). (b) Target binding to the recognition module stabilizes the transducer module, which enables the fluorescence module to fold and activate the fluorescence of the fluorophore

mixture and centrifugation. Wash twice with 1 TE buffer and dry the column by centrifugation before usage. 12. Urea-denatured 10% polyacrylamide gel. 13. 5 Broccoli folding buffer: 200 mM HEPES, 500 mM KCl, 0.5% DMSO, pH 7.4. 14. DFHBI-1T stock solution: Dissolve DFHBI-1T in 100% DMSO to make 20 mM stock solution. Further dilute into 2 mM DFHBI-1T stock by 70–100% DMSO (see Note 4). Store under 20
C. 2.2 Molecular Cloning

1. One shot™ BL21 Star™ (DE3) chemically competent E. coli (Invitrogen, Carlsbad, CA, USA). Stored under 80
C. 2. BglII and XhoI restriction enzymes. Store under 20
C. 3. Double digestions buffer. Store under 20
C. 4. pET28c plasmid. Store under 20
C. 5. TOP10 chemically competent cell: Self-prepared by adding CaCl2, MgCl2 and glycerol into concentrated TOP10 E. coli cells. Detailed protocol can be found here [20]. Store under 80
C. 6. Lysogeny broth (LB) medium: Following manufacturer’s instructions, dissolve the powder in 1 L of ultrapure water. Autoclave at 121
C for 30 min and store at room temperature. 7. Kanamycin stock solution: Dissolve solid kanamycin in water to make 50 mg/mL stock solution. Aliquot and store under 20
C. 8. Antarctic phosphatase. Store under 20
C.

144

Qikun Yu et al.

9. Phosphatase reaction buffer. 10. DNA gel extraction kit. Store at room temperature. 11. 1% agarose gel. 12. T4 DNA ligase. Store under 20
C. 13. S.O.C. medium. Store at room temperature. 14. Petri dishes containing LB medium in 1% agarose supplemented with 50 μg/mL kanamycin. 15. Plasmid miniprep kit. Store at room temperature. 2.3 Intracellular Imaging

1. Poly-L-lysine solution 0.01%. Aliquot and store under 20
C. 2. 1 M isopropyl β-D-1-thiogalactopyranoside (IPTG) stock solution. Aliquot and store under 20
C. 3. Dulbecco’s phosphate buffered saline (DPBS) buffer, no calcium, no magnesium.

3

Methods

3.1 In Silico Design of Fluorogenic RNA-Based Tetracycline Sensors

1. Literature search and identify the correct RNA sequences for Broccoli and tetracycline aptamer (see Note 5), maybe also transducer sequences (see Note 6). 2. Determine the stem regions that play only structural roles (see Note 7) in both Broccoli (see Note 8) and tetracycline aptamer. Start with sequences that are next to the target-binding pockets. 3. Fuse Broccoli, transducer, and tetracycline aptamer into one RNA sequence (see Note 9) in the order from 50 - to 30 -site: Broccoli (50 -part) - transducer (50 -part) - tetracycline RNA aptamer - transducer (30 -part) - Broccoli (30 -part) (see Note 10). 4. Estimate in vitro behavior of each RNA sensor design by computational simulation using RNA secondary structure prediction software, like Mfold (see Note 11). 5. Convert RNA sensor candidates into corresponding DNA sequences, add the T7 promoter sequence (50 -TAATACGACT CACTATA-30 ) at the very beginning of 50 -site (see Note 12).

3.2 In Vitro Optimization of Fluorogenic RNA-Based Tetracycline Sensors

1. Dissolve all the DNA oligonucleotides in TE buffer to make 100 μM stock. Follow manufacturer’s instruction, mix appropriate amount of high-fidelity DNA polymerase, DNA templates, forward and reverse primers and other required reagents in the PCR tubes. PCR amplify in a thermocycler. 2. Purify the DNA product after 30 PCR cycles with commercial PCR clean-up kit. Measure the concentration of purified DNAs

Fluorogenic RNA Sensors

145

Fig. 2 Example in vitro characterization and optimization of allosteric fluorogenic RNA sensors. (a) Normalized fluorescence signal of Broccoli and three different sensor designs. Here, Design 2 is the optimal sensor with a large target-induced fluorescence enhancement and high absolute fluorescence intensity. (b) The target detection range of the optimal sensor

using Nanodrop. Follow manufacturer’s instruction, mix appropriate amount of T7 RNA polymerase and purified DNA with other required reagents at 20 μL scale in a microcentrifuge tube. Incubate under 37
C for 4 h in a thermomixer. 3. Add appropriate amount of DNase I and buffer directly to the above T7 transcription mixture and incubate under 37
C for 30 min in the thermomixer. Denature all the enzymes at 72
C for 8 min using a dry bath. 4. Purify the in vitro-transcribed RNAs using a homemade Sephadex G-25 column (see Note 13) and measure the concentration of RNAs using Nanodrop. Validate the size of RNAs by ureadenatured 10% polyacrylamide gel electrophoresis. 5. To test the fluorescence response of the tetracycline sensor, add 1 μM RNA in 1 Broccoli folding buffer in the presence of 1 mM MgCl2 (see Note 14), 20 μM DFHBI-1T, and different concentrations (e.g., 10 μM, 100 μM, and 1 mM) of tetracycline. Incubate (see Note 15) at room temperature in dark for 2 h and measure the fluorescence signal using a fluorometer. The excitation and emission peak of Broccoli/DFHBI-1T is at 480 nm and 503 nm, respectively. 6. Analyze all the data and identify the optimal tetracycline sensor (see Note 16) based on a large fold of tetracycline-induced fluorescence enhancement and comparable signal with original Broccoli (Fig. 2). 7. (Optional) further optimize the fold enhancement of the tetracycline sensor by changing the length and making point mutations of the transducer region.

146

Qikun Yu et al.

8. Determine the detection range of the optimal tetracycline sensor by incubating with 8–10 different concentrations of tetracycline under the same condition as Step 5 (Fig. 2). 9. Test the selectivity of the sensor under the same condition as step 5 by changing tetracycline into its analog antibiotics, including tobramycin, gentamicin, neomycin, kanamycin, ampicillin, etc. 10. It is also desirable to perform kinetic studies by mixing 1 μM optimal tetracycline sensor RNA, 1 Broccoli folding buffer with 1 mM MgCl2 and 20 μM DFHBI-1T. After adding 1 mM tetracycline, monitor the fluorescence change until the signal reaches to a plateau and stays stable for several minutes. 3.3 Molecular Cloning of Sensors into E. coli Cells

1. In this example, our goal is to clone a sensor-incorporated pET28c plasmid into a BL21 Star™ (DE3) E. coli strain (see Note 17). We use a pair of BglII and XhoI restriction enzymes (see Note 18) as an example for the plasmid double digestion. 2. Prepare and purify pET28c plasmid for cloning. Transform the plasmid into TOP10 competent cells (see Note 19) and grow overnight in a 10 mL LB medium with 50 μg/mL kanamycin at 37
C with shaking at 200 rpm. Follow the protocol of the plasmid purification kit to obtain purified pET28c plasmid DNA. 3. Design and PCR amplify the insert DNA fragment by incorporating a BglII and XhoI restriction site, a T7 promoter, optimal tetracycline sensor and a T7 terminator sequence. The procedure of PCR and clean-up has been shown in Subheading 3.2. 4. Digest the PCR product of DNA insert and pET28c plasmid with BglII and XhoI. Mix 1 μg of DNA (see Note 20), 10 units of both BglII and XhoI enzymes in the double digestion buffer at 50 μL scale. Incubate at 37
C for 4 h. 5. Purify the digested DNA insert products using a PCR clean-up kit. For the pET28c plasmid, add 10 units Antarctic phosphatase and 1 phosphatase reaction buffer first to reduce the selfligation. After incubation at 37
C for 30 min and heat inactivation at 80
C for 2 min, purify the digested plasmid using 1% agarose gel and gel extraction kit. Measure the concentrations of both purified digested DNA insert and plasmid products using Nanodrop. 6. Ligate the digested pET28c plasmid with the insert at a molar ratio of 1:3 (see Note 21). For the positive ligation, mix 100 ng pET28c plasmid, appropriate amount of sensor DNA insert, and 2 μL T4 DNA ligase within 1 T4 DNA ligase buffer at 20 μL scale. Meanwhile, as a negative control, mix 50 ng pET28c plasmid in another 20 μL solution with 1 μL T4

Fluorogenic RNA Sensors

147

DNA ligase and 1 T4 DNA ligase buffer, but without sensor DNA insertion. Incubate both samples at 16
C overnight in a thermomixer with 500 rpm shaking. Heat inactive the samples at 65
C for 10 min the second day. 7. Thaw two tubes of TOP10 chemical competent cells on ice, each contains 50 μL cells. Transfer 40 μL of TOP10 cells to the bottom of each 15 mL transformation tube without generating bubbles. Add 4 μL of positive ligation product or negative control product directly into the TOP10 cell droplets in the transformation tube. Mix by gently shaking on ice and incubate on ice for 20 min (see Note 22). 8. Heat shock the cells in a 42
C water bath for around 1 min. Place the tube back on ice immediately after heat-shock treatment and chill them for at least 90 s. Add 300–500 μL S.O.C. medium to the tubes and incubate at 37
C for 2 h with shaking at 200 rpm. 9. Plate positive and negative ligation cell mixture, respectively, on a Petri dish containing LB medium supplemented with 50 μg/mL kanamycin. Leave the Petri dishes dry for 15 min and then invert the plate. Incubate the plates at 37
C for 12–14 h until individual colonies can be easily identified (see Note 23). 10. The positive ligation dish should contain significantly more colonies than the negative one. Take 6–10 single colonies from the positive ligation dish and inoculate in the separated sterile culture tubes containing 5 mL LB medium and 50 μg/ mL kanamycin. Grow at 37
C for 12–14 h with shaking at 200 rpm. 11. Extract the plasmid from each inoculation solution using a plasmid extraction kit. Perform Sanger sequencing to confirm the correct sensor insertion in the plasmid. 12. Transform the sensor-incorporated pET28c plasmid into BL21 Star™ (DE3) cells and plate onto a Petri dish following a similar procedure as shown in steps 7–9. 3.4 Fluorescence Imaging of Tetracycline in Live E. coli Cells

1. Pick a single colony and grow sensor-incorporated BL21 Star™ (DE3) cells at 37
C for 12–14 h with shaking at 200 rpm in 5 mL LB medium and 50 μg/mL kanamycin. 2. Add 300 μL poly-L-lysine (PLL) solution into each well of a glass-bottomed 8-well chamber imaging plate. Incubate the plate at 37
C for 3 h. Remove the PLL solution and rinse each well twice with 500 μL sterile water to remove excess PLL. Add 500 μL sterile water and store the plate at room temperature before usage (see Note 24).

148

Qikun Yu et al.

Fig. 3 Fluorescence imaging of sensor-expressing E. coli cells after adding different concentrations of tetracycline. Scale bar, 5μm

3. Take a 1:10 dilution of the overnight cell culture (step 1) and measure its optical density at 600 nm (OD600). Dilute it into OD600 ¼ 0.1–0.2 unit/mL at the final volume of 5 mL and add 50 μg/mL kanamycin. 4. Regrow the diluted cells at 37
C with shaking until OD600 ¼ 0.4 (~45 min if starting OD600 ¼ 0.2 or ~90 min if starting OD600 ¼ 0.1). 5. Add 1 mM IPTG and continue growing at 37
C with shaking for another 2 h. 6. Transfer 200 μL cell aliquot into a 1.5 mL microcentrifuge tube. Spin the cells at 5000  g (g ¼ rcf) for 2 min at room temperature to pellet. Remove the supernatant and resuspend the cell pellet in 1 mL DPBS buffer. 7. Remove any remaining sterile water in the PLL-treated 8-well chamber (step 2). Add 100 μL cell suspension sample to each well (see Note 25) and incubate the 8-well chamber at 37
C for 45 min to allow cells to adhere to the glass bottom. 8. Remove suspended cells from each well and wash twice with 500 μL DPBS buffer to remove unattached cells. Add 100 μL DPBS with 200 μM DFHBI-1T to each well and incubate the chamber at 37
C for 30 min. 9. Add different concentration of tetracycline into each well (see Note 26) and incubate the chamber at room temperature for 2 h without light exposure. 10. Image under fluorescence microscope with 60 oil objective (Fig. 3). The DFHBI-1T fluorescence can be visualized with 488 nm laser excitation. Adjust the laser power and exposure time for the better fluorescence signal. Take bright-field and fluorescence images from at least three randomly selected places in every well. 11. Analyze the images using software like NIS-Elements and Image J.

Fluorogenic RNA Sensors

4

149

Notes 1. Disodium EDTA2H2O will not be fully dissolved until the pH of the solution is ~8 by adding NaOH. 2. The Sephadex™ G-25 medium beads cannot be dissolved in 1 TE buffer but immersed in the buffer for packing the column. Mix the beads with the TE buffer every time and leave it under 4
C overnight before usage. Swirl the mixture right before usage to resuspend the beads in the TE buffer. 3. The purpose for putting glass wool at the bottom is to avoid leakage of the beads from the column. The glass wool needs to be siliconized to reduce the adherence of RNA. We use Sigmacote™ (Sigma Diagnostics, Livonia, MI, USA) to siliconize the glass wool by immersing overnight and then dry inside a hood. 4. DFHBI-1T can be easily dissolved in DMSO to make 20 mM stock. Since high concentration DMSO may be toxic to the cells, we suggest to have the final imaging solution to be at least 100-fold diluted from the stock solution. 5. There may be several different sequences of RNA aptamers available for the same target molecule. It could be better to choose aptamers of known secondary and tertiary structures and importantly, exhibit good binding affinity and target selectivity. If there is no available RNA aptamer, a SELEX (systematic evolution of ligand by exponential enrichment) procedure may be needed to identify new RNA aptamers. It may be quite obvious, the sequence of a DNA aptamer can’t be transferred to an RNA aptamer by simply replacing thymine to uridine. 6. There may be some previous studies on developing allosteric RNA sensors for the same target by using different fluorogenic RNAs, ribozymes, and so on. The same transducer sequence can be potentially directly applied in your new RNA sensor design. 7. Some stem regions play only structural roles, that is, these stem sequences are not essential for the function of aptamer, the only requirement for this region is to form a duplex or a specific structure. One way to test if a stem is sequence-independent is by changing the original base pairs with other base pair sequences. If the mutated RNAs still have comparable functions as the original aptamer, then this stem region likely plays only structural roles, otherwise it’s sequence-dependent. 8. The stem regions that play only structural roles in Broccoli has been identified [7]. Normally, this stem can be directly used as the transducer region of the sensor. 9. To design a new transducer sequence, first directly fuse Broccoli with the target-binding aptamer, without adding

150

Qikun Yu et al.

additional transducer sequences. Based on the sensor simulation results, elongate the transducer if no stable Broccoli or target-binding aptamer structure is shown. If both Broccoli and target-binding aptamer stably form in the simulation, try to reduce the number of base pairs in the transducer. The optimal transducer will likely allow Broccoli and target-binding aptamer partially fold, but not that stable without target binding. 10. This order is not always correct for all the RNA sensor designs. If the only region to design transducer is in a middle stem of the target-binding aptamer, then Broccoli can be fused in this region after linking the 50 - and 30 -end of Broccoli using a stable tetraloop, for example, 50 -GCAA-30 . The revised order of the sensor will be 50 -target aptamer (50 -part)–transducer (50 -part)– Broccoli–transducer (30 -part)–target aptamer (30 -part) -30 . 11. Existing online software can only predict the secondary structures of pure RNA sequences without considering the binding of target. These results, though, are useful in estimating the background signal (in the absence of target), which is a critical parameter for the RNA sensor design. 12. T7 RNA polymerase-based transcription will be more efficient if starts with two or three G bases after the T7 promoter. 13. The RNA purification can be performed with commercially available kits as well. The cut-off molecular weight of the column should be smaller than that of the RNA sensor. 14. Most efficient folding of Broccoli requires ~1 mM magnesium ion, which is similar as the intracellular magnesium concentration (1–5 mM). 15. It is better to measure both negative and positive control at the same time with the RNA sensor. The negative control only contains DFHBI-1T in 1 Broccoli folding buffer and MgCl2, without adding any targets or RNAs. In the positive control sample, Broccoli RNA is used instead of the sensor RNA. Different concentrations of targets are added in the positive control as well. This is because target molecules with heterocycle rings may compete with DFHBI-1T to bind with Broccoli, which may quench the Broccoli signal at high concentrations. 16. If no sensor design exhibit noticeable target-induced fluorescence enhancement, try to further change the transducer sequence and length, as well as other stem region. 17. BL21 Star™ (DE3) cells carry a mutated rne gene that encodes a truncated RNase E enzyme. It lacks the ability for RNA degradation, resulting in high RNA sensor concentration, which can potentially exhibit high fluorescence intensity.

Fluorogenic RNA Sensors

151

18. When choosing restriction enzymes, make sure the enzyme recognition sites only exist at the desired location. Meanwhile, better to check if both enzymes can work efficiently in the same buffer system using tool like NEB Double Digest Finder. 19. TOP10 cells are used for the amplification and purification of plasmid with high-copy number. TOP10 cell itself is streptomycin resistant, it’s not suitable for the selection of any streptomycin resistant plasmid. 20. For short PCR DNA insert of less than 200 bases, a smaller amount of DNA (e.g., 300 ng) can be used to ensure the efficient digestion, especially when afterward the DNA is only purified by a spin column. 21. The molar ratio of vector-to-insert can be 1:3 or 1:5 depending on the size of DNA insert. If the insert is shorter than 200 bases, a molar ratio of 1:5 can increase the insertion efficiency as compared to 1:3. 22. Do not pipet to mix them. This 20 min incubation on ice is essential for the successful transformation of the constructed plasmid into bacterial cells. 23. The growth of individual colonies may take longer than 14 h depending on the strain. The negative ligation control should grow less colonies than the positive one. Plates containing transformed bacterial cells can be sealed with Parafilm and store at 4
C for up to ~4 weeks. 24. Similar procedure can be applied for 24-well plate as well. The as-prepared PLL-coated plate can be stored with sterile water for 24 h at room temperature or up to 1 week at 4
C. In other words, you can prepare the plate the same day or several days before imaging. 25. When using a 24-well plate, ~200 μL cell suspension is needed to cover the whole well. 26. Dyes that can stain dead cells, for example, propidium iodide or SYTOX Blue, can be used along with the target if the target, for example, tetracycline, is toxic to the cell. You can further study the correlation between the target cellular concentration and cell death.

Acknowledgments The authors gratefully acknowledge the UMass Amherst start-up grant, NIH R01AI136789, NSF CAREER, and Alfred P. Sloan Research Fellowship to M. You. The authors also thank other members in the You Lab for useful discussion and valuable comments.

152

Qikun Yu et al.

References 1. Zhang J, Campbell RE, Ting AY et al (2002) Creating new fluorescent probes for cell biology. Nat Rev Mol Cell Biol 3:906–918 2. Frommer WB, Davidson M, Campbell RE (2009) Genetically encoded biosensors based on engineered fluorescent proteins. Chem Soc Rev 38:2833–2841 3. Miyawaki A, Niino Y (2015) Molecular spies for bioimaging-fluorescent protein-based probes. Mol Cell 58:632–643 4. Ni Q, Mehta S, Zhang J (2018) Live-cell imaging of cell signaling using genetically encoded fluorescent reporters. FEBS J 285:203–219 5. Miyawaki A (2011) Development of probes for cellular functions using fluorescent proteins and fluorescence resonance energy transfer. Annu Rev Biochem 80:357–373 6. Paige JS, Wu KY, Jaffrey SR (2011) RNA mimics of green fluorescent protein. Science 333:642–646 7. Filonov GS, Moon JD, Svensen N et al (2014) Broccoli: rapid selection of an RNA mimic of green fluorescent protein by fluorescencebased selection and directed evolution. J Am Chem Soc 136:16299–16308 8. Paige JS, Nguyen-Duc T, Song W et al (2012) Fluorescence imaging of cellular metabolites with RNA. Science 335:1194 9. Mudiyanselage APKKK, Wu R, Leon-Duque MA et al (2019) “Second-generation” Fluorogenic RNA-based sensors. Methods 161:24–34 10. Sun Z, Nguyen T, McAuliffe K et al (2019) Intracellular imaging with genetically encoded RNA-based molecular sensors. Nanomaterials 9:233 11. Song W, Strack RL, Jaffrey SR (2013) Imaging bacterial protein expression using genetically

encoded RNA sensors. Nat Methods 10:873–875 12. Kellenberger CA, Wilson SC, Sales-Lee J et al (2013) RNA-based fluorescent biosensors for live cell imaging of second messengers cyclic di-GMP and cyclic AMP-GMP. J Am Chem Soc 135:4906–4909 13. You M, Litke JL, Jaffrey SR (2015) Imaging metabolite dynamics in living cells using a spinach-based riboswitch. Proc Natl Acad Sci U S A 112:e2756–e2765 14. Yu Q, Shi J, Mudiyanselage APKKK et al (2019) Genetically encoded RNA-based sensors for intracellular imaging of silver ions. Chem Commun 55:707–710 15. Dolgosheina EV, Jeng SCY, Panchapakesan SSS et al (2014) RNA mango aptamerfluorophore: a bright, high-affinity complex for RNA labeling and tracking. ACS Chem Biol 9:2412–2420 16. Song W, Filonov GS, Kim H et al (2017) Imaging RNA polymerase III transcription using a photostable RNA-fluorophore complex. Nat Chem Biol 13:1187–1194 17. Chen X, Zhang D, Su N et al (2019) Visualizing RNA dynamics in live cells with bright and stable fluorescent RNAs. Nat Biotechnol 37:1287–1293 18. Arora A, Sunbul M, Jaschke A (2015) Dualcolour imaging of RNAs using quencher-and fluorophore-binding aptamers. Nucleic Acids Res 43:e144 19. Sunbul M, Jaschke A (2018) SRB-2: a promiscuous rainbow aptamer for live-cell RNA imaging. Nucleic Acids Res 46:e110 20. Making Calcium Competent Cells. http:// mcb.berkeley.edu/labs/krantz/protocols/cal cium_comp_cells.pdf. Accessed 24 March 2020

Chapter 12 Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell Imaging of Bacteria Cordelia A. Weiss and Wade C. Winkler Abstract Riboswitches are a class of noncoding RNAs that regulate gene expression in response to changes in intracellular metabolite concentrations. When riboswitches are placed upstream of genetic reporters, the degree of reporter activity reflects the relative abundance of the metabolite that is sensed by the riboswitch. This method describes how reporters for live cell imaging, such as yellow fluorescent protein (YFP), can be placed under genetic control by metabolite-sensing riboswitches in the bacterium Bacillus subtilis. Specifically, a protocol for generating a fluorescent YFP reporter, based on a c-di-GMP responsive riboswitch, is outlined below. Key words Fluorescence microscopy, Live cell imaging, Cyclic di-GMP, Riboswitch, Bacillus subtilis, Metabolite biosensor

1

Introduction Riboswitches are cis-acting regulatory RNAs that modulate downstream gene expression in response to changes in metabolite or ion concentrations [1]. In general, riboswitches are usually located in the 50 leader regions of mRNAs, where they typically regulate gene expression by controlling formation of transcription termination sites or by affecting the efficiency of translation initiation [2]. This is achieved via the concerted action of two domains: a sensor domain and an expression platform. The sensor domain (aptamer), folds into a complex three-dimensional shape that binds its target ligand with exquisite affinity and selectivity and is capable of discriminating against closely related ligand variants. The expression platform couples the ligand-induced conformational changes to control of downstream gene expression. Around 40 distinct classes of riboswitches have been discovered and experimentally validated (Table 1); from these discoveries, it has been extrapolated that there are numerous riboswitches encoded among bacteria, including many that still await discovery [3, 4]. A significant subset of

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_12, © Springer Science+Business Media, LLC, part of Springer Nature 2021

153

154

Cordelia A. Weiss and Wade C. Winkler

Table 1 Diversity of Riboswitch Ligands. The table groups representative riboswitch classes by ligand type Ligand Type

Riboswitch Ligands

Coenzymes

Thiamin pyrophosphate (TPP), Adenosylcobalamin (AdoCbl), Aquacobalamin (AqCbl), Flavin mononucleotide (FMN), S-Adenosylmethionine (SAM), SAdenosylhomocysteine (SAH), Tetrahydrofolate (THF), Tungsten cofactor (Wco), Molybdenum cofactor (Moco), Nicotinamide adenine dinucleotide (NAD+)

Nucleotide derivatives

Guanine, Adenine, Prequeuosine-1 (Pre-Q1), 20 -Deoxyguanosine (20 -dG)

Signaling molecules

Cyclic di-GMP, cyclic di-AMP, cyclic GMP-AMP, 5-aminoimidizole-4-carboxamide ribonucleoside-50 -triphosphate (ZTP), guanosine tetraphosphate (ppGpp)

Ions

Magnesium, manganese, nickel/cobalt, fluoride

Other metabolites Guanidine, Glucosamine-6-phosphate (GlcN6P), Azaaromatics Amino acids

Glycine, lysine, glutamine

riboswitch ligands incorporates RNA components (such as cofactors, metabolites, and second messenger signals) and are widespread among bacteria. From these observations it has been argued that some riboswitches are likely to be modern relics of an ancient, primordial RNA World, where RNA molecules comprised all of the necessary catalytic and genomic functions [5]. However, in extant bacteria, riboswitches act as metabolite-sensing genetic elements that tune downstream gene expression to a specific dynamic range in response to metabolite abundance. Therefore, given their abundance, diversity, and specificity, riboswitches can be used as “molecular biosensors” for fluorescent imaging of metabolic changes within individual bacterial cells. While, in recent years, new advancements in mass spectrometry have increased the ability to quantify changes in metabolite concentrations, there is still an urgent need to develop new genetic tools for analyzing metabolite fluctuations within individual cells [6]. Such tools would allow for live imaging of metabolite dynamics within single cells. For example, several different fluorescent biosensor constructs have recently been developed for quantifying the spatiotemporal regulation of the bacterial signaling molecule cyclic di-GMP (c-di-GMP) [7]. For many bacteria, c-di-GMP regulates the transition from a unicellular motile state to a multicellular sessile community [8]. C-di-GMP has also been shown to regulate a number of other phenotypes, such as cell cycle control and virulence [9, 10]. By adapting a c-di-GMP-binding protein with fluorescent protein domains, binding of the nucleotide ligand can be signaled by Fo¨rster resonance energy transfer (FRET). This FRET-based biosensor has been used to interrogate c-di-GMP

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . .

