Faculty of Pharmacy Tabriz University of Medical Sciences

A Dissertation submitted for Ph.D. degree in Pharmaceutical Biotechnology

Entitled: Cloning and sequencing of of Alishewanella tabrizica strain RCRI4 and comparative studies on similar riboswitches in other

By: Elnaz Mehdizadeh Aghdam

Supervisors: Dr Mohammad Saeid Hejazi Dr Abolfazl Barzegar Advisor: Dr Jörg S. Hartig

September 1026 Thesis No: 00

Clarification and Copyright Declaration

I hereby declare that I am the sole author of this thesis. I take full responsibility of this work and declare that the results of this study are original and based on my personal work. All scientific materials used in this thesis are fully referenced.

All rights are transferred to Faculty of pharmacy, Tabriz University of Medical sciences. The Faculty of pharmacy is authorized to lend or reproduce this thesis, in total or in part.

Professors’ names: Dr. Mohammad Saeid Hejazi; Dr Abolfazl Barzegar

Signature:

Name: Elnaz Mehdizadeh Aghdam Student No: 0020210102

Date:

Signature:

I would like to dedicate this thesis to my dear husband

Dr Hamid Reza Heidari

For his endless patience, kindness and love Acknowledgments I would like to express my deepest acknowledgments to my supervisor, Prof. Mohammad Saeid Hejazi, for the patient guidance, encouragement and advice he has provided throughout my time as his student. I have been extremely lucky to have a supervisor who cared so much about my work, and who responded to my questions and queries so promptly. I believe he has been more than a supervisor to me, as he has guided me patiently and constantly all through my academic life and I am and will be grateful to him for the rest of my life. My special thanks also go to my second supervisor Dr Abolfazl Barzegar for all the incredible opportunities he made for me and his constant kind support and motivation. I would like to thank Prof. Jörg Hartig for his extremely kind support during my research stay in the University of Konstanz. Completing this work would have been difficult if it were not the support and friendship provided by him and the entire lab members of his group. In particular, I would like to thank Malte Sinn for his great patience in teaching me the techniques and all his support and help during my stay in Hartig group. I would also like to thank all the members of department of pharmaceutical biotechnology for their kind help and support during my PhD. My special thanks to Dr Parvin Akbarzadeh, Dr Leila Molavi and Dr Azita Dilmaghani for their kind help and encouragements during my PhD. I must express my gratitude to Dr HamidReza Heidari, my husband, for his continued support and encouragement. I was continually amazed by his endless willingness to help and support me by any means and in every aspect of our shared life. My deepest gratitude goes to my mother, father and brother for their patience and inspirations throughout my life. My sincere appreciation goes to my mother-in-law, father-in-law and sister and brothers-in-law who experienced all of the ups and downs of my research. This never-ending support of my family has been the most valuable God's blessings to me. Last but not least, I would like to express my thanks to my friends, old and new, my kind colleagues in the faculty of pharmacy, who made everything easy for me during this memorable journey. My special thanks to Dr. Vahideh Tarhriz, Dr Simin Sharifi, Dr Hamideh Mahmmodzadeh Hosseini, Dr Sanam Arami, Dr Samin Hamidi and Ms Nazanin Lakestani and all the past and current students in the Facultz of Pharmacy.. . List of Publications

I. Mehdizadeh Aghdam, E., A. Barzegar, and M.S. Hejazi, Evolutionary origin and conserved structural building blocks of riboswitches and ribosomal : riboswitches as probable target site for aminoglycosides interaction. GENE, In press II. Mehdizadeh Aghdam, E., A. Barzegar, and M.S. Hejazi, Evolutionary origin and conserved structural building blocks of riboswitches and ribosomal RNAs: riboswitches as probable target site for aminoglycosides interaction. Adv Pharm Bull, 1022. 2(0): p. 111-101. DOI: 2061602/apb.10226000. III. Mehdizadeh Aghdam, E., et al., Riboswitches as potential targets for aminoglycosides compared with rRNA molecules: in silico study. J Microbial Biochem Technol, 1022. 0(1). IV. Mehdizadeh Aghdam, E., et al., TPP characterization in Alishewanella tabrizica and Alishewanella aestuarii and comparison with other TPP riboswitches. Submitted Abbreviations I

Abbreviations 1D: Two dimensional 0D: Three dimensional FMN: Flavin mononucleotide GFP: Green Fluorescent Protein GlcN6P: glucosamine-6-phosphate glmS: glucosamine-6-phosphate synthetase GMP: Guanosine monophosphate HTS: High throughput screening MDR: Multiple drug resistance mRNA: messenger RNA PDB: Protein Data Bank PreQ2: Pre-queuosine 2 RMSD: Root Mean Square Deviation RNA: Ribonucleic Acid rRNA: Ribosomal RNA SAM: S-adenosylmethionine SELEX: Systematic Evolution of Ligands by Exponential Enrichment SPR: Surface Plasmon Resonance THF: Tetrahydrofuran TPP: Thiamine pyrophosphate UTR: Untranslated region

Table of Contents II

Table of contents

Abstract 1

Chapter 1: General Introduction 2

262 INTRODUCTION ...... 0 261 OBJECTIVES ...... 1 260 PROJECT HYPOTHESES ...... 1 262 SCOPE AND OUTLINES ...... 6 261 REFERENCES ...... 0 Chapter 2: Riboswitches: from living biosensors to novel targets of antibiotics 11

162 ABSTRACT ...... 22 161 INTRODUCTION ...... 21 160 HISTORY OF RIBOSWITCHES ...... 21 162 MECHANISMS OF RIBOSWITCHES ...... 21 16262 glmS riboswitch/ribozyme ...... 20 161 METHODS FOR RIBOSWITCH DISCOVERY ...... 10 16162 In silico approaches ...... 12 16161 Biochemical verification of riboswitch-ligand binding ...... 11 16160 Genetic validation of riboswitch-mediated gene regulation ...... 12 16162 Finding riboswitches using genome-wide mapping of transcriptional termination ... 12 166 ENGINEERED RIBOSWITCHES ...... 11 16662 Background ...... 11 16661 Riboswitch engineering ...... 16 1666162 Selection methods...... 16 1.66161 Artificial Riboswitches and applications ...... 11 16660 Riboswitch-based biosensors ...... 10 16662 Perspective ...... 01 161 RIBOSWITCHES AS THERAPEUTICS TARGETS ...... 01 16162 Background ...... 01 16161 Development of riboswitch-targeted antibiotics ...... 02 1616162 Lysine riboswitch ...... 01 1616161 riboswitch ...... 01 1616160 glmS riboswitch ...... 01 1616162 FMN riboswitch ...... 00 16160 Screening of compounds ...... 00 1616062 glmS riboswitch ...... 00 1616061 c-di-GMP riboswitch ...... 00 1616060 SAM-I riboswitch ...... 00 1616062 T-box riboswitch ...... 00 1616061 preQ2-I riboswitch ...... 00 1616066 TPP riboswitch ...... 20 Table of Contents III

16160.1 In silico HTS ...... 20 16162 Limitations and perspective ...... 20 160 Riboswitch-related patents ...... 22 16062 Patents of novel riboswitches, structures, methods, cognate and non-cognate ligands ...... 22 16061 Patents of riboswitch-based drug design ...... 21 1606162 glmS riboswitch/ribozyme...... 21 1606161 Fluoride riboswitch ...... 21 1606160 Guanine riboswitch ...... 21 1606162 T-box riboswitch ...... 26 16060 Patents of Engineered riboswitches ...... 26 160 CONCLUSION ...... 26 1620 REFERENCES ...... 20 Chapter 3: Evolutionary Origin and Conserved Structural Building Blocks of Riboswitches and Ribosomal RNAs: Riboswitches as Probable Target site for Aminoglycosides Interaction 86

062 ABSTRACT ...... 60 061 INTRODUCTION ...... 10 060 MATERIALS AND METHODS ...... 11 0.062 Databases and Programs ...... 11 06061 Functional and structural riboswitch alignments ...... 11 06060 Sequences alignment ...... 10 06062 Motifs categorization ...... 10 06061 Docking ...... 12 0606162 Preparation of the macromolecules...... 12 0606161 Preparation of the ligand ...... 12 0606160 Preparation of the grid files ...... 12 0606162 Preparation of the docking files ...... 11 0606161 Statistical Analyses ...... 11 062 RESULTS AND DISCUSSION ...... 16 06262 Structural riboswitch alignments ...... 16 06261 Motifs categorization ...... 10 06260 Sequences alignment ...... 00 06262 Functional properties of riboswitches-rRNAs motifs ...... 00 061 CONCLUSION ...... 00 066 REFERENCES ...... 00 Chapter 4: Riboswitches as potential targets for aminoglycosides compared with rRNA molecules: in Silico study 44

262 ABSTRACT ...... 01 261 INTRODUCTION ...... 06 Table of Contents IV

260 METHODS ...... 00 26062 Molecular docking ...... 00 2606262 Preparation of Macromolecules for AutoDock Vina ...... 00 2606261 Preparation of the ligand for AutoDock Vina ...... 00 2606260 Docking procedure ...... 00 26061 Docking validation...... 00 2606162 Docking preparation for rDock ...... 00 2606161 Docking processing for rDock ...... 200 2606160 Post-processing and analysis of results ...... 200 26060 Molecular dynamics simulation ...... 200 262 RESULTS ...... 202 26262 Docking of riboswitches with various aminoglycosides ...... 202 26261 Docking validation...... 200 26260 Molecular dynamics simulation ...... 220 261 DISCUSSION ...... 220 266 CONCLUSION ...... 221 261 REFERENCES ...... 221 Chapter 5: TPP riboswitch characterization in Alishewanella tabrizica and Alishewanella aestuarii and comparison with other TPP riboswitches 121

162 ABSTRACT ...... 211 161 INTRODUCTION ...... 210 160 MATERIALS AND METHODS ...... 211 16062 Preparation of RNA sequences from different bacteria ...... 211 1606262 Providing TPP riboswitch template for in vitro transcription ...... 211 160626262 A. aestuarii TPP riboswitch ...... 211 160626261 A. tabrizica TPP riboswitch ...... 211 160626260 E. coli and B. subtilis TPP riboswitch...... 210 160626262 Secondary structures ...... 210 1606261 In vitro transcription ...... 210 16061 In-line probing ...... 210 1606162 RNA labeling ...... 210 1606161 In-line probing ...... 210 16060 SPR assay...... 210 1606062 Immobilization of biotinylated single stranded DNA on SA chip ...... 210 1606061 Riboswitch loading on SA-chips ...... 200 1606060 Real-time binding experiment for TPP Riboswitch-TPP...... 200 1606062 Kd determination ...... 202 162 RESULTS ...... 202 16262 Sequence and secondary structure analysis of TPP riboswitch in A. aestuarii and A. tabrizica ...... 202 16260 Surface Plasmon Resonance Spectroscopy ...... 222 Table of Contents V

161 DISCUSSION ...... 220 166 CONCLUSION ...... 226 161 REFERENCES ...... 221 Chapter 8: Conclusion, Perspectives and Recommendations 151

662 CONCLUSION ...... 211 661 PERSPECTIVES ...... 210 660 RECOMMENDATIONS ...... 212 662 REFERENCES ...... 211 Appendix A

L چکیده فارسی List of Tables VI

List of Tables Table 162 Studies related to biosensor design based on riboswitches ...... 02 Table 161 Studies related to drug discovery based on riboswitches ...... 06 Table 160 Some important patents regarding riboswitches ...... 20 Table 062 Types of riboswitches which share more than 105 secondary structure identity with ribosomal RNA motifs...... 02 Table 061 Binding energy of paromomycin interactions with different types of riboswitches and “26S rRNA A site” as receptors...... 01 Table 162 Oligos used in order to obtain target TPP riboswitches in studied bacteria...... 216 Table 161: Affinity and kinetic parameters of TPP binding to TPP riboswitch aptamer domain using In-line probing and SPR ...... 201 Table S2 Pairwise sequence alignment between lysine riboswitch (261 bp)with below sequence and similar rRNA PDB structures...... B Table S1 Pairwise sequence alignment between riboswitch with below sequence and similar rRNA PDB structures...... C Table S0 Pairwise sequence alignment between glycine riboswitch with below sequence and similar rRNA PDB structures...... D Table S2 Pairwise sequence alignment between TPP riboswitch with below sequence and similar rRNA PDB structures...... E Table S1 Pairwise sequence alignment between SAM riboswitch with below sequence and similar rRNA PDB structures...... F Table S6 Pairwise sequence alignment between FMN riboswitch with below sequence and similar rRNA PDB structures...... G Table S1 Pairwise sequence alignment between c-di-GMP I riboswitch with below sequence and similar rRNA PDB structures ...... H Table S0 Pairwise sequence alignment between c-di-GMP II riboswitch with below sequence and similar rRNA PDB structures ...... I Table S0 Pairwise sequence alignment between THF riboswitch with below sequence and similar rRNA PDB structures ...... J Table S20 Pairwise sequence alignment between preQ2 riboswitch with below sequence and similar rRNA PDB structures ...... K

List of Figures VII

List of figures Figure 162 Riboswitches schematic structures ...... 22 Figure 161 Riboswitch gene control mechanisms ...... 20 Figure 062 Correlation of secondary structure identity and 0D structure of rRNAs with the mentioned riboswitches...... 11 Figure 061 Correlation of ribosomal RNA motifs with 20 different types of riboswitches...... 10 Figure 060 Average of global pairwise alignment similarity percentage of rRNA sequences with structurally-based similar riboswitches via Needle program ...... 01 Figure 062 RMSD vs. binding energy for 20 types of riboswitches and “26S rRNA A site” based on Autodock results...... 06 Figure 061 Three dimensional representation of docked paromomycin and corresponding RNAs...... 00 Figure 262 Docking of paromomycin ...... 201 Figure 261 Docking of gentamicin ...... 200 Figure 260 Docking of neomycin ...... 202 Figure 262 Docking of kanamycin ...... 201 Figure 261 Docking of amikacin ...... 206 Figure 266 Docking of sisomicin ...... 201 Figure 261 Docking of tobramycin ...... 200 Figure 260 rDock results of paromomycin and gentamicin docking ...... 220 Figure 260 Molecular dynamic simulation results of lysine riboswitch- paromomycin interaction ...... 221 Figure 162 Multiple alignment of TPP riboswitches ...... 200 Figure 161 Secondary structure of TPP riboswitch ...... 201 Figure 160 In-line probing of TPP riboswitch A. tabrizica ...... 200 Figure 162 In-line probing of TPP riboswitch A. aestuarii ...... 200 Figure 161 In-line probing of TPP riboswitch B. subtilis ...... 220 Figure 166 SPR Affinity analysis of the TPP riboswitch aptamer from B. subtilis, E. coli, A. tabrizica and A. aestuarii...... 221 Abstract 2

Abstract

Introduction: Riboswitches, as cis acting non-coding RNA elements, regulate gene expression via specific binding of various small molecules. Their conserved structures are the key characteristics for binding specific molecules and gene regulation. Also, the contribution of riboswitches in antibiotic targeting is possible. In this study, we aimed first to find out the structural and functional level of similarity among riboswitches and rRNA structures including binding of aminoglycosides to riboswitches through computational tools. On the other side, riboswitches are rarely characterized in environmental bacteria. In this study, TPP riboswitch sourced from novel characterized specie Alishewanella tabrizica, was identified. Methods: After PDB structures’ selection, multiple sequence and structural alignment were carried out. Highly similar rRNA motifs with riboswitches (including “A site”) were sorted out. Subsequently, the probable interaction of riboswitches with aminoglycosides ere studied using docking studies. For identifying of TPP riboswitch of A. tabrizica, kinetic and affinity analysis of TPP binding to different TPP riboswitch aptamer domains sourced from A. tabrizica, Alishewanella aestuarii, E. coli, B. subtilis were studied and compared using In- line probing and Surface Plasmon Resonance (SPR) methods. Results: Docking analysis showed significant high binding energies between riboswitches-aminoglycosides. Accordingly, lysine, glycine and SAM-I riboswitches were recognized as the best RNA targets for all of the aminoglycosides. Docking results were also validated through rDock program and MD simulation. According to the KD values from experimental analysis, the affinity of TPP binding was the highest in A. tabrizica. In addition, the order of TPP binding affinity was TPP aptamer domains sourced from A. tabrizica > A. aestuarii > E. coli > B. subtilis. Conclusion: Taking together, it can be proposed that riboswitches have this potential to be the target of aminoglycosides. The observed variation TPP riboswitches affinity could be referred to the studied bacteria adaptation to diverse environmental conditions.

Keywords: Riboswitch, Docking, Simulation, Thiamine pyrophosphate, Alishewanella .

Chapter 1: General Introduction

Chapter 2 0

1.1 Introduction

It is been a while that RNA molecules are considered more as regulatory agents in the cell. Many types of riboregulators have been introduced in recent years [reviewed in 2]. RNAs play important roles in many processes of adaptations and stress responses of bacteria [2-0]. Besides, RNA can fold into special 0D structures capable of making active sites as well as binding to specific ligands [1]. In addition, many clinically used antibiotics targets ribosome, especially the ribosomal RNA such as aminoglycosides which binds to A site motif of 26S rRNAs in bacteria [2, 1].

One of the wide distributed cis-acting regulating RNAs are located in 1’ UTR of mRNAs called “Riboswitches” [reviewed in 6]. They consisted of two different parts: aptamer domain and expression platform. The aptamer domain could generate a conformational change in the expression platform with an allosteric binding to a small ligand and lead to the regulation of downstream gene [1]. Nowadays, more than 00 classes of riboswitches such as lysine, glycine, purine,

FMN, TPP, B21 riboswitches have been identified and entitled based on their affinity to interact with the specific ligands [0]. The conserved structures of their aptamer domain in specific and selective binding to very vital metabolites of microorganisms as well as their protein independent mechanism of gene regulation make them unique targets for drug design [0].

Multiple drug resistance (MDR) in microorganisms is a demanding problem in the last decade. The discovery of riboswitches as regulators of cell surviving genes in bacteria has been made a promising target for designing of novel antibiotics. Since the first riboswitch was introduced by Nahvi et al. [20], the possibility of finding new antibiotics inhibiting riboswitches was declared by the same group. They worked on the new compounds which can target lysine riboswitch and inhibit the growth of Bacillus subtilis [22]. Chapter 2 2

The idea of using riboswitches led the structural biologists to elucidate the 0D structures as well as the binding conformations of these interesting antibiotic targets. Now the presence of crystallographic structures of riboswitches in databases makes it easier to employ computational studies and screenings of the compounds for finding the novel active ligands. Riboswitches can be simply checked out for specific ligand binding in various in vitro, in vivo, and in silico high-throughput screening (HTS) methods [21].

The aminoglycoside antibiotics are important therapeutic compounds for treatment of severe bacterial infections. They exert their effects via binding to 26S rRNA in the 00S ribosomal subunit at the A site and cause interference in mRNA [2, 20, 22]. However, it has been shown that aminoglycosides could bind to several other RNA types [21-20]. It is established that conformational changes in the RNA molecules can be occurred by drug binding at such sites [20, 10]. In addition, designing artificial riboswitches for aminoglycosides have been conducted in the past decade [12-12]. There have been some reports about the structural similarity between riboswitches and rRNAs which raises the possibility of functional connection between these two types of RNA molecules [11]. Regarding the structural similarity between rRNAs and riboswitches, the possibility of binding of aminoglycosides to different types of riboswitches were also investigated.

Environmental microorganisms have been rarely used as a target to study riboswitches. These bacteria due to their different conditions as well as possible diverse metabolic pathways could show different pattern of gene regulation such. Genus Alishewanella from the phylum of gamma Proteobacteria was first introduced in 1000 and the first bacterium in this genus was named Alishewanella fetalis. Alishewanella tabrizica, a Gram-negative, aerobic, motile and rod-shaped bacterium, was identified by our group in 1021 isolated from Qurugöl Lake [16]. Apart from some annotations regarding bioinformatic prediction of cobalamin and c-di-GMP riboswitches, no experimental identification or confirmation of riboswitches have been carried out in this genus. Chapter 2 1

1.2 Objectives

Regarding the importance of identifying novel riboswitch representatives from the available novel sources and studying the sequence and structure of the riboswitches, the main aim of this project was:

2. Studying the structural and functional properties of different types of riboswitches especially any similarity between riboswitches and ribosomal RNAs. 1. Comparison of the sequences and structures of similar riboswitches in various bacteria including the interaction of aminoglycosides with riboswitches. 0. Cloning and sequencing of riboswitches of Alishewanella tabrizica strain RCRI2 (focused on TPP riboswitch) and comparative studies on the achieved element

1.3 Project hypotheses

2. There is a structural similarity between riboswitches and ribosomal RNAs motifs 1. Aminoglycosides could bind to different classes of riboswitches (based on computational studies) 3. TPP riboswitch from Alishewanella tabrizica can be identified using cloning and affinity experiments.

Chapter 2 6

1.4 Scope and outlines

In this study, PDB structures of different riboswitches were extracted from the database and their sequence similarity were investigated computationally. The results showed that there is a moderate to high similarity between riboswitches and rRNA motifs. Docking and MD simulation studies were applied to study the possible affinity of aminoglycosides toward riboswitches.

In the experimental section, TPP riboswitches of newly discovered specie Alishewanella tabrizica were detected and amplified. Using in-line probing and Surface Plasmon Resonance (SPR), TPP affinity to this riboswitch was compared with those interactions observed in the TPP riboswitches of Escherichia coli, B. subtilis and Alishewanella Aestuarii.

Chapter 2 provides a comprehensive review regarding all aspects of riboswitches including history, discovery steps, engineered riboswitches and studies regarding drug design based on riboswitches structures.

In Chapter 3, we aimed to investigate relation between rRNAs and riboswitches structures and their subsequent functions such as binding to specific antibiotics. In this regard, in current study we attempted to survey structural similarity including primary, 1D, 0D and motifs as well as interaction similarity among rRNAs and riboswitch elements through bioinformatics and computational tools.

The structural similarity between rRNAs and riboswitches and the possibility of binding of paromomycin, as a representative of aminoglycosides, to different types of riboswitches were studied in previous chapter. As a result, in chapter 4 we aimed to evaluate and validate the binding potential of seven aminoglycosides including paromomycin, gentamicin, amikacin, kanamaycin, neomycin, tobramaycin, sisomicin against nine types of riboswitches through computational methods.

In Chapter 5 we intended to identify and characterize the TPP riboswitch of thiC operon in A. aestuarii and A. tabrizica and compare with the same type of Chapter 2 1

riboswitch in E. coli and B. subtilis. For this purpose, the sequences of TPP riboswitches were amplified and TPP binding assay was examined using in-line probing and Surface Plasmon Resonance (SPR) methods to determine the affinity of TPP to these motifs.

Chapter 2 0

1.5 References

2. Waters L S, Storz G (1000). Regulatory RNAs in bacteria, Cell. 206(2), 621-10. 1. Toledo-Arana A, Repoila F, Cossart P (1001). Small noncoding RNAs controlling pathogenesis, Curr Opin Microbiol. 20(1), 201-0. 0. Wachter A (1020). Riboswitch-mediated control of gene expression in eukaryotes, RNA Biol. 1(2), 61-16. 2. Fourmy D, Recht M I, Blanchard S C, Puglisi J D (2006). Structure of the A site of Escherichia coli 26S ribosomal RNA complexed with an aminoglycoside antibiotic, Science. 112(1102), 2061-12. 1. Serganov A A, Masquida B, Westhof E, Cachia C, Portier C, Garber M, et al. (2006). The 26S rRNA binding site of Thermus thermophilus ribosomal protein S21: comparison with Escherichia coli S21, minimum site and structure, RNA. 1(22), 2212-00. 6. Winkler W C (1001). Riboswitches and the role of noncoding RNAs in bacterial metabolic control, Curr Opin Chem Biol. 0(6), 102-601. 1. Mandal M, Boese B, Barrick J E, Winkler W C, Breaker R R (1000). Riboswitches control fundamental biochemical pathways in Bacillus subtilis and other bacteria, Cell. 220(1), 111-06. 0. Mandal M, Breaker R R (1002). Gene regulation by riboswitches, Nat Rev Mol Cell Biol. 1(6), 212-60. 0. Blount K F, Breaker R R (1006). Riboswitches as antibacterial drug targets, Nat Biotechnol. 12(21), 2110-62. 20. Nahvi A, Sudarsan N, Ebert M S, Zou X, Brown K L, Breaker R R (1001). Genetic control by a metabolite binding mRNA, Chem Biol. 0(0), 2020- 2020. 22. Blount K F, Wang J X, Lim J, Sudarsan N, Breaker R R (1001). Antibacterial lysine analogs that target lysine riboswitches, Nat Chem Biol. 0(2), 22-0. 21. Penchovsky R, Stoilova C C (1020). Riboswitch-based antibacterial drug discovery using high-throughput screening methods, Expert Opin Drug Discov. 0(2), 61-01. 20. Carter A P, Clemons W M, Brodersen D E, Morgan-Warren R J, Wimberly B T, Ramakrishnan V (1000). Functional insights from the structure of the 00S ribosomal subunit and its interactions with antibiotics, Nature. 201(6001), 020-0. Chapter 2 0

22. Davies J, Davis B D (2060). Misreading of ribonucleic acid code words induced by aminoglycoside antibiotics. The effect of drug concentration, J Biol Chem. 120(21), 0021-6. 21. Rogers J, Chang A H, von Ahsen U, Schroeder R, Davies J (2006). Inhibition of the self-cleavage reaction of the human hepatitis delta virus ribozyme by antibiotics, J Mol Biol. 110(1), 026-11. 26. Mikkelsen N E, Brannvall M, Virtanen A, Kirsebom L A (2000). Inhibition of RNase P RNA cleavage by aminoglycosides, Proc Natl Acad Sci U S A. 06(22), 6211-60. 21. Stage T K, Hertel K J, Uhlenbeck O C (2001). Inhibition of the hammerhead ribozyme by neomycin, RNA. 2(2), 01-202. 20. von Ahsen U, Davies J, Schroeder R (2002). Antibiotic inhibition of group I ribozyme function, Nature. 010(6021), 060-10. 20. Murchie A I, Davis B, Isel C, Afshar M, Drysdale M J, Bower J, et al. (1002). Structure-based drug design targeting an inactive RNA conformation: exploiting the flexibility of HIV-2 TAR RNA, J Mol Biol. 006(0), 611-00. 10. Davis B, Afshar M, Varani G, Murchie A I, Karn J, Lentzen G, et al. (1002). Rational design of inhibitors of HIV-2 TAR RNA through the stabilisation of electrostatic "hot spots", J Mol Biol. 006(1), 020-16. 12. Weigand J E, Sanchez M, Gunnesch E B, Zeiher S, Schroeder R, Suess B (1000). Screening for engineered neomycin riboswitches that control translation initiation, RNA. 22(2), 00-01. 11. Morse D P (1001). Direct selection of RNA beacon aptamers, Biochem Biophys Res Commun. 010(2), 02-202. 10. Schmidtke S R, Duchardt-Ferner E, Weigand J E, Suess B, Wohnert J (1020). NMR resonance assignments of an engineered neomycin-sensing riboswitch RNA bound to ribostamycin and tobramycin, Biomol NMR Assign. 2(2), 221-0. 12. Vandenengel J E, Morse D P (1000). Mutational analysis of a signaling aptamer suggests a mechanism for ligand-triggered structure-switching, Biochem Biophys Res Commun. 010(2), 12-6. 11. Barrick J E, Breaker R R (1001). The distributions, mechanisms, and structures of metabolite-binding riboswitches, Genome Biol. 0(22), 2-20. 16. Tarhriz V, Nematzadeh G, Zununi Vahed S, Hejazi M A, Hejazi M S (1021). Alishewanella tabrizica sp. nov., isolated from Qurugol Lake, Int J Syst Evol Microbiol. 61(Pt 0), 2006-02.

Chapter 1: Riboswitches: from living biosensors to novel targets of antibiotics

Chapter 1 22

2.1 Abstract

Riboswitches are generally located in 15-UTR region of mRNAs and specifically bind small molecular partners. Following ligand binding, gene expression is controlled mostly by transcription termination, translation inhibition or mRNA degradation processes. More than 00 classes of known riboswitches have been identified by now. Most riboswitches consist of an aptamer domain and an expression platform. The aptamer domain of each class of riboswitch is a conserved structure and stabilizes specific structures of the expression platforms through binding to specific compounds. In this review, we are highlighting most aspects of riboswitch research including the history of riboswitch discoveries, routine methods for discovering and investigating riboswitches along with newly discovered classes and mechanistic principles of riboswitch-mediated gene expression control. Moreover, we will give an overview about the potential of riboswitches as therapeutic targets for antibiotic design and also their utilization as biosensors for metabolites detection.

Chapter 1 21

2.2 Introduction

There are several control mechanisms to modulate the expression of genes in response to numerous cellular conditions. The combination of protein signaling and factors which control transcription, translation and messenger RNA (mRNA) processing or degradation is the most common known mechanism of gene expression regulation. Furthermore, there is a developing idea that RNA has a definite role in posttranscriptional control mechanisms. Trans-acting riboregulators such as microRNAs (miRNA) and short interfering RNAs (siRNA) regulate gene expression by forming base paired structures with target mRNA and through various protein mediated pathways. On the other hand, cis-acting regulating RNAs are usually located in non-coding parts of mRNA [reviewed in 2]. RNA based regulation provides several advantages. First, since mRNA is the direct target, a rapid response is ensured. Second, for small RNAs (sRNAs) acting through binding to mRNAs, specific and fast recognition can be gained with a low number of base pairings [1, 0]. Third, specificity may evolve more rapidly since one nucleotide change can be sufficient to change specificity [2]. Finally, the capacity of RNAs to modulate their conformation upon binding increases the number of contacts with a given ligand or makes the recognition of multiple targets easier. Riboswitches are genetic regulatory elements usually found in the 15-UTR (untranslated regions) of mRNA [1-0] (Figure 1.2A) and are widely distributed in bacteria [reviewed in 0]. The term riboswitch is greatly associated with binding to specific small molecules without protein factors involvement and eventually controlling gene expression by allosterically altering dimensional structures of the mRNA. They are mostly located in the upstream regions of bacterial mRNAs that encode biosynthetic enzymes or metabolite transporters. These sequences comprise of two parts: aptamer domain and expression platform. Generally, ligand binding to the aptamer domain of a riboswitch stabilizes specific structural elements of joining expression platform, which modulates the expression of Chapter 1 20

downstream genes. As a result, they regulate the process of transcription and/or translation (Figure 162B). Riboswitches can bind and sense cellular metabolites such as amino acids and derivatives [e.g. lysine, glycine, S-adenosyl methionine (SAM), S-adenosyl homocysteine (SAH)], carbohydrates [glucosamine-6-phosphate (Glcn6P)], coenzymes [flavin mononucleotide (FMN), thiamin pyrophosphate (TPP), coenzyme B21], nucleobases and their derivatives [, guanine, c-di-GMP, c- di-AMP , pre-queuosine (preQ2) [20, 22] and ions like Mg1+ [21, 20]. Some of the recent discovered riboswitches are c-AMP-GMP [22, 21], 1-Amino-2- imidazolecarboxamide riboside 1'-monophosphate (ZMP) [26], cations Mn1+ [21, 20] and Ni1+/Co1+ [20] binding motifs and anions such as fluoride [10]. The particular interest about anion binding to RNA is how a negatively charged RNA could bind to a negatively charged ligand. Ren and colleagues found out that bound fluoride could be encapsulated with three Mg1+ ions which are coordinated to water molecules and RNA backbone phosphates [12]. Gene regulation mechanisms by riboswitches are in various levels of transcription and translation [2]. The first mechanism involves the ligand-dependent formation of a Rho-independent termination which is a GC-rich stem followed by a poly-U tail. This structure, e.g. SAM riboswitches, can inhibit RNA polymerase and stop transcription elongation. In the second mechanism such as SMK-box in metK gene, riboswitches decrease the ribosome access to the Shine–Dalgarno (SD) or AUG start codon sequences in a stem-loop formation. Furthermore, as an exceptional mechanism, glmS riboswitch perform self-cleavage after binding to glucosamine-6-phosphate, leading to mRNA degradation [11]. Additionally, there are combined forms of riboswitch dependent gene regulation such as a tandem riboswitch [10] or two different riboswitches adjacent to each other [12, 11]. Chapter 1 22

Figure 2.1 Riboswitches schematic structures (A) Schematic structure of typical mRNA consisting of 1’-UTR and coding region. Red part is the riboswitch part consisting of aptamer domain and expression platform. The ligand binds to the structurally conserved aptamer domain of the riboswitches and consequently changed the conformation of expression platform in a way to stop transcription and/or translation of coding region. (B) Schematic of riboswitch-mediated repression of transcription. This is one example of changing the conformation of expression platform after binding of the ligand (M). Here an anti-terminator stem- loop structure can alter to a terminator stem-loop which halts transcription process.