155

dynamics in Gram-negative organisms such as Salmonella typhimurium, Caulobacter crescentus, Pseudomonas aeruginosa, and Escherichia coli [11–15]. A related strategy was used to couple binding of c-di-GMP to bioluminescence (BRET); this biosensor protein was shown to be active in Vibrio cholerae lysates [16, 17]. However, c-di-GMP signaling has been less intensively studied in many other organisms, where it is unclear if these particular reporter proteins will function. Also, FRET-based biosensors exhibit lower fluorescence intensities than individual fluorescent proteins, which can result in an overall narrowing of their dynamic range. In many instances, output signals that are stronger than can be offered by the FRET reporter may be desired. Moreover, these particular reporter proteins are designed specifically to respond to c-di-GMP; they were not created by a general approach that can easily lead to other metabolite-responsive reporters. In contrast, there are dozens of ligand-sensing riboswitches that have been discovered; therefore, the coupling of riboswitches to fluorescent outputs offers a wide range of biosensing possibilities. One class of fluorescent elements that has been used to tag and image RNA transcripts in living cells are collectively referred to as fluorogenic light-up aptamers (FLAPs) [18]. These are RNA sequences that bind to small-molecule fluorogens; upon binding to their cognate fluorogen, the intrinsic fluorescence of this ligand is greatly activated [19]. Several FLAPs have been discovered through in vitro evolution experiments—Spinach and its variants (Spinach2, iSpinach, Broccoli) bind and activate fluorescence of 3,5-difluoro-4-hydroxybenzylidene imidazolinone (DFHBI) [20– 23]. Other aptamers, such as Mango and Corn, bind and activate thiazole orange derivatives and 3,5-difluoro-4-hydroxybenzylidene imidazolinone-2-oxime (DFHO), respectively [24–26]. Newly optimized aptamers, such as DIR2s, Chili, and Peppers, continue to be developed and expand the spectral range of FLAPs [27– 30]. FLAPs can also be employed as sensors of small molecules by allosterically coupling the fluorogenic aptamer domain to a second, metabolite-sensing aptamer domain. For example, Spinach variants can be carefully fused to riboswitch aptamer domains through a stem loop called a “transducer module,” such that fluorescence of the Spinach/DFHBI complex is achieved only in response to binding of the riboswitch ligand, as allosterically signaled through the transducer module [31–33]. When optimized through these sorts of molecular engineering efforts, FLAPs can act as highly dynamic biosensors, capable of detecting aptamer analytes over several orders of magnitude. However, while this strategy exhibits ideal performance characteristics in vitro, technical optimization is required for routine usage in bacterial cells. Sometimes the inherent fold of the riboswitch may not always accommodate incorporation of the transducer module, thus requiring rational optimization of the design of the biosensor [34]. Also, the allosteric RNA biosensor

156

Cordelia A. Weiss and Wade C. Winkler

must be tuned to an equilibrium affinity value that is relevant to the aptamer domain [35]. Furthermore, any nonnative RNA sequence may face unexpected issues with RNA stability in bacterial cells, where the half-life of RNAs are typically only a few minutes. In an effort to combat this problem, prior expression studies in E. coli have shown that a tRNALys scaffold sequence can be added to aptamer-fused Spinach sequences to aid in their overall stability [36, 37]. Perhaps the greatest challenge to the in vivo bacterial use of FLAPs is that, unlike the intrinsic fluorescence provided by fluorescent proteins, FLAPs rely upon the binding of exogenous fluorogens. For example, permeability of DFHBI is likely to fluctuate among bacterial species, depending on the composition of the cell membrane or cell wall. And in certain instances, perhaps even a majority of instances, excessive amounts of DFHBI may have to be added to cells and allowed to incubate for a prolonged interim, in order to allow for sufficient quantities of the chromophore to passage through the cell membrane. Due to these challenges, parallel approaches are needed. One such alternative approach is to place genes encoding for fluorescent proteins immediately downstream of riboswitch sequences. Since riboswitches have been finetuned by natural selection to respond to their cognate ligands within the intracellular environment, they do not face some of the challenges listed above. It is also technically easy to subclone riboswitches upstream of a wide variety of fluorescent reporter genes, including proteins with high fluorescence quantum yields, such as “superfolder” fluorescent proteins [38]. In contrast, a potential disadvantage of a genetic-based reporter system is that the dynamics of the reporter protein, which has its own rate of synthesis, maturation and decay, can only be viewed as representative of the metabolite dynamics sensed by the riboswitch RNA. Nonetheless, riboswitch-based fluorescent reporters offer a powerful approach for assessing the relative changes in metabolite levels of bacterial cells, as visualized at the level of single cells. The method herein describes the development of a c-di-GMP-sensing riboswitch-yfp reporter fusion, which provides useful information on relative c-di-GMP abundance. Given that ensemble cultures of B. subtilis are comprised of several distinct subpopulations of differentiated cell types [39] (motile cells, competent cells, exopolysaccharide-producing cells, sporulating cells, etc.), the c-diGMP riboswitch-yfp reporter allows for the relative analysis of c-diGMP levels between cellular subpopulations. YFP and its relatives are versatile proteins that demonstrate higher fluorescence than other systems. Furthermore, fusion of yfp to a riboswitch allows for rapid imaging of live cells without the need for further manipulation. This method provides a streamlined approach for creating a fluorescent yfp reporter out of any riboswitch of interest.

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . .

2

157

Materials

2.1 Equipment and Supplies

1. Micropipettors and tips. 2. Benchtop vortexer. 3. Microcentrifuge. 4. 1.5 mL microcentrifuge tubes 5. Nuclease free water. 6. Sterile filter bottles. 7. 14 mL round-bottom culture tubes (17  100 mm). 8. Petri dishes—standard (35  10 mm).

(100



15

mm)

and

small

9. Hotplate. 10. Standing and shaking incubators set to 37  C. 11. Glass bottom dishes (Willco Wells). 12. Microscope (Zeiss Axio-Observer Z1 inverted fluorescence microscope (Zeiss, Jena, Germany)) equipped with a Rolera EM-C2 electron-multiplying charge-coupled (EMCC) camera (QImaging, British Columbia, Canada), and a mercury light source X-Cite 120Q (Excelitas Technologies Corp., Massachusetts, USA). To measure YFP fluorescence, a filter set (46 HE) was used with an excitation of BP 500/20, FT 515 beamsplitter, and emission of BP 535/30. 13. Image analysis software (see Note 1). 2.2

Reagents

1. LB Broth or 2xYT Broth, LB Agar. 2. pDG1662 plasmid Columbus, Ohio).

(Bacillus

Genetic

Stock

Center,

3. Competent E. coli DH5α. 4. Plasmid purification kit. 5. B. subtilis PY79, 168, or other recombination-proficient B. subtilis strain. 6. B. subtilis transformation medium: 25 g/L K2HPO4·3H2O, 6.0 g/L KH2PO4, 1.0 g/L trisodium citrate, 0.2 g/L MgSO4·7H2O, 2.0 g/L Na2SO4, 50μM FeCl3, 2μM MnSO4, 0.4% glucose, 0.2% glutamate. 7. Minimal Salts glycerol glutamate (MSgg) medium: 5 mM potassium phosphate buffer pH 7.0, 100 mM MOPS pH 7.0, 2 mM MgCl2, 0.05 mM MnCl2, 0.001 mM ZnCl2, 0.002 mM thiamine, 50μg/mL phenylalanine, 50μg/mL threonine, 50μg/mL tryptophan, 0.5% glycerol, 0.5% glutamate, 0.05 mM FeCl3, 0.7 mM CaCl2.

158

Cordelia A. Weiss and Wade C. Winkler

8. Low Melting Point Agarose. 9. Immersion oil such as Carl Zeiss™ Immersol™ 518F.

3

Method

3.1 Construction of a Riboswitch-yfp Reporter

Over 500 examples of c-di-GMP riboswitches have been identified among many bacterial species, and are predicted to regulate expression of diverse gene categories [40]. To date, two distinct structural classes of c-di-GMP riboswitches, each characterized by a GEMM (Genes for the Environment, Membranes, and Motility) motif, have been discovered [41, 42]. These riboswitch RNAs are cataloged on Rfam (https://rfam.xfam.org). Rfam is a database that collects sequence and structural information about various noncoding RNA elements, including but not limited to riboswitches. It is therefore an excellent source for determining which organisms encode a riboswitch of interest. In this study, we wanted to build a riboswitch-yfp reporter that responds to c-di-GMP riboswitch and that could be expressed in B. subtilis. However, a c-di-GMP riboswitch has not been identified in B. subtilis, although several are found within the order Bacillales, including in the genomes of other Bacillus species. For example, we proceeded with a c-di-GMP riboswitch located within the untranslated leader region of the lch gene cluster of the closely related organism Bacillus licheniformis [43]. Many riboswitches couple the detection of their target signal with transcription attenuation [44]. Oftentimes, an unbound riboswitch adopts a conformation in which an antiterminator helix is created from a portion of the terminator and a portion of the upstream aptamer sequence. When bound to its cognate ligand, the riboswitch will be structurally rearranged in such a way that this antiterminator helix is disrupted or prevented, thereby promoting formation of the default intrinsic terminator. Manual inspection of the B. licheniformis lch leader region revealed a candidate terminator that was preceded by a c-di-GMP-sensing aptamer and was therefore likely to control gene expression through a transcription attenuation mechanism. In general, when creating a riboswitch-yfp reporter from a riboswitch that controls gene expression via transcription attenuation, the entire riboswitch sequence should be simply placed downstream of a promoter and upstream of a gene encoding a fluorescent reporter protein. For example, for our c-diGMP riboswitch-yfp reporter, we included the entire DNA sequence from 25 nucleotides upstream of the aptamer until just before the start of the ribosomal binding site of lchAA, which is downstream of the intrinsic terminator [43]. This ensured that the entire riboswitch was present to exert regulatory control over yfp (see Note 2). Under conditions of high intracellular c-di-GMP, the

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . .

159

Fig. 1 Construction of the lchAA leader-yfp reporter. Based on the transcription attenuation mechanism of regulation by the lchAA riboswitch, it is expected that fluorescence of the lchAA leader-yfp reporter should only be observed under conditions of low intracellular c-di-GMP. Expression of a constitutive yfp reporter should be independent of c-di-GMP levels. Schematic was used previously [57]

riboswitch terminator is expected to inhibit expression of yfp. Conversely, yfp fluorescence should only be observed under low intracellular c-di-GMP conditions, when an antiterminator is formed instead (Fig. 1). 3.1.1 Assembly of the lchAA Leader-yfp Sequence

The riboswitch-yfp reporter sequence must be placed in a plasmid that is compatible with the targeted bacterium. For this study, the riboswitch-yfp reporter sequence was inserted into the E. coli-B. subtilis shuttle plasmid, pDG1662 (Bacillus Genetic Stock Center, Columbus, OH), using standard restriction-free ligation of PCR products [45]. Following expression in E. coli, pDG1662 allows the integration of the cloned insert into the nonessential locus amyE of the B. subtilis genome, via double-homologous recombination of DNA sequences corresponding to the “left” portion and “right” portion of amyE. This therefore transfers the intervening sequences into the genome between these regions of homologous recombination (Fig. 2). Within the DNA sequence that is transferred to the genome is a chloramphenicol resistance gene cassette; flanking the inserted DNA sequence is a spectinomycin cassette. Therefore, selection of chloramphenicol-resistant, spectinomycin-sensitive colonies indicates that the plasmid is likely to have successfully transferred its insert into the amyE locus via double homologous recombination. In general, expression of the riboswitch-yfp reporter requires several sequence elements: [1] a promoter that allows for constitutive expression of the transcript; our construct herein uses a synthetic, highly active promoter called “Pconst”, [2] the sequence of the entire riboswitch of interest, followed by [3] the coding sequence of yfp preceded by a strong ribosome binding site. Lastly, the sequence ends with [4] an intrinsic terminator, for termination of the transcript. Our studies use a synthetic

160

Cordelia A. Weiss and Wade C. Winkler

Fig. 2 Assembly of the lchAA leader-yfp sequence. Schematic diagram of sequence components of B. subtilis plasmid pDG1662, as well as the components required to make a c-di-GMP riboswitch-yfp reporter. The “X” marks denote locations of homologous recombination between the plasmid and genomic DNA. cat encodes for chloramphenicol resistance and spec encodes for spectinomycin resistance

bidirectional terminator sequence. Each of these modular components can be replaced as desired. The sequences of many such options are available as Biobricks (https://igem.org). Shown below is the sequence of the construct that was used in this study (Pconst-lchAA leader-yfp-terminator). Pconst is underlined, the lchAA leader is italicized between two hashmarks, and the coding sequence of yfp is in bold. The synthetic bidirectional terminator is double underlined: 50 -GTAGCCCTTGCCTACCTAGCTTCCAAGAAAGATAT CCTTACAGCACAAGAGCGG. AAAGATGTTTTGTTCTACATCCAGAACAACCTCTGCTAAAATTCCTGAAAAATTTTCGAAAAAGTTGTTGACTTTATCTACAAGGTGTGGCATAATGTGTGTGCA GCAGAAAA TGAATTTATATCAAGAAAAGCAGATAAAGG CAAACCTGCGGAAACGCAGGGACGCAAAGCCATGGCC TAAGGTGCTGACGGTGCTACGGTTGACAGGTTGCCGAA TAAACAGGGAGTTCGCCCGTTTTTATTCGGGCGGGCTCT TTTCTTTTTATTTCCAATATAATGTTTTATTGGAAACGACA AATCTGTGACAGCGTTTTTCGCTCATCGCAAAACCGCAA CATTGCATTGCGGCTTGGCTGTTCGCATCGTCATACATAA CAAGAGAT AAGCTTAAGGAGGAAAGTCACATT. ATGAGCAAAGGTGAAGAACTGTT-

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . .

161

CACCGGCGTTGTGCCAATTCTGGTTGAGCTGGATGGTGACGTGAATGGCCACAAATTTTCCGTGTCTGGTGAAGGCGAGGGTGATGCTACTTATGGCAAACTGACTCTGAAACTGATCTGTACCACCGGCAAACTGCCTGTTCCGTGGCCAACTCTGGTCACTACTCTGGGTTACGGCCTGATGTGTTTTGCGCGTTACCCGGATCACATGAAACAGCATGACTTCTTCAAATCTGCCATGCCGGAAGGCTATGTCCAAGAACGTACGATCTTTTTCAAGGACGACGGCAACTATAAAACCCGTGCCGAAGTTAAATTCGAGGGTGACACCCTGGTCAACCGCATCGAACTGAAAGGCATTGACTTCAAAGAGGACGGCAACATTCTGGGTCACAAGCTGGAATACAACTACAACTCCCACAACGTTTACATTACTGCTGACAAGCAGAAAAACGGCATCAAAGCAAACTTCAAGATCCGTCACAACATTGAAGATGGTGGCGTACAGCTGGCAGATCACTACCAGCAGAACACTCCAATCGGTGATGGCCCAGTACTGCTGCCAGATAACCATTACCTGTCCTACCAGAGCAAACTGTCTAAAGACCCGAACGAAAAACGTGACCACATGGTACTGCTGGAATTTGTTACCGCGGCAGGCATTACCCACGGTATGGACGAACTGTATAAATAAGCTAGCAAAAACCCCGCCCCTGACAGGGCGGGG TTTTTTTT-30 . Within this sequence can be found the c-di-GMP aptamer (AGCAGATAAAGGCAAACCTGCGGAAACGCAGGGACGCAAAGCC. ATGGCCTAAGGTGCTGACGGTGCTACGGTTGACAGGTTGCCGAA) and the lchAA riboswitch intrinsic terminator (GAGTTCGCCCGTTTTTATTCGGGCGGGCTC. TTTTCTTTTT ). The sequence of the ribosome binding site preceding yfp is AAGGAGGAAAG. As a control, a constitutively fluorescent yfp reporter (Pconst-yfp) should also be subcloned into pDG1662 for integration into B. subtilis amyE. For this, use the sequence above but without the lchAA leader sequence (italicized within hashmarks). 3.1.2 Transformation of B. subtilis

Domesticated strains of B. subtilis such as PY79 or 168 exhibit natural competence under certain growth conditions [46]. The following method has been empirically shown to result in an increase in the number of competent cells. This protocol relies on a glucose minimal medium with glutamate as the sole nitrogen source. These nitrogen-limiting conditions were previously shown to induce expression of competence genes in this organism [47]. For our analysis, we chose to transform two cellular background strains: PY79 and PY79 ΔpdeH. The latter contains a markerless deletion of a gene encoding a c-di-GMP

162

Cordelia A. Weiss and Wade C. Winkler

phosphodiesterase; this strain is known to exhibit higher default levels of c-di-GMP [43, 48, 49]. Step 1: Method to transform B. subtilis (Estimated time: 1.5 days). 1. Streak for isolation on LB agar plates with the desired recipient strains. In this study, we used PY79 and PY79 ΔpdeH. Incubate overnight at 37  C. 2. Use an inoculation loop to extract a small sample of isolated colonies for inoculation of 5.0 mL B. subtilis transformation medium. Vortex briefly. Incubate standing at 37  C overnight. Make sure the culture tube cap is loose, to allow for oxygenation of the culture (see Note 3). 3. The following morning, transfer the standing culture to a shaking 37  C incubator. Incubate the culture, shaking, until it reaches an OD600 0.4–0.8. This should take several hours. 4. Once the culture has reached an OD600 of at least 0.4, but not higher than 0.8, remove 1 mL aliquots and transfer to new sterile culture tubes. 1 mL culture is required for each transformation. Add 0.5–4μg miniprepped plasmid DNA to each 1 mL culture. As a negative control, maintain a 1 mL culture with no DNA added. 5. Incubate shaking at 37  C for 40 min. 6. If pDG1662 is being used as the base vector for the riboswitchyfp reporter, chloramphenicol is the antibiotic that is needed to select for successful DNA integration. Add 1 mL 2xYT or LB supplemented with 0.1μg/mL chloramphenicol to begin to induce chloramphenicol resistance. Incubate at 37  C for an additional 45 min. 7. Spread 100μL of the various cultures on LB plates supplemented with 5μg/mL chloramphenicol. If transformation efficiencies are low, collect the remaining volume and briefly (1–2 min) pellet the cells using a microcentrifuge. Resuspend in 100μL of broth and spread onto a selective agar plate. Transformation efficiencies can vary; therefore, the number of transformants will vary. This protocol may require optimization for the plasmid DNAs that are used. Incubate the plates overnight at 37  C. Step 2: Checking for successful double homologous integration (Estimated time: 1.5 days). 1. From the plate of B. subtilis transformants, pick 4 colonies with an inoculation loop. Patch a short streak from each colony first on an LB agar plate supplemented with 5μg/mL chloramphenicol and then, without sterilizing the loop between plates, an LB agar plate supplemented with 100μg/mL spectinomycin.

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . .

163

As a control, the base strain (in this case, PY79) should not be resistant to either. Incubate the plates overnight at 37  C. 2. Double homologous integration at the amyE locus is likely to have occurred if the transformant strain is resistant to chloramphenicol and sensitive to spectinomycin. With a sterile loop, pick one or two of the appropriate patches and restreak for isolated colonies on an LB plate supplemented with 5μg/mL chloramphenicol. Incubate overnight at 37  C. From this plate, incubate overnight cultures for preparation of glycerol freezer stocks. The sequence of the inserted DNA can be easily verified by isolating genomic DNA and using PCR to amplify the amyE sequence, which can be assessed by Sanger sequencing. 3.2 Live Cell Imaging of B. subtilis Harboring the lchAA Leader-yfp Reporter

Microscopy is an indispensable tool for the visualization of cell morphology. For example, the use of microscopy for live cell imaging allows for direct visualization of phenotypic heterogeneity among an isogenic population of bacterial cells. In our study, live cell imaging of B. subtilis cells that expressed a fluorescent c-diGMP riboswitch-yfp reporter revealed that c-di-GMP levels are markedly different among B. subtilis subpopulations [43]. The riboswitch-yfp reporter revealed a bimodal distribution of fluorescence in wild-type cells (Fig. 3b, d). Furthermore, the riboswitchyfp reporter displayed significantly decreased yfp expression in the ΔpdeH background, suggesting that the riboswitch reduces downstream gene expression in response to elevated c-di-GMP (Fig. 3b, d). Conversely, the constitutive yfp reporter remained unchanged in both cellular backgrounds (Fig. 3a, c). The yellow fluorescent protein YFP described in this method has proven to be useful in multiple studies for the analysis of B. subtilis cell differentiation [50, 51]. The sequence of YFP that is commonly used for imaging experiments in B. subtilis contains a few mutations that give it a shortened half-life. This lowers the fluorescent signal but improves the ability to report relative changes in YFP levels (e.g., such as reporting relative differences in gene expression). However, it remains to be determined whether YFP, or other fluorescent proteins, are sufficient to measure rapid dynamics in B. subtilis.

3.2.1 Growth of Bacterial Strains for Fluorescence Microscopy

Step 3: Bacterial Growth Conditions (Estimated time: 16–20 h). 1. Inoculate 3 mL MSgg (see Note 4) supplemented with 5μg/ mL chloramphenicol from a single colony of the transformed B. subtilis cells. Incubate the cultures at 37  C shaking overnight. In our experiment, we inoculate 5 cultures each with: (a) PY79 WT. (b) PY79 WT amyE::Pconst- yfp cat. (c) PY79 ΔpdeH amyE::Pconst -yfp cat.

164

Cordelia A. Weiss and Wade C. Winkler

Fig. 3 Expression of the lchAA leader-yfp reporter in vivo. (a and b) Representative microscopy images of B. subtilis PY79 wild-type (WT) or ΔpdeH (elevated c-di-GMP strain) cells expressing the (a) constitutive Pconst-yfp reporter or (b) Pconst-lchAA leader-yfp reporter. (c and d) Histograms of the quantification of fluorescence intensity per cell comparing B. subtilis PY79 WT or ΔpdeH cells expressing the (c) constitutive Pconst-yfp reporter or (d) Pconst-lchAA leader-yfp reporter (n ~ 300). Histograms were used previously [43]

(d) PY79 WT amyE::Pconst-lchAA leader-yfp cat. (e) PY79 ΔpdeH amyE::Pconst-lchAA leader-yfp cat. 2. In the morning, sub-culture 3 mL fresh MSgg supplemented with 5μg/mL chloramphenicol with your stationary phase overnight cultures, such that the new OD600 ~ 0.1. Incubate the cultures shaking at 37  C until the culture reaches an OD600 ~ 1.0 (or until the desired OD600 is reached). 3.2.2 Use of Agarose Pads for Fluorescence Microscopy

Fluorescence microscopy of live, motile, bacteria requires that cells be immobilized prior to imaging. Poly-L-lysine is a positively charged synthetic amino acid that can be used to coat plastic or glass surfaces to promote cell attachment. However, this technique

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . .

165

is not always sufficient to fully immobilize motile bacteria. Agarose pads present a superior alternative that prevent both motility of live bacteria, as well as Brownian motion of fixed cells. The agarose polymer provides a matrix for trapping cells. However, the nutrient and gas exchange within the agarose matrix ensures that bacteria remain viable, thereby permitting time-lapse imaging. Step 4: Preparation of agarose pads (Estimated time: 2 h). 1. Fill a glass beaker (250 mL) with water and begin heating the beaker on a hot plate. Heating the beaker between 100–200  C will result in fully dissolved agarose in about 1 h. 2. In a 15.0 mL conical tube, add 10.0 mL MSgg medium. Add 0.15 g low-melting point agarose, for a final concentration of 1.5%. Resuspend the solution briefly by inverting the conical tube. Place the conical tube in the water-filled beaker. 3. Heat the MSgg solution until the agarose has fully dissolved (see Note 5). 4. Pour melted agarose pad in a small petri dish until it is a few mm in height. Individual preferences may vary between 0.3 and 10 mm. Let the agarose pad dry with the dish covered for at least 20 min. Then remove the lid of the petri dish and expose the agarose to surrounding air for an additional 20 min (see Note 6). 3.2.3 Fluorescence Microscopy Experiments

Step 5: Preparation of samples for microscopy (Estimated time: 5 min). 1. Pipette a small volume of each bacterial culture on to the agarose pad. Evenly space the spots so that there is no crosscontamination of cultures. Since the agarose pad has been allowed to “dry” for 20 min prior to inoculation, the bacterial culture should rapidly diffuse into the pad within a few minutes. We find that spotting 3μL of bacterial culture that is at approximately OD600 of 0.5–1.0 provides an appropriate number of cells for imaging. However, the total amount of cells that are added to the agarose pad will depending on the needs of the individual experiment. Make sure the spot is dry before proceeding. 2. Using a sterile razor blade, cut a small square section of agarose around each spot, and pick the pad up carefully using a corner of the blade. Flip the pad over onto a glass bottom dish so that the bacteria are compressed in a monolayer between the pad and the glass. Step 6: Imaging cells by fluorescence microscopy (Estimated time: 1 h).

166

Cordelia A. Weiss and Wade C. Winkler

1. Technical procedures for setting up the fluorescence microscope will vary between microscope systems. Once immersion oil has been dabbed onto the underside of the glass-bottom dish and it has been securely placed onto the microscope stage, the experimentalist can begin to focus on the bacterial cells using differential interference contrast (DIC) or phase contrast microscopy (see Note 7). 2. The cellular sample that is expected to exhibit the highest fluorescence should be examined first, in order to identify exposure conditions that can be used for all of the biological samples. For example, we generally like to begin our experiments by assessing the fluorescence of a strain containing a constitutive yfp gene (PY79 WT amyE::Pconst- yfp). 3. Once a field of view has been chosen for fluorescence analysis and the targeted cells are in focus, one will need to vary exposure times to find appropriate conditions (see Note 8). Once the appropriate YFP fluorescence exposure conditions have been identified, such that the fluorescence signal is as bright as it can be without being saturated, note the exposure time and use this for your other strains, so that their measurements of cellular fluorescence can be directly compared. 4. For each sample, at least three different fields of view should be imaged (preferably more). Make sure to image the cells using the DIC/phase and yellow fluorescence filter channels. Image as many fields of view as needed to obtain measurements for at least 300 individual cells. Save each file to be analyzed as a TIFF file. Step 7: Analysis of microscopy data (Estimated time: 2 h). 1. Open your saved TIFF files in any image analysis software. Individual preferences for this step may vary between software such as Oufti, Zen, or ImageJ. 2. Quantification of the fluorescent signal relies on determining the mean fluorescence intensity per cell. Either manually outline individual cells in the phase image or allow your software program to detect cells in the image with cell segmentation parameters. 3. To quantify fluorescence intensity per cell, overlay the fluorescence image on top of the phase image, so that pixel intensity of fluorescence is quantified. Obtain the mean fluorescence intensity per cell and record the measurements. 4. To obtain background fluorescence intensity, quantify the mean fluorescence intensity in three separate areas of an image that have no cells. Subtract the background intensity from the mean fluorescence intensity of each cell.

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . .

167

5. Graph the mean intensities per cell of each sample using appropriate spreadsheet analysis software, such as Prism GraphPad or Excel, to generate histograms.

4

Notes 1. Oufti (Jacobs-Wagner Lab, Stanford University) is an opensource software package designed to perform automated detection of cells as well as quantification of fluorescent signals from microscopy images [52]. The software enables highthroughput image analysis and is broadly accessible to users, irrespective of their level of knowledge about computational programming. However, software options such as Zen (Zeiss), ImageJ, or others can also be used [53]. 2. To ensure that the riboswitch of interest is solely regulating expression of yfp, several control strains should be made. For example, deleting the constitutive promoter from your reporter construct should result in no observable fluorescent signal. If, however, this control reporter exhibits fluorescence, it would suggest that the sequence of the riboswitch region might include one or more cryptic promoter sites. An additional negative control should include deletion of an essential part of the aptamer sequence. 3. Allowing the culture to stand overnight rather than to incubate on an orbital shaker ensures that the cells do not enter stationary phase before the following morning. 4. MSgg is a defined minimal medium that has been shown to promote biofilm formation in B. subtilis [54]. Because it is a nonfluorescent medium, MSgg can also be used as the base of the agarose pad for live cell imaging, and cells can be transferred directly from the culture tube to the agarose pad. If you choose to grow your cultures in LB, 2xYT, or other yellow medium, the cultures must be pelleted and washed three times with PBS to remove any residual growth medium. The agarose pad can be made with PBS instead of MSgg if needed. 5. Heating the solution in a closed falcon tube immersed in a hot water bath ensures that the percentage of agarose remains constant during heating and prevents evaporation. 6. While this general approach works well for most of our laboratory experiments, the method of pouring agarose pads can be individualized per the needs of the experimentalist. For example, some researchers find it useful to use a biopsy punch to extricate small sections of the agarose pad after it has cooled and solidified. Additional protocols also exist that do not require the use of small petri dishes.

168

Cordelia A. Weiss and Wade C. Winkler

7. To avoid photobleaching, one should avoid exposing the cells to the reflective light source until they are ready to collect images for analysis. Correspondingly, one should focus the cells using phase-contrast or DIC. 8. Exposure times of one’s sample will be affected by a wide variety of variables, such as the sensitivity of the microscope camera, the fluorescent proteins being utilized, and the background fluorescence exhibited by the bacteria and the medium that surrounds them. It may be useful to perform a photobleaching experiment to determine the degree to which the cellular fluorescence is reduced during collection of the micrographs. More details on these considerations can be found elsewhere [55, 56].

Acknowledgments The work described in this chapter was supported by NIH T32 AI89621-7 to C.A.W. and NIH R01AI110432 to W.C.W. References 1. Winkler WC, Breaker RR (2005) Regulation of bacterial gene expression by riboswitches. Annu Rev Microbiol 59:487–517 2. Serganov A, Nudler E (2013) A decade of riboswitches. Cell 152:17–24 3. McCown PJ, Corbino KA, Stav S et al (2017) Riboswitch diversity and distribution. RNA 23:995–1011 4. Sherlock ME, Breaker RR (2020) Former orphan riboswitches reveal unexplored areas of bacterial metabolism, signaling, and gene control processes. RNA 26:675–693 5. Nelson JW, Breaker RR (2017) The lost language of the RNA World. Sci Signal 10: eaam8812 6. Greenwald EC, Mehta S, Zhang J (2018) Genetically encoded fluorescent biosensors illuminate the spatiotemporal regulation of signaling networks. Chem Rev 118:11707–11794 7. Petchiappan A, Naik SY, Chatterji D (2020) Tracking the homeostasis of second messenger cyclic-di-GMP in bacteria. Biophys Rev 12:719–730 8. Romling U, Galperin MY, Gomelsky M (2013) Cyclic di-GMP: the first 25 years of a universal bacterial second messenger. Microbiol Mol Biol Rev 77:1–52 9. Jenal U, Reinders A, Lori C (2017) Cyclic di-GMP: second messenger extraordinaire. Nat Rev Microbiol 15:271–284

10. Hall CL, Lee VT (2018) Cyclic-di-GMP regulation of virulence in bacterial pathogens. Wiley Interdiscip Rev RNA 9. https://doi.org/10. 1002/wrna.1454 11. Christen M, Kulasekara HD, Christen B et al (2010) Asymmetrical distribution of the second messenger c-di-GMP upon bacterial cell division. Science 328:1295–1297 12. Kulasekara BR, Kamischke C, Kulasekara HD et al (2013) C-di-GMP heterogeneity is generated by the chemotaxis machinery to regulate flagellar motility. Elife 2:e01402 13. Mills E, Petersen E, Kulasekara BR et al (2015) A direct screen for c-di-GMP modulators reveals a Salmonella Typhimurium periplasmic L-arginine–sensing pathway. Sci Signal 8:ra57 14. Ho CL, Chong KSJ, Oppong JA et al (2013) Visualizing the perturbation of cellular cyclic di-GMP levels in bacterial cells. J Am Chem Soc 135:566–569 15. Petersen E, Mills E, Miller SI (2019) Cyclic-diGMP regulation promotes survival of a slowreplicating subpopulation of intracellular salmonella typhimurium. PNAS 116:6335–6340 16. Dippel AB, Anderson WA, Evans RS et al (2018) Chemiluminescent biosensors for detection of second messenger cyclic di-GMP. ACS Chem Biol 13:1872–1879 17. Dippel AB, Anderson WA, Park JH et al (2020) Development of Ratiometric bioluminescent

Riboswitch-Mediated Detection of Metabolite Fluctuations During Live Cell. . . sensors for in Vivo detection of bacterial signaling. ACS Chem Biol 15:904–914 18. Hennig S, Neubacher S (2019) Fluorescent RNA tags: current and future applications. Future Med Chem 11:2483–2485 19. Truong L, Ferre´-D’Amare´ AR (2019) From fluorescent proteins to fluorogenic RNAs: tools for imaging cellular macromolecules. Protein Sci 28:1374–1386 20. Paige JS, Wu KY, Jaffrey SR (2011) RNA mimics of green fluorescent protein. Science 333:642–646 21. Strack RL, Disney MD, Jaffrey SR (2013) A superfolding Spinach2 reveals the dynamic nature of trinucleotide repeat RNA. Nat Methods 10:1219–1224 22. Autour A, Westhof E, Ryckelynck M (2016) iSpinach: a fluorogenic RNA aptamer optimized for in vitro applications. Nucleic Acids Res 44:2491–2500 23. Filonov GS, Moon JD, Svensen N et al (2014) Broccoli: rapid selection of an RNA mimic of green fluorescent protein by fluorescencebased selection and directed evolution. J Am Chem Soc 136:16299–16308 24. Dolgosheina EV, Jeng SCY, Panchapakesan SSS et al (2014) RNA mango aptamerfluorophore: a bright, high-affinity complex for RNA labeling and tracking. ACS Chem Biol 9:2412–2420 25. Autour A, Jeng SCY, Cawte AD, Abdolahzadeh A et al (2018) Fluorogenic RNA Mango aptamers for imaging small non-coding RNAs in mammalian cells. Nat Commun 9:656 26. Song W, Filonov GS, Kim H et al (2017) Imaging RNA polymerase III transcription using a photostable RNA-fluorophore complex. Nat Chem Biol 13:1187–1194 27. Tan X, Constantin TP, Sloane KL et al (2017) Fluoromodules consisting of a promiscuous RNA aptamer and red or blue Fluorogenic cyanine dyes: selection, characterization, and bioimaging. J Am Chem Soc 139:9001–9009 28. Steinmetzger C, Palanisamy N, Gore KR et al (2019) A multicolor large stokes shift Fluorogen-activating RNA aptamer with cationic chromophores. Chem Eur J 25:1931–1935 29. Chen X, Zhang D, Su N et al (2019) Visualizing RNA dynamics in live cells with bright and stable fluorescent RNAs. Nat Biotechnol 37:1287–1293 30. Wu J, Zaccara S, Khuperkar D et al (2019) Live imaging of mRNA using RNA-stabilized fluorogenic proteins. Nat Methods 6:862–865 31. Paige JS, Nguyen-Duc T, Song W et al (2012) Fluorescence imaging of cellular metabolites with RNA. Science 335:1194–1194

169

32. Nakayama S, Luo Y, Zhou J et al (2012) Nanomolar fluorescent detection of c-di-GMP using a modular aptamer strategy. Chem Commun 48:9059–9061 33. Kellenberger CA, Wilson SC, Sales-Lee J et al (2013) RNA-based fluorescent biosensors for live cell imaging of second messengers cyclic di-GMP and cyclic AMP-GMP. J Am Chem Soc 135:4906–4909 34. Su Y, Hickey SF, Keyser SGL et al (2016) In vitro and in vivo enzyme activity screening via RNA-based fluorescent biosensors for S-Adenosyl-l-homocysteine (SAH). J Am Chem Soc 138:7040–7047 35. Hallberg ZF, Su Y, Kitto RZ et al (2017) Engineering and in vivo applications of riboswitches. Annu Rev Biochem 86:515–539 36. Ponchon L, Dardel F (2007) Recombinant RNA technology: the tRNA scaffold. Nat Methods 4:571–576 37. Kellenberger CA, Hallberg ZF, Hammond MC (2015) Live cell imaging using riboswitch-spinach tRNA fusions as metabolite-sensing fluorescent biosensors. Methods Mol Biol 1316:87–103 38. Pe´delacq J-D, Cabantous S, Tran T et al (2006) Engineering and characterization of a superfolder green fluorescent protein. Nat Biotechnol 24:79–88 39. Lopez D, Vlamakis H, Kolter R (2009) Generation of multiple cell types in Bacillus subtilis. FEMS Microbiol Rev 33:152–163 40. Lee ER, Sudarsan N, Breaker RR (2010) Riboswitches that sense cyclic Di-GMP. In: Wolfe AJ, Visick KL (eds) The second messenger cyclic Di-GMP. ASM Press, Washington, DC, p 215–229 41. Sudarsan N, Lee ER, Weinberg Z et al (2008) Riboswitches in eubacteria sense the second messenger cyclic Di-GMP. Science 321:411–413 42. Lee ER, Baker JL, Weinberg Z et al (2010) An allosteric self-splicing ribozyme triggered by a bacterial second messenger. Science 329:845–848 43. Weiss CA, Hoberg JA, Liu K et al (2019) Single-cell microscopy reveals that levels of cyclic di-GMP vary among Bacillus subtilis subpopulations. J Bacteriol 201:e00247–e00219 44. Henkin TM (2008) Riboswitch RNAs: using RNA to sense cellular metabolism. Genes Dev 22:3383–3390 45. Gibson DG, Young L, Chuang R-Y et al (2009) Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 6:343–345 46. Dubnau D (1991) Genetic competence in Bacillus subtilis. Microbiol Rev 55:395–424

170

Cordelia A. Weiss and Wade C. Winkler

47. Jarmer H, Berka R, Knudsen S et al (2002) Transcriptome analysis documents induced competence of Bacillus subtilis during nitrogen limiting conditions. FEMS Microbiol Lett 206:197–200 48. Chen Y, Chai Y, Guo J -h et al (2012) Evidence for cyclic Di-GMP-mediated signaling in Bacillus subtilis. J Bacteriol 194:5080–5090 49. Gao X, Mukherjee S, Matthews PM et al (2013) Functional characterization of Core components of the Bacillus subtilis cyclic-DiGMP signaling pathway. J Bacteriol 195:4782–4792 50. Su¨el GM, Kulkarni RP, Dworkin J et al (2007) Tunability and noise dependence in differentiation dynamics. Science 315:1716–1719 51. Campo N, Rudner DZ (2007) SpoIVB and CtpB are both Forespore signals in the activation of the sporulation transcription factor K in Bacillus subtilis. J Bacteriol 189:6021–6027 52. Paintdakhi A, Parry B, Campos M et al (2016) Oufti: an integrated software package for high-

accuracy, high-throughput quantitative microscopy analysis. Mol Microbiol 99:767–777 53. Schindelin J, Arganda-Carreras I, Frise E et al (2012) Fiji—an open source platform for biological image analysis. Nat Methods 9:676–682 54. Branda SS, Gonza´lez-Pastor JE, Ben-Yehuda S et al (2001) Fruiting body formation by Bacillus subtilis. Proc Natl Acad Sci U S A 98:11621–11626 55. Meyer P, Dworkin J (2007) Applications of fluorescence microscopy to single bacterial cells. Res Microbiol 158:187–194 56. van Teeffelen S, Shaevitz JW, Gitai Z (2012) Image analysis in fluorescence microscopy: bacterial dynamics as a case study. BioEssays 34:427–436 57. Orr MW, Weiss CA, Severin GB et al (2018) A subset of exoribonucleases serve as degradative enzymes for pGpG in c-di-GMP signaling. J Bacteriol 200:e00300–e00318

Chapter 13 FRET Analysis of RNA–Protein Interactions Using Spinach Aptamers Laura Gerhard and Sven Hennig Abstract The method development to analyze direct RNA–protein interaction is of high importance as not many homogeneous assay formats are available. The discovery of fluorescent light-up aptamers (FLAPs), short RNA aptamers that switch the fluorescence of small, cell-permeable, and nontoxic organic chromophores on, paved the road for their utilization in direct RNA–protein interactions. The combination with fluorescent proteins as biological fluorophores enabled the development of homogeneous assays that are in principle even encodable on genomic level. Here the rules and methods to design a homogeneous in vitro assay for the detection and quantification of a direct RNA–protein interaction will be described. The design and application of a homogeneous assay to observe and quantify the interaction of the Pseudomonas aeruginosa bacteriophage coat protein 7 (PP7) with its binding RNA sequence (pp7-RNA) will be shown. For this, the Spinach-DFHBI aptamer as RNA fusion and the red fluorescent mCherry as protein fusion was used. The methods presented here do not require any chemical modification of proteins or RNAs which make them relatively easy to use and to adopt on other systems. As all fluorophores are fusion tags to the according biomolecules, standard cloning strategies and molecular biology technologies are sufficient and make this method available for a broad community of researchers. Key words Fo¨rster Resonance Energy Transfer (FRET), Fluorescence Light-up Aptamers (FLAPs), RNA–protein interaction, REMSA, Fluorescence quenching, PP7 protein, mCherry, Spinach aptamer, DFHBI

1

Introduction Since next generation sequencing technologies gave us deeper insights into the world of transcriptomes (RNA-Seq), it is known that the number of RNAs are bigger than ever estimated [1]. These plain numbers subsequently cause the immense need for new methods to investigate and understand the roles of RNAs. This includes their role in complex with other biomolecules such as protein partners. Transient or relatively stable, tissue specific or ubiquitous, the functions of RNA–protein complexes are manifold and methods to understand them in detail and depth are urgently needed.

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_13, © Springer Science+Business Media, LLC, part of Springer Nature 2021

171

172

Laura Gerhard and Sven Hennig

Fluorescent Light-up Aptameres (FLAPs) are aptameric RNA sequence stretches that are able to bind a small organic molecule and thereby increase its intrinsically very low fluorescence by several orders of magnitude (Fig. 1a) [2]. They “start” to fluoresce. This effect is known for some time already, since Babendure and coworkers developed an RNA aptamer that was able to form a FLAP with the organic dye malachite green [3]. However, due to its severe cytotoxic side effects on cells, FLAPs got into the focus of interdisciplinary researchers, until 2011. It was Paige et al. who developed the nontoxic, cell-permeable organic dye (Z)-4(3,5-difluoro-4-hydroxybenzylidene)-1,2-dimethyl-1H-imidazol5(4H)-one (DFHBI), based on the chromophore present in the green fluorescent protein (GFP), and established an RNA aptamer sequence as a FLAP for it [4]. The complex of the folded RNA bound to DFHBI was called “Spinach,” due to its green light emission. Ever since a great variety of follow-up chromophores (either derived from DFHBI or as new developed scaffolds) and their accompanying RNA aptamers where published, increasing the available range of spectral characteristics (for further reading: Neubacher et al.) [2]. Fo¨rster Resonance Energy Transfer (FRET) is a process first described by Theodor Fo¨rster (1948) [5], in which the energy of a excited electron of a fluorescent molecule (donor) is transferred to a neighboring fluorescence molecule (acceptor) without any light emission. Therefore, the donor emission spectra needs to overlap with the excitation spectra of the acceptor molecule and the orientation of acceptor absorption dipole moment is ideally colinear to the donor emission dipole moment [6]. The result of this FRET process is twofold: (1) the acceptor molecule now hosts an excited electron that—upon to the return into the ground state—is able to emit a red-shifted emission light and (2) the resulting donor emission will be decreased by the energy proportion that was transferred to the acceptor molecule (Fig. 1b). This FRET system is highly distance dependent and therefore only molecules in close proximity will undergo FRET. Many FRET pairs based on genetically encodable biological fluorophores (GFP-RFP, CFP-YFP, and many more) are available and became tools to study complex biological systems using a broad variety of methods ranging from in vitro methods to fluorescent microscopy, life-cell imaging, and nanoscopy [7, 8]. Here the setup of a proximity based FRET RNA–protein assay that utilizes genetically encodable biological fluorophores is described (Fig. 1). It is shown how the Spinach RNA-FLAP and the mCherry protein can be used to investigate RNA–protein interactions. The Pseudomonas aeruginosa bacteriophage coat protein 7 (PP7) and its binding RNA sequence (pp7-RNA) [9, 10] to demonstrate REMSA studies using direct Spinach Fluorescence as well as a proximity based homogeneous FRET assay is used [11]. These protocols serve as a guide to adopt the system to other RNA–protein interactions.

Detecting RNA-Protein Interactions using Spinach-DFHBI

173

Fig. 1 Principle of an RNA–protein FRET assay using genetically encodable fluorescent RNA and protein tags. (a) Principle of Spinach-DFHBI fluorescence. The folding of the Spinach-RNA is stabilized by integration in the tRNALys scaffold. After binding of the small molecule DFHBI the fluorescence is enhanced. Unbound DFHBI in solution is almost nonfluorescent. (b) The pp7-RNA is integrated into the P2-loop of Spinach, thereby not disturbing its own or the folding of the Spinach-RNA. At the same time the PP7-Protein can bind to its RNA analog, thereby bringing mCherry in approximation to DFHBI and allowing donor quenching

2

Materials All materials and reagents are listed in the order of their appearance in chapter 3. Also, all entries are only mentioned once, although they might be needed in multiple protocol steps.

2.1 Equipment and Supplies

1. 0.5–10 μl, 2–20 μl, 20–200 μl, and 100–1000 μl pipettes.