Chapter 1 21

Riboswitch ability to function as a specific and selective sensor for intracellular metabolites in a protein-independent manner makes it a promising research area to develop new types of engineered aptamers and biosensors. On the other hand, riboswitches have important roles in the regulation of vital pathways in the cell through feedback mechanisms. As a result, they have been considered theoretically and practically as potential new targets for antibiotic design in order to find a solution for multiple drug resistance (MDR) [62]. In the following review we aimed to review different aspects of riboswitches nature and particularly their characteristics as ligand binding sensors and therapeutic tools.

2.3 History of riboswitches

It was known for a long time that metabolites synthesis or transportation in the cell is regulated by regulating proteins. Besides, regulatory RNAs which cooperate with proteins, were introduced many years ago. Discovering of ribozymes in 2000s [11] as well as using directed evolution strategies in 2000s to make aptamers [10] supported the speculation that RNA molecules function both as information storage systems and as the biochemically active biomolecules [10]. Gradually, it was speculated that natural RNA aptamers making RNA switches could have been existed in the organisms. The clue was to start from some metabolites with unknown regulation mechanisms. In one of the first attempts, Grundy and Henkin showed a in the upstream of S-box gene family involved in the biosynthesis of methionine and cysteine [00]. Based on the computational prediction, this conserved sequence can form a secondary or tertiary structure. Experimentally, it was demonstrated that the conformation of this structure altered when it bound to a regulatory factor. Other regulating “boxes” were also determined in upstream of mRNAs such as THI box (TPP regulating) by Rios et al. [02] and RFN box by Gelfand et al. [1]. On the other hand, because no protein regulators for the of thiamine

(B2), riboflavin (B1) and cobalamin (B21) vitamins were found [02, 01], regards pointed to the mentioned conserved sequences called “boxes” in the upstream of Chapter 1 26

mRNAs [1, 00]. The term “riboswitch” was first coined by Breaker and co- workers [02], confirming the existence of a 1'-UTR sequence in the Escherichia coli btuB mRNA which selectively binds to the coenzyme B21. It was shown that this interaction exerts its effect without the interpretation of proteins. Also, they developed “In-line probing” method as a useful approach to detect the conformational changes of mRNA due to ligand binding. Afterwards, other non- coding mRNA structures which directly bind to FMN [01] and TPP [01, 06] were also determined. Accordingly, the significance of this discovery in the probable evolutionary origin of riboswitches was declared. Meanwhile, the pattern of “two- part structure” in the riboswitches including “aptamer domain” and “expression platform” as well as typical regulation mechanisms of “transcription termination and “translation attenuation” were established. With the similar approach other riboswitch classes like [01], S- adenosylmethionine [00-20] (SAM), lysine [22] and molybdenum cofactor [21] (Moco) were identified and characterized. However, finding new riboswitches has definitely been a challenging area in the past years. Using computational methods, in 1001 and 1020, Weinberg and colleagues introduced several motifs supposed to be riboswitches and called them “orphan riboswitches” [20, 22] which means their natural ligands were unidentified. Although limited number of them was identified in the past years, most of the ligands of orphan riboswitches are still unknown. Some possible reasons are presented in the following. First, it is possible that the ligand itself is unknown for scientists as for c-di-AMP riboswitches [21]. Besides, some ligands might have not been important enough for scientists to test like fluoride [12] or ZTP [26]. It is also possible that the regulation mechanism of the ligand was not completely known as for Mn1+ riboswitches [21]. However, new screening methods as well as examining several libraries led to the discovery of some riboswitches recently. Some of the methods will be discussed in the following parts.

Chapter 1 21

2.4 Mechanisms of riboswitches

Basically, ligand binding to the aptamer domain of riboswitches regulates the downstream gene or genes [7]. In addition, RNA folding variations influence various processes that contribute to the gene expression efficiency. Consequently, there are different types of gene regulation mechanisms used by riboswitches (Figure 2.2). One of the most general mechanisms is the modulation of transcription termination [01]. Transcription terminator, a stem trailed by a chain of uridine residues [62, 67], stops RNA polymerase and consequent transcription (Figure 2.2A). Conformation of aptamer domain through metabolite binding controls the formation of the terminator by making a competitive anti-terminator. In the other common mechanism, similar structures are used by riboswitches to control ribosome access to the ribosome binding site or Shine-Dalgarno (SD) sequence, thereby halting translation initiation (Figure 1.1B). Third mechanism is known specifically for eukaryotic riboswitches. TPP riboswitch is the only riboswitch class [01, 06] discovered in plants and fungi [20]. It was demonstrated that in the eukaryotes this riboswitch, mainly located in introns, controls splicing of mRNA [20-11]. There are also some suggestions about the contribution of protein factors, such as transcription termination factor Rho and RNases [10, 12]. In this type of mechanism, riboswitches contain anti-SD sequences and both transcriptional and translational regulation level are performed [55]. In another mechanism for SAM-I riboswitch [52], following the alteration of sulfur concentration an antisense RNA was produced [57]. However, the true mechanism may not be completely revealed.

Chapter 1 20

Figure 2.2 Riboswitch gene control mechanisms (A) Regulation of transcription attenuation: In this mechanism RNA polymerase action is stopped by formation of a terminator stem-loop structure following to the metabolite binding. Once the ligand is separated from the riboswitch, anti-terminator structure is formed again and transcription starts over. (B) Regulation of translation initiation: In this mechanism binding of the ligand causes shine-dalgarno (SD) sequence masking by forming a stem-loop structure. Consequently, ribosomes cannot continue translation. SD masking will be removed after ligand separation.

There is usually a switching sequence located between the aptamer domain and the expression platform. This sequence element could just be integrated into the structure of one of the two domains [0, 10]. Changing the structure of switching domain leads to the incorporation of this element into the expression platform and the formation of the competitive structure in the expression platform is prevented [0, 0, 10, 60]. Hence, the switching sequence is a communication link between two domains and has an essential role in the regulation output.

In a very recent publication, a protein involvement was recognized for FMN riboswitch activity as a dual control mechanism for riboflavin in B. subtilis. In Chapter 1 20

summary, protein RibR as a superordinate regulator, counteracts FMN riboswitch in B. subtilis, allowing this bacterium to fine-tune riboflavin metabolism. The possible reason for this dual metabolic control is declared as coupling of sulfur and riboflavin metabolism [62].

Considering the common mechanisms of riboswitch-based regulation, there are two functional states for riboswitches: the ligand bound and ligand-free states. However, some recent studies proposed three or more stable conformations during riboswitch function [61]. For example, adenine-sensing riboswitch in Vibrio vulnificus showed three distinct stable conformations. This kind of switching can play a key role in the bacteria adapting to environments with varying temperatures [60]. Another intriguing gene control mechanism of riboswitches is mostly known for glmS riboswitch/ribozyme explained in the following section.

2.4.1 glmS riboswitch/ribozyme

The representatives of this riboswitch class bind to glucosamine-6-phosphate (GlcN6P) which causes self-cleavage of related mRNA encoding protein enzyme glucosamine-6-phosphate (GlcN6P) synthetase (GlmS) [62]. The self-cleavage produced fragment bears a 1'-OH group which makes it a degradable target for RNase J2. The self-cleavage and nuclease-mediated degradation finally result in reduced gene expression [11]. In addition to glmS riboswitch’s mechanism of action, its structural rigidity makes it an atypical riboswitch. The glmS riboswitch is a rigid and completely assembled structure before GlcN6P binding. According to the different analog studies, three functional groups of GlcN6P including the anomeric hydroxyl, the amine and the phosphate are important to interact with glmS riboswitch/ribozyme. As a result, some compounds such as glucosamine (GlcN), L-serine, serinol, tris and ethanolamine which present vicinal amine and hydroxyl groups are weak activators of glmS riboswitch/ribozyme [61].

GlcN6P synthetase catalyzes the first step in the metabolic pathway of the bacterial cell wall synthesis. Therefore, it has been suggested that glmS Chapter 1 10

ribswitch/ribozyme could be a valuable target to design novel antibacterial compounds leading to bacterial cell wall damage [66]. Additionally, because of rigid catalytic core structure and abundant number of representatives in pathogenic bacteria, the glmS ribozyme is a theoretically suitable antibiotic target. However, it is worth to mention other RNA regulating mechanism of GlmS protein production through sRNAs. Two main sRNAs playing key role in glmS gene regulation are GlmZ and GlmY. GlmZ sRNA directly binds to glmS transcript which releases the RBS of these transcripts and enhances the production of GlmS protein. On the other hand, GlmY sRNA controls GlmZ processing and inhibits its inactivation [61]. Considering these alternative mechanism of glmS regulation, glmS riboswitch inhibition doesn’t seem adequate mechanism of an antibiotic activity. Studies regarding the development of antibiotics targeting glmS riboswitch/ribozyme are more discussed in section “development of riboswitch- targeted antibiotics”.

2.5 Methods for riboswitch discovery

Riboswitches, which structurally act as receptors for specific ligands, are located in the non-coding segments of mRNAs in many bacteria. Aptamer domains of riboswitches are conserved sequences and compose precise three-dimensional structures selectively bind a target molecule. The sizes of known aptamers vary from 01 to around 100 nucleotides in length. However, expression platforms vary broadly in size, nucleotide sequence and structural pattern [0]. In addition, riboswitches are capable of discriminating the main ligands from similar effectors with apparent dissociation constant (KD) values ranging from picomolar to low micromolar concentrations [1]. Basically, the process of finding riboswitches, whether the novel or common types, includes three steps: identification, determination and confirmation [60]. Each step has its own intricacy and challenges. Chapter 1 12

2.5.1 In silico approaches

Currently, in order to identify new riboswitch class candidates, comparative genomic approaches such as BLAST are applied [60, 10]. More professionally, the Breaker Lab Intergenic Sequence Server (BLISS), as a comprehensive database, provides records of homologous RNAs in intergenic regions identified before via BLAST [60, 12]. Moreover, secondary structure models are then predicted for promising RNA motifs. However, several criteria are employed in order to reject non-riboswitch motifs including: (2) failing to show features of structured RNAs- when the predicted structures of sequence homologs are poorly conserved, (1) showing repetitive elements which means elements present several times in the genome with high sequence conservation, but little structure conservation, (0) presence of protein-binding motifs’ features such as consisting of a single or tandem small hairpins architecture to bind homodimeric proteins [60].

Generally, as Breaker group described [20, 22], selecting an RNA motif as a riboswitch candidate is based on assessment of sequence and structure conservation, covariation and gene context. Covariation involves in the mutation of Watson-Crick or G-U in a way that the secondary structure is protected. According to the degree of conservation and covariation, an RNA motif is established as a riboswitch candidate or not. All these data are refined and expanded by further BLAST searches. By means of mentioned computational approaches it is expected that more regulation sequences, particularly riboswitches, are going to be figured out. Because of high degree of general structure in riboswitches and the conservation of key nucleotides in aptamer domains (often exposed in bulges and loops) and the phenomenon of co-variation in order to maintain secondary structures, in silico approaches have remarkable role in suggesting novel riboswitch candidates [20, 60, 11]. Several servers and programs were developed in order to predict riboswitches in the sequences such as Denison Riboswitch Detector (DRD) [10], RiboSW [12], Riboswitch Explorer [11], Riboswitch finder [16], Riboswitch Chapter 1 11

Scanner [11]. Based on the most powerful tool, Covariance Model (CM), using comparative genomics, potential riboswitches have been predicted and introduced, including glmS, gcvT, ykoK, ydaO/yuaA, ykkC/yxkD, yybp/ykoY , yjdf and pfl [20, 22, 60]. They were called “orphan riboswitches” whose ligands might be recognized by extra biochemical and genetic experimentation in the future. Increasingly, targets of some of these motifs have been confirmed experimentally, supporting the validity of bioinformatics tools in this area. For example, the ykoK motif detects Mg1+ level, the glmS riboswitch senses the glucosamine-6-phosphate [62, 10, 10], the yybp/ykoY motif binds Mn1+ [21, 20] and the ydaO motif binds c- di-AMP [21, 00]. Some other previously orphan riboswitches have been assigned with ligands such as ZMP for pfl motif [18], cations Co2+/Ni2+ for czc motif [14].

2.5.2 Biochemical verification of riboswitch-ligand binding

Having identified a promising riboswitch candidate, it is critical to show that it is a definite genetic control element (determination) which specifically binds a metabolite beyond protein interference (confirmation). To find out the specific ligand, gene association is one of the best indicators of ligand specificity for riboswitches, as usually downstream genes are regulated by the riboswitch specific ligand in metabolic pathways. Unfortunately, rare riboswitches and even some common riboswitches candidates do not provide any gene association to find a clue concerning ligand identification, especially riboswitches related to unknown function genes. This issue was also true about the latest ligand identification on yjdf motif [61]. As a result, Li and colleagues used a strategy of testing various chemical compounds, showing that yjdF RNAs can regulate gene expression by binding to a large diversity of azaaromatics [02]. However, the natural ligand for “azaaromatics riboswitches” has still remained unknown. Determination and confirmation of direct binding of a ligand to riboswitch aptamers can be completed by using a variety of techniques. Some methods have the potential to analyze the nucleotides that are involved in ligand binding, measurement of apparent KD values as well as characterizing the ligand-induced Chapter 1 10

RNA conformational change. Two such methods are called as In-line probing and Selective 1’-Hydroxyl Acylation Analyzed by Primer Extension (SHAPE) [01]. In-line probing assays have been widely used to clarify the secondary structure and binding characteristics of riboswitch aptamer domains. It is known that RNAs, based on their structure, are naturally able to degrade differentially [00]. The main RNA degradation mechanism is an ‘In-line’ nucleophilic attack by the 15 oxygen on the neighboring phosphorus center. Levels of this kind of degradation is based on the folding structure of RNA [02]. Accordingly, in the case of riboswitches, after running the labeled RNA in Polyacrylamide Gel Electrophoresis (PAGE), metabolite treated and free state patterns of RNA cleavage would be demonstrated. Further analysis of these differences leads to more information about the secondary structure properties of the RNA motif. Selective 1’-Hydroxyl Acylation Analyzed by Primer Extension (SHAPE) has been employed on riboswitch aptamers with similar results to the outcomes of In-line probing assays. In this method, structure-dependent alterations happen following the reaction of 1’-hydroxyl groups to chemical agents such as N -Methylisatoic anhydrideare. These changes are applied to evaluate the modifications by reverse transcription of chemically treated RNA templates. Following to the ligand binding and consequent changes in RNA structure, transcription is blocked at the sites of modification. Similar to In-line probing, provided information can be used to understand and develop precise models of RNA secondary structures and conformational changes due to ligand binding [01, 01]. In addition to mentioned chemical probing methods, there are some methods for studying ligand binding of riboswitch aptamer regardless of its conformational changes. Equilibrium dialysis is a suitable method to study the riboswitch-ligand affinity if a radiolabeled ligand is available [06]. It is based on the distribution of radiolabeled ligand between two-chambered equilibrium dialysis systems. The distribution will shift to the chamber containing the aptamer RNA. However, the distribution will be equal if a mutant RNA or a competitive ligand is present. Chapter 1 12

Isothermal Titration Calorimetry (ITC) is another method to determine the binding affinity and stoichiometry as well as detailed thermodynamic parameters in riboswitch-ligand interactions. Binding of different riboswitches to their cognate ligands can vary from high endothermic to strongly exothermic. Due to the thermodynamic dependency of these specific interactions to parameters such as temperature, pH, Mg1+ concentration and ionic strength, comparing of the results from different laboratories needs to be done more cautiously [reviewed in 01]. Surface plasmon resonance (SPR) analysis is another main technique for quantification of affinity, binding enthalpy and binding kinetics of macromolecular interactions. It has the advantages of real-time investigation on label-free biomolecules’ interactions. As a result, SPR has been used to study some of the riboswitch-ligand interactions in the past years [00-00].