2.1.1 RNA and Protein Cloning

3. Transparent 1.5 ml tubes.

2. Apparatus for gel electrophoresis. 4. 15 ml Cell-culture tubes. 5. Erlenmeyer flasks (5 l, 1 l, 250 ml, 50 ml).

174

Laura Gerhard and Sven Hennig

6. PCR purification kit. 7. Vortexer. 8. Tabletop centrifuge. 9. Household microwave. 10. Block Heater. 11. Alphaimager (Alphainnotech). 12. Milli-Q water (18 MΩ cm at 25
C, 2 ppb TOC). 2.1.2 In Vitro Transcription and Purification of RNA

1. HiScribe T7 High Yield RNA Synthesis Kit (NEB, E2040S). 2. Exclusion chromatography (Superdex 200 10/300 GL or equivalent). ¨ kta Pure FPLC (or equivalent). 3. A 4. Fraction tubes, RNase-free. 5. Spin concentrator 500 μl; 10 kDa cutoff. 6. Rotator waver for gel staining. 7. Biorad ChemiDoc (or equivalent).

2.1.3 Expression and Purification of Protein

MP

Gel

Documentation

System

1. Sterile pipettes 5 ml, 10 ml, and 25 ml. 2. Propipette. 3. Ultracentrifuge. 4. Microfluidizer (Model 110S, Microfluidics Corporation). 5. 1 ml Ni2+-NTA column HisTrap HP 6. Akta FPLC system (or equivalent). 7. Spin concentrator 500 μl; 10 kDa and 30 kDa cutoff. 8. NanoDrop One (Thermo Fisher Scientific) or equivalent. 9. 25 ml GSH column. 10. HiLoad 16/600 Superdex 200 pg (GE Healthcare, 28-989335). 11. HiLoad 26/60 Superdex 75 prep grade (GE Healthcare, 17-1070-01).

2.1.4 REMSAs

1. Horizontal electrophoresis system. 2. Power supply. 3. ChemiDoc MP (Bio-Rad).

2.1.5 Donor Quenching Assay and Competition Assay

1. 384-well plate (low-volume, black, PS, round bottom). 2. Safire2 fluorescence reader (Tecan) or equivalent. 3. Rotor A-2-DWP (for Eppendorf 5804 R centrifuge).

Detecting RNA-Protein Interactions using Spinach-DFHBI

2.2

Reagents

2.2.1 RNA and Protein Cloning

175

1. gBlock containing RNA sequence of interest. 2. pUC19I. 3. pGEX-6P-2 vector. 4. pOPIN-mCherry (pOPIN E-based vector). 5. PCR primer as mentioned in main text. 6. LE Agarose. 7. Tris–HCl. 8. Acetic acid. 9. EDTA. 10. Horizontal electrophoresis system. 11. Restriction enzymes (NEB): BglII, XhoI, KpnI-HF, HindIIIHF, BamHI-HF, EcoRI-HF, EagI-HF. 12. Restriction enzyme buffers (NEB): SmartCut buffer, NEBBuffer 3.1. 13. T4-DNA-Polymerase. 14. T4-DNA-Polymerase Buffer. 15. dCTP, 16. RecA. 17. Gel extraction kit. 18. Ethidium bromide. 19. 6 DNA loading dye. 20. 2-Log DNA Ladder. 21. T4-DNA-Ligase. 22. T4-DNA-Ligase buffer. 23. Chemical competent E. coli DH5α strain. 24. Ultracompetent E. coli OmniMax strain (Thermo Fisher Scientific) or equivalent. 25. LB-Medium. 26. LB-Agar for agar plates. 27. Ampicillin. 28. PCR clean-p columns.

2.2.2 In Vitro Transcription and Purification of RNA

1. DEPC (see Note 1). 2. NaCl. 3. KCl. 4. MgCl2. 5. Spermidin. 6. HiScribe T7 High Yield RNA Synthesis Kit (Thermo Fisher Scientific, E2040S).

176

Laura Gerhard and Sven Hennig

2.2.3 Expression and Purification of Protein

1. E. coli strain BL21(DE3). 2. IPTG. 3. Mercaptoethanol. 4. Imidazole. 5. DNAse I. 6. Lysozyme. 7. PMSF. 8. Glutathione. 9. PreScission Protease. 10. Urea.

2.2.4 REMSAs

1. 6 DNA loading dye. 2. DFHBI (synthesis according to Paige et al. [4], or equivalent).

3

Method In order to measure a direct FRET signal of a protein of interest and an RNA of interest using genetically encodable fluorophores, suitable FRET pairs need to be identified. Bajar et al. deliver a guide with the perspective on protein–protein interactions [7]. A lot of literature is published on what determines good FRET pairs, mainly: distance, orientation and donor emission-to-acceptor excitation overlap. Still many requirements apply of which no full control in the setup of FRET systems using biological molecules and genetically encodable tags was given. With raising numbers of FLAPs and a broad variety of fluorescent proteins available. More combinations might be suitable and would need to be tested during future assay optimizations. Here the focus was on Spinach-DFHBI as donor fused to the RNA of interest and the mCherry protein as acceptor fused to the protein of interest. For the direct measurement of RNA–protein interactions, the detail design rules and methods are explained step by step on the model system PP7-protein:pp7-RNA (Fig. 1b). The Pseudomonas phage coat protein 7 dimer (PP7) is known to bind to a 25 nt stretch of RNA (pp7-RNA). Subheading 3.1 will show all detailed steps from construction to actual synthesis and purification of the pp7-RNA fusions. Subheading 3.2 will guide you through the PP7-protein fusion design, production, and purification steps. In Subheading 3.3 details for the measurement of the direct interaction between PP7 and pp7-RNA are shared in a fluorescent REMSA and a proximity based quenching assay.

Detecting RNA-Protein Interactions using Spinach-DFHBI

3.1 Construct Design, Subcloning, Synthesis, and Purification Spinach-pp7-RNA Fusion 3.1.1 Construction of Spinach-pp7-RNA

177

Both RNAs—the Spinach aptamer and the RNA of interest (pp7-RNA)—need to be able to fold properly, which is essential for their correct function. For Spinach it is known that an insertion of its sequence into a tRNALys sequence stabilizes the Spinach stem region. For fusions with the Spinach aptamer one could either attach the RNA of interest to the 30 - or 50 -end of the stabilizing tRNALys or insert it into the P2- flexible loop of Spinach (Fig. 2a). For other FLAPs suitable insertion sites might need to be identified. For smaller and more compact FLAPs, a 30 - or 50 -fusion might work already, and no insertion is needed. For some FLAPs more than one chromophore is available which might need to be tested (e.g., DFHBI and DFHBI1-T). For overview see Neubacher et al. [2]. For the pp7-RNA as a model system for an RNA of interest, a crystal structure in complex with the PP7 protein dimer, was available [10]. The hairpin structure can be substituted with the P2-loop of Spinach (Fig. 2a). For other RNAs of interest a careful design of the construct is necessary. Elongated, double stranded RNA structures could be, similar to the design for the pp7-RNA, inserted into the P2-loop of Spinach. If no structural information (RNase protection data or crystal structures) are available, it might be helpful to use a structural prediction tools such as RNAfold or mfold [12, 13]. If in doubt, subcloning as many constructs as design options are available is recommended. Set of design rules: 1. Use Spinach on tRNALys support structure. 2. If target sequence has hairpin-like double stranded regions: insert them into Spinach-P2 loop region. 3. For more options: attach RNA of interest on 30 - and 50 -region of tRNALys. 4. Check for novel, alternative chromophores (e.g., DFHBI-1 T for Spinach). If other than Spinach FLAPs are used: 1. Check spectral characteristics for FRET in combination with the FRET partner. 2. Check for novel, alternative chromophores. 3. (a) Insertion regions might need to be identified. (b) In case of globular FLAPs: attach RNA of interest to 30 and 50 -end of FLAP. 4. Linker regions might need to be applied. Keeping in mind: FRET signal is distance dependent. For the construct used in Subheading 3.3, the P2 flexible loop RNA sequence is substituted with the pp7-RNA sequence (Fig. 2b, c).

178

Laura Gerhard and Sven Hennig

Fig. 2 Overview of RNA construct design options. (a) Schematic drawing tRNALys-Spinach-tRNALys. 50 - and 30 -end as labeled, gray: Stabilizing tRNALys-region, green: Spinach, green star: bound DFHBI chromophore, dotted line: flexible P2-loop region that could be used for RNA insertions, orange: pp7-RNA, which is substituted with P2-loop. (b) Subcloning strategy for Spinach tagged pp7-RNA. Vector map of pUC19Spinach-pp7. Indicated features: multiple cloning site with unique cutters, BglII/XhoI insertion position of the Spinach construct, ori: origin of replication, AmpR: ampicillin resistance 3.1.2 Subcloning of Spinach-pp7-RNA

For the production of the Spinach-tagged pp7-RNA, the coding DNA-sequence is subcloned into a high copy plasmid vector (pUC19 derivate, Fig. 2b). In addition to the construct design in Subheading 3.1.1, a T7-promotor sequence was added upstream of the RNA construct sequence in vitro transcription. The construct was commercially ordered as synthetic, double stranded DNA construct (see Note 2 for full sequence). These can be purchased via

Detecting RNA-Protein Interactions using Spinach-DFHBI

179

various vendors. Here gBlocks (Integrated DNA Technologies Inc.) are used. The target vector is then linearized via restriction and used as template for in vitro transcription (see Subheading 3.1.3). Additionally, by this, the coding sequence is easy to amplify, simply by vector plasmid amplification in E. coli and Mini/Midi preparation. Alternatively, any kind of molecular biology technique can be used for the generation of a DNA template for in vitro transcription (see Subheading 3.1.3). 1. PCR-amplify the Spinach-pp7 insert from synthetic dsDNA template (gBlock) using the following PCR oligo primer (underlined: BglII and XhoI restriction sites). Spinach-pp7_for: 50 -ATCGAGATCT CGATCCCGCGAAATTAATAC GACTCACTATAGG-30 . Spinach-pp7_rev: 50 -ATCGCTCGAGCAAAAAACCCCTCAAGACCCGTT TAGAGG-30 . 2. Purify PCR product by agarose gel electrophoresis and subsequent gel extraction. 3. Digest all Spinach-pp7 product from step 2 with 1 μl of BglII in NEB3.1 1 buffer (total volume 50 μl) for 1 h at room temperature. 4. Purify digest by PCR cleanup columns. 5. Digest all Spinach-pp7 product from step 4 with 2 μl of XhoI in SmartCut 1 buffer (total volume 50 μl) for 1 h at room temperature. 6. Purify digest by PCR cleanup columns. 7. Digest 1 μg target vector using the same procedure described in steps 3–6. 8. Purify linearized vector by agarose gel electrophoresis and subsequent extraction. 9. Ligate 25 ng vector from step 6 with insert from step 4 (25 ng, 75 ng and 250 ng, respectively) in 10 μl total volume ligase buffer (1 T4 DNA ligase buffer) and 1 μl T4 DNA Ligase. Incubate for 1 h at room temperature or overnight at 16
C. 10. Five microliters of the mixture is transformed into chemical competent E. coli DH5α. 11. Plate and incubate on LB (+50 μg/ml Amp) agar plates overnight at 37
C. 12. Pick several clones and check by Sanger sequencing.

180

Laura Gerhard and Sven Hennig

3.1.3 Synthesis of Spinach-pp7-RNA—In Vitro Transcription and Purification Step 1: Linearization of pUC19-Spinach-pp7 (see Subheading 3.1.2, Fig. 2b)

RNase-free working conditions apply (see Note 1).

1. Prepare restriction digest mix. (a) 20 μg pUC19-Spinach-pp7. (b) 2.0 μl CutSmart buffer (10, NEB). (c) 2.0 μl XhoI (20 U/μl, NEB). (d) add H2O (Milli-Q) up to 20 μl. 2. Incubate overnight at room temperature. 3. Purification of linearized vector by agarose gel electrophoresis and subsequent extraction.

Step 2a: In Vitro Transcription of Spinach-pp7 RNA

RNase-free working conditions apply (see Note 1). 1. In vitro transcription is performed following the instructions of HiScribe T7 High Yield RNA Synthesis Kit. (a) 2.0 μg of linearized and purified Spinach-pp7 vector from Subheading “Step 1: Linearization of pUC19-Spinachpp7”. (b) 2.0 μl T7-reaction buffer (10). (c) 2.0 μl T7 RNA Polymerase Mix (see NEB manual). (d) 2.0 μl ATP (100 mM). (e) 2.0 μl GTP (100 mM). (f) 2.0 μl UTP (100 mM). (g) 2.0 μl CTP (100 mM). (h) add H2O to 20 μl (DEPC treated Milli-Q water). 2. Incubate for 3 h at 37
C (for process monitoring follow Subheading “Step 2b: Reaction Process Monitoring”). 3. Immediately proceed to Subheading “Step 3: Purification of In Vitro Transcribed RNA”.

Step 2b: Reaction Process Monitoring

RNase free working conditions apply (see Note 1). 1. Add 20 μl of the reaction mixture of “Step 2a: In Vitro Transcription of Spinach-pp7 RNA” into one well of a 384-well plate. 2. Track the in vitro transcription every 5 min. With a Safire2 Fluorescence reader at 462 nm excitation wavelength and 506 nm emission wavelength with a preset gain of 60 (Fig. 3a).

Detecting RNA-Protein Interactions using Spinach-DFHBI

181

Fig. 3 Spinach-pp7 RNA synthesis and purification. (a) In vitro transcription of Spinach-pp7-RNA can be observed in real time. RFU506: Relative Fluorescence Units at 506 nm emission. (b) SEC chromatogram (Absorbance measured at 260 nm) of in vitro transcribe crude mix (10/300 S75, A¨tka Pure FPLC, GE Healthcare). Different species are indicated. A260: Absorbance at 260 nm Step 3: Purification of In Vitro Transcribed RNA

1. Prepare 1 RNA-buffer under RNase free condition (50 mM Tris–HCl pH 7.5, 300 mM KCl, 15 mM MgCl2, 2 mM spermidine in DEPC-treated Milli-Q water). 2. Preequilibrate size exclusion column (Superdex 200 10/300 ¨ kta Pure FPLC). GL on A 3. Inject reaction mix of Subheading “Step 2a: In Vitro Transcription of Spinach-pp7 RNA” (see step 3) to SEC column and collect 0.5 ml fractions. 4. Fractions containing the purified Spinach-pp7 RNA are identified on chromatogram (Fig. 3b, measured absorbance at 260 nm) and combined (see Note 3). 5. Combined fractions are concentrated up to 2–5 μM by ultrafiltration using a spin concentrator, 30 kDa cutoff. 6. Concentrated RNA is aliquoted, snap-frozen in liquid nitrogen, and stored at 80
C.

Step 4: Quality Check of Purified RNA.

Right before usage of Spinach-pp7-RNA in interaction assays (see Subheading 3.3.3) the quality of the RNA is checked via a fluorescence agarose gel electrophoresis: 1. Check concentration of RNA at 260 nm (NanoDrop OneC). 2. Cast a 2% agarose gel using 1 TAE-buffer (40 mM Tris–Cl (pH ¼ 8), 20 mM HAc, 1 mM EDTA). 3. Prepare 10 μl sample of a 250 nM RNA solution in 1 RNA-buffer (see Subheading “Step 3: Purification of In Vitro Transcribed RNA”, step 1).

182

Laura Gerhard and Sven Hennig

4. Electrophoresis is performed for 50 min at 100 V. Stain the gel in 1 RNA buffer containing 5 μM DFHBI for 1 h at room temperature. 5. Specific Spinach fluorescence is detected using ChemiDoc MP gel documentation system. 6. Subsequently stain gel in 1 RNA-buffer containing 1 μg/ml ethidium bromide for 30 min at room temperature. 7. Use UV-light excitation of AlphaImager to visualize total nucleic acid content of sample including potential DNA or RNA impurities. 3.2 Construction and Transformation of mCherry Fused PP7 Constructs 3.2.1 Construction of mCherry Fused PP7 Proteins

3.2.2 Cloning of mCherry Tagged PP7 Constructs

For the fluorescent labeling of the protein of interest (here the PP7 coat protein), mCherry is chosen as it fits spectroscopically as a FRET acceptor to Spinach as a FRET donor (Fig. 6). The fulllength PP7 coat protein is used as a non–capsid forming variant [14]. As FRET efficiency highly depends on the distance between the fluorophores, it is recommended to use N- and C-terminal fusions during assay development. Unlabeled PP7 as well as unaccompanied mCherry is chosen as negative control for further binding experiments. The sequences of the final proteins used can be found in Note 4. Standard microbiology techniques are used to generate the fusion protein plasmids. Resulting vector maps and important coding regions are mentioned in Fig. 4. The N-terminally mCherry tagged PP7-mCherry is subcloned by an cloning and protein facility using a combination of in vivoand Sequence Ligation Independent Cloning (SLIC) into a derivative of the pOPINe vector carrying a mCherry coding region (Fig. 4a) [15, 16]. 1. Digest 1 μg pOPIN-mCherry(N) with 1 μl of each restriction enzyme (KpnI-HF and HindIII-HF) in SmartCut buffer (1, total volume 50 μl) for 1 h at 37
C. 2. Purify linearized vector by agarose gel electrophoresis and subsequent extraction. 3. Amplify PP7 protein coding region [10] from template (e.g. gBlock) using the following oligo primer (underlined: pOPIN-mCherry specific sequence for in vivo-SLIC cloning, italic: stop codon). mCherry_PP7_for: 50 -GAAGTTCTGTTTCAGGGTCCC TCCAAAAC CATCGTTCTTGCGGT-30 . mCherry_PP7_rev: 50 -TAAACTGGTCTAGAAAGCT TTA ACGGCC CAGCGGCAC-30 .

Detecting RNA-Protein Interactions using Spinach-DFHBI

183

Fig. 4 Vector maps of N- and C-terminally mCherry tagged PP7 fusion proteins as well as unaccompanied mCherry and PP7

4. Purify PCR product by PCR cleanup columns. 5. Calculate required volume X of insert (clean PCR product from step 4): stoichiometric ratio of PCR:target vector ¼ 1:1. 6. Prepare Insert Mix. (a) X μl insert. (b) 0.033 μl T4-polymerase (3 U/μl).

184

Laura Gerhard and Sven Hennig

(c) 2.0 μl T4-polymerase buffer (10). (d) add H2O (Milli-Q) to 20 μl. 7. Incubate for 10 min at 22
C, then add 2 μl dCTP (10 mM). 8. Prepare target Vector Mix (parallel to step 5). (a) Y μl vector (25 ng linearized vector from step 2) (b) 0.33 μl T4-polymerase (3 U/μl). (c) 1.0 μl T4-polymerase buffer (10). (d) add H2O (Milli-Q) to 20 μl. 9. Incubate for 10 min, 22
C and add 2 μl dCTP (10 mM). 10. Prepare in vivo-SLIC Mix. (a) 0.02 μl RecA (2 μg/μl). (b) 2.0 μl 10 T4-DNA ligase buffer. (c) 9.0 μl Insert Mix (step 6). (d) 9.0 μl Vector Mix (step 8). (e) 20.02 μl total volume. 11. Incubate for max. 30 min, 37
C. Start with transformation of 1 μl SLIC Mix into ultracompetent OmniMax E. coli cells after 20 min incubation. Transformation into bacteria cells should be finished after 30 min total incubation time. 12. Plate and incubate on LB (+ 50 μg/ml Amp) agar plates overnight at 37
C. 13. Pick several clones and check by Sanger sequencing. The C-terminally mCherry-tagged PP7 is subcloned via a BamHI-HF/EcoRI-HF restriction-ligation procedure into the GST encoding pGEX-6P-2 backbone (Fig. 4b). 1. Amplify PP7-mCherry coding region from template (e.g., gBlock) using the following oligo primer (BamHI-HF and EcoRI-HF restriction sites are underlined, italic: stop codon): PP7_for: 50 -ATCGGGATCC TCCAAAAC 0 CATCGTTCTTGCGGTC-3 . mCherry_rev: 50 -ATCGGAATTCTTAGTGATGGTGATGGTGATGTT TAAACTGCTTGTAC -30 . 2. Purify PCR product using the Cycle Pure Kit. 3. Digest all product from step 2. with 1 μl of each restriction enzyme (BamHI-HF and EcroRI-HF) in SmartCut buffer (1, total volume 50 μl) for 1 h at 37
C. 4. Purify PCR product using the Cycle Pure Kit.

Detecting RNA-Protein Interactions using Spinach-DFHBI

185

5. Digest 1 μg pGEX-6P-2 with 1 μl of each restriction enzyme (BamHI-HF and EcroRI-HF) in SmartCut buffer (1, total volume 50 μl) for 1 h at 37
C. 6. Purify linearized vector by agarose gel electrophoresis and subsequent extraction. 7. Ligate 25 ng pGEX-6P2 vector and PP7-mCherry insert (use 1:1, 1:3, and 1:10 vector–insert ratio) with 1 μl T4 DNA ligase (NEB) in 10 μl total volume 1 T4 DNA ligase buffer for 1 h at room temperature or overnight at 16
C. 8. Five microliters of the mixture is transformed into chemical competent E. coli DH5α. 9. Plate and incubate on LB (+50 μg/ml Amp) agar plates overnight at 37
C. 10. Pick several clones and check by Sanger sequencing. The same strategy is used for subcloning mCherry (via BamHI/EcoRI) and PP7 (via BamHI/EagI) without fusion part as negative controls (Fig. 4c, d). The following oligonucleotides are used for PCR amplification (restriction sites are underlined, italic: stop codon): mCherry_for: 50 -ATCGGGATCC AGCGGTGTGAG 0 CAAGGGC -3 . mCherry_rev: 50 -ATCGGAATTCTTAGTGATGGTGATGGT GATGTTTAAACTGCTTGTAC -30 . PP7_for: 50 -ATCGGGATCC TCCAAAAC 0 CATCGTTCTTGCGGTC-3 . PP7_rev: 50 -ATCGCGGCCGTTAACGGCCCAGCGGCAC0 3. All resulting constructs are quality-checked via Sanger sequencing. 3.2.3 Expression and Purification of PP7 Constructs Step 1: Expression of PP7 Fusion Proteins

SDS-samples have been taken after every step and analyzed via SDS-Polyacrylamide-Gelelectrophoresis (SDS-PAGE). 1. Freshly transform the according expression plasmid into E. coli BL21(DE3). 2. Inoculate sterile 25 ml LB (+50 μg/ml Amp) overnight culture (volume should be about 1/40 of the planned main culture volume) and shake at 37
C 160 rpm. 3. Inoculate a sterile 1 L LB (+50 μg/ml Amp) main culture in a 5 L Erlenmeyer flask by adding the preculture (this should not exceed OD600 ¼ 0.1). 4. Shake the culture at 120 rpm, 37
C until the culture reaches a OD600 of 0.7.