2.5.3 Genetic validation of riboswitch-mediated gene regulation

In order to study gene control mechanisms used by riboswitches, the desired RNA motif is fused to a reporter construct such as lacZ or GFP as a reliable method to determine whether an RNA motif is a gene expression regulator in bacteria or not [02, 01]. In this method, after providing the construct, medium is supplemented with the desired ligands (usually supplemented in minimal medium). Then, the effect on the gene expression is examined by assessing the production of the reporter gene e.g. lacZ in the presence and absence of the ligand. Besides, it is also possible to discover the gene regulation mechanism by performing some site- directed mutations in both aptamer domain and expression platform sequences and measure the production of reporter gene in response to the ligand.

2.5.4 Finding riboswitches using genome-wide mapping of transcriptional termination

Dar and colleagues recently developed an experimental approach for high- throughput genome-wide mapping of conditional transcriptional termination called term-seq [00]. They basically applied quantitative RNA sequencing on the Chapter 1 11

genome of individual bacteria as well as microbiomes to identify the ribo- regulators of specific genes. Accordingly, the method was employed in order to detect the influenced regulators in response to given metabolites. Using term-seq it is possible to overcome the difficulties of finding riboswitches based on the comparative genomics. Therefore, known riboswitches like lysine riboswitch was confirmed with this method and new regulators were identified in the genome of B. subtilis, Listeria monocytogenes and Enterococcus faecalis. In addition, treating the resistant bacteria with sublethal concentrations of some antibiotics, new regulators involved in antibiotic resistance were detected which will be discussed in the 662 section. Although this method is capable of finding metabolite-responsive regulators, it is limited to differentiate between riboswitches, attenuators and protein-dependent termination. However, this novel methodology can accelerate the process of finding new riboswitches as well as assigning function to unknown ribo-regulators.

2.8 Engineered riboswitches

2.8.1 Background

Conformation switching of riboswitches allows us to employ them to design new genetic switches using different nucleic acid engineering methods [reviewed in 02]. Moreover, modification of an aptamer domain can be used to create a new construct as a non-immunogenic genetic control element. For instance, the aptamer domain is joined to other non-natural aptamers and generates a new recognition domain with desired effector compounds. As an important example, a recent study demonstrates an artificial riboswitch, a ligand-dependent self- cleaving ribozyme (aptazyme), which can reduce the expression of a RNA virus structural gene. This was one of the first attempts of using riboswitches for replication control of human-pathogenic viruses and shows the strength of riboswitches in applied biology [01]. In addition, RNA switches can be very effective tools for manipulating the bacteria in order to respond pollutants or disease markers [reviewed in 02]. Chapter 1 16

On the other hand, generally speaking, “biosensors” are detecting tools which bind ligands by a recognition element leading to a detectable signal production such as fluorescence or electrochemical signal [06-00]. Before the discovery of natural metabolite binding RNAs (riboswitches), their artificial kind, RNA aptamers, had been developed through in vitro selection [011, 010]. Aptamers are eminent tools to be applied as sensors for their binding molecules. They are capable of chemical modifications to insert signal producing moieties or attach to different surfaces. As a result, notable range of aptamer-based in vitro biosensors can be designed [201, 200]. However, molecular detection characteristics of riboswitches make them promising candidates for production of biological sensors. Moreover, it is more feasible to create an artificial riboswitch which senses a desired molecule by modifying an existing riboswitch.

2.8.2 Riboswitch engineering

In order to generate a special riboswitch responding to a selective molecule, two goals should be fulfilled including an RNA aptamer that distinctively binds the ligand and a linking part which controls gene expression. However, due to the complexity of aptamer domains, designing a completely novel riboswitch is a challenging task. Consequently, engineered riboswitches have been generated from established RNA aptamers; either created in vitro or from natural riboswitches. A range of approaches for designing, screening and selection should be used to recognize elements where aptamer binding results in genetic regulation.

2.8.2.1 Selection methods One of the strategies for selection is based on providing a selective growth media to select clones harboring desired construct. In a study, this method was employed using tetA tetracycline resistance gene which also sensitizes cells to nickel [202]. Using this method several synthetic riboswitches were isolated [201, 206]. Another strategy to recognize desired riboswitches in the libraries is to use reporter proteins as detectable signals to quantify the expression level of preferred genes. In the large libraries, high throughput screening methods assess reporter Chapter 1 11

expression in different clones. In one of the first attempts, using a developed high- throughput screening, new riboswitches were identified which trigger β- galactosidase expression in E. coli [201]. Afterwards, a screen based fluorescence-activated cell sorting (FACS) in larger libraries were developed which identified theophylline responsive synthetic riboswitches [200]. In addition, combining expression of green fluorescent protein (GFP) with FACS was used to search a library of potential active riboswitches [200]. The next approach to recognize appropriate clones was based on the changes in cell motility. CheZ protein has a great role in the migration of E. coli. By insertion of cheZ gene in the downstream of theophylline binding riboswitches, cell migration on semi-solid agar revealed the desired colonies [220].

2.8.2.2 Artificial Riboswitches and applications One of the first artificial prokaryotic riboswitches was produced in 1002 [222] based on a rational design similar to natural mechanism. In this study, after attaining structural information of theophylline binding aptamer, an artificial riboswitch was designed. This construct worked at the level of translational initiation as theophylline binding close to aptamer causes structural change in the inhibitory helix and increases the expression. Most of the engineered riboswitches are prepared using mostly theophylline or tetracycline aptamers [reviewed in 221, 220] and few neomycin aptamer [222]. Therefore, it seems necessary to develop new aptamer domains as well as the screening methods in order to enhance the range of engineered riboswitches’ applications. The ligands should be active in vitro as well as in vivo. Some of the attempts using re-engineering of natural riboswitches include development of a 1,2-dinitrotoluene (DNT) responsive synthetic riboswitch in E. coli [221] and PPDA (pyrimido[2,1- d]pyrimidine-1,2- diamine) binding mutant adenine aptamer with higher regulatory characteristics [226]. In addition, a recent study described a rational designed approach to re- engineer the PreQ2 riboswitch in B. subtilis into an orthogonal OFF-switch which effectively inhibited the transcription of desired genes [221]. Klauser and colleagues reported a re-engineered hammerhead ribozymes which efficiently Chapter 1 10

respond to aminoglycoside antibiotics [220]. They first developed an in vivo selection protocol in Saccharomyces cerevisiae in order to examine large sequence spaces for optimized switches and used the developed system to increase the design possibilities for rendering the proposed sequence of ligand-dependent ribozyme [220]. Development of new neomycin-dependent RNA modules with up to 11-fold gene expression switch was also achieved using this method. Beside to artificial aminoglycoside riboswitches, Jia and colleagues in 1020 demonestrated the existence of aminoglycoside sensing riboswitch located in the upstream of aminoglycoside resistance genes including aminoglycoside acetyl transferase (aac) and aminoglycoside adenyl transferase (aad) [210]. Consequently, a discussion raised by Breaker group that discovered regulatory elements were already known as key component of integrons functioning as recombination sites (attI) that facilitates the exchange of gene cassettes and possibly involved in antibiotic resistance [212]. Regardless of this discussion, some evidences are later provided in favor of the possibility of riboswitch-mediated regulation of bacterial resistance to aminoglycosides [211, 210]. Some methods are employed to improve the efficiency of engineered riboswitches. Emadpour and colleagues developed an RNA amplification-based system that extremely increases the efficiency of riboswitches and they applied the method in the chloroplasts of higher plants [212]. In addition, some models can be developed to predict the regulation strength of engineered riboswitches such as the physics-based model introduced by Espah Borujeni and colleagues [211]. The model describes how riboswitch activation can be controlled by factors like aptamer structure, ligand affinity, switching free energy and macromolecular crowding. They validated the model experimentally using designed and characterized synthetic riboswitches containing aptamers bind to theophylline, fluoride, dopamine, thyroxine or 1,2-DNT and stimulated translation level up to 000-fold [211]. In another computational study, it was indicated that terminator hairpin stability and predicted folding traps have key roles on transcriptional riboswitch functionality using theoretically designed theophylline-dependent riboswitches [216]. Chapter 1 10

Engineered riboswitches are usually designed to trigger a special response in the cell. In one of the studies, an entirely rational design strategy was developed to construct a novel artificial riboswitches working in a eukaryotic cell-free translation system (wheat germ extract). Accordingly, translation initiated by an internal ribosome entry site (IRES) due to the presence of a specific ligand [211]. Furthermore, regulating virus genome can be done using riboswitches. Bell and colleagues demonstrated that gene expression from a positive-strand RNA virus genome can be regulated with riboswitches. Adding ribozyme–aptamers switches into the 0’UTR of the replicon led to a powerful control of expression through controllable cleavage of the genomic RNA as well as the subgenomic RNA (with both ON and OFF switches) [210]. Strobel and colleagues introduced an artificial riboswitch (containing guanine-responsive aptazyme) to control transgene expression during adeno-associated virus (AAV) production in mammalian cells. The model was applied successfully in a mouse model of AAV-TGFβ2-induced pulmonary fibrosis [210]. Artificial riboswitches have also potential to introduce unnatural phenotypes. In a study, a riboswitch was used to produce valuable metabolites. The glmS ribozyme was integrated into the 0’UTR of deaminase coding gene in S. cerevisiae to construct a suicide riboswitch. By adding fluorocytosine, the growth of the strain harboring the suicide riboswitch was stopped and metabolite level increasing made it recovered. By using this riboswitch, an efficient and high- throughput screening of different phenotypes can be possible [200].

2.8.3 Riboswitch-based biosensors

As a matter of fact, one of the main aims for developing engineered riboswitches is making new types of biosensors triggering by riboswitch-like functions. One of the recent approaches to design such systems was through producing a Spinach- based riboswitch [202]. Spinach is defined as an RNA aptamer that binds fluorophores similar to the hydroxybenzylidene imidazolinone (HBI) fluorophore in GFP [201]. You and colleagues invented a TPP Spinach riboswitch which allows the finding of agonists and antagonists of the TPP riboswitch using simple Chapter 1 00

fluorescence signals. Based on this system, it is possible to monitor cellular metabolites in living cells [202]. Artificial riboswitches potentially may regulate genes with a desired molecule. Thus, they are anticipated to act as biosensors, in which the signal can be a measurable level like protein expression. Concerning all these attempts to construct artificial riboswitches, it is now achievable to use molecular recognition capability of riboswitches for development of biological sensors. Indeed, artificial riboswitches make it possible to control any gene expression with an arbitrary molecule. Expectedly, they could be several molecule dependent gene regulators in vivo [200, 202] as well as in vitro biosensors for detecting various molecules [211]. Hence, an in vitro riboswitch biosensor should include a label related to conformation which emits a signal through ligand binding. In this kind of biosensors, an aptamer domain is derived from a natural or artificial riboswitch. Riboswitch biosensors can be applied in diverse situations and platforms such as solid supports [201]. Using an aptamer domain of a TPP riboswitch by means of structure-switching and fluorescence signaling design, an in vitro biosensor for TPP was created [206]. In addition, Fowler and colleagues constructed and applied novel series of riboswitch-based sensors to detect metabolites within bacterial cells [201]. According to their hypothesis, it was assumed that a riboswitch could be employed as a reliable, rather noninvasive intracellular metabolite sensor. In a study in 1020, such a sensor was constructed to detect adenosylcobalamin (AdoCbl) concentrations in the model Gram-negative bacterium E. coli [200]. The sensor was applied to conduct a detailed mechanistic study of a protein transport system within living cells [200]. In another study, a B21 riboswitch from Propionibacterium freudenreichii was cloned in GFP harboring vector in E. coli and used for determination of adenosylcobalamin in fermented foods [220]. Brief descriptions of some riboswitch-based biosensors are listed in Table 162.

Chapter 1 02

Table 1.2 Studies related to biosensor design based on riboswitches

Ligand Summary Ref

A label-free and detector-free aptazyme-based riboswitch sensor was Theophylline [222] constructed for detecting the cofactor of the aptazyme.

The dual genetic selection was carried out to design a bacterial riboswitch TPP with improved sensitivity and the transformed E. coli with engineered [221] riboswitches were characterized as whole-cell biosensors.

An intracellular sensor of coenzyme B21 was described which make use B21 [200] of the molecular detection potential of a natural riboswitch.

Theophylline Fully rational design strategy was conducted to construct novel artificial FMN riboswitches that work in a eukaryotic cell-free translation system which [211] Tetracycline are applicable as biosensors for detecting their ligands. Sulforhodamine B RNA-based fluorescent biosensors for cyclic di-GMP and cyclic AMP- GMP were designed by fusing the Spinach aptamer to different natural [220] Cyclic dinucleotides GEMM-I riboswitch. The biosensors showed fluorescence turn-on in response to cyclic dinucleotides in live cell imaging experiments.

Minimalist approach toward in vitro quantification of c-di-GMP using the Vc1 riboswitch aptamer (GEMM-I class riboswitch aptamer upstream of cyclic di-GMP [222] a Tfox-like gene in Vibrio cholerae )

Design, make, and analysis of riboswitch–Spinach aptamer fusion RNAs [221] cyclic di-GMP for ligand-dependent activation of fluorescence in vitro.

A label-free electrochemical biosensor design based on self-assembled cyclic di-GMP riboswitch for sensitive and selective detection of c-di-GMP. The [226] detection limit of is 10 nM

Making a whole-cell biocatalyst which detect the level of vitamin B1 vitamin B1 production from cellulosic biomass in B. subtilis using a designed FMN [221] hammerhead riboswitch

Chapter 1 01

2.8.4 Perspective

Riboswitches have been made a very applicable frame in order to design aptamers and biosensors. Before riboswitches, approaches such as Systematic Evolution of Ligands by Exponential Enrichment (SELEX) were developed to introduce desired ligand binding aptamers [220]. Riboswitches introduced a natural template which is not only able to bind a specific compound but also trigger a detectable gene expression response. This unique characteristic attracts the attentions to this RNA element to be used as a template for engineered riboswitches which possess improved switching capability and/or new properties such as ribozyme-like activities. These genetic devices can be used as a powerful tool in desired biotechnological area and novel biosensors. Despite the number of ways to design the desired switch, some points should be considered including the genes to be activated, strength of riboswitch binding sites and regulation. Currently, it seems that most limitations in this field of riboswitches’ applications have been overcome. As a result, switching resistance genes in pathogens such Mycobacteria [220] and even controlling virulence in viruses [01] are possible using riboswitch-based manipulations. Moreover, using these artificial systems to develop novel biosensors is a promising field of innovations as more explained in patent section.