186

Laura Gerhard and Sven Hennig

5. Quickly cool down the culture to 20
C (cold room, or on ice) and add 0.1 μM IPTG for protein induction. 6. Shake the culture overnight at 120 rpm, 20
C. 7. Harvest the cells by centrifugation at 3,000  g for 15 min (Beckmann Coulter Avanti JXN-26, JLA-8.1000 rotor). 8. Wash the cells (resuspension and centrifugation of the cells) three times with 100 ml lysis buffer (for GST-tagged fusion proteins: 50 mM Tris–Cl pH 7.5, 100 mM NaCl, 14 mM. 9. β-Mercaptoethanol, 1 mM EDTA; for His6-tagged fusion proteins: 50 mM Tris–Cl pH 7.5, 100 mM NaCl, 15 mM imidazole). 10. Resuspend the cells in 100 ml lysis buffer either immediately proceed to Subheading “Step 2: Cell Lysis” or store at 20
C (up to a few weeks). Step 2: Cell Lysis

From here, all steps are performed at 4
C. Take samples for SDS-PAGE analysis of the whole purification process, minimal samples to decide how to proceed during protein purification are mentioned in the following steps. More samples will help you understand how the purification proceeds. 1. Use freshly resuspended cells from Subheading “Step 1: Expression of PP7 Fusion Proteins” or thaw cells. 2. Add a tiny bit (tip of a spatula) of DNaseI and lysozyme powder to suspension. 3. Add PMSF to a final concentration of 0.1 mM. 4. Incubate for 1 h on a stirring plate (possible while thawing the cells). 5. Lyse cells on Microfluidizer).

ice

by

mechanical

shear

force

(e.g.,

6. Clear the lysate by ultracentrifugation at 51,250  g, 1 h, 4
C (Beckmann Coulter Avanti JXN-26, JA-25.50 rotor). 7. Decant supernatant and keep on ice. Step 3a: Purification of His6-Tagged mCherry-PP7

1. Preequilibrate 1 ml Ni2+-NTA affinity chromatography column ¨ kta Pure, flow (using peristaltic pump or sample pump on A rate: 1 ml/min) with wash buffer (50 mM Tris–Cl pH ¼ 7.5, 100 ml NaCl, 15 mM imidazole). 2. Apply supernatant to preequilibrated column (using peristaltic ¨ kta Pure). pump or sample pump on A 3. Wash with wash buffer until OD280 chromatogram reaches ¨ kta Pure). baseline (on A

Detecting RNA-Protein Interactions using Spinach-DFHBI

187

4. Elute column using a gradient of 5 column volumes (CV) ranging from 15 mM to 500 mM imidazole. Use 1 ml fractions. 5. Analyze purification process via Coomassie stained SDS-PAGE gels and pool target protein containing fractions. 6. Regenerate column using the supplier’s manual. 7. Concentrate the protein by ultrafiltration using a spin concentrator, 10 kDa cutoff and determine the concentration via OD280 and specific extinction coefficients (see Note 4, NanoDrop OneC). Alternative Step 2b: Purification of GST-tagged PP7-mCherry, PP7, and mCherry

1. Preequilibrate 25 ml GSH affinity chromatography column ¨ kta Pure, flow (using peristaltic pump or sample pump on A rate: 1 ml/min) wash buffer (50 mM Tris–HCl pH ¼ 7.5, 100 mM NaCl, 14 mM mercaptoethanol, 1 mM EDTA). 2. Apply supernatant to preequilibrated column (using peristaltic ¨ kta Pure). pump or sample pump on A 3. Wash with wash buffer until OD280 chromatogram reaches ¨ kta Pure). baseline (on A 4. Add 150 μl PreScission Protease to 25 ml of wash buffer and circulate the buffer overnight over the column with 0.5 ml/ min rate (overnight on column digest of target protein). 5. Collect all cut protein solution by washing the protein off the column (use wash buffer until OD280 is down to base line or use about 1.5 CV). 6. Regenerate the column (a) using elution buffer (wash buffer +10 mM GSH, pH ¼ 7.5) and (b) using 8 M Urea followed by at least 5 CV water (or use supplier’s manual). Concentrate the protein by ultrafiltration using a spin concentrator, 10 kDa cutoff or 30 kDa cutoff respectively and determine the concentration via OD280 and specific extinction coefficients (see Note 4, NanoDrop OneC).

Step 3: Size exclusion Chromatography

Size exclusion chromatography (SEC) is performed for (1) further purification of the proteins and (2) for quantitative buffer exchange into RNA-buffer and into RNase free conditions (see Note 1) for future assays. Depending on the protein amount, SEC column size and number of runs are determined. HiLoad 16  600 pg

Max. capacity

30% of total bacterial RNA) [13]. High-level expression also facilitated the purification of target BERAs, which were demonstrated to be biologically and pharmacologically active in the control of target gene expression, modulation of cellular processes, and management of specific diseases in preclinical models [12, 13, 18–21]. Herein, we describe a protocol for the expression and purification of recombinant or bioengineered RNAs using the optimal tRNA/pre-miR-34a carrier. In particular, the miR-34a duplexes may be substituted by target miRNAs, siRNAs (e.g., GFP-siRNA), or other forms of sRNAs along with respective complementary sequences (Fig. 1). Meanwhile, RNA aptamers (theophylline aptamer (TPA), EpCAM aptamer (EpCAMA), etc.) or other single-stranded RNAs can be directly inserted into the tRNA/pre-miRNA carrier to achieve overexpression (Fig. 1). Detailed instructions are provided in the following sections for reproducible production of high-purity BERA/sRNAs ready for basic and applied studies.

2

Materials All solutions were made by using distilled and deionized (dd) water and biological grade reagents. Prepare and store all reagents at room temperature (unless stated otherwise). All procedures are carried out by following standard RNase-free practices (see Note 1).

Expression and Purification of tRNA/pre-miRNA-Based Recombinant Noncoding RNAs

251

Fig. 1 Design of bioengineered noncoding RNA agents (BERAs) using a tRNA/pre-miRNA carrier and construction of target BERA/sRNA-expressing plasmid. (a) Target miRNA or siRNA (red) and complementary sequence (green) may be designed to replace miR-34a within the tRNA/pre-miR-34a carrier to achieve highyield production of BERA/miRNA or siRNA. In addition, an aptamer or sRNA can be inserted between tRNA and pre-miRNA toward the expression of target BERA/sRNA. (b) Cloning primers are designed to span the 15-nt from restriction enzyme sites (blue), overlapping pre-miR-34a spanning toward the inserted sRNA on both ends. The tRNA sequence is highlighted in yellow, and pre-miRNA sequence is in black in which miRNA duplex sequences are substituted with target siRNA or miRNA (red) and complementary sequence (green). (c) Coding sequence of BERA/sRNA is cloned into the pBSTNAV vector consisting of a strong lipoprotein (lpp) gene promotor, a ribosomal RNA operon transcription terminator (rrnC). Positive plasmids could be selected through ampicillin resistance and verified by DNA sequencing 2.1 Cloning, Bacterial Transformation and Culture 2.1.1 Laboratory Equipment

1. Thermocycler. 2. Water bath. 3. Agarose Horizontal Electrophoresis System. 4. Incubator. 5. Shaking Incubator. 6. Magnetic stirrer. 7. Laboratory balances. 8. NanoDrop.

252

Mei-Juan Tu et al.

9. Gel Imaging System. 10. Microcentrifuge. 2.1.2 Bacterial Culture and Transformation

1. Diethyl pyrocarbonate (DEPC) water: Add 1 mL of DEPC (>97%) to 1 L dd water and stir at room temperature for 2 h (see Note 2). Autoclave at 121
C for 30 min. Store at 4
C. 2. 10 TAE buffer: Dissolve 48.4 g of Tris base, 57.1 mL of glacial acetic acid (100%), and 9.4 g of EDTA in 800 mL autoclaved DEPC water. Adjust volume to 1 L with autoclaved DEPC water. Store at 4
C. 3. 3% Agarose gel: Add 1.50 g of agarose to 50 mL of 1 TAE buffer, and heat the mixture until agarose is dissolved. When the solution cools to ~55
C, add 1 μL of ethidium bromide. Immediately pour the solution into the previously prepared gel mold. 4. Ampicillin stock solution (100 mg/mL, 1000): Dissolve 5 g of ampicillin powder in 50 mL of 70% ethanol. Store at 20
C. 5. LB or 2YT medium: Add 20 g of LB Broth powder or 31 g of 2YT medium broth to 1 L of dd water and sterilize by autoclaving at 121
C for 15 min. If required, antibiotics (ampicillin was used in this study) can be added to a final concentration of 100 mg/L, after the autoclaved medium is cooled (~55
C). The medium can be stored at 4
C for up to 1 month. 6. LB agar plate: Add 15 g of agar, 20 g of LB Broth powder to 1 L of dd water. Autoclave the mixture at 121
C for 15 min. After ampicillin was added, pour a layer of solution into the Petri dishes to cover the plate (about 15 mL/plate). Leave the plates on the bench for at least 1 h to solidify. Store at 4
C. 7. Stellar™ Competent Cells (E. coli HST08 strain) (Takara Bio, Mountain View, CA). 8. E. coli DH5α competent cells. 9. Sequences of model BERAs and corresponding cloning primers are listed in Table 1. 10. pBSTNAV plasmid containing tRNA/pre-miR-34a scaffold [13] may be utilized as a template for PCR cloning (see Note 3). 11. PCR Premix: CloneAmp HiFi PCR Premix (Takara). 12. Vector digestion enzymes and buffer: restriction enzyme EagIHF® (20,000 units/mL, New England Biolabs, Ipswich, MA), SacII (20,000 units/mL, New England Biolabs), 10 CutSmart® Buffer (New England Biolabs).

180

180

246

199

BERA/GFPsiRNA

BERA/ TPA30 + 50

BERA/ EpCAMA30

GGCUACGUAGCUCAGUUGG UUAGAGCAGCGGCCGGGCCAGCUGUGAGUGUUUC UUUGGCAGUGUCUUAGCUGGUUGUUGUGAGCAA UAGUAAGGAAGCAAUCAGCAAGUAUACUGCCC UAGAAGUGCUGCACGUUGUUGGCCCGCGACUGG UUACCCGGUCGCCGCGGGUCACAGGUUCGAA UCCCGUCGUAGCCACCA

GGCUACGUAGCUCAGUUGG UUAGAGCAGCGGCCGGGCGA UACCAGCCGAAAGGCCCUUGGCAGCGUCGGCCAGC UGUGAGUGUUUCUUUGGCAGUGUCUUAGCUGG UUGUUGUGAGCAAUAGUAAGGAAGCAAUCAGCAAG UAUACUGCCCUAGAAGUGCUGCACGUUG UUGGCCCGGCGAUACCAGCCGAAAGGCCC UUGGCAGCGUCCCGCGGGUCACAGGUUCGAA UCCCGUCGUAGCCACCA

GGCUACGUAGCUCAGUUGG UUAGAGCAGCGGCCGGGCCAGCUGUGAGUGUUUC UUAGUUGUACUCCAGCUUGUGCCCUGUGAGCAA UAGUAAGGAAGGGCACAAGUGGUAGUACAACC UAGAAGUGCUGCACGUUGUUGGCCCCCGCGGG UCACAGGUUCGAAUCCCGUCGUAGCCACCA

GGCUACGUAGCUCAGUUGG UUAGAGCAGCGGCCGGGCCAGCUGUGAGUGUUUC UUUGGCAGUGUCUUAGCUGGUUGUUGUGAGCAA UAGUAAGGAAGCAAUCAGCAAGUAUACUGCCC UAGAAGUGCUGCACGUUGUUGGCCCCCGCGGG UCACAGGUUCGAAUCCCGUCGUAGCCACCA

Length (nt) Sequence (50 !30 )

BERA/miR34a-5p

BERA

F GTTAGAGCAGCGGCCGGGCCAGCTGTGAGTGTTTCT R TCGAACCTGTGACCCGCGGCGACCGGGTAACCAG TCGCGGGCCAACAACGTGCAGCAC

F GTTAGAGCAGCGGCCGGGCGA TACCAGCCGAAAGGCCCTTGGCAGCGTCGGCCAGC TGTGAGTGTTT R TCGAACCTGTGACCCGCGGGACGCTGCCAAGGGCC TTTCGGCTGGTATCGCCGGGCCAACAACGTGCAGC

F TCGAACCTGTGACCCGCGGGGGCCAACAACG TGCAGCACTTCTAGGTTGTACTACCACTTGTGCCC TTCCTTACTATTGC R GTTAGAGCAGCGGCCGGGCCAGCTGTGAGTGTTTC TTAGTTGTACTCCAGCTTGTGCCCTGTGAGCAATAG TAA

F GTTAGAGCAGCGGCCGGGCCAGCTGTGAGTGTTTC TTTG R TCGAACCTGTGACCCGCGGGGGCCAACAACG TGCAGC

Cloning primers (50 !30 )

Table 1 Sequences of target BERAs and respective cloning primers. The pBSTNAV plasmid containing the optimum tRNA/pre-miR-34a carrier [13] was used as a template during PCR cloning, except for BERA/GFP-siRNA whose cloning primers were extended to have about 15 nt overlaps

Expression and Purification of tRNA/pre-miRNA-Based Recombinant Noncoding RNAs 253

254

Mei-Juan Tu et al.

13. Infusion enzyme and buffer: 5 In-Fusion HD Enzyme Premix (Takara). 14. Gel cleanup kit: NucleoSpin gel and PCR cleanup kit (Takara). 15. Plasmid extraction: QlAprep Spin Miniprep Kit (QIAGEN, Germantown, MD).

2.2 RNA Extraction and Denaturing Urea Polyacrylamide Gel Electrophoresis (PAGE)

1. Vertical Electrophoresis Cell. 2. Mini-PROTEAN® Tetra Cell Casting Module and comb, 15-well, 1.0 mm, 26 μL. 3. Platform Rocker.

2.2.1 Laboratory Equipment 2.2.2 RNA Extraction and Denaturing Urea PAGE

1. 10 mM magnesium acetate–Tris-hydrochloride (HCl) solution: Add 214.45 mg of magnesium acetate and 121.14 mg of Tris–HCl to 90 mL of autoclaved DEPC water. Adjust pH to 7.4 and adjust volume to 100 mL. Store at 4
C. 2. 5 M sodium chloride (NaCl): Add 29.22 g of NaCl to 90 mL of autoclaved DEPC water. Mix and adjust volume to 100 mL. Store at 4
C. 3. 10% Ammonium persulphate solution (APS): Dissolve 1 g of APS in 10 mL of autoclaved DEPC water. Store aliquots (1 mL of each) at 20
C. 4. Denaturing urea (8 M) polyacrylamide (8%) gel solution (one gel): Add 0.6 mL of 10  TAE, 1.2 mL of 40% acrylamide–bis (19:1, 1610144, Bio-Rad), and 3.0 g urea to a 50 mL conical tube. Adjust volume to 6 mL with autoclaved DEPC water. Dissolve the urea by vortex and shaking. Mix the solution with 30 μL of 10% APS, 6 μL TEMED, and cast gel with a Cell Casting Module (see Note 4).

2.3

RNA Purification

2.3.1 Laboratory Equipment

1. NGC Quest™ 10 FPLC System. 2. BioFrac™ Fraction Collector. 3. Enrich™ Q 10  100 column. 4. Centrifuge. 5. Amicon® Ultra-2 centrifugal filter concentrator (2 mL, 30K, MilliporeSigma, St. Louis, MO).

2.3.2 Solutions

1. Buffer A: 10 mM NaH2PO4. Dissolve 1.38 g of NaH2PO4 in 900 mL of DEPC water. Adjust pH to 7.0 and volume to 1000 mL. Store at 4
C.

Expression and Purification of tRNA/pre-miRNA-Based Recombinant Noncoding RNAs

255

2. Buffer B: 10 mM NaH2PO4, 1 M NaCl. Dissolve 1.38 g of NaH2PO4 and 58.44 g NaCl in 900 mL of DEPC water. Adjust pH to 7.0 and volume to 1000 mL. Store at 4
C. 2.4 RNA Purity Analyses 2.4.1 Laboratory Equipment

1. Shimadzu LC-20 AD Prominence Ultra-Fast Liquid Chromatography system equipped with binary pumps, an on-line degassing unit, an autosampler a UV photodiode array detector, and a column oven (Shimadzu, Kyoto, Japan). 2. XBridge OST C18 column (2.1  50 mm, 2.5 μm particle size; Waters, Milford, MA). 3. Microplate reader.

2.4.2 Solutions and Reagents

1. Hexafluoro-2-propanol (HFIP). 2. Buffer C: Add 0.6 mL of TAE (8.6 mM) and 5.3 mL of HFIP (100 mM) to 450 mL of HPLC grade water. Adjust volume to 500 mL with water. 3. Buffer D: Add 0.6 mL of TAE (8.6 mM) and 5.3 mL of HFIP (100 mM) to 450 mL of methanol. Adjust volume to 500 mL with methanol. 4. Endotoxin determination kit: Pyrogent-5000 kinetic LAL assay (Lonza, Walkersville, MD).

3

Methods

3.1 Design and Construction of BERA/ sRNA-Expressing Plasmid

1. Define target miRNA, siRNA, or sRNA sequences (see Fig. 1 and Table 1).

3.1.1 Design of Target BERA/sRNA and Corresponding Cloning Primers

3. Design the forward and reverse cloning primers by spanning upstream and downstream 15 nt from restriction site (see Note 5).

2. Add two restriction sites, EagI and SacII, to the 50 and 30 end of the target BERA coding sequence, respectively, to infuse the target BERA to the expression plasmid, pBSTNAV.

4. Extend the primers to overlap pre-miR-34a spanning toward the target ncRNA on both ends (see Fig. 1, Note 6, and Table 1). 5. The mature miR-34a-5p and complementary sequences within the tRNA/pre-miR-34a carrier are replaced by target miRNA, siRNA or sRNA sequences (see Fig. 1, Note 7). In this chapter, GFP-siRNA is used as a model to illustrate the expression and purification of target BERA/sRNA (Table 1). 6. An RNA aptamer or sRNA can be inserted to the 30 or 50 of the pre-miR-34a (see Fig. 1a). For example, theophylline aptamer (TPA) is inserted to both the 30 and 50 of pre-miR-34a to construct BERA/TPA30 + 50 , and EpCAM aptamer

256

Mei-Juan Tu et al.

(EpCAMA) is inserted to the 30 of pre-miR-34a to construct BERA/EpCAMA30 (Table 1). 3.1.2 PCR Amplification of Target Insert

1. Add 2 μL of forward primer (10 μM) and reverse primer (10 μM), 25 μL of PCR Premix to 21 μL dd water to make a 50-μL PCR reaction system. 2. Run PCR under the following condition: (a) 95.0
C for 30 s. (b) 95.0
C for 10 s. (c) 68.0
C for 30 s (see Note 8). (d) 72.0
C for 30 s. (e) Repeat steps (b)–(d) for 30 cycles. (f) 72.0
C for 10 min. (g) 4
C forever.

3.1.3 Vector Preparation, DNA Product Isolation, and Ligation

1. Add 1 μg of pBSTNAV-tRNA/pre-miR-34a plasmid [13], 1 μL of restriction enzyme EagI, 1 μL of restriction enzyme SacII, and 5 μL of 10 CutSmart Buffer to dd water to a 50-μL reaction mixture. 2. Incubate at 37
C for 2 h. 3. Load the PCR products or digested vector products into an agarose gel (50 μL/well). 4. Separate the target band from others by running gel electrophoresis (100 V for about 20 min). 5. Cut and collect the band containing target fragment in an open UV detector box. 6. Isolate target DNA from the collected gel using a gel cleanup kit. 7. Elute the sample with 25 μL (PCR product) or 50 μL dd water (digestion product). 8. Mix 7 μL of PCR cleanup product and 1 μL of vector isolated above with 2 μL of 5  In-Fusion HD Enzyme Premix. 9. Incubate at 50
C for 15 min.

3.1.4 Transformation

1. Add 5 μL of ligation product to 20 μL of Stellar™ Competent Cells, and mix gently with pipette tips. Incubate on ice for 30 min. 2. Incubate the mixture in water bath at 42
C for 45 s (do not shake the tubes). 3. Put tubes back on ice for 2 min. 4. Add 800 μL of LB medium (without antibiotics) and mix with pipette gently.

Expression and Purification of tRNA/pre-miRNA-Based Recombinant Noncoding RNAs

257

5. Shake the cultures at 225 rpm at 37
C for 60 min. 6. Plate the transformation on a LB-ampicillin agar plate and incubate overnight at 37
C. 3.1.5 Plasmid Amplification, Mini Preparation, and Sequence Verification

1. Pick up 3–5 colonies from the LB-agar plate and transfer each into a 50-mL conical tube containing 15 mL LB-ampicillin medium. 2. Shake overnight at 225 rpm at 37
C. 3. Extract plasmids from the overnight culture (see Note 9). 4. Send out the plasmids for sequencing. The plasmids with correct sequences are thus used for RNA expression.

3.2 Fermentation Production of Target BERA/sRNA

1. Mix 50–100 ng of sequence-confirmed plasmids with 20 μL of Stellar™ Competent Cells and incubate on ice for 30 min (see Note 10).

3.2.1 Small-Scale Expression of BERA/sRNAs

2. Follow the rest steps described in Subheading 3.1.4. 3. Add the transformation product to 15 mL of ampicillincontaining 2  YT medium in a 50 mL conical tube (see Note 11). 4. Shake overnight at 225 rpm and 37
C.

3.2.2 Large-Scale Expression of BERA/sRNAs

1. Mix 100–200 ng of sequence-confirmed plasmids with 30 μL Stellar™ Competent Cells and incubate on ice for 30 min (see Note 10). 2. Follow the rest steps described in Subheading 3.1.4. 3. Add the transformation product to 600 mL ampicillincontaining 2  YT medium in a 2-L flask. 4. Shake overnight at 225 rpm at 37
C.

3.2.3 Isolation of Total Bacterial RNA

1. Transfer the broth to multiple 250-mL round bottom bottles, centrifuge at 9000  g and 4
C for 6 min (for small-scale RNA expression, centrifuge the tube directly, see Note 11), and then remove the supernatant. 2. Add 4 mL (400 μL for small-scale expression) of 10 mM magnesium acetate—Tris–HCl solution to resuspend the pellet. 3. Combine and transfer the resuspension to a 50-mL tube (about 20 mL in each tube; about 500 μL in a 2-mL tube for small-scale), and then add equal volume of phenol (about 20 mL; around 500 μL for small-scale). 4. Shake the tubes gently on a rocker for 20–60 min. 5. Centrifuge at 10,000  g at 4
C for 10 min. Transfer the aqueous phase (about 16 mL; ~400 μL for small-scale) to fresh tube(s).

258

Mei-Juan Tu et al.

6. Add 10% volume (about 1.6 mL; ~40 μL for small-scale) of 5 M NaCl, and centrifuge at 10,000  g and 4
C for 10 min. 7. Carefully transfer the supernatant to new 50-mL conical tube (s) (2-mL tube(s) for small-scale). 8. Add 2 volumes (about 30 mL; ~800 μL for small-scale) of pure ethanol (97%) by a single run on Enrich-Q 10  100 column. If sample does not reach to desired purity, repurification may be conducted using the same or different columns [14]. 15. To identify the elution time of target BERA peak, total RNA of wild type HST08 bacteria may be separated by the same FPLC method ahead of time. 16. To remove a large amount of salt within the FPLC fractions, it is necessary to wash three times. Make sure that the filter is not overloaded with RNA. 17. We usually store the purified RNA at a high concentration (e.g., >5 mg/mL) at 80
C. Dilute it to a proper working solution and aliquot it for regular use. Even though we have found that BERAs are stable over 50 freeze–thaw cycles, it is still recommended to minimize freeze–thaw cycles for long term storage and avoiding possible contamination.

Acknowledgments This study was supported by National Cancer Institute (grant No. R01CA225958) and National Institute of General Medical Sciences (R01GM113888), National Institutes of Health. References 1. Ambros V (2004) The functions of animal microRNAs. Nature 431:350–355 2. Cech TR, Steitz JA (2014) The noncoding RNA revolution-trashing old rules to forge new ones. Cell 157:77–94 3. Setten RL, Rossi JJ, Han SP (2019) The current state and future directions of RNAi-based therapeutics. Nat Rev Drug Discov 18:421–446 4. Yu AM, Jian C, Yu AH et al (2019) RNA therapy: are we using the right molecules? Pharmacol Ther 196:91–104 5. Bennett CF (2019) Therapeutic antisense oligonucleotides are coming of age. Annu Rev Med 70:307–321 6. Levin AA (2019) Treating disease at the RNA level with oligonucleotides. N Engl J Med 380:57–70 7. Bramsen JB, Kjems J (2012) Development of therapeutic-grade small interfering RNAs by chemical engineering. Front Genet 3:154 8. Khvorova A, Watts JK (2017) The chemical evolution of oligonucleotide therapies of clinical utility. Nat Biotechnol 35:238–248

9. Liu YP, Berkhout B (2011) miRNA cassettes in viral vectors: problems and solutions. Biochim Biophys Acta 1809:732–745 10. Ponchon L, Beauvais G, Nonin-Lecomte S et al (2009) A generic protocol for the expression and purification of recombinant RNA in Escherichia coli using a tRNA scaffold. Nat Protoc 4:947–959 11. Ponchon L, Dardel F (2007) Recombinant RNA technology: the tRNA scaffold. Nat Methods 4:571–576 12. Chen Q-X, Wang W-P, Zeng S et al (2015) A general approach to high-yield biosynthesis of chimeric RNAs bearing various types of functional small RNAs for broad applications. Nucleic Acids Res 43:3857–3869 13. Ho PY, Duan Z, Batra N et al (2018) Bioengineered noncoding RNAs selectively change cellular miRNome profiles for cancer therapy. J Pharmacol Exp Ther 365:494–506 14. Petrek H, Batra N, Ho PY et al (2019) Bioengineering of a single long noncoding RNA molecule that carries multiple small RNAs. Appl Microbiol Biotechnol 103:6107–6117

Expression and Purification of tRNA/pre-miRNA-Based Recombinant Noncoding RNAs 15. Li M-M, Addepalli B, Tu M-J et al (2015) Chimeric microRNA-1291 biosynthesized efficiently in Escherichia coli is effective to reduce target gene expression in human carcinoma cells and improve chemosensitivity. Drug Metab Dispos 43:1129–1136 16. Li M-M, Wang W-P, Wu W-J et al (2014) Rapid production of novel pre-microRNA agent hsa-mir-27b in Escherichia coli using recombinant RNA technology for functional studies in mammalian. Cell 42:1791–1795 17. Wang W-P, Ho PY, Chen Q-X et al (2015) Bioengineering novel chimeric microRNA34a for prodrug cancer therapy: high-yield expression and purification, and structural and functional characterization. J Pharmacol Exp Ther 354:131–141 18. Jilek JL, Zhang QY, Tu MJ et al (2019) Bioengineered let-7c inhibits Orthotopic

265

hepatocellular carcinoma and improves overall survival with minimal immunogenicity. Mol Ther Nucleic Acids 14:498–508 19. Li PC, Tu MJ, Ho PY et al (2018) Bioengineered NRF2-siRNA is effective to interfere with NRF2 pathways and improve chemosensitivity of human cancer cells. Drug Metab Dispos 46:2–10 20. Tu MJ, Ho PY, Zhang QY et al (2019) Bioengineered miRNA-1291 prodrug therapy in pancreatic cancer cells and patient-derived xenograft mouse models. Cancer Lett 442:82–90 21. Yi WR, Tu MJ, Liu Z et al (2020) Bioengineered miR-328-3p modulates GLUT1mediated glucose uptake and metabolism to exert synergistic antiproliferative effects with chemotherapeutics. Acta Pharm Sin B 10:159–170

Chapter 19 Synthetic Biology Medicine and Bacteria-Based Cancer Therapeutics Jaehyung Lee, Andrew C. Keates, and Chiang J. Li Abstract Spontaneous tumor regression following bacterial infection has been observed for hundreds of years. These observations along with anecdotal medical findings in 1890s led to the development of Coley’s “toxins,” consisting of killed Streptococcus pyogenes and Serratia marcescens bacteria, as the first cancer immunotherapy. The use of this approach, however, was not widely accepted at the time especially after the introduction of radiation therapy as a treatment for cancer in the early 1900s. Over the last 30–40 years there has been renewed interest in the use of bacteria to treat human solid tumors. This is based on the observation that various nonpathogenic anaerobic bacteria can infiltrate and replicate within solid tumors when given intravenously. Bacteria tested as potential anticancer agents include the Gram-positive obligate anaerobes Bifidobacterium and Clostridium, as well as the gram-negative facultative anaerobe Salmonella. Recent advances in synthetic biology and clinical success in cancer immunotherapy provide renewed momentum for developing bacteria-based cancer immunotherapy for cancer treatment and should allow greater potential for the development of novel therapeutic approaches for this devastating disease. Key words Transkingdom gene silencing, tkRNAi, Bacterial RNAi, Functional genomics, RNAitherapy

1

Introduction Recent advances in cancer immunotherapy have revolutionized the treatment of a wide variety of tumors [1, 2]. Although this approach is typically thought of as a recent medical advance, the origins cancer immunotherapy can be traced back to antiquity. In particular, attempts to induce tumor regression through deliberate bacterial infection have been reported anecdotally for hundreds of years. For example, in writings ascribed to the Egyptian physician Imhotep (c 2600 BC) the recommended treatment for tumors was a poultice followed by an incision to induce infection and tumor regression [3]. Moreover, in the 1700 and 1800s it was commonplace to apply septic dressings to ulcerated tumors, or to deliberately infect tumors to cause erysipelas or gangrene [4, 5]. In the

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0_19, © Springer Science+Business Media, LLC, part of Springer Nature 2021

267

268

Jaehyung Lee et al.

1860s, Fehleisen and Busch reported significant tumor regression in patients after erysipelas infection [6, 7]. The bacterial strain responsible for the erysipelas and tumor regression has been previously identified as Streptococcus pyogenes [8]. William Coley MD is now widely recognized as the father of modern immunotherapy. As an orthopedic surgeon at New York Memorial Hospital in the 1890s, he noticed that a number of postoperative cancer patients underwent spontaneous tumor regression after developing erysipelas, suggesting that these recoveries were mediated by the immune system [9]. Starting in 1891, Coley began a systematic investigation of the ability of different mixtures of live and attenuated Streptococcus pyogenes and Serratia marcescens to induce antitumor responses, eventually administering what became known as “Coley’s toxin,” the first example of a cancer immunotherapeutic, to over 1000 patients [9]. Using this approach, Coley achieved complete remission in several types of malignancies including sarcoma, lymphoma, and testicular carcinoma [9]. However, the use of Coley’s toxin was controversial and not widely accepted by the medical community. This was mainly due to its unknown mechanism of action, and the risk associated with deliberately infecting cancer patients with pathogenic bacteria. Moreover, the technique was labor intensive and difficult to standardize. Ultimately, the development of radiation therapy in the early 1900s, which was more predictable and less labor intensive, led oncologists to largely abandon the use of Coley’s toxin as a treatment for cancer. Over the last 30 years, there has been renewed interest in the use of bacteria to treat cancer [10–12]. This is based on the observation that various nonpathogenic anaerobic bacteria can infiltrate and replicate within solid tumors when given intravenously. Solid tumors generally have a dysregulated vasculature that generates areas of hypoxia that surround a central necrotic core in the tumor mass. These poorly perfused areas are resistant radiotherapy and chemotherapy interventions and can, therefore, lead to tumor regrowth and relapse following treatment. It is these same hypoxic/necrotic regions, however, that can be exploited by anaerobic bacteria for enhanced tumor targeting. Bacteria tested as potential anticancer agents include the grampositive obligate anaerobes Bifidobacterium and Clostridium, as well as the gram-negative facultative anaerobe Salmonella [10, 12]. In addition to having varying degrees of intrinsic antitumor activity, many of these bacteria have been genetically engineered to improve their safety and/or endow them with improved therapeutic potential. For example, Clostridium novyi was attenuated by deletion of phage encoding α-toxin, to yield Clostridium novyi-NT [13]. Similarly, deletion of the msbB gene from Salmonella, which prevents myristoylation of the lipid A component of LPS, reduced bacterial toxicity by 10,000-fold [14]. More recently,

Transkingdom Gene Silencing

269

tumor targeting bacteria also have be engineered to successfully deliver a range of therapeutic molecules such as cytotoxic agents, pro-drug converting enzymes, immunomodulators, gene silencers and synthetic gene circuits [10, 12, 15]. Although attenuated tumor-targeting bacteria are well-suited for the selective delivery of various payloads to solid tumors, there have been relatively few clinical trials investigating this approach. Phase I studies with Salmonella have largely focused on VNP20009, a genetically modified strain of S. typhimurium in which the msbB and purI genes have been deleted. Although cancer patients tolerated intravenous doses of VNP20009 up to 3
108 CFU/m2, consistent tumor colonization and objective tumor regression were not observed [16–20]. However, VNP20009 expressing E. coli cytosine deaminase (S. typhimurium TAPET-CD) has proven effective for the selective conversion of the pro-drug 5-fluorocytosine to 5-fluorouracil [21]. Furthermore, a phase I/II clinical trial investigating treatment of cancer patients with advanced solid tumors with Bifidobacterium longum expressing cytosine deaminase (B. longum APS001F) is ongoing [22]. Intravenous or intratumoral administration of C. Novyi-NT spores to cancer patients has also been investigated in several phase I trials. These studies have demonstrated evidence of tumor colonization as well as objective tumor regression [23–25]. A phase Ib trial investigating combination therapy with pembrolizumab and intratumoral injection C. Novyi-NT spores is currently ongoing [26]. The ability to enhance cellular capabilities through synthetic biology has emerged as an important approach for the development of novel therapies for a variety of conditions, including cancer [27]. Previously, we have reported that bacteria can be engineered using synthetic biology technology to enable modified nonpathogenic bacteria to express gene silencer, invading transformed cells, escaping endosome and selective silence oncogenes in mammalian cells (termed transkingdom gene silencing or tkRNAi) in vitro and in vivo (Fig. 1) [15]. To translate this basic science technology toward clinical application, we successfully engineered nonpathogenic E. coli or attenuated Salmonella typhimurium to encode RNAi against beta-catenin or mutant K-Ras, and demonstrated that oral administration of these engineered bacteria can mediate potent and specific beta catenin gene silencing in intestinal epithelial cells, and significantly reduce polyp formation in APCmin/+ mice [15]. We further designed an optimized tkRNAi-based therapy for treating familial adenomatous polyposis (FAP) patients. This tkRNAi therapeutic received FDA approval for phase Ia/II trials, and was the first synthetic biology–based therapeutic to receive FDA clearance for clinical testing as well as the first oral gene-therapy to enter clinical trials [28].

270

Jaehyung Lee et al.

invasin shRNA

Inducible single-target shRNA expression (E.coli)

Endosome

Dicer

RISC

Target mRNA recognition

mRNA cleavage

Fig. 1 Schematic representation of transkingdom gene silencing using inducible single target bacterial shRNA expression

Since the publication of our original study, we have investigated the possibility of developing a next generation transkingdom RNAi gene silencing technology. As part of these studies, we have designed a novel transkingdom gene silencing vector capable of constitutively expressing long double stranded RNA. Cleavage of long dsRNA by RNase III in bacteria should generate a “cocktail” of silencing RNA thereby significantly enhancing target gene silencing. Moreover, this approach can be easily adapted to

Transkingdom Gene Silencing

invasin

271

Constitutive multi-target long dsRNA expression (E.coli)

cleaved long dsRNA

Endosome

Dicer?

RISC Target mRNA recognition

mRNA cleavage

Fig. 2 Schematic representation of transkingdom gene silencing using constitutive multitarget bacterial long dsRNA expression

multitarget gene silencing (Fig. 2). Using this approach, we have shown that next generation transkingdom RNAi gene silencing can mediate specific dual target gene silencing in vitro.

2 2.1

Materials Bacterial Culture

1. Brain Heart Infusion (BHI) Broth (Remel, Thermo Fisher Scientific): 37 g BHI is dissolved in 1 L of tissue culture water, dispensed into appropriate containers and sterilized by autoclaving at 121  C for 15 min. Store at room temperature.

272

Jaehyung Lee et al.

2. Luria-Bertani broth and agar: 35 g of LB powder is dissolved in 1 L of distilled water and sterilized by autoclaving at 121  C for 15 min. Store at 4  C. To prepare agar plates, 25 g of LB agar powder is added to 1 L of distilled water followed by autoclaving at 121  C for 15 min. After appropriate antibiotics are added, the mixture is then dispensed to bacterial culture dishes to solidify at room temperature. Store at 4  C. 3. Sterile disposable round-bottom plastic tubes (14-ml); baffled culture flasks (500-ml). 4. Ampicillin was dissolved in distilled water at 100 mg/ml, passed through a 0.22 μm filter and stored in aliquots at 20  C. Working solution was prepared by diluting stock 1000-fold. 5. Isopropyl β-D-1-thiogalactopyranoside (IPTG) is dissolved in tissue culture water at 1 M, passed through a 0.22 m filter and stored in aliquots at 20  C. Working solution is prepared by diluting stock 1000-fold. 2.2

Cell Culture

1. SW480 and HT-29 colonic epithelial cells (American Type Culture Collection) are maintained in complete growth media in an atmosphere of 5% CO2 and 95% air. For long-term storage, cells are resuspended in complete growth medium supplemented with 5% (v/v) dimethyl sulfoxide and placed in liquid nitrogen. 2. Roswell Park Memorial Institute 1640 (RPMI 1640) medium supplemented with 10% fetal bovine serum. Store at 4  C. 3. Dulbecco’s Modified Eagle Medium (DMEM) medium supplemented with 10% fetal bovine. Store at 4  C. 4. Sterile plastic tissue culture flasks (75 cm2) and dishes (6 cm). 5. 0.25% Trypsin-EDTA solution is stored in aliquots at 20  C. 6. Penicillin-Streptomycin solution (10,000 U/ml penicillin, 10 mg/ml streptomycin in 0.9% sodium chloride) is stored in aliquots at 20  C. Working solution is prepared by diluting stock 1000-fold. 7. Amphotericin B solution (250 mg/ml in tissue culture water; Sigma) is stored in aliquots at 20  C. Working solution is prepared by diluting stock 1000-fold. 8. Gentamicin solution (10 mg/ml in tissue culture water) is stored at 4  C. Working solution is prepared by diluting stock 1000-fold. 9. Ofloxacin is dissolved in tissue culture water at 10 mg/ml, passed through a 0.22 μm filter and stored in aliquots at 20  C. Working solution is prepared by diluting stock 1000-fold.

Transkingdom Gene Silencing

2.3 Oral Administration of Bacteria to Mice

273

1. Female C57BL/6 mice (Charles River Laboratories) are housed under conventional conditions in isolator cages (4 mice per cage). Mice are fed standard chow (Harlan Teklad) and provided with tap water ad libitum. 2. 10
phosphate buffered saline is stored at room temperature. A 1
working solution is prepared by diluted the stock solution 10-fold with sterile water. This solution is stored at 4  C. 3. Feeding needle (Cadence Science). 4. 1 ml Norm-Ject tuberculin syringes (Henke Sass Wolf).

2.4 Intravenous Administration of Bacteria to Mice

1. Female nude Balb/c mice (Nu/Nu; Charles River Laboratories) are housed under specific pathogen free conditions in sterile isolator cages (4 mice per cage). Mice are fed with irradiated chow (Harlan Teklad) and provided with sterile water ad libitum. 2. 10
phosphate buffered saline is stored at room temperature. A 1
working solution is prepared by diluted the stock solution 10-fold with sterile water. This solution is stored at 4  C. 3. 1 ml Norm-Ject tuberculin syringes (Henke Sass Wolf). 4. 26G1/2 PrecisionGlide needles (Becton Dickinson).

3

Methodology Transkingdom gene silencing is one of the first applications of synthetic biology directed toward the development of a human therapeutic. In this approach, interfering shRNA or long double stranded RNA are produced inside nonpathogenic bacteria that have also been engineered to invade target cells in vitro and in vivo. For potential therapeutics development, this novel approach offers several advantages. Foremost is clinical safety since bacteria do not integrate genetic material into the human genome. Transkingdom gene silencing also abolishes the need to chemically synthesize siRNA, and may mitigate host immune responses since the silencing RNA are produced inside target cells. Furthermore, the need for attenuation of bacteria and the risk of environmental release of modified/mutated bacterial vectors can be addressed by methods already developed for bacteria-based interventions. Transkingdom gene silencing may, therefore, provide a practical and clinically compatible way to achieve RNAi for medical indications, particularly for diseases of the GI tract and other human organs and tissues colonized with normal bacteria flora.

274

Jaehyung Lee et al.

Methodology for conducting in vitro and in vivo transkingdom gene silencing studies has been published previously in this series [29]. In the sections below, we have updated our original methods using inducible tkRNAi vectors (where necessary) and also describe the use of constitutive tkRNAi vectors for in vitro bacterial RNAi gene silencing studies. 3.1 In Vitro Transkingdom Gene Silencing 3.1.1 Preparation of E. coli Inducible tkRNAi Vectors

1. Chemically competent E. coli BL21 (DE3) are transformed with inducible tkRNAi plasmids. Bacteria are then grown on BHI plates containing 100 μg/ml ampicillin overnight at 37  C. A single colony is then inoculated into BHI medium containing 100 μg/ml ampicillin, and grown overnight at 37  C (see Note 1). 2. The next day, 5 ml of each overnight culture is diluted 1:40 into fresh BHI medium containing 100 μg/ml ampicillin and grown for a further 2–4 h (until the OD600 ¼ 0.5). Each culture is then treated with IPTG (1 mM final concentration) for 2–4 h to induce transcription of interfering shRNA. 3. After IPTG induction, the total number of bacteria in each culture is calculated by measuring the OD600 value (8
108 bacteria/ml culture has an OD600 ¼ 1). The number of bacteria for cell treatment is then calculated (MOI; 20:1 to 2000:1, bacteria to cells). 4. The required volume of bacteria culture is then centrifuged at 2500
g for 10 min at 4  C and the pellet is washed once with serum-free, RPMI 1640 medium containing 100 μg/ml ampicillin and 1 mM of IPTG, and resuspended in the same medium at the required density for bacterial infection.

Constitutive tkRNAi Vectors

1. Chemically competent E. coli BL21 (DE3) are transformed with constitutive tkRNAi plasmids and cultured on LB agar plates containing 100 μg/ml ampicillin. A single colony is then inoculated into 30 ml of LB broth with 100 μg/ml ampicillin and cultured at 37  C with shaking at 200 rpm overnight. 2. The next day, the bacterial culture is centrifuged at 3000
g for 10 min and the pellet is washed with 5 ml of serum-free and antibiotic-free DMEM 3 times (see Note 2). 3. After washing, the pellet is resuspended with 3 ml of serumfree and antibiotic-free DMEM. 4. The total number of bacteria is then calculated by measuring the optical density at 600 nm (OD600 ¼ 1 is approximately 8
108 bacteria/ml). 5. The number of bacteria for infection is then determined by the colon cancer cell number to calculate the multiplicity of infection (MOI) for an infection volume.

Transkingdom Gene Silencing

275

3.1.2 Preparation of Attenuated S. typhimurium (See Note 3)

1. aroA attenuated S. typhimurium (SL7207) are inoculated in 3 ml of Brain Heart Infusion (BHI) broth with 50 μg/ml of kanamycin at 37  C without shaking for overnight.

Constitutive tkRNAi Vectors

2. The next day, the bacterial culture is centrifuged at 3000
g at 4  C for 10 min, and the pellet is washed three times with cold 10 mM HEPES. 3. The bacteria are then resuspended in 200 μl of cold 10% glycerol and kept on ice for 30 min. 4. The bacteria are then transformed with tkRNAi plasmids by electroporation using 2 mm gap width cuvettes at 2500 V for 5 ms (see Note 4). 5. Immediately after electroporation, 250 μl of SOC media is added and the mixture is placed in a 37  C shaking incubator for 45 min. The mixture is the spread on the BHI agar plate with 100 μg/ml ampicillin and grown over night at 37  C. 6. The next day, a single colony is inoculated into 30 ml of BHI broth with 100 μg/ml ampicillin and cultured at 37  C without shaking for overnight (see Note 5). 7. Preparation of attenuated S. typhimurium then proceeds according to steps 2–5 in Subheading 3.1.1 for E. coli constitutive tkRNAi vectors.

3.1.3 Bacterial Infection and Assessment of Target Gene Silencing Inducible tkRNAi Vectors

1. SW480 human colon cancer cells are cultured in an atmosphere of 95% air, 5% CO2 at 37  C in RPMI 1640 medium containing 10% FBS, 10 U/ml penicillin G, 10 μg/ml streptomycin and 250 μg/ml amphotericin. 24 h before bacterial infection, cell cultures are trypsinized, resuspended in complete RPMI 1640 medium and plated on 6-cm tissue culture dishes at 20–30% confluency. 2. Thirty minutes before bacterial infection, the cell culture medium is replaced with 2 ml of fresh serum-free RPMI 1640 medium containing 100 μg/ml of ampicillin and 1 mM IPTG. 3. Bacteria prepared in Subheading 3.1.1 above are then added to the cells at the desired MOI for 2 h at 37  C. 4. After the infection period, the cells are washed three times using serum-free RPMI 1640 medium (see Note 6). The cells are then incubated with 2 ml of fresh complete RPMI 1640 medium containing 100 μg/ml of ampicillin and 150 μg/ml of gentamicin for 2 h to kill any remaining extracellular bacteria. 5. After treatment with ampicillin and gentamicin, the cells are incubated with 3 ml of fresh complete RPMI 1640 medium containing 10 μg/ml of ofloxacin to kill any remaining bacteria.

276

Jaehyung Lee et al.

6. The cells are then harvested at various time points (from 24 to 96 h) in order to assess the extent of target gene silencing by real-time PCR (for mRNA) and western blotting (for protein). Constitutive tkRNAi Vectors

1. SW480 or HT29 human colon cancer cells are cultured in DMEM with 10% FBS and 10 μg/ml of gentamicin at 37  C in humidified 5% CO2 incubator. 2. One day before infection, cell cultures are trypsinized and seeded onto 12-well cell culture plates at 30–70% confluency with DMEM without antibiotics. 3. The next day, bacteria prepared in Subheadings 3.1.1 or 3.1.2 are added to the cells in 2 ml of serum-free and antibiotic-free DMEM for 2 h (see Note 7). 4. After bacterial infection, the media is removed and the cells are incubated with 2 ml of serum-free DMEM containing 100 μg/ ml of gentamicin for 1 h to kill remaining bacteria. 5. The cells are then washed three times with PBS and incubated with 2 ml of DMEM containing 10% FBS and 10 μg/ml of gentamicin. 6. Cells are then harvested at various time points (from 24 to 96 h) in order to assess the extent of target gene silencing by real-time PCR (for mRNA) and western blotting (for protein). An example of the results produced is shown in Fig. 3.

3.2

In-vivo tkRNAi

3.2.1 Preparation of E. coli

1. Transformed E. coli BL21 (DE3) bacteria containing inducible transkingdom gene silencing plasmids are grown in BHI medium containing 100 μg/ml ampicillin for 37  C until they reach early log phase (OD600 ¼ 0.5). The bacteria are then centrifuged at 2500
g for 10 min at 4  C, resuspended in 25 ml of BHI medium, aliquoted and stored in 80  C freezer as 15% glycerol stock (see Note 8). 2. One day prior to treatment the bacteria stocks are thawed, inoculated into 50 ml of fresh BHI medium containing 100 μg/ml ampicillin, and incubated overnight with shaking at 37  C. 3. Overnight cultures are inoculated into fresh BHI medium (at a 1:40 ratio) containing 100 μg/ml ampicillin, and grown for a further 2–4 h (until the OD600 ¼ 0.5). IPTG is then added (final concentration 1 mM), and the bacteria are incubated at 37  C with shaking for another 2–4 h. 4. After IPTG induction, the total number of bacteria is determined by measuring the OD600 (8
108 bacteria/ml culture has an OD600 ¼ 1). The volume of bacterial culture required

Transkingdom Gene Silencing

a

b

SW480

HT29

Salmonella Infection Non

Wild

GFP

100

100

Salmonella Infection

PD-L1/β-cat 50

75

277

Non Wild GFP

PD-L1/β-cat

100 MOI 500 500

PD-L1

PD-L1

β-Catenin

β-Catenin

10

50 100 200 300 500 MOI

α-Tubulin

α-Tubulin

c

SW480 Cell Proliferation (WST1)

O.D 450-690 nm

3 2.5 Non

2 1.5

Wild

1

GFP PD-L1/bcat

0.5 0 0hr

24hr

48hr

72hr

96hr

Fig. 3 Salmonella-mediated transkingdom gene silencing using multitarget bacterial long dsRNA expression. A next generation transkingdom gene silencing vector was designed to silence the colon cancer oncogene β-catenin as well as the immune check point modulator PD-L1. (a) Western blot analysis of beta-catenin and PD-L1 gene silencing in SW480 colon cancer cells. (b) Western blot analysis of beta-catenin and PD-L1 gene silencing in HT29 colon cancer cells. (c) Analysis of cell proliferation in SW480 colon cancer cells. Attenuated S. typhimurium treatments: nontransformed (non); wild-type S. typhimurium (wild); transkingdom gene silencing vector against GFP (GFP); transkingdom gene silencing vector against PD-L1 and beta-catenin (PD-L1/β-Cat)

for animal treatment is then centrifuged at 2500
g for 10 min at 4  C and the pellet is washed once with 1
PBS. The bacterial pellet is then resuspended at the required density in 1
PBS for oral administration or intravenous injection. 3.2.2 Oral Treatment of Normal Mice

1. Age-matched female C57BL/6 mice are divided into control or treatment groups (typically consisting of 6–8 animals per group). 2. Animals are given 5
108 to 5
1010 c.f.u. of tkRNAi E. coli (in 200 μl PBS) via an oral feeding needle fitted to a 1 ml syringe. Control animals are treated with E. coli containing an empty transkingdom gene silencing vector, or an inactive (e.g., scrambled) sequence. 3. Oral administration of bacteria is then performed 5 days per week for a total of 4 weeks (see Note 9). Two days after the final treatment, the animals are sacrificed and colonic tissues are collected for analysis of target gene silencing.