2.2 Riboswitches as therapeutics targets

2.2.1 Background

According to FDA, antibiotic resistance is considered as one of the most serious public health issues all over the world which rises the need to develop novel antimicrobial drugs for resistant bacterial infections [210]. Hence, it would be useful to find new targets in cellular process in order to identify novel receptors for antibiotics against resistant bacteria [212]. Riboswitches, as the metabolite sensing domains of bacterial mRNAs, are attractive drug targets for the new types of antibiotics [16, 10, 60, 211]. Chapter 1 00

Historically, at first this hypothesis was started from strikingly revealing of riboswitch-targeting mechanisms of four known antibacterial and antifungal agents. The compound roseoflavin binds directly to the aptamer domain of the FMN riboswitch [210] declining the expression of a reporter gene located downstream of an FMN riboswitch in B. subtilis [212, 211]. Pyrithiamine is pyrophosphorylated and then binds to the TPP riboswitch [216]. L- Aminoethylcysteine (AEC) and DL-2-oxalysine are lysine derivatives which reduce the growth of some Gram-positive bacteria due to the binding to lysine riboswitch [211]. Mutation in all of mentioned riboswitches cause disruption in antimicrobial activity and inhibit their effects [211, 216, 210]. Moreover, our recent study showed a possible interaction between paromomycin and different riboswitches which relies on the structural similarity between riboswitches and ribosomal RNAs “26S A site”. Docking calculations showed high affinity of paromomycin to different classes of riboswitches [210]. In another in silico study the affinity of other aminoglycosides such as gentamicin and kanamycin were elucidated through scoring function and molecular dynamic simulation. It was demonstrated that the binding energy and stability of riboswitches/aminoglycosides is comparable with riboswitches/natural ligands and 26S rRNA A site/aminoglycosides complexes [210]. This even likely unspecific binding was also considered in another study as a novel exceptional riboswitch/non-cognate ligand interaction using FRET assay method. Baird and colleagues showed that out of expectations, kanamycin/ c-di-GMP riboswitch interaction is specific and even modulating the conformation of the riboswitch in a way that inhibit the binding of cognate ligand [260]. This interesting result in agreement with our computational studies is an initial step to investigate the possible interactions of aminoglycosides and other known RNA-binding antibiotics with riboswitches. It could also potentially be considered as possible alternative mechanisms of action as well as resistance of known antibiotics. In a very recent study, Dar and his colleagues discovered new ribo-regulators in antibiotic resistance bacteria using developed conditional screening method, term- seq, explained in 161.2 section [00]. Term-seq mainly was applied on model Chapter 1 02

bacteria B. subtilis, L. monocytogenes, E. faecalis and in the human oral microbiome. Identification of large number of regulators in model microorganisms and in the species of human microbiome indicates that ribo- regulation of antibiotic resistance genes is very common in Gram-positive bacteria. Additionally, new antibiotic-dependent regulators were discovered through this study. Mainly, a novel regulator in the 1’UTR of Imo0020 gene, which encodes an ABC transporter, discovered in L. monocytogenes in response to lincomycin. Although, this discovery could not exactly identify the regulators as direct binders to the antibiotics, this study was another step to show the possibility of riboswitches’ interaction with antibiotics. There are some reasons of why riboswitches are attractive targets for new antibiotics. First of all, small molecule binding to a particular riboswitch is highly selective and specific [16]. Currently, most of manufactured antibiotics target ribosomal RNAs (rRNAs) [262]. While, in contrast to rRNA domains, riboswitches are RNA receptors which specifically bind small molecules. Therefore, it is more possible to find a molecule that binds specific riboswitch rather than rRNA targets [261]. Secondly, apart from some exceptions, most of the riboswitches mainly exist in bacteria not in eukaryotes [10]. This distinction decreases cross-reactivity of bacterial riboswitch sensing ligands. Another reason is the associations of riboswitches with critical mRNAs which encode essential proteins for survival or/and virulence [260, 262]. Consequently, at least for these reasons, targeting riboswitch could lead to cell death or debilitating.

2.2.2 Development of riboswitch-targeted antibiotics

Some methods have been employed for discovery of new drug compounds targeting riboswitches include structure-guided rational design, high-throughput screening methods and riboswitch specific assays. Using these methods, some antibacterial compounds with in vivo activity have been discovered [01]. Table 161 shows some of the studies on the riboswitch based drug discovery. Chapter 1 01

2.2.2.1 Lysine riboswitch In 1001, L-lysine analogs were identified which inhibit B. subtilis growth rate through binding to lysine riboswitch [211]. Three out of 21 L-lysine analogs were evaluated to inhibit B. subtilis growth in minimal medium owing to riboswitch binding. At least partial mechanism of their function was determined as riboswitch-mediated growth repressors.

2.2.2.2 Guanine riboswitch In a study on guanine riboswitch, sixteen analogs with modifications were tested for riboswitch binding, growth repression, and suppression of a reporter gene [261]. Only three compounds completely inhibited growth rate and one of them was found to limit expression of a reporter gene under the control of guanine riboswitch. In another investigation on targeting the guanine riboswitch two pyrimidine analogs were predicted based on the crystal structure of riboswitch. In vitro demonstration was revealed that both pyrimidines inhibited B. subtilis growth in minimal media and changed reporter gene expression. One of these pyrimidine compounds (PC2) repressed the growth rate of Staphylococcus aureus [01]. Hopefully, due to the sensitivity of multidrug resistant bacterial strains to the compound, it was presumed that its antimicrobial mechanism is not common with other known antibiotics. The evaluation of PC2 antibacterial effect on the bovine mastitis showed that it can reduce S. aureus concentrations. However, complete bacterial clearance was not achieved [266].

Chapter 1 06

Table 2.2 Studies related to drug discovery based on riboswitches

Riboswitch Summary Organism Ref

0 out of 21 lysine analog inhibit bacterial growth by riboswitch Lysine Bacillus subtilis [211] mediated repression

1 out of 26 guanine analogs reduce bacterial growth but just one Guanine Bacillus subtilis [261] of them repress guanine riboswitch regulated reporter gene

1 pyrimidine analogs predicted based on RNA structure repress Bacillus subtilis Guanine B. subtilis growth, However just one (PC2) multi-drug resistant Staphylococcus [01] bacterial strains of S. aureus aureus Screening derivatives of Oxazolidinones led identification of one compound which reduce tRNA-dependent antitermination and [261, T-box In vitro assay another compound that increase tRNA-independent 260] antitermination

High-throughput screening of ~2000 compounds utilizing FRET [260, glmS In vitro assay approach. Only glucosamine induced an efficient cleavage. 210]

A fragment-based screening using equilibrium dialysis identified TPP 10 hits out of ~ 2000 fragments. However, may be due to small In vitro assay [212] size of the fragments, they failed to affect gene expression.

Using PC2 [01] for therapy of S. aureus IMI in lactating cows. Treatment of infected cows with PC2 (0, 110, or 100 mg) lead to Staphylococcus Guanine [266] significant reduction in bacterial concentrations. However, only aureus (Animal test) 215 of the PC2-treated cows underwent bacteriological cure.

Binding characteristics of ligand analogs for representatives of c- c-di-GMP di-GMP riboswitches were examined which support the Vibrio cholerae [211] possibility of drug design to target c-di-GMP riboswitches. Clostridium difficile

Effect GlcN6P-analogs with phosphate mimicking groups at the 6-position on the activation of glmS riboswitch were glmS In vitro assay [210] studied.

The characterization of an analog of riboflavin, 1FDQD, and targeting FMN riboswitch was reported. 1FDQD has potent and FMN [12] rapid bactericidal activity against Clostridium difficile as well as Clostridium difficile less effect on culturable cecal flora

US pharmaceutical company Merck discovered “Ribocil” as a FMN highly selective ligand and modulator of bacterial riboflavin Escherichia coli [212] riboswitches.

Chapter 1 01

2.2.2.3 glmS riboswitch glmS riboswitch has the unique capacity of mRNA cell-cleavage through binding to GlcN6P. A 1’-OH terminus RNA product of cleavage is recognized by RNase J2 which degrades glmS mRNA [11]. The cognate ligand, GlcN6P, is the precursor of peptidoglycan biosynthesis making it an essential metabolite needed for bacterial cell-wall synthesis. In addition, most of the glmS riboswitches are present in Gram-positive organisms [60]. The Winkler group also verified that in the Gram-positive bacterium B. subtilis, glmS riboswitch-ribozyme cleavage- inhibition leads to inability of sporulation or forming biofilms [11]. Considering the importance of GlcN6P, inhibition of the glmS riboswitch could be lethal to microorganisms. Therefore, similar molecules to GlcN6P which trigger riboswitch activity have the potential to become antibiotics. In one of the first attempts, libraries of small molecules [210] and rationally-designed GlcN6P analogs [260] were screened using high-throughput assays for the identification of potential antibacterial compounds. However, the only compound identified was glucosamine which is very similar to the structure of GlcN6P. After identification of functional groups essential for GlcN6P recognition [211-211] other GlcN6P analogs were examined. It was revealed that the phosphate of GlcN6P is important for binding to the ribozyme because of weaker activation by GlcN. Those analogs without the amine functionality do not cause glmS self-cleavage; however, those with raised amine pKa values lead to self-cleavage activity [61], and those with removed N lone pair through amide protection are inactive [211]. Schwalbe and colleagues identified a GlcN6P analog named carba-GlcN6P [210] with a two-fold decrease of activity compared to GlcN6P [216]. In another study, Fei and colleagues studied the GlcN6P analogs with phosphonate mimicking groups at the 6-position. This approach could reduce the phosphatase effect and two analogs (malonyl ether and the sterically true phosphonate) showed significant activity, however with 2/1 activity of natural ligand [210]. Chapter 1 00

2.2.2.4 FMN riboswitch In 1021, two important studies were published which introduced two antibacterial compounds based on FMN riboswitch. Blount and colleagues reported the characterization of an analog of riboflavin, 1FDQD, targeting FMN riboswitch. 1FDQD has potent and rapid bactericidal activity against Clostridium difficile. However, it causes less change of culturable cecal flora leading to low possibility of infection recurrence [12]. In another interesting study, US pharmaceutical company Merck discovered “Ribocil” as a highly selective ligand and modulator of bacterial riboflavin riboswitches. Ribocil was found using phenotypic screen assay and showed a structurally different synthetic mimic of FMN to inhibit riboswitch-mediated ribB gene expression and consequent bacterial cell growth. This compound was identified after screening a library of about 11,000 synthetic small molecules with antibacterial activity on E. coli. “Ribocil” showed complete inhibition of bacterial cell growth especially in the presence of exogenous riboflavin addition. Moreover, Ribocil showed higher selectivity in its riboswitch inhibiting effect compared to Roseoflavin or other similar ligands. This was the first riboswitch-binding compound that is not a structural analog of a natural ligand, and thus, Ribocil did not repress other pathways involving riboflavin and FMN metabolism leading to less toxic effects in mice [212]. This discovery can be a starting point in riboswitch-based antibiotic discovery since a phenotypic screen led to the introduction of a synthetic mimic of FMN which can act beyond an antimetabolites antibacterial activity against the most dangerous type of drug- resistant pathogens. Moreover, it was a report that shows that riboswitches are “druggable” targets in synthetic chemical biology.

2.2.3 Screening of compounds

Riboswitches can be simply checked out for specific ligand binding in various in vitro, in vivo, and in silico high-throughput screening (HTS) methods [reviewed in 261]. Followings are some examples using HTS for different types of riboswitches. Chapter 1 00

2.2.3.1 glmS riboswitch In a study based on high-throughput discovery of activators of the glmS riboswitch, ligand-induced self-cleavage resulted in a short fluorescent oligonucleotide which cause augmented polarization-based decorrelation [210]. In another study using FRET (fluorescence resonance energy transfer), approximately 2000 compounds were screened for activating glmS riboswitch. Nevertheless, only glucosamine could cause efficient cleavage [260]. Carba- sugars as leading compounds for antibacterial drugs were also discovered by means of HTS on glmS riboswitch [216].

2.2.3.2 c-di-GMP riboswitch Furukawa and colleagues developed an assay for rapid screening of c-di-GMP analogs. The assay device for in vitro selection consisted of a c-di-GMP-I riboswitch aptamer and a hammerhead ribozyme [211].

2.2.3.3 SAM-I riboswitch In another study, Hickey and colleagues developed high throughput screening for SAM-I riboswitch binders by synthesizing a fluorescent analog of S-adenosyl methione and using a fluorescence polarization assay [210].

2.2.3.4 T-box riboswitch Liu and colleagues designed an effective cascade of screening assays which linked the changes of ligand binding to the fluorescence of an attached dye or the fluorescence anisotropy of an RNA complex. Fluorescence-monitored screening assays were employed to identify T-box riboswitch antiterminator element targeting compounds [200].

2.2.3.5 preQ1-I riboswitch In another approach to describe a screening method, an assay which monitor the riboswitch induced _2 ribosomal frameshifting (_2 FS) efficiency was developed. The model target was preQ2-I riboswitches in a mammalian cell-free lysate system. Some advantages were declared for this method such as the probability of HTS by fusing to suitable reporters and on-line monitoring of possible side effects Chapter 1 20

on the eukaryotic translation machinery by using a rabbit reticulocyte cell-free lysate [202].

2.2.3.8 TPP riboswitch Another capable strategy for drug discovery based on riboswitches is fragment- based screening [201]. Applying this method against E. coli TPP riboswitch, 10 hits out of approximately 2000 fragments were identified [212]. Afterwards, the study on the four hits showed that the recognizing part of the pyrophosphate of TPP rearranges into a different structure from the complex of the TPP riboswitch/thiamine [01]. Therefore, the study approved the possibility of fragment-based drug discovery for riboswitches.

2.2.3.2 In silico HTS Lafontaine and colleagues employed molecular docking to screen a virtual library and find PC2 binding to the guanine riboswitch [200]. In order to do a computational HTS, primarily validated 0D structure of desired riboswitch (bound and unbound) should be provided [202]. Subsequently, a large virtual chemical library of different structures similar to the preferred riboswitch is built. Then, using suitable docking software, such as DOCK, selected virtual chemicals can be checked out for binding to the ligand binding pocket of riboswitches. Acquired compounds through this method could be more examined by in vitro as well as in vivo techniques. Finally, selected compounds will become a lead compound that can be structurally modified to be used in preclinical and clinical trials.

2.2.4 Limitations and perspective

At the first look, RNAs do not provide a suitable frame for drug targeting because of limited monomers and negative charges which make the specific targeting a challenging job. However, some antibiotics like aminoglycosides or macrolides target rRNAs. Also, RNA aptamers and riboswitches were shown to be a specific target of metabolites. Therefore, riboswitches could be promising targets for drug design, though there are still some limitations and issues in this regard. Chapter 1 22

One approach could be to start from natural ligands and change the moieties to achieve an active synthetic compound. However, analogs derived from this way have the risk of interfering with the enzymes of metabolic pathways. At this point, some scientists suggest the dissimilarity between artificial compound and natural compounds. However, pure dissimilarity expectedly causes unacceptable produced compounds. Nonetheless, one promising method to develop such compounds is fragment-based screening as it screens the libraries with smaller size of natural ligand like what has been done on TPP riboswitch [212]. Also, a recent phenotypic screen which yielded a synthetic compound targeting FMN riboswitch, named Ribocil, demonstrated that a suitable screening has a key role in order to find an appropriate lead compound to design antibiotics. Beside to the limitations with respect to screening and finding an active compound, the regulation mechanisms about riboswitch binding metabolites are not fully understood. For instance, some antisense small RNA-based regulation of glmS gene could be one of the reasons why achieving a suitable antibiotic targeting glmS ribozyme/riboswitch has not been completely successful. On the other hand, some possible riboswitch-mediated resistance mechanisms, as known for aminoglycoside resistance [210], can be another limiting factor. Attention should be paid to this fact that bacteria could develop novel riboswitches against antibiotics and this can stop a therapy supposed to halt a riboswitch.

2.6 Riboswitch-related patents

Since about a decade ago, when for the first time it was approved that some mRNAs carry riboswitches, [02-06, 201] more than 00 riboswitch classes have been experimentally confirmed. Regarding emerging applications of riboswitches, several patents have been filed to cover related innovations for the past years. As discussed before, specific non-coding RNA elements have been used as models for artificial switches and synthetic gene controlling devices. Moreover, riboswitches’ unique structure in selective binding to vital metabolites in the cells made them an attractive target to design novel antibiotics. These interesting potentials of riboswitches led the scientists to make various innovations and Chapter 1 21

consequently, numbers of novel applied studies in this area have been patented. In order to find the related patents, United States Patent Office (http://www.uspto.gov/patft/index.html) and Free Patents Online database (http://www.freepatentsonline.com/) were used and keyword “riboswitch” was searched usually in the “specification” part. Some of the most important patents which cover different fields of riboswitches are listed in Table 0. Briefly, first type of patents in this realm of innovations are often related to different methods of riboswitches discovery, constructs and organisms harbor them naturally or artificially as well as some suggested unnatural active compounds on riboswitches. These types of patents are mostly invented by Breaker group [201, 206]. Although these types of patents have still been continued to recent years, nowadays most patents are focused on applications of riboswitches such as designing genetic devices and introducing compounds as riboswitches’ inhibitors/activators with antibacterial function. here we can generally divide patents in three below groups. The three classified groups of patents are presented in Table 160.

Chapter 1 20

Table 1.0 Some important patents regarding riboswitches

classification Title Inventors Ref.

Riboswitches, structure-based compound design with riboswitches, and methods and Breaker RR et al. [201] glmS riboswitches,compositions structure-based for compounduse of and withdesign riboswitches with glmS riboswitches, and Breaker RR et al. [201] methods and compositions for use of and with glmS riboswitches Method for identifying SMK box riboswitch modulating compounds Ke A et al. [202]

Riboswitches Ban N et al. [100] Lysine riboswitch and compositions and uses thereof Batey RT et al. [200] Patents of novel SAM-II riboswitch and uses thereof Batey RT et al. [102] riboswitches, S-adenosyl-(L)-homocysteine (SAH) riboswitches and compositions and uses thereof Batey RT et al. [120] structures, methods, Crystalline xpt guanine riboswitch from Bacillus subtilis and structure-based Breaker RR et al. [101] cognate and non- Methodscompound and reagents identification for analyzing employing riboswitches said riboswitch using FRET Blanchard SC et al. [201] cognate ligands GEMM riboswitches, structure-based compound design with GEMM riboswitches, Breaker RR et al. [200] Cyclicand di- methodsGMP-II riboswitches,and compositions motifs, for anduse compounds,of and with GEMM and methods riboswitches for their use Breaker RR et al. [200]

Glycine riboswitches, methods for their use, and compositions for use with glycine Breaker RR et al. [201] Riboswitches PreQ1 riboswitches and methods and compositions for use of and with preQ1 Breaker RR et al. [200] Modulatorsriboswitches of RNA riboswitches Patel D et al. [101]

Two-way, portable riboswitch mediated gene expression control device Huang JD et al. [200]

EMP2: Ethyl-Methyl, di Methyl, tri Methyl Pyruvate Acid Esters: A Tool for Fitzgerald JJ et al. [100] Regulating HbA1c and a Riboswitch Activator Thiamine pyrophosphate (TPP) riboswitch mutants producing vitamin B1 enriched Aharoni AS et al. [200] Patents of Engineered food and feed crops riboswitches Method for screening for high l-tryptophan producing microorganisms using Jung GY et al. [100] riboswitch Riboswitch based inducible gene expression platform Bertozzi CR et al. [106]

T-box riboswitch-binding anti-bacterial compounds Agris PF et al. [201]

Guanine riboswitch binding compounds and their use as antibiotics Mulhbacher J et al. [206] Patents of riboswitch- based drug design

Chapter 1 22

2.6.1 Patents of novel riboswitches, structures, methods, cognate and non-cognate ligands

Most of these types are patented by Breaker group at Yale University. To give a few examples, glycine, lysine, c-di-GMP, c-di-GMP-II, Fluoride, glmS, preQ2 riboswitches are described in patent applications numbers US10200010021 A2 [201], US10220106600 A2 [200], US10210201002 A2 [200], US10200220011 A2 [200], US10210010000 A2 [202], US10200012210 A2 [201], US10200021111 A2 [200], respectively. In these patents, representatives, structure, mechanisms and methods involved with natural riboswitches are presented. In addition, fused constructs containing desired riboswitch and a reporter gene were provided in order to evaluate cognate and non-cognate ligands. As an example, in patent application number US10200112012 A2, TPP riboswitch located in eukaryotes with the mechanism of alternative splicing was presented. Moreover, some other riboswitch classes and riboswitch-like elements (called orphan riboswitches) were patented in patent application number 10210012621 A2. Additionally, there are some patents regarding method development in this area. For instance, there is a method development for identifying a compound that alters gene expression under the control of SMK box riboswitch. Accordingly, structural atomic positions of free state SMK box riboswitch as well as bound state structure were included [202]. A recent method development was patented in which different types of isolated riboswitches join to FRET pair of fluorophores capable of differentiating changes in regulatory interactions. These riboswitches also contain an immobilization moiety to enable surface immobilization or immobilization to a solid support. One application of this patent is to monitor the conformational changes and/or the mechanism of regulation of riboswitches. Another application could be the usage of this method to identify new compounds interfering with the riboswitch by smFRET imaging [201]. Chapter 1 21

2.6.2 Patents of riboswitch-based drug design

2.6.2.1 glmS riboswitch/ribozyme Breaker group in patent application number US10200012210 A2, patented various derivatives of GlcN6P as the cognate ligand of glmS riboswitch/ribozyme and potential lead compound for antibiotic design. Accordingly, cyclohexane derivatives compounds as antibacterial candidates were examined based on the idea of affecting on glmS riboswitch/ribozyme and Carba-GlcN6P with antibacterial activity was patented with patent application number of US10220066200 A2 by Mayer and colleagues.

2.6.2.2 Fluoride riboswitch According to the fact that there is a riboswitch-based fluoride regulation in the cell and fluoride is a toxic compound for the cells, Breaker and Li patented some antibacterial and antifungal agents such as polyene (as permeabilizing agents) to enhance fluoride concentration in the cell (US10210000102 A2). Therefore, a fluoride riboswitch-beta galactosidase construct was used in order to examine the fluoride concentration increase by the effect of the compounds on the expression of the reporter gene.

2.6.2.3 Guanine riboswitch Patent application number US10210210000 A2 considered novel antimicrobial compounds derivatives of non-ribosylable ligand of guanine riboswitch with the mechanism of guaA gene inhibition through binding to a guanine riboswitch. These compounds were used for preventing or treating the infection caused by a pathogen harboring guanine riboswitch-mediated guaA gene regulation. Also, disinfection, sterilization and antisepsis of the same type of bacteria as well as treating MDR infections were evaluated by invented compounds [206]. As an important point, it was shown that S. aureus did not develop resistance toward compound 1,1,6-triaminopyrimidine-2-one (known as PC2). Chapter 1 26

2.6.2.4 T-box riboswitch In patent application number WO1021020202 A2, inventors developed compounds target tRNA-dependent riboswitches of aaRS gene mostly found in Gram-positive bacteria. Compounds were screened computationally first and then tested to measure MIC and MBC on S. aureus and B. subtilis [201]. Compounds PKZ6 and PKZ20 had the most effective antibacterial activity.

2.6.3 Patents of Engineered riboswitches

Chimeric riboswitches including an aptamer domain from one source and expression platform from another source are interesting switching devices to design biosensors. Using such an approach, in patent number US0111000-A, so called two-way engineered riboswitch was designed which is in ON state by default and changed to OFF state with theophylline and go back to ON state with IPTG [200]. An engineered riboswitch can also be used to modify metabolism of an organism. Following this idea, TPP riboswitch of thic gene of Arabidopsis thaliana was engineered in order to be blocked and thiamine level of the plant increased. These modified plants can be used as enriched food supply [200]. In patent number USRE20000 E2 [100], an engineered riboswitch was constructed to regulate pre-mRNA splicing. High affinity theophylline binding aptamer was inserted into the pre-mRNA. In the presence of theophylline, RNA splicing can be modulated. Based on this invention, there are other innovations in this patent such as theophylline-dependent riboswitches which alter RNA splicing, identifying theophylline-dependent riboswitches and procedures of treating abnormal RNA splicing caused diseases.

2.4 Conclusion

More than a decade after first riboswitch discovery, we know more about these RNA regulators. More riboswitch candidates have been assigned with definite molecules just recently and new gates are open now to find out more about these special elements in live world. Structure-based analysis, both computationally and Chapter 1 21

experimentally, have been carried out to understand the mechanism of riboswitches more efficiently. In addition several applications are developed using natural or artificial riboswitches. Considering specificity and selectivity of riboswitches to bind different kinds of metabolites, they are attractive tools to be applied as living biosensors. Besides, artificial riboswitches can be employed to regulate the expression of target genes through binding the molecules [202]. Pioneer efforts in this area make it easier to develop new probes employed to detect important molecules. Riboswitch-based sensors are simple, flexible and reliable enough to exploit a large list of possible applications for them. Therefore, there is a bright future for them as precious tools in biological studies. In the other side, riboswitches are very special drug targets since they have been considered as small molecule sensors in the cell [16]. They seem to be very potential targets for the discovery and development of new antibiotics against resistant bacteria strains. A problem in riboswitch sensing drug discovery is to find out whether in vitro effective compounds could exert their effects in vivo. Another issue is cross-reactivity and toxicity for potential riboswitch targeting drugs. Compounds such as pyrithiamine [102] and AEC [101] which target riboswitches have been revealed to interact with mammalian proteins. However, different recognition mechanisms of riboswitches and proteins as well as absence of known riboswitches in human reduce the probability of cross-reactivity of new riboswitch-targeting compounds. Antibiotic resistance could be another concern because of riboswitch mutations. However, this resistance may be less occurred in riboswitches rather than target proteins. Recent publication of US pharmaceutical Merck Inc. reporting “Ribocil” [212] as a FMN riboswitch targeted antibiotic with strong bactericidal effect and less side effects shed light to the pathway of riboswitch-based antibiotic discovery and it is hopefully achievable to develop new types of antibiotic compounds against resistant infections due to the presence of multiple riboswitch classes in different pathogens. Selective targeting of riboswitches with synthetic compounds is still is in its initial stages. After about 21 years of riboswitches introduction to molecular biology Chapter 1 20

world, very few compounds are known to be active as antibacterial compounds in vivo. Moreover, the interaction possibility between known antibiotics such as aminoglycosides and riboswitches make them a complicated platform for drug design. Enlightening studies on the mechanisms and structures of riboswitches as well as investigation on their probable involvement in antibiotic resistance can accelerate the process of riboswitch-based antibiotic design.

Chapter 1 20

2.11 References

2. Waters L S, Storz G (1000). Regulatory RNAs in bacteria, Cell. 206(2), 621-10.

1. Aiba H (1001). Mechanism of RNA silencing by Hfq-binding small RNAs, Curr Opin Microbiol. 20(1), 202-0.

0. Vogel J (1000). A rough guide to the non-coding RNA world of Salmonella, Mol Microbiol. 12(2), 2-22.

2. Mandal M, Breaker R R (1002). Adenine riboswitches and gene activation by disruption of a transcription terminator, Nat Struct Mol Biol. 22(2), 10- 01.

1. Gelfand M S, Mironov A A, Jomantas J, Kozlov Y I, Perumov D A (2000). A conserved RNA structure element involved in the regulation of bacterial riboflavin synthesis genes, Trends Genet. 21(22), 200-21.

6. Henkin T M (1000). Transcription termination control in bacteria, Curr Opin Microbiol. 0(1), 220-10.

1. Mandal M, Breaker R R (1002). Gene regulation by riboswitches, Nat Rev Mol Cell Biol. 1(6), 212-60.

0. Nudler E, Mironov A S (1002). The riboswitch control of bacterial metabolism, Trends in Biochemical Sciences. 10(2), 22-1.

0. Winkler W C, Breaker R R (1001). Regulation of bacterial gene expression by riboswitches, Annu Rev Microbiol. 10, 201-121.

20. Barrick J E, Breaker R R (1001). The distributions, mechanisms, and structures of metabolite-binding riboswitches, Genome Biol. 0(22),2-20.

22. Sudarsan N, Lee E R, Weinberg Z, Moy R H, Kim J N, Link K H, et al. (1000). Riboswitches in eubacteria sense the second messenger cyclic di- GMP, Science. 012(1001), 222-0.

21. Cromie M J, Shi Y, Latifi T, Groisman E A (1006). An RNA sensor for intracellular Mg(1+), Cell. 211(2), 12-02.

20. Dann C E, 0rd, Wakeman C A, Sieling C L, Baker S C, Irnov I, Winkler W C (1001). Structure and mechanism of a metal-sensing regulatory RNA, Cell. 200(1), 010-01. Chapter 1 10

22. Nelson J W, Sudarsan N, Phillips G E, Stav S, Lunse C E, McCown P J, et al. (1021). Control of bacterial exoelectrogenesis by c-AMP-GMP, Proc Natl Acad Sci U S A. 221(21), 1000-02.

21. Kellenberger C A, Wilson S C, Hickey S F, Gonzalez T L, Su Y, Hallberg Z F, et al. (1021). GEMM-I riboswitches from Geobacter sense the bacterial second messenger cyclic AMP-GMP, Proc Natl Acad Sci U S A. 221(21), 1000-0.

26. Kim P B, Nelson J W, Breaker R R (1021). An ancient riboswitch class in bacteria regulates purine biosynthesis and one-carbon metabolism, Mol Cell. 11(1), 021-10.

21. Dambach M, Sandoval M, Updegrove T B, Anantharaman V, Aravind L, Waters L S, et al. (1021). The ubiquitous yybP-ykoY riboswitch is a manganese-responsive regulatory element, Mol Cell. 11(6), 2000-200.

20. Price I R, Gaballa A, Ding F, Helmann J D, Ke A (1021). Mn(1+)-sensing mechanisms of yybP-ykoY orphan riboswitches, Mol Cell. 11(6), 2220- 10.

20. Furukawa K, Ramesh A, Zhou Z, Weinberg Z, Vallery T, Winkler W C, et al. (1021). Bacterial riboswitches cooperatively bind Ni(1+) or Co(1+) ions and control expression of heavy metal transporters, Mol Cell. 11(6), 2000-00.

10. Baker J L, Sudarsan N, Weinberg Z, Roth A, Stockbridge R B, Breaker R R (1021). Widespread genetic switches and toxicity resistance proteins for fluoride, Science. 001(6061), 100-1.

12. Ren A, Rajashankar K R, Patel D J (1021). Fluoride ion encapsulation by Mg1+ ions and phosphates in a fluoride riboswitch, Nature. 206(1202), 01-0.

11. Collins J A, Irnov I, Baker S, Winkler W C (1001). Mechanism of mRNA destabilization by the glmS ribozyme, Genes Dev. 12(12), 0016-60.

10. Mandal M, Lee M, Barrick J E, Weinberg Z, Emilsson G M, Ruzzo W L, et al. (1002). A glycine-dependent riboswitch that uses cooperative binding to control gene expression, Science. 006(1602), 111-0.