278

Jaehyung Lee et al.

3.2.3 Intravenous Treatment of Nude Mice Bearing Colon Cancer Xenografts

1. Age-matched female nude Balb/c mice are divided into control or treatment groups (typically consisting of 6–8 animals per group). 2. Three weeks before bacterial treatment, animals in each experimental group are subcutaneously implanted in the right flank with 1
107 colon cancer cells (in 100 μl PBS) using a 1 ml syringe fitted with a 26G needle. 3. Bacterial treatments are initiated when the xenograft tumors reach approximately 10 mm in diameter. Animals are treated intravenously with 1
108 c.f.u. of tkRNAi E. coli (in 100 μl PBS) by tail vein injection using a 1 ml syringe fitted with a 26G needle. Control animals are treated with E. coli containing an empty transkingdom gene silencing vector, or an inactive (e.g., scrambled) sequence. 4. Animals are treated every 5 days for a total of three treatments (see Note 10). Five days after the final treatment, the animals are sacrificed and tumor tissues are collected for analysis of target gene silencing.

3.2.4 Assessment of Target Gene Knockdown

1. For analysis of target gene mRNA levels by real-time PCR, colon and xenograft tissue is frozen and stored at 80  C. Total RNA isolation and real-time PCR analysis are then performed according to standard protocols. 2. For analysis of target gene protein levels by immunohistochemistry, colon and xenograft tissue is fixed in paraformaldehyde, paraffin-embedded, sectioned and stained according to standard procedures.

4

Notes 1. As transkingdom gene silencing plasmids are relatively large (~8.9 kb), the transformation efficiency using E. coli BL21 (DE3) is quite low. To ensure successful transformation, transkingdom gene silencing plasmids should be transformed into high-efficiency competent BL21 bacteria. 2. Bacteria must be washed using prewarmed serum-free and antibiotic-free DMEM medium. Washing with cold DMEM increases toxicity to the host cells. 3. Do not use aliquoted transformed attenuated S. typhimurium. Freshly transformed Salmonella should be used for each experiment. 4. Due to the relatively large size of transkingdom gene silencing plasmids (~8.9 kb), the efficiency of electroporation using aroA attenuated S. typhimurium is quite low. To ensure successful transformation, give the electric pulse twice.

Transkingdom Gene Silencing

279

5. Transformed Salmonella should be cultured in 25 ml of BHI broth for 16 h to give the best infection rate. 6. Cells must be washed using serum-free RPMI 1640 medium. Washing the cells with PBS can cause the bacteria to attach to the cell surface which can adversely affect cell viability and increase the probability of erroneous results. 7. When bacterial infection is performed using 12-well culture plates, add 1 ml of DMEM first followed by 1 ml of bacteria in DMEM to give final volume of 2 ml at the desired MOI. 8. The use of aliquoted bacterial stocks gives more reproducible results for in vivo tkRNAi than growing single colonies from BHI-ampicillin plates. 9. Oral administration is well tolerated with no gross or microscopic signs of epithelial damage or ulcerations. 10. Intravenous injection is generally well tolerated without adverse effects. Mice should be closely monitored during treatment. If the nude mice get sick during treatment, 40 mg/kg of ofloxacin can be applied. Alternatively, one treatment with bacteria can be omitted. References 1. Zaidi N, Jaffee EM (2019) Immunotherapy transforms cancer treatment. J Clin Invest 129(1):46–47 2. Yang Y (2015) Cancer immunotherapy: harnessing the immune system to Battle cancer. J Clin Invest 125(9):3335–3337 3. Ebbell B (1937) The papyrus Ebers: the greatest Egyptian medical document. Oxford University Press, London 4. Tanchou S (1844) Recherches sur le traitement me´dical des tumeurs cance´reuses du sein. Ouvrage practique base´ sur trois cents observations (extraits d’un grand nombre d’auteurs). Paris: G Baillie`re, 5. Dussaussoy (1787). Dissertations et observations sur la gangre`ne dans les hoˆpitaux. Lyon 6. Busch W (1868) Niederrheinische Gesellshaft fur Natur und Heilkunde in Bonn. Berlin Klin Wochenschr 5:137–138 7. Fehleisen F (1886) On erysipelas. In: Cheyne WW (ed) Recent essays on bacteria in relation to disease. New Sydenham Society, London, pp 263–286 8. Fehleisen F (1883) Die Aetiologie des Erysipels. Theodor Fischer, Berlin 9. McCarthy EF (2006) The toxins ofWilliam B. Coley and the treatment of bone and softtissue sarcomas. Iowa Orthop J 26:154–158

10. Zhou S, Gravekamp C, Bermudes D, Liu K (2018) Tumour-targeting bacteria engineered to fight cancer. Nat Rev Cancer 18 (12):727–743 11. Sedighi M, Bialvaei AZ, Hamblin MR et al (2019) Therapeutic bacteria to combat cancer; current advances, challenges, and opportunities. Cancer Med 8(6):3167–3181 12. Laliani G, Sorboni SG, Lari R et al (2020) Bacteria and cancer: different sides of the same coin. Life Sci 246:117398 13. Dang LH, Bettegowda C, Huso DL et al (2001) Combination bacteriolytic therapy for the treatment of experimental tumors. Proc Natl Acad Sci U S A 98:15155–15160 14. Low KB, Ittensohn M, Le T et al (1999) Lipid a mutant salmonella with suppressed virulence and TNFalpha induction retain tumortargeting in vivo. Nat Biotechnol 17:37–41 15. Xiang S, Fruehauf J, Li CJ (2006) Short hairpin RNA-expressing bacteria elicit RNA interference in mammals. Nat Biotechnol 24 (6):697–702 16. Toso JF, Gill VJ, Hwu P et al (2002) Phase I study of the intravenous administration of attenuated salmonella typhimurium to patients with metastatic melanoma. J Clin Oncol 20:142–152

280

Jaehyung Lee et al.

17. Heimann DM, Rosenberg SA (2003) Continuous intravenous administration of live genetically modified salmonella typhimurium in patients with metastatic melanoma. J Immunother 26:179–180 18. US National Library of Medicine (2013). ClinicalTrials.gov http://www.clinicaltrials. gov/ct2/show/NCT00004216 19. US National Library of Medicine (2013). ClinicalTrials.gov http://www.clinicaltrials. gov/ct2/show/NCT00006254 20. US National Library of Medicine (2008). ClinicalTrials.gov http://www.clinicaltrials. gov/ct2/show/NCT00004988 21. Nemunaitis J, Cunningham C, Senzer N et al (2003) Pilot trial of genetically modified, attenuated salmonella expressing the E. coli cytosine deaminase gene in refractory cancer patients. Cancer Gene Ther 10:737–744 22. US National Library of Medicine (2017). ClinicalTrials.gov http://www.clinicaltrials. gov/ct2/show/NCT01562626

23. US National Library of Medicine (2016). ClinicalTrials.gov. http://www.clinicaltrials. gov/ct2/show/NCT00358397 24. US National Library of Medicine (2016). ClinicalTrials.gov http://www.clinicaltrials. gov/ct2/show/NCT01118819 25. US National Library of Medicine (2018). ClinicalTrials.gov http://www.clinicaltrials. gov/ct2/show/NCT01924689 26. US National Library of Medicine (2018). ClinicalTrials.gov https://clinicaltrials.gov/ ct2/show/NCT03435952 27. Wu MR, Jusiak B, Lu TK (2019) Engineering advanced cancer therapies with synthetic biology. Nat Rev Cancer 19(4):187–195 28. Trieu V, Hwang L, Ng K et al (2017) First-inhuman phase I study of bacterial RNA interference therapeutic CEQ508 in patients with familial adenomatous polyposis (FAP). Ann Oncol 28(suppl_5):v158–v208 29. Xiang S, Keates AC, Fruehauf J et al (2009) In vitro and in vivo gene silencing by TransKingdom RNAi (tkRNAi). Methods Mol Biol 487:147–160

INDEX A Acetic acid...................................21, 69, 81, 82, 124, 252 Acrylamide.................................................. 41, 45, 46, 82, 89, 94, 111–113, 115, 116, 223, 224, 254 Affinity chromatography............................. 186, 187, 225 Algorithms .................................... 3–5, 42, 51–56, 62, 63 All-atom structures ...................................................2, 7–9 Allosteric .............................................................. 141–151, 155, 208, 213 Allosteric self cleaving ribozyme (aptazyme) ........................................213–216, 219 Alphaimager ......................................................... 174, 182 Alternative structures ....................................... 49–51, 219 AMBER ........................................ 55, 56, 58, 62, 63, 103 Ammonium persulphate solution (APS) ............... 33, 35, 254, 262 Amphotericin B .................................................... 215, 272 Ampicillin .......................................................... 40, 68–70, 82, 86, 101, 102, 105, 146, 175, 178, 235, 251, 252, 262, 272, 274–276 Antarctic phosphatase ......................................... 143, 145, 236, 239, 245 Antibiotics ................................................ 13, 40, 42, 105, 125, 137, 146, 162, 215, 217, 252, 256, 272, 276 Anticodon........................................................... 40–42, 68 Antisense RNAs (asRNAs) ........................................... 249 Aptamers..................................................... 68, 72, 75–77, 122, 123, 139, 141, 142, 144, 149, 150, 153, 155, 156, 158, 161, 166, 172, 199–201, 204, 213, 215, 216, 218–220, 249–251, 255 Armored tRNA (AtRNA) .........................................67, 68 aRNA, see Small artificial RNA (aRNA) Atomic force microscope (AFM) ........................ 228, 229 ATP ..................................................................14, 81, 123, 124, 180, 225, 236, 245 AtRNA, see Armored tRNA (AtRNA) Autoclaving.............................................................. 40, 69, 81, 102, 203, 236, 252, 261, 271, 272 Avsunviroidae .............................................. 100, 103, 106

B Bacillus subtilis ...............................................28, 156–166 Bacteria ..............................................................68, 71, 99, 104–106, 122, 153–168, 184, 209, 234, 236, 242, 250, 260, 262, 264, 268–270, 273–279

Bacto agar ..................................................................40, 69 Balances ......................................................................... 251 Balb/c mice .......................................................... 273, 278 BamHI-HF ........................................................... 175, 184 Barium chloride............................................................... 45 Base pairings .......................................................... 1–3, 28, 50, 78, 84, 94, 233, 234 BD Influx v7 cell ........................................................... 124 Benzoyl cyanide (BzCN) ................................... 15–17, 21 β-galactosidase............................................. 100, 101, 122 BglII..................................................................... 124, 129, 143, 145, 175, 178, 179 Bifidobacterium.................................................... 268, 269 Biocomputing................................................................ 199 Bioengineered RNA agent (BERA) ................... 250, 251, 253, 255–264 Bioluminescence............................................................ 155 Biosensors ............................................121–139, 154, 155 Biotin .................................................................. 77, 90, 94 BL21 .................................................................... 100, 105, 131, 133, 136, 147, 278 BL21 (DE3) ....................................................72, 93, 100, 105, 125, 131, 132, 135, 143, 145, 147, 150, 176, 183, 274, 276, 278 BL21 Star ............................................................ 133, 136, 143, 146, 147, 150 Bovine serum albumin (BSA)............................... 81, 112, 114, 115, 236, 246 Bradford...............................................111, 113, 224, 227 Brain Heart Infusion (BHI) ...............271, 274–276, 279 Broccoli.............................. 123, 141–146, 149, 150, 155 Bromophenol blue (BPB)......................... 41, 83, 87, 224 BsaI .............................................................. 100, 101, 104 BstEII .............................................. 77–79, 81, 85, 93, 94

C Calf intestinal alkaline phosphatase........................ 81, 85, 124, 129 Cancer therapeutics.............................................. 267–279 Cantilevers ....................................................224, 229–231 Capillary electrophoresis...........................................16, 18 Capillary sequencer ...................................................16, 22 Carbenicillin ................................................ 124, 203, 208 Catalytic ................................................ 84, 85, 90, 94–96, 154, 201, 209, 211, 216, 217

Luc Ponchon (ed.), RNA Scaffolds: Methods and Protocols, Methods in Molecular Biology, vol. 2323, https://doi.org/10.1007/978-1-0716-1499-0, © Springer Science+Business Media, LLC, part of Springer Nature 2021

281

RNA SCAFFOLDS: METHODS

282 Index

AND

PROTOCOLS

Caulobacter crescentus ................................................... 155 Centrifugation columns............................................40, 45 C57BL/6 mice..................................................... 273, 277 Chili ............................................................................... 155 Chimeras............................................................ 40, 42–45, 67, 72, 84, 85, 111–118 Chimeric ............................................... 68, 76–78, 83–90, 92, 93, 95, 96, 100, 250 Chloramphenicol.............................................68–70, 101, 102, 105, 159, 160, 162–164, 223, 225 Chloroform ...................................... 17, 81, 94, 103, 105 Chromophores ........................................... 122, 123, 156, 172, 177, 178, 188 Circular recombinant RNA ....................................99–106 Clostridium ................................................................... 268 Coarse grain..................................................................... 51 Cobra venom nuclease V1.............................................. 14 Co expression .................................................................. 99 Cofactors ..........................................................14, 21, 154 Connected path............................................................... 53 Corn............................................................................... 155 Counterions..................................................................... 26 Crystallization ....................................... 26–28, 31, 39–46 Crystals .....................................9, 25–35, 45, 50, 56, 177 CTP.............................................................. 124, 180, 225 Cyclic di GMP............................................. 122, 123, 154 Cyclo-N0 -[2-(Nmethylmorpholino)ethyl] carbodiimide-p-toluenesulfonate (CMCT)............................................................... 15 Cytosine deaminase....................................................... 269

D dATP ................................................................................ 16 dCTP ..............................................................16, 175, 184 ddATP.............................................................................. 16 ddCTP ............................................................................. 16 ddTTP ............................................................................. 16 Deoxynucleotide triphosphate (dNTP)....................... 142 Deoxyribonucleic acid (DNA) .........................18, 25, 26, 43, 81, 83, 87, 94, 96, 101, 104, 117, 124, 125, 128–130, 138, 144, 145, 147, 149, 151, 158–160, 162, 163, 175, 176, 178, 179, 182, 185, 189, 194, 215, 222, 223, 225, 230, 236, 239, 240, 242, 243, 245, 246, 251, 256 Deoxyribozymes.............................................................. 77 DEPC, see Diethylpyrocarbonate (DEPC) Designs ......................................................... 1, 27, 40, 41, 83–85, 101, 103, 104, 127, 141–151, 155, 176–182, 190, 194, 200, 216, 219, 224, 225, 234, 236–238, 251, 255–257, 262 DFHBI, see Difluoro hydroxybenzylidene D4-PA .............................................................................. 16 DH5α................................................. 100, 104, 175, 179, 185, 236, 239, 245, 252, 263

Diethylpyrocarbonate (DEPC) ........................26, 33, 93, 175, 180, 191, 194, 252, 254, 255, 258, 259, 261 Difluoro hydroxybenzylidene.............122, 126, 155, 172 DiMethyl Sulfate (DMS) .............................14–17, 20–22 DIR2s ............................................................................ 155 DisconnectionDPS.......................................................... 55 Discrete path sampling (DPS)..................................52, 63 dITP................................................................................. 16 DMEM, see Dulbecco’s Modified Eagle Medium (DMEM) DNA, see Deoxyribonucleic acid (DNA) DNase I....................................................... 142, 145, 176, 237, 242, 243, 247 DNAzymes .............................................................. 77, 78, 83, 84, 90, 92, 94, 96 dNTPs, see Deoxynucleotide triphosphate (dNTPs) Double stranded RNA (dsRNA)........................ 177, 270, 271, 273, 277 Doubly-nudged elastic band (DNEB).....................53, 62 DpnI ............................................................ 202, 205, 206 DSL, see Designed and selected ligase (DSL) DTT ....................................... 40, 44, 111, 112, 115, 236 dTTP................................................................................ 16 D2-PA .............................................................................. 16 Dulbecco’s Modified Eagle Medium (DMEM)......................................... 111, 112, 215, 217, 272, 274, 276, 278, 279 Dulbecco’s phosphate buffered saline (DPBS)...................................................... 144, 148

E EagI...................................... 72, 129, 130, 185, 255, 256 EagI HF...............................................124, 130, 175, 252 EasyXtal 15-Well Tool .................................................... 26 E. coli OmniMax strain ................................................ 175 EcoRI................................................................40, 43, 185 EcoRI-HF............................................................. 175, 184 EGFP, see Enhanced green fluorescent protein (EGFP) Eggplant latent viroid (ELVd) ................... 100, 103, 104 Electromobility gel-shift assays (EMSA) ............. 14, 222, 224, 227, 228, 230 Electrophoresis ........................................... 44, 69, 71, 72, 77, 82, 83, 87, 89, 90, 92, 95, 101, 104, 105, 111, 116, 145, 173–175, 179–182, 185, 189, 202, 204, 223, 224, 227, 228, 235, 239, 245, 247, 251, 254, 256, 258, 263 Electrophoretic mobility shift assay (EMSA) ............................................ 188, 222, 227 Electroporator ......................................... 69, 78, 102, 203 EMSA, see Electrophoretic mobility shift assay (EMSA) Encephalomyocarditis virus (EMCV) .......................... 110 Energy barriers ............................................ 50, 51, 54, 56 Energy landscapes .....................................................49–64

RNA SCAFFOLDS: METHODS Enhanced green fluorescent protein (EGFP).....................................214, 216, 218, 219 Environmental conditions .............................................. 51 EpCAM aptamer .................................................. 250, 255 Escherichia coli (E. coli).............................. 40, 67–73, 77, 79, 85–89, 92–94, 99–106, 111, 112, 122, 124, 125, 129, 131, 132, 143, 145, 147, 148, 155–157, 159, 175, 176, 179, 183–185, 215, 216, 225, 230, 233–236, 239, 240, 242, 245, 246, 252, 258, 260, 262, 263, 269, 274–278 Ethanol ................................................. 15–18, 33, 44, 69, 71, 80, 81, 86, 102, 111, 112, 125, 132, 201, 203, 206, 223, 226, 252, 258, 259 Ethidium bromide .................................40, 86, 175, 182, 202, 206, 209, 252, 258, 261 eukaryotic initiation factor (eIF).................................. 110 Exclusion chromatography........................................... 174 Exonucleases........................................................... 99, 100 Expression .........................................................40, 41, 68, 70–72, 76, 77, 79, 93, 95, 100, 105, 109, 122, 125, 127–132, 135, 137, 138, 153, 156, 158–160, 162, 164, 166, 174, 176, 183–188, 199–201, 205, 208, 214, 219, 225, 234–236, 238–242, 250, 251, 255, 257, 258, 260, 262, 263, 270, 271, 277

F Familial Adenatomous Polyposis (FAP) ...................... 269 Fast protein liquid chromatography (FPLC)............................................ 174, 181, 188, 254, 259, 260, 263, 264 FeCl3 ............................................................................. 157 Fermentation ................................ 76, 250, 257–258, 260 Fetal bovine serum (FBS) ........................... 215, 217, 272 Filter binding................................................................... 14 Filtration ........................................... 41, 72, 82, 102, 235 5-Bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) .............................................. 102, 105, 106 5-Difluoro-4-hydroxybenzylidene imidazolinone (DFHBI)......................................... 122, 123, 126, 127, 131, 134, 135, 138, 139, 155, 156, 172, 173, 176–178, 182, 188–191 5S rRNA derived scaffold .........................................75–96 5S rRNA terminator ............................................ 101, 104 Flow cytometry ................................................... 122, 131, 132, 135–139 Fluorescence microscopy .................................... 122, 131, 132, 134, 162, 164, 165 Fluorescence module ........................................... 142, 143 Fluorescence quantum yields ....................................... 156 Fluorescence reader.............................................. 174, 180 Fluorescence resonance energy transfer (FRET)............................................ 121, 154, 155, 172, 173, 176, 177, 182, 189, 190, 192

AND

PROTOCOLS Index 283

Fluorescent ........................................................16, 80, 90, 121–139, 141, 154–156, 158, 161, 162, 166, 172, 176, 182, 188, 189 Fluorescent light-up aptamer (FLAP) ......................... 188 Fluorescent microscopy ...............................132–135, 172 Fluorescent proteins................................... 121, 154–156, 162, 168, 172, 176 Fluorescent RNAs ................................................ 123, 173 Fluorogenic .........................................123, 141–151, 155 Fluorogenic light-up aptamer (FLAP)................ 172, 177 Fluorophores ....................................................... 139, 141, 172, 176, 182, 189, 190 Folding ......................................................... 2, 15, 33, 41, 42, 50, 59, 61, 143, 145, 146, 150, 173, 221, 222 Foot-and-mouth disease virus (FMDV) ...................... 110 Footprinting ..................................................... 13–22, 237 Force fields ................................................................56, 63 Formamide .................................................. 16, 41, 82, 83 Fragment based ................................................................. 2 Free energies.......................................... 50, 52, 55, 59–61 FRET, see Fluorescence resonance energy transfer (FRET)

G gBlocks.................................................175, 179, 182, 184 Gel electrophoresis equipment....................................... 40 Gel imaging system .............................................. 252, 258 Gel-shift .................................................................. 14, 116 Gene expression ........................................... 93, 109, 153, 154, 158, 162, 199–211, 213, 215, 216, 219, 233, 236, 237, 243–245, 249, 250 Gene knockdown .......................................................... 278 Gene silencing ............................................ 234, 235, 237, 238, 240–242, 269–271, 273–278 Gentamicin ..........................................146, 272, 275, 276 G. kaustophilus ............................................................... 41 Glucose ......................................................... 82, 102, 125, 126, 157, 160, 203, 215 Glutathione (GSH) ..................................... 174, 176, 187 Glycerol..............................................................69, 77, 81, 86, 87, 102, 111, 113, 126, 143, 157, 163, 202, 223, 224, 275, 276 Glycogen....................................................................16, 17 GPU...........................................................................55, 64 G-quadruplexe ................................................................ 50 GROMACS ...............................................................55, 56 GST................................................................................ 184 GTP .....................................................123, 124, 180, 225

H Half-life...........................................................99, 156, 162 Hammerhead ribozyme (HHR) ................ 103, 104, 200 Hand held UV lamp ....................................................... 69

RNA SCAFFOLDS: METHODS

284 Index

AND

PROTOCOLS

HEK293 cell................................................ 203, 208, 217 HeLa cells .................................................... 111, 112, 115 Hela229 cell .................................................................. 203 Helix ..............................................................3, 4, 7–9, 20, 41, 42, 77–79, 93, 158, 222 Hemocytometer ............................................................ 208 Hexafluoro-2-propanol (HFIP) ................................... 255 High-fidelity DNA polymerase ............................ 96, 142, 144, 201, 205 High-resolution.....................................49, 222, 228, 231 Hind III ........................................................................... 81 Hind III-HF .................................................................. 175 HiRE-RNA...................................................................... 55 Horizontal gel electrophoresis ....................................... 78 Hrluc reporter gene ............................................. 200, 205 HT-29 colonic epithelial cells....................................... 272 Hybridization ....................................................... 225–227