12. Spinelli S V, Pontel L B, Garcia Vescovi E, Soncini F C (1000). Regulation of magnesium homeostasis in Salmonella: Mg(1+) targets the mgtA transcript for degradation by RNase E, FEMS Microbiol Lett. 100(1), 116-02. Chapter 1 12

11. Sudarsan N, Hammond M C, Block K F, Welz R, Barrick J E, Roth A, et al. (1006). Tandem riboswitch architectures exhibit complex gene control functions, Science. 022(1101), 000-2.

16. Blount K F, Breaker R R (1006). Riboswitches as antibacterial drug targets, Nat Biotechnol. 12(21), 2110-62.

11. Kruger K, Grabowski P J, Zaug A J, Sands J, Gottschling D E, Cech T R (2001). Self-splicing RNA: autoexcision and autocyclization of the ribosomal RNA intervening sequence of Tetrahymena, Cell. 02(2), 221- 11.

10. Ellington A D, Szostak J W (2000). In vitro selection of RNA molecules that bind specific ligands, Nature. 026(6101), 020-11.

10. Joyce G F (2002). The rise and fall of the RNA world, New Biol. 0(2), 000-201.

00. Grundy F J, Henkin T M (2000). The S box regulon: a new global transcription termination control system for methionine and cysteine biosynthesis genes in gram-positive bacteria, Mol Microbiol. 00(2), 101- 20.

02. Miranda-Rios J, Navarro M, Soberon M (1002). A conserved RNA structure (thi box) is involved in regulation of thiamin biosynthetic gene expression in bacteria, Proc Natl Acad Sci U S A. 00(21), 0106-22.

01. Nou X, Kadner R J (2000). Coupled changes in translation and transcription during cobalamin-dependent regulation of btuB expression in Escherichia coli, J Bacteriol. 200(12), 6120-10.

00. Stormo G D, Ji Y (1002). Do mRNAs act as direct sensors of small molecules to control their expression?, Proc Natl Acad Sci U S A. 00(21), 0261-1.

02. Nahvi A, Sudarsan N, Ebert M S, Zou X, Brown K L, Breaker R R (1001). Genetic control by a metabolite binding mRNA, Chem Biol. 0(0), 2020.

01. Mironov A S, Gusarov I, Rafikov R, Lopez L E, Shatalin K, Kreneva R A, et al. (1001). Sensing small molecules by nascent RNA: a mechanism to control transcription in bacteria, Cell. 222(1), 121-16.

06. Winkler W, Nahvi A, Breaker R R (1001). Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression, Nature. 220(6020), 011-6. Chapter 1 11

01. Mandal M, Boese B, Barrick J E, Winkler W C, Breaker R R (1000). Riboswitches control fundamental biochemical pathways in Bacillus subtilis and other bacteria, Cell. 220(1), 111-06.

00. Epshtein V, Mironov A S, Nudler E (1000). The riboswitch-mediated control of sulfur metabolism in bacteria, Proc Natl Acad Sci U S A. 200(0), 1011-6.

00. McDaniel B A, Grundy F J, Artsimovitch I, Henkin T M (1000). Transcription termination control of the S box system: direct measurement of S-adenosylmethionine by the leader RNA, Proc Natl Acad Sci U S A. 200(6), 0000-0.

20. Winkler W C, Nahvi A, Sudarsan N, Barrick J E, Breaker R R (1000). An mRNA structure that controls gene expression by binding S- adenosylmethionine, Nat Struct Biol. 20(0), 102-1.

22. Grundy F J, Lehman S C, Henkin T M (1000). The L box regulon: lysine sensing by leader RNAs of bacterial lysine biosynthesis genes, Proc Natl Acad Sci U S A. 200(12), 21011-61.

21. Regulski E E, Moy R H, Weinberg Z, Barrick J E, Yao Z, Ruzzo W L, et al. (1000). A widespread riboswitch candidate that controls bacterial genes involved in molybdenum cofactor and tungsten cofactor metabolism, Mol Microbiol. 60(2), 020-01.

20. Weinberg Z, Barrick J E, Yao Z, Roth A, Kim J N, Gore J, et al. (1001). Identification of 11 candidate structured RNAs in bacteria using the CMfinder comparative genomics pipeline, Nucleic Acids Res. 01(22), 2000-20.

22. Weinberg Z, Wang J X, Bogue J, Yang J, Corbino K, Moy R H, et al. (1020). Comparative genomics reveals 202 candidate structured RNAs from bacteria, archaea, and their metagenomes, Genome Biol. 22(0), 2- 21.

21. Ren A, Patel D J (1022). c-di-AMP binds the ydaO riboswitch in two pseudo-symmetry-related pockets, Nat Chem Biol. 20(0), 100-6.

26. Yarnell W S, Roberts J W (2000). Mechanism of intrinsic transcription termination and antitermination, Science. 102(1222), 622-1.

21. Gusarov I, Nudler E (2000). The mechanism of intrinsic transcription termination, Mol Cell. 0(2), 201-102.

20. Sudarsan N, Barrick J E, Breaker R R (1000). Metabolite-binding RNA domains are present in the genes of eukaryotes, RNA. 0(6), 622-1. Chapter 1 10

20. Wachter A, Tunc-Ozdemir M, Grove B C, Green P J, Shintani D K, Breaker R R (1001). Riboswitch control of gene expression in plants by splicing and alternative 0' end processing of mRNAs, Plant Cell. 20(22), 0201-10.

10. Cheah M T, Wachter A, Sudarsan N, Breaker R R (1001). Control of alternative RNA splicing and gene expression by eukaryotic riboswitches, Nature. 221(1220), 201-100.

12. Croft M T, Moulin M, Webb M E, Smith A G (1001). Thiamine biosynthesis in algae is regulated by riboswitches, Proc Natl Acad Sci U S A. 202(11), 10110-1.

11. Kubodera T, Watanabe M, Yoshiuchi K, Yamashita N, Nishimura A, Nakai S, et al. (1000). Thiamine-regulated gene expression of Aspergillus oryzae thiA requires splicing of the intron containing a riboswitch-like domain in the 1'-UTR, FEBS Lett. 111(0), 126-10.

10. Hollands K, Proshkin S, Sklyarova S, Epshtein V, Mironov A, Nudler E, et al. (1021). Riboswitch control of Rho-dependent transcription termination, Proc Natl Acad Sci U S A. 200(22), 1016-02.

12. Caron M P, Bastet L, Lussier A, Simoneau-Roy M, Masse E, Lafontaine D A (1021). Dual-acting riboswitch control of translation initiation and mRNA decay, Proc Natl Acad Sci U S A. 200(10), E0222-10.

11. Breaker R R (1021). Riboswitches and the RNA world, Cold Spring Harb Perspect Biol. 2(1). 2-21.

16. Rodionov D A, Vitreschak A G, Mironov A A, Gelfand M S (1002). Comparative genomics of the methionine metabolism in Gram-positive bacteria: a variety of regulatory systems, Nucleic Acids Res. 01(22), 0020-10.

11. Andre G, Even S, Putzer H, Burguiere P, Croux C, Danchin A, et al. (1000). S-box and T-box riboswitches and antisense RNA control a sulfur metabolic operon of Clostridium acetobutylicum, Nucleic Acids Res. 06(20), 1011-60.

10. Grundy F J, Henkin T M (1006). From ribosome to riboswitch: control of gene expression in bacteria by RNA structural rearrangements, Critical Reviews in Biochemistry and Molecular Biology. 22(6), 010-00.

10. Winkler W C (1001). Riboswitches and the role of noncoding RNAs in bacterial metabolic control, Curr Opin Chem Biol. 0(6), 102-601.

60. Winkler W C, Breaker R R (1000). Genetic control by metabolite-binding riboswitches, Chembiochem. 2(20), 2012-01. Chapter 1 12

62. Pedrolli D B, Kuhm C, Sevin D C, Vockenhuber M P, Sauer U, Suess B, et al. (1021). A dual control mechanism synchronizes riboflavin and sulphur metabolism in Bacillus subtilis, Proc Natl Acad Sci U S A. 221(21), 22012-0.

61. Furtig B, Nozinovic S, Reining A, Schwalbe H (1021). Multiple conformational states of riboswitches fine-tune gene regulation, Curr Opin Struct Biol. 00, 221-12.

60. Reining A, Nozinovic S, Schlepckow K, Buhr F, Furtig B, Schwalbe H (1020). Three-state mechanism couples ligand and temperature sensing in riboswitches, Nature. 200(1210), 011-0.

62. Winkler W C, Nahvi A, Roth A, Collins J A, Breaker R R (1002). Control of gene expression by a natural metabolite-responsive ribozyme, Nature. 210(6000), 102-6.

61. McCarthy T J, Plog M A, Floy S A, Jansen J A, Soukup J K, Soukup G A (1001). Ligand requirements for glmS ribozyme self-cleavage, Chem Biol. 21(22), 2112-6.

66. Ferre-D'Amare A R (1020). The glmS ribozyme: use of a small molecule coenzyme by a gene-regulatory RNA, Q Rev Biophys. 20(2), 210-21.

61. Urban J H, Vogel J (1000). Two seemingly homologous noncoding RNAs act hierarchically to activate glmS mRNA translation, PLoS Biol. 6(0), 602-621.

60. Regulski E E, Breaker R R (1000). In-line probing analysis of riboswitches, Methods Mol Biol. 220, 10-61.

60. Barrick J E, Corbino K A, Winkler W C, Nahvi A, Mandal M, Collins J, et al. (1002). New RNA motifs suggest an expanded scope for riboswitches in bacterial genetic control, Proc Natl Acad Sci U S A. 202(21), 6212-6.

10. Corbino K A, Barrick J E, Lim J, Welz R, Tucker B J, Puskarz I, et al. (1001). Evidence for a second class of S-adenosylmethionine riboswitches and other regulatory RNA motifs in alpha-proteobacteria, Genome Biol. 6(0), R10.2-R10620.

12. Blount K F, Megyola C, Plummer M, Osterman D, O'Connell T, Aristoff P, et al. (1021). Novel riboswitch-binding flavin analog that protects mice against Clostridium difficile infection without inhibiting cecal flora, Antimicrob Agents Chemother. 10(0), 1106-26.

11. Kazanov M D, Vitreschak A G, Gelfand M S (1001). Abundance and functional diversity of riboswitches in microbial communities, BMC Genomics. 0, 021-016. Chapter 1 11

10. Havill J T, Bhatiya C, Johnson S M, Sheets J D, Thompson J S (1022). A new approach for detecting riboswitches in DNA sequences, Bioinformatics. 00(12), 0021-0.

12. Chang T H, Huang H D, Wu L C, Yeh C T, Liu B J, Horng J T (1000). Computational identification of riboswitches based on RNA conserved functional sequences and conformations, RNA. 21(1), 2216-00.

11. Abreu-Goodger C, Merino E (1001). RibEx: a web server for locating riboswitches and other conserved bacterial regulatory elements, Nucleic Acids Res. 00, W600-W601.

16. Bengert P, Dandekar T (1002). Riboswitch finder--a tool for identification of riboswitch RNAs, Nucleic Acids Res. 01, W212-W210.

11. Mukherjee S, Sengupta S (1021). Riboswitch Scanner: an efficient pHMM-based web-server to detect riboswitches in genomic sequences, Bioinformatics, 01(1), 116-0.

10. Klein D J, Ferre-D'Amare A R (1006). Structural basis of glmS ribozyme activation by glucosamine-6-phosphate, Science. 020(1102), 2111-6.

10. Cochrane J C, Lipchock S V, Smith K D, Strobel S A (1000). Structural and chemical basis for glucosamine 6-phosphate binding and activation of the glmS ribozyme, Biochemistry. 20(21), 0100-26.

00. Gao A, Serganov A (1022). Structural insights into recognition of c-di- AMP by the ydaO riboswitch, Nat Chem Biol. 20(0), 101-01.

02. Li S, Hwang X Y, Stav S, Breaker R R (1026). The yjdF riboswitch candidate regulates gene expression by binding diverse azaaromatic compounds, RNA. 11(2), 100-22.

01. Wakeman C A, Winkler W C (1000). Analysis of the RNA backbone: structural analysis of riboswitches by in-line probing and selective 1'- hydroxyl acylation and primer extension, Methods Mol Biol. 120, 210-02.

00. Soukup G A, Breaker R R (2000). Relationship between internucleotide linkage geometry and the stability of RNA, RNA. 1(20), 2000-11.

02. Soukup G A, DeRose E C, Koizumi M, Breaker R R (1002). Generating new ligand-binding RNAs by affinity maturation and disintegration of allosteric ribozymes, RNA. 1(2), 112-06.

01. Warner K D, Homan P, Weeks K M, Smith A G, Abell C, Ferre-D'Amare A R (1022). Validating fragment-based drug discovery for biological RNAs: lead fragments bind and remodel the TPP riboswitch specifically, Chem Biol. 12(1), 102-1. Chapter 1 16

06. Poiata E, Meyer M M, Ames T D, Breaker R R (1000). A variant riboswitch aptamer class for S-adenosylmethionine common in marine bacteria, RNA. 21(22), 1026-16.

01. Zhang J, Jones C P, Ferre-D'Amare A R (1022). Global analysis of riboswitches by small-angle X-ray scattering and calorimetry, Biochim Biophys Acta. 2000(20), 2010-2010.

00. Schaffer M F, Choudhary P K, Sigel R K (1022). The AdoCbl-riboswitch interaction investigated by in-line probing and surface plasmon resonance spectroscopy (SPR), Methods Enzymol. 120, 261-00.

00. Mayer G, Raddatz M S, Grunwald J D, Famulok M (1001). RNA ligands that distinguish metabolite-induced conformations in the TPP riboswitch, Angew Chem Int Ed Engl. 26(2), 111-60.

00. Jenkins J L, Krucinska J, McCarty R M, Bandarian V, Wedekind J E (1022). Comparison of a preQ2 riboswitch aptamer in metabolite-bound and free states with implications for gene regulation, J Biol Chem. 106(10), 12616-01.

02. Block K F, Hammond M C, Breaker R R (1020). Evidence for widespread gene control function by the ydaO riboswitch candidate, J Bacteriol. 201(21), 0000-0.

01. Mulhbacher J, Brouillette E, Allard M, Fortier L C, Malouin F, Lafontaine D A (1020). Novel riboswitch ligand analogs as selective inhibitors of guanine-related metabolic pathways, PLoS Pathog. 6(2), 2-22.

00. Dar D, Shamir M, Mellin J R, Koutero M, Stern-Ginossar N, Cossart P, et al. (1026). Term-seq reveals abundant ribo-regulation of antibiotics resistance in bacteria, Science. 011(6101), aad0021-aad00111.

02. Hartig J S (1001). Teaching bacteria new tricks--with RNA switches, Angew Chem Int Ed Engl. 26(22), 1122-0.

01. Ketzer P, Kaufmann J K, Engelhardt S, Bossow S, von Kalle C, Hartig J S, et al. (1022). Artificial riboswitches for gene expression and replication control of DNA and RNA viruses, Proc Natl Acad Sci U S A. 222(1), E112-61.

06. Wang R E, Zhang Y, Cai J, Cai W, Gao T (1022). Aptamer-based fluorescent biosensors, Curr Med Chem. 20(11), 2211-02.

01. Ilkhani H, Ganjali M R, Arvand M, Hejazi M S, Azimi F, Norouzi P (1022). Electrochemical spectroscopic investigations on the interaction of an ytterbium complex with DNA and their analytical applications such as biosensor, Int J Biol Macromol. 20(1), 2221-10. Chapter 1 11

00. Pournaghi-Azar M H, Ahour F, Hejazi M S (1020). Direct detection and discrimination of double-stranded oligonucleotide corresponding to hepatitis C virus genotype 0a using an electrochemical DNA biosensor based on peptide nucleic acid and double-stranded DNA hybridization, Anal Bioanal Chem. 001(0), 0102-1.