I ImageJ..................................................124, 134, 135, 166 Imaging............................................... 121–139, 141–151, 154, 156, 162, 164, 165, 224, 228–231 Imidazole ........................................ 70, 71, 176, 187, 223 Incubators..........................................................78, 80, 96, 112, 124, 132, 134, 135, 162, 202, 203, 208, 209, 215, 217, 240, 251, 276 Infusion enzyme............................................................ 254 Interensemble transitions ............................................... 63 Interfering shRNA ............................................... 273, 274 Internal ribosome entry site (IRES) .............................................. 109, 110, 116 Intracellular imaging ............................................ 141, 144 Inverted microscope ..................................................... 215 In vitro transcription................15, 40, 44, 178–182, 224 In vivo ......................................................... 13, 67, 68, 76, 77, 121, 122, 125, 127, 138, 156, 164, 199–211, 234, 245, 269, 273, 274, 276–279 Ion exchange .............................................................44, 82 Ion replacement .............................................................. 33 IRES transacting factor (ITAF).................................... 110 Irradiated chow ............................................................. 273 Isopropyl β-D-1-thiogalactopyranoside (IPTG) .......................................... 68–71, 93, 134, 135, 138, 144, 148, 176, 186, 223, 225, 235, 238, 242, 272, 274–276 Ispinach.......................................................................... 155

J JM109.................................................... 77, 85–87, 92, 94

K Kanamycin .................................................. 111, 112, 125, 143–148, 223, 225, 275

KCl ................................................. 15, 16, 26, 27, 82, 83, 90, 102, 126, 143, 175, 181, 190, 203, 223, 224 Kd..................................................................................... 14 KH2PO4 ....................................102, 125, 126, 157, 203 Kinetics ..................... 50, 51, 54, 56, 146, 234, 255, 261 Kink turn (K turn) ..........................................27–29, 221, 222, 225, 227, 229 KpnI-HF........................................................................ 175 K2HPO4 .............................................................. 102, 157

L LacZ .............................................. 77, 101, 234, 235, 237 Laemmli buffer...............................................72, 112, 115 Large RNAs ........................................................ 2, 30, 230 Large-scale (recombinant) expression ........................... 40 LB agar ....................................................... 40, 69, 82, 86, 105, 157, 162, 175, 215, 216, 236, 252, 257, 272, 274 LB Broth...................................................... 157, 252, 274 LE Agarose .................................................................... 175 Life-cell imaging............................................................ 172 Ligands ....................................................... 15, 21, 26, 45, 49–51, 96, 122, 127, 138, 149, 153–156, 158, 199, 200, 208–211, 213, 215, 216, 218–220 Ligase ...........................................96, 106, 124, 129, 179, 185, 207, 236, 239, 245, 246 Live cell imaging ................................................. 121, 122, 131–137, 141, 153–168 Local minimum ............................................................... 55 Logic gates............................................................ 213–220 Long noncoding RNAs (lncRNAs) ............................. 249 L7Ae protein ................................ 28, 221, 224, 225, 227 Luciferase........................... 118, 200, 201, 205, 208–211 Luminometer ....................................................... 204, 210 Luria broth (LB) ........................................ 40, 69, 82, 86, 94, 101, 102, 105, 111, 112, 124, 129, 131, 132, 134, 135, 138, 143–145, 147, 162, 163, 166, 179, 183–185, 215, 217, 223, 225, 236, 240, 242, 252, 256, 272 Lysozyme.............................................176, 186, 237, 242

M Macromolecules .............................................25, 222, 227 Malachite green ..............................................68, 123, 172 Mammalian cells......................................... 111, 112, 200, 203, 213–220, 222, 269 Mango............................................................................ 155 Mass spectrometry ..............................110, 116, 117, 154 mcherry....................................................... 172, 173, 176, 182–192, 194, 196, 218, 219 MDtraj ............................................................................. 57 Medicine ............................................................... 267–279 Mercaptoethanol ........................................................... 176

RNA SCAFFOLDS: METHODS Messenger RNA (mRNA) ........................... 76, 213, 214, 233–235, 237, 238, 240, 241, 245, 276, 278 Metabolites .......121, 122, 131, 137, 139, 141, 153–168 Methanol ....................................................................... 255 Methylbisacrylamide ....................................................... 41 Mfold ..................................... 42, 84, 144, 177, 241, 245 MgCl2 ...................................................15, 16, 26, 33, 40, 44, 45, 70, 81, 82, 102, 111, 143, 145, 146, 175, 181, 190, 203, 223, 224, 236 Mica disk........................................................................ 224 Microcentrifuge.......................................... 80, 85, 86, 90, 123, 124, 145, 148, 157, 162, 235, 252 Microfluidizer....................................................... 174, 186 MicroLoops ...............................................................26, 32 MicroMounts ............................................................26, 32 Microplate reader ................................215, 218, 224, 255 microRNA (miR) ................................................. 249, 251 MicroSaws .................................................................26, 32 Microscope ................................................. 124, 131, 134, 138, 148, 157, 166, 168, 204 MicroSieves................................................................26, 31 Minimal Salts glycerol glutamate (MSgg).....................................157, 162, 164–166 Minimizations ................................................................. 54 Miniprep ..........................................................71, 93, 105, 144, 215, 217, 254 Misfoldings ................................................................39, 42 MMLV Reverse Transcriptase ..................................16, 18 MnCl2 ........................................................................83, 91 Models ...........................................................2, 3, 5–9, 45, 50–52, 55, 56, 58, 63, 127, 174, 176, 177, 196, 241, 250, 252, 255, 262 Molecular dynamic (MD).................................51, 57, 59, 60, 63, 254, 255, 268 Motif based................................................................2, 3, 5 Motifs...............................................................2–8, 27, 84, 109–112, 122, 158, 221, 222, 225, 227 MPD ................................................................................ 45 MsbB gene..................................................................... 268 MS2 operator hairpin ...............................................67, 68 Multifunnel energy landscapes ....................................... 50

N Na cacodylate ............................................................41, 45 NaCl...................................................................41, 70, 71, 81–83, 90, 101, 102, 111, 112, 125, 126, 175, 186, 187, 196, 203, 223, 236, 240, 254, 255, 258 Nanodrop ................................................... 145, 174, 181, 187, 188, 215, 223, 225, 226, 235, 240, 251, 258, 260 Nanoscopy ..................................................................... 172 Nanostructures ..................................................... 221–231 Nanotechnology............................................................ 221 NdeI................................................................72, 124, 130

AND

PROTOCOLS Index 285

NEBuilder............................................................. 101, 104 Nickel nitrilotriacetic acid (NiNTA) ........................68, 71 Nitrogen ....................................... 32, 160, 181, 188, 272 NMR, see Nuclear magnetic resonance (NMR) N,N-dimethylformamide.............................................. 102 N,N,N0 ,N0 -tetramethylethylenediamine (TEMED) ........................... 83, 89, 254, 261, 262 Noncoding RNA (ncRNA) ................................. 251, 255 NTP ..................................................................40, 44, 194 Nuclear magnetic resonance (NMR) ................ 50, 52, 61 Nucleases ..................................14, 41, 67, 157, 230, 236

O Ofloxacin ..................................................... 272, 275, 279 Oligonucleotides ................. 26, 125, 141, 144, 237, 249 1-Methyl-7-nitroisatoic anhydride (1M7)..................... 15 OPTIM .........................................................52, 58, 61–63 Overlapping PCR .............................................43, 81, 138

P PACYCT2..................................................................68, 72 PANTHER software ..................................................... 112 Pathogen infection ........................................................ 100 PATHSAMPLE .............................. 52, 55, 58, 59, 61–63 Path sampling .................................................................. 52 PBS, see Phosphate buffered saline (PBS) PBSKrna .......................................................................... 68 PBSMrnaStrep...................................................... 111, 112 PBSTNAV...............................69, 72, 251–253, 255, 262 pCA2c .................................................... 79, 81, 85–87, 93 pCMV-mCherry................................................... 215, 217 pCOLADuet........................................125, 126, 130, 131 PCR, see Polymerase chain reaction (PCR) pDG1662 .....................................................157, 159–162 pEGFP-N1 .................................................................... 215 Pembrolizumab ............................................................. 269 Penicillin ..............................................203, 215, 272, 275 Peppers .......................................................................... 155 pET24a ........................................................ 125, 130, 131 pET28c ........................................................ 143, 145, 147 pET31...........................................................125–131, 137 P15LZ.......................................................... 101, 104, 105 pGEX-6P-2.................................................. 175, 184, 185 Phase determination .........................................................vi Phenol......................................................... 16, 17, 69, 70, 72, 81, 93, 103, 105, 111, 112, 223, 226, 246, 257, 261 Phenylalanine................................................................. 157 Phenol extraction ............................................... 15, 17, 21 Phenol GTC ..............................................................16, 17 Phosphate buffered saline (PBS)........................ 111, 112, 126, 135, 138, 166, 215, 218, 273, 276–279 Phusion DNA polymerase ...........................124, 128–130

RNA SCAFFOLDS: METHODS

286 Index

AND

PROTOCOLS

Plants ...................................................100, 101, 103–106 Plasmid DNAs ..................................................... 145, 162, 215, 236, 239, 240, 245 Plasmids .............................................................40, 42, 43, 46, 67–69, 72, 77, 79, 81, 85–87, 93–95, 100, 101, 103–106, 111, 112, 118, 122, 125, 127, 129–132, 137, 138, 143–145, 147, 151, 157, 159, 160, 178, 179, 182, 183, 196, 200, 202–205, 207–211, 214–218, 223, 225, 236, 238–240, 242, 246, 250–257, 260, 262, 263, 274–276, 278 PMSF .................................................................... 176, 186 Polylinker...............................................42, 100, 101, 104 Poly-L-lysine (PLL) .................................... 144, 147, 164 Polymerase.........................................................26, 75, 76, 205, 209, 222, 223, 225, 246 Polymerase chain reaction (PCR) ....................17, 26, 40, 43–46, 80, 81, 85, 90, 96, 101, 104, 124, 127–130, 138, 142, 144, 145, 151, 159, 163, 174, 175, 179, 183–185, 216, 218, 223, 225, 230, 235, 242–247, 252–254, 256, 262, 263, 276, 278 pOPIN-mCherry.................................................. 175, 182 Pospiviroidae ................................................................. 106 Post crystallization treatment ..................... 26, 30, 31, 33 Predictions ..................................................... 1–3, 6–9, 18, 42, 51, 84, 144, 177, 241, 245 Pre-miRNA........................................................... 250, 251 Pre-mir-34a ......................................................... 250–253, 255, 256, 262, 263 PreScission Protease............................................. 176, 187 Primers........................................... 16, 18, 21, 22, 43, 96, 104, 125, 128–130, 142, 144, 179, 182, 184, 204, 205, 215, 216, 218, 222, 230, 237, 241, 243, 244, 246, 251–253, 255, 256, 262, 263 pRNA .........................................................................43, 45 Probing .........................................................13, 15–17, 20 Protein expression .............................................. 68, 69, 72 Pseudoknots ........................................................... 3, 7, 42 Pseudomonas aeruginosa ............................................... 155 Pseudomonas aeruginosa bacteriophage coat protein 7 (PP7)............................................... vii, 172, 176, 177, 179, 180, 182, 184, 189–195 psi CHECK2 vector ...................................................... 200 pUC19......................................40, 42, 43, 178, 215, 217 pUC19I ......................................................................... 175 Pull-down ............................................110, 111, 113–116 pULZ ...................................................100, 101, 104, 105 Pure RNA ............................................................. 150, 260 Purification ................................................. 17, 39, 46, 68, 70–72, 76, 78, 81, 86, 88, 89, 100, 102, 105, 111, 112, 115, 142, 145, 150, 151, 157, 174–188, 202, 204, 207, 223, 225, 236, 237, 239, 240, 242, 249–264

PurI genes ...................................................................... 269 Python .......................................................................57, 63

Q Quantitative polymerase chain reaction (qPCR).....................................235, 237, 243–247 Quenching..................................173, 174, 176, 192, 196

R Reactivities ....................................................13, 15, 18–20 Real time............................ 121, 181, 243, 244, 276, 278 RecA...................................................................... 175, 184 Recognition modules........................................... 142, 143 Recombinant RNAs ................................... 67, 75–96, 99, 100, 102, 103, 105, 106, 250, 263 Refolding ..................................................... 15, 41, 45, 50 Regulation .................................................. 1, 77, 93, 100, 109, 154, 159, 213, 214, 235, 249 REMSAs .....................................172, 174, 176, 188, 190 Replication.................................... 68, 100, 101, 125, 178 Reporters .................................................... 122, 155, 156, 158–166, 200, 201, 204, 208, 214–216, 219, 234, 235 Restriction endonuclease ................................... 81, 86, 94 Reverse transcriptase .............................96, 237, 243, 247 Reverse transcription polymerase chain reaction (RT-PCR) ................................................. 104, 138 Ribonucleases .................................................................. 76 Ribonucleic acid (RNA) ................................. 1–9, 13–22, 25–35, 39–46, 49–51, 55–57, 60, 62, 67–72, 75–79, 83–85, 87–93, 95, 96, 99–101, 103, 104, 106, 109–118, 122, 127, 131, 134, 138, 141, 142, 144–146, 149, 150, 154–156, 158, 171–196, 213, 216, 221–231, 233–247, 249, 250, 254–255, 257–261, 263, 264, 270, 273, 278 Ribonucleoprotein (RNP) ........................... 99, 100, 221, 222, 224, 225, 228, 230, 231 Ribosomal RNA (rRNA) .................................. 77–79, 84, 88, 93, 94, 100, 115, 221, 244, 251, 262 Ribosomes .................................................. 13, 27, 39, 40, 77, 159, 161, 221, 235, 237 Riboswitches......................................................27, 28, 49, 121–139, 153–156, 158, 159, 161, 162, 166, 207, 213–220 Riboswitch-yfp .............................................156, 158–162 Ribozymes .......................................................44, 75, 103, 106, 149, 200, 201, 209, 211, 213, 214, 216, 217, 219, 220 Rigid bodies..................................................................... 34 RNA, see Ribonucleic acid (RNA) RNA binding protein.............................. 20, 27, 109–118 RNA degradation ................................................... 39, 150 RNAdraw...................................................................83, 84

RNA SCAFFOLDS: METHODS RNA expression.......................................... 39, 69, 76, 77, 85, 86, 93, 100, 131, 138, 238, 257, 263 RNA folds ............................................19, 25, 42, 84, 177 RNAi ....................................................269–271, 273, 274 RNA loading buffer .........................................41, 44, 258 RNApdbee ....................................................................... 57 RNA-protein interaction ............................... vii, 171–196 RNase-free .................................................. 40, 44, 46, 86, 110–115, 142, 174, 180, 181, 187–191, 193, 236, 250, 261 RNA sensors ......................................................... 141–151 RNA 7SK ............................................................ 56, 60, 61 RNA-Seq ................................................................ 96, 171 RNases .............................................................67, 99, 127, 150, 177, 245, 261, 270 Root mean square deviation (RMSD) ............... 4, 5, 7–9, 34, 56, 63 Root mean square fluctuation (RMSF) ...................56, 57 Rosetta(DE3) pLysS pRARE2 ............................ 223, 225 Roswell Park Memorial Institute 1640 (RPMI 1640)...........................272, 274, 275, 279 Rotor........................................................ 80, 90, 174, 186 Run-off transcription ...................................................... 43

S SacII ............................ 72, 124, 129, 130, 252, 255, 256 Salmonella............................................268, 269, 278, 279 Salmonella typhimurium..................................... 155, 269, 275, 277, 278 Scaffolds...........................................................1–9, 41, 42, 45, 67, 75–97, 99–106, 122, 127, 156, 172, 173, 222, 234, 235, 237–238, 245, 250, 252 Screening .................................................... 100, 101, 106, 122, 199–211, 214, 220, 263 SDS-polyacrylamide-gelelectrophoresis (SDS-PAGE)................................... 112, 116, 117, 183, 186, 187, 223, 227 Secondary structures .................................... 2, 18, 19, 28, 41–43, 46, 49–51, 57, 68, 76, 78, 84, 110, 144, 150, 201, 217, 234, 241, 245 Sensor domain............................................................... 153 Sephadex............................................................... 142, 145 Sequence alignment ..................................................18, 27 Shaking incubators...............................69, 157, 223, 235, 240, 242, 251, 275 SHAPEs ...........................................................3, 4, 14–17, 19–22, 59, 153, 224 SHORTCUT BARRIER ................................................ 54 SibC terminator............................................................. 234 Silencing ..................................................... 100, 118, 234, 237, 245, 270, 273 Simulations ................................................. 51, 52, 54–56, 58, 63, 144, 150 Single strandedness ..................................... 234, 237, 245

AND

PROTOCOLS Index 287

Size exclusion chromatography (SEC) .......................................181, 187–188, 194 SL3a ............................................................ 111, 112, 114, 115, 117, 118 SmaI ................................................ 40, 43, 236, 238, 239 Small artificial RNA (aRNA) ................ 77–79, 87, 93, 94 Small interfering RNA (siRNA) ......................... 118, 251, 255, 262, 273 Small molecules .................................................13, 14, 45, 51, 121–123, 155, 173 SOC, see Super Optimal broth with Catabolite repression (SOC) SOC medium ........................ 82, 85, 104, 203, 207, 208 Sodium hydroxide .....................................................16, 21 Sonicator................................................................. 70, 223 Spermidine.................................................................40, 44 Spermine ...................................................... 26, 28, 45, 83 Spinach aptamer ..................................122, 131, 177, 190 Spinach2 ..............................................122, 123, 139, 155 Spin concentrator ......................... 28, 174, 181, 187, 188 Spring constants ..................................................... 54, 231 SsRNA ............................................................................. 50 Stem-loop .................................................... 155, 217, 238 Sterile .................................................................40, 69, 80, 85, 86, 102, 110–112, 124, 147, 148, 151, 157, 162, 163, 165, 174, 183, 203, 208, 210, 258, 259, 272, 273 Streptavidin............................................77, 110, 115, 116 Streptavidin agarose .................................... 90, 92, 94, 96 Streptavidin-aptamer....................................110, 113–116 Streptavidin-coated magnetic beads........... 111, 113, 114 Streptomycin ..............................151, 203, 215, 272, 275 Structural rearrangement................................... 13, 19, 20 Structures.................................................. 1–9, 14, 19, 20, 22, 25–27, 31, 41, 42, 45, 49–52, 56, 57, 60–63, 67, 68, 84, 100, 110, 122, 149, 150, 177, 216, 219–221, 228, 230, 234, 238 Sub-cloning ..................................................177–182, 185 Super optimal broth with catabolite repression (SOC)................................................................. 102 SW480 ..........................................................272, 275–277 SYBR green ................................................. 224, 228, 230

T TAE, see Tris acetate EDTA buffer (TAE) Tag ........................................................... 76–78, 155, 189 Targets ........................................................ 39–46, 50, 51, 61, 76, 93, 121, 123, 141–143, 145, 149–151, 153, 158, 177, 179, 183, 184, 187, 188, 233–235, 237, 238, 240, 241, 244, 245, 249–251, 253, 255–264, 270, 271, 273, 275–278 TBE, see Tris borate EDTA buffer (TBE) TBM, see Tris borate magnesium buffer (TBM) T box RNA ......................................................... 27, 31, 35

RNA SCAFFOLDS: METHODS

288 Index

AND

PROTOCOLS

Template library ........................................................3–5, 9 TEN, see Tris EDTA NaCl buffer (TEN) Terrific Broth (TB)............................................... 102, 105 Tertiary interactions...................................................... 201 Tertiary structures ......................................................... 149 T4 DNA ligase ................................................81, 85, 124, 129, 144, 145, 147, 175, 179, 184, 185, 218, 236, 238 T4 DNA Polymerase..................................................... 175 T4 polynucleotide kinase ...............................96, 236, 238 Theophylline................................................ 123, 215, 218 Theophylline aptamer (TPA)...................... 217, 250, 255 Theophylline dependent aptazyme .............................. 200 Therapeutics ................................................. 75, 199, 234, 249, 268, 269, 273 Thermal cyclers ...............................................16, 78, 124, 128–130, 144, 222, 224, 235, 238, 251 Thermodynamics................................................ 27, 51, 59 Thermophile .................................................................... 41 Thiamine............................................................... 123, 157 Thiazole orange derivate ..................................... 123, 155 3 Aminopropyltriethoxy silane 3,5-difluoro-4hydroxybenzylidene imidazolinone-2-oxime (DFHO)............................................................. 155 Three dimensional structures ............................ 41, 54, 67 3 ( N Morpholino)propanesulfonic acid (MOPS) ................................................83, 91, 157 Three way junction ....................................................... 3, 8 Threonine ...................................................................... 157 TkRNAi ........................................................269, 274–279 TNF α, see Tumor necrosis factor alpha (TNF α) TOP10........................................................ 124, 129, 143, 145, 147, 151, 215, 216 Toxins ............................................................................ 268 Transcription termination...................................... 76, 153 Transducer modules.................................... 142, 143, 155 Transfection reagents ........................................... 215, 219 Transfer RNA (tRNA) ............................... 16, 17, 26–28, 31, 33, 35, 40–45, 67, 72, 95, 112, 115, 121–139, 249–264 Transition states .................................... 51–55, 58, 59, 62 Transkingdom ..................................... 269–271, 273–278 Translation..............................................34, 77, 109, 110, 118, 153, 199–201, 213, 233, 235, 238 Tris Acetate EDTA (TAE) ..............................83, 90, 124, 252, 254, 255, 258, 263 Tris borate EDTA (TBE)..................................40, 41, 82, 83, 86, 89, 94, 223, 224, 226 Tris EDTA buffer (TE).....................................80, 86, 88, 102, 105, 142–144, 149, 223, 237 Tris EDTA NaCl buffer (TEN)......................... 83, 90, 92 Tris-HCl ....................................40, 44, 69, 70, 102, 103, 111, 112, 142, 175, 181, 187, 190, 254, 257 Triton X-100 .............................................................40, 44

tRNA, see Transfer RNA (tRNA) tRNAGly............................................................. 28, 33, 41 tRNA ligases ........................................100, 101, 103–106 tRNALys..............................................156, 173, 177, 194 tRNA scaffold............................................. 39–46, 67, 68, 110, 114, 115, 125, 127, 128, 138, 262 Trypsin ......................................................... 203, 208, 215 Tryptone .............................. 69, 101, 102, 126, 203, 236 Tryptophan.................................................................... 157 T7 bacteriophage terminator ....................................... 101 T7 RNA polymerase ........... 26, 27, 40, 44, 93, 145, 180 T7 RNA polymerase promoter ...................................... 42 Tumors .........................................................267–269, 278 Two dimensional structure................................................v 2 x TY medium ............................................................... 69

U Ultra-Fast Liquid Chromatography system................. 255 Urea ..............................41, 44, 45, 82, 89, 95, 111, 112, 115, 176, 187, 223, 226, 254, 258, 262, 263 Urea polyacrylamide electrophoresis ........................... 254 UTP ...................................................................... 180, 225 UV spectrophotometer.......................................... 80, 215 UV transilluminator ..................................................78, 87

V Vertical gel electrophoresis ............................................. 78 Vfold ...................................................................................v Vfold3D......................................................................... 1–9 VfoldLA ......................................................................... 1–9 Vibrio cholerae................................................................ 155 Vibrio proteolyticus .............................................. 77, 79, 93 Viroids .....................................................................99–106 Visualization .............................8, 55, 131, 138, 162, 190

W Water bath .................. 46, 147, 166, 235, 239, 251, 256

X XhoI ........... 72, 124, 129, 130, 143, 145, 175, 178–180 XL1 Blue............................................................. 69–71, 93 X-ray crystallography ...................................................... 25 Xylene cyanol............................................. 41, 69, 95, 202

Y Yeast extract...................69, 82, 101, 102, 126, 203, 236 Yellow fluorescent protein (YFP) .......156, 157, 162, 166

Z Zeiss 200M AxioVert microscope ................................ 124 ZnCl2 ............................................................................. 157