00. Pournaghi-Azar M H, Hejazi M S, Alipour E (1006). Developing an electrochemical deoxyribonucleic acid (DNA) biosensor on the basis of human interleukine-1 gene using an electroactive label, Anal Chim Acta. 110(1), 222-10.

200. Wilson D S, Szostak J W (2000). In vitro selection of functional nucleic acids, Annu Rev Biochem. 60, 622-21.

202. Edwards T E, Klein D J, Ferre-D'Amare A R (1001). Riboswitches: small- molecule recognition by gene regulatory RNAs, Curr Opin Struct Biol. 21(0), 110-0.

201. Sefah K, Phillips J A, Xiong X, Meng L, Van Simaeys D, Chen H, et al. (1000). Nucleic acid aptamers for biosensors and bio-analytical applications, Analyst. 202(0), 2161-11.

200. Navani N K, Li Y (1006). Nucleic acid aptamers and enzymes as sensors, Curr Opin Chem Biol. 20(0), 111-02.

202. Nomura Y, Yokobayashi Y (1001). Dual selection of a genetic switch by a single selection marker, Biosystems. 00(2), 221-10.

201. Sharma V, Nomura Y, Yokobayashi Y (1000). Engineering complex riboswitch regulation by dual genetic selection, J Am Chem Soc. 200(20), 26020-1.

206. Nomura Y, Yokobayashi Y (1001). Reengineering a natural riboswitch by dual genetic selection, J Am Chem Soc. 210(21), 20022-1.

201. Lynch S A, Desai S K, Sajja H K, Gallivan J P (1001). A high-throughput screen for synthetic riboswitches reveals mechanistic insights into their function, Chem Biol. 22(1), 210-02.

200. Lynch S A, Gallivan J P (1000). A flow cytometry-based screen for synthetic riboswitches, Nucleic Acids Res. 01(2), 202-01.

200. Fowler C C, Brown E D, Li Y (1000). A FACS-based approach to engineering artificial riboswitches, Chembiochem. 0(21), 2006-22.

220. Topp S, Gallivan J P (1000). Random walks to synthetic riboswitches--a high-throughput selection based on cell motility, Chembiochem. 0(1), 120-0. Chapter 1 10

222. Suess B, Fink B, Berens C, Stentz R, Hillen W (1002). A theophylline responsive riboswitch based on helix slipping controls gene expression in vivo, Nucleic Acids Res. 01(2), 2620-2.

221. Wittmann A, Suess B (1021). Engineered riboswitches: Expanding researchers' toolbox with synthetic RNA regulators, FEBS Lett. 106(21), 1016-00.

220. Berens C, Suess B (1021). Riboswitch engineering - making the all- important second and third steps, Curr Opin Biotechnol. 02, 20-1.

222. Weigand J E, Sanchez M, Gunnesch E B, Zeiher S, Schroeder R, Suess B (1000). Screening for engineered neomycin riboswitches that control translation initiation, RNA. 22(2), 00-01.

221. Davidson M E, Harbaugh S V, Chushak Y G, Stone M O, Kelley- Loughnane N (1020). Development of a 1,2-dinitrotoluene-responsive synthetic riboswitch in E. coli cells, ACS Chem Biol. 0(2), 102-22.

226. Robinson C J, Vincent H A, Wu M C, Lowe P T, Dunstan M S, Leys D, et al. (1022). Modular riboswitch toolsets for synthetic genetic control in diverse bacterial species, J Am Chem Soc. 206(00), 20621-12.

221. Wu M C, Lowe P T, Robinson C J, Vincent H A, Dixon N, Leigh J, et al. (1021). Rational Re-engineering of a Transcriptional Silencing PreQ2 Riboswitch, J Am Chem Soc. 201(10), 0021-12.

220. Klauser B, Atanasov J, Siewert L K, Hartig J S (1021). Ribozyme-based aminoglycoside switches of gene expression engineered by genetic selection in S. cerevisiae, ACS Synth Biol. 2(1), 126-11.

220. Klauser B, Rehm C, Summerer D, Hartig J S (1021). Engineering of ribozyme-based aminoglycoside switches of gene expression by in vivo genetic selection in Saccharomyces cerevisiae, Methods Enzymol. 110, 002-10.

210. Jia X, Zhang J, Sun W, He W, Jiang H, Chen D, et al. (1020). Riboswitch control of aminoglycoside antibiotic resistance, Cell. 211(2-1), 60-02.

212. Roth A, Breaker R R (1020). Integron attI2 sites, not riboswitches, associate with antibiotic resistance genes, Cell. 210(1), 2221-0.

211. He W, Zhang X, Zhang J, Jia X, Sun W, Jiang H, et al. (1020). Riboswitch control of induction of aminoglycoside resistance acetyl and adenyl- transferases, RNA Biol. 20(0), 2166-10. Chapter 1 10

210. Chen D, Murchie A I (1022). An aminoglycoside sensing riboswitch controls the expression of aminoglycoside resistance acetyltransferase and adenyltransferases, Biochim Biophys Acta. 2000(20), 012-0.

212. Emadpour M, Karcher D, Bock R (1021). Boosting riboswitch efficiency by RNA amplification, Nucleic Acids Res. 20(20), e66.

211. Espah Borujeni A, Mishler D M, Wang J, Huso W, Salis H M (1026). Automated physics-based design of synthetic riboswitches from diverse RNA aptamers, Nucleic Acids Res. 22(2), 2-20.

216. Wachsmuth M, Domin G, Lorenz R, Serfling R, Findeiss S, Stadler P F, et al. (1021). Design criteria for synthetic riboswitches acting on transcription, RNA Biol. 21(1), 112-02.

211. Ogawa A (1022). Rational design of artificial riboswitches based on ligand-dependent modulation of internal ribosome entry in wheat germ extract and their applications as label-free biosensors, RNA. 21(0), 210- 00.

210. Bell C L, Yu D, Smolke C D, Geall A J, Beard C W, Mason P W (1021). Control of alphavirus-based gene expression using engineered riboswitches, Virology. 200, 001-22.

210. Strobel B, Klauser B, Hartig J S, Lamla T, Gantner F, Kreuz S (1021). Riboswitch-mediated Attenuation of Transgene Cytotoxicity Increases Adeno-associated Virus Vector Yields in HEK-100 Cells, Mol Ther. 10(20), 2101-02.

200. Lee S W, Oh M K (1021). A synthetic suicide riboswitch for the high- throughput screening of metabolite production in Saccharomyces cerevisiae, Metab Eng. 10, 220-10.

202. You M, Litke J L, Jaffrey S R (1021). Imaging metabolite dynamics in living cells using a Spinach-based riboswitch, Proc Natl Acad Sci U S A. 221(12), E1116-61.

201. Song W, Strack R L, Svensen N, Jaffrey S R (1022). Plug-and-play fluorophores extend the spectral properties of Spinach, J Am Chem Soc. 206(2), 2200-102.

200. Suess B, Weigand J E (1000). Engineered riboswitches: overview, problems and trends, RNA Biol. 1(2), 12-0.

202. Topp S, Gallivan J P (1020). Emerging applications of riboswitches in chemical biology, ACS Chem Biol. 1(2), 200-20. Chapter 1 60

201. Breaker R R, Lim J, Blount K F, Puskarz I, Batey R (1000). Riboswitches, Structure-Based Compound Design with Riboswitches, and Methods and Compositions for Use of and with Riboswitches. US PATENT 10000160110 A2. Availabe from https://www.google.com/patents/US10000160110?dq=us10000160110+ A2&hl=en&sa=X&ved=0ahUKEwjat6LLl21PAhVDDxoKHZHjCo0Q6A EIIDAA, Internet, accessed September 22, 1026.

206. Lau P S, Coombes B K, Li Y (1020). A general approach to the construction of structure-switching reporters from RNA aptamers, Angew Chem Int Ed Engl. 20(20), 1000-21.

201. Fowler C C, Li Y (1022). Construction and application of riboswitch- based sensors that detect metabolites within bacterial cells, Methods Mol Biol. 2200, 211-01.

200. Fowler C C, Brown E D, Li Y (1020). Using a riboswitch sensor to examine coenzyme B21 metabolism and transport in E. coli, Chem Biol. 21(1), 116-61.

200. Fowler C C, Sugiman-Marangos S, Junop M S, Brown E D, Li Y (1020). Exploring intermolecular interactions of a substrate binding protein using a riboswitch-based sensor, Chem Biol. 10(21), 2101-21.

220. Zhu X, Wang X, Zhang C, Gu Q (1021). A riboswitch sensor to determine vitamin B21 in fermented foods, Food Chem. 211, 110-0.

222. Ogawa A, Maeda M (1001). Aptazyme-based riboswitches as label-free and detector-free sensors for cofactors, Bioorg Med Chem Lett. 21(22), 0216-60.

221. Muranaka N, Sharma V, Nomura Y, Yokobayashi Y (1000). Efficient design strategy for whole-cell and cell-free biosensors based on engineered riboswitches, Analytical Letters. 21(2), 200-211.

220. Kellenberger C A, Wilson S C, Sales-Lee J, Hammond M C (1020). RNA- based fluorescent biosensors for live cell imaging of second messengers cyclic di-GMP and cyclic AMP-GMP, J Am Chem Soc. 201(20), 2006-0.

222. Kellenberger C A, Sales-Lee J, Pan Y, Gassaway M M, Herr A E, Hammond M C (1021). A minimalist biosensor: Quantitation of cyclic di- GMP using the conformational change of a riboswitch aptamer, RNA Biol. 21(22), 2200-01.

221. Kellenberger C A, Hammond M C (1021). In vitro analysis of riboswitch- Spinach aptamer fusions as metabolite-sensing fluorescent biosensors, Methods Enzymol. 110, 221-11. Chapter 1 62

226. Xie Q, Zhao F, Liu H, Shan Y, Liu F (1021). A label-free and self- assembled electrochemical biosensor for highly sensitive detection of cyclic diguanylate monophosphate (c-di-GMP) based on RNA riboswitch, Anal Chim Acta. 001, 11-16.

221. Meyer A, Pellaux R, Potot S, Becker K, Hohmann H P, Panke S, et al. (1021). Optimization of a whole-cell biocatalyst by employing genetically encoded product sensors inside nanolitre reactors, Nat Chem. 1(0), 610-0.

220. Stoltenburg R, Reinemann C, Strehlitz B (1001). SELEX--a (r)evolutionary method to generate high-affinity nucleic acid ligands, Biomol Eng. 12(2), 002-200.

220. Seeliger J C, Topp S, Sogi K M, Previti M L, Gallivan J P, Bertozzi C R (1021). A riboswitch-based inducible gene expression system for mycobacteria, PLoS One. 1(2), 2-6.

210. Lewis R (2001). The rise of antibiotic-resistant infections, FDA Consum. 10(1), 22-1.

212. Theuretzbacher U, Toney J H (1006). Nature's clarion call of antibacterial resistance: are we listening?, Curr Opin Investig Drugs. 1(1), 210-66.

211. Barrick J E, Breaker R R (1001). The power of riboswitches, Sci Am. 106(2), 10-1.

210. Serganov A, Huang L, Patel D J (1000). Coenzyme recognition and gene regulation by a flavin mononucleotide riboswitch, Nature. 210(1101), 100-1.

212. Ott E, Stolz J, Lehmann M, Mack M (1000). The RFN riboswitch of Bacillus subtilis is a target for the antibiotic roseoflavin produced by Streptomyces davawensis, RNA Biol. 6(0), 116-00.

211. Lee E R, Blount K F, Breaker R R (1000). Roseoflavin is a natural antibacterial compound that binds to FMN riboswitches and regulates gene expression, RNA Biol. 6(1), 201-02.

216. Sudarsan N, Cohen-Chalamish S, Nakamura S, Emilsson G M, Breaker R R (1001). Thiamine pyrophosphate riboswitches are targets for the antimicrobial compound pyrithiamine, Chem Biol. 21(21), 2011-01.

211. Blount K F, Wang J X, Lim J, Sudarsan N, Breaker R R (1001). Antibacterial lysine analogs that target lysine riboswitches, Nat Chem Biol. 0(2), 22-0. Chapter 1 61

210. Patte J C, Akrim M, Mejean V (2000). The leader sequence of the Escherichia coli lysC gene is involved in the regulation of LysC synthesis, FEMS Microbiol Lett. 260(2), 261-10.

210. Mehdizadeh Aghdam E, Barzegar A, Hejazi M S (1022). Evolutionary origin and conserved structural building blocks of riboswitches and ribosomal RNAs: riboswitches as probable target site for aminoglycosides interaction, Adv Pharm Bull. 2(0), 111-101.

260. Baird N J, Inglese J, Ferre-D'Amare A R (1021). Rapid RNA-ligand interaction analysis through high-information content conformational and stability landscapes, Nat Commun. 6(0000), 2-20.

262. Sutcliffe J A (1001). Improving on nature: antibiotics that target the ribosome, Curr Opin Microbiol. 0(1), 102-21.

261. Penchovsky R, Stoilova C C (1020). Riboswitch-based antibacterial drug discovery using high-throughput screening methods, Expert Opin Drug Discov. 0(2), 61-01.

260. Zhang C T, Zhang R (1000). Gene essentiality analysis based on DEG, a database of essential genes, Methods Mol Biol. 226, 002-200.

262. Zhang R, Ou H Y, Zhang C T (1002). DEG: a database of essential genes, Nucleic Acids Res. 01, D112-D111.

261. Kim J N, Blount K F, Puskarz I, Lim J, Link K H, Breaker R R (1000). Design and antimicrobial action of purine analogues that bind Guanine riboswitches, ACS Chem Biol. 2(22), 021-11.

266. Ster C, Allard M, Boulanger S, Lamontagne Boulet M, Mulhbacher J, Lafontaine D A, et al. (1020). Experimental treatment of Staphylococcus aureus bovine intramammary infection using a guanine riboswitch ligand analog, Journal of Dairy Science. 06(1), 2000-0.

261. Anupam R, Nayek A, Green N J, Grundy F J, Henkin T M, Means J A, et al. (1000). 2,1-Disubstituted oxazolidinones: High affinity molecular effectors of RNA function, Bioorg Med Chem Lett. 20(21), 0122-2.

260. Means J, Katz S, Nayek A, Anupam R, Hines J V, Bergmeier S C (1006). Structure-activity studies of oxazolidinone analogs as RNA-binding agents, Bioorg Med Chem Lett. 26(20), 0600-2.

260. Blount K, Puskarz I, Penchovsky R, Breaker R (1006). Development and application of a high-throughput assay for glmS riboswitch activators, RNA Biol. 0(1), 11-02. Chapter 1 60

210. Mayer G, Famulok M (1006). High-throughput-compatible assay for glmS riboswitch metabolite dependence, Chembiochem. 1(2), 601-2.

212. Cressina E, Chen L, Abell C, Leeper F J, Smith A G (1022). Fragment screening against the thiamine pyrophosphate riboswitch thiM, Chemical Science. 1(2), 211-261.

211. Furukawa K, Gu H, Sudarsan N, Hayakawa Y, Hyodo M, Breaker R R (1021). Identification of ligand analogues that control c-di-GMP riboswitches, ACS Chem Biol. 1(0), 2206-20.

210. Fei X, Holmes T, Diddle J, Hintz L, Delaney D, Stock A, et al. (1022). Phosphatase-inert glucosamine 6-phosphate mimics serve as actuators of the glmS riboswitch, ACS Chem Biol. 0(21), 1011-01.

212. Howe J A, Wang H, Fischmann T O, Balibar C J, Xiao L, Galgoci A M, et al. (1021). Selective small-molecule inhibition of an RNA structural element, Nature. 116(1111), 611-1.

211. Lim J, Grove B C, Roth A, Breaker R R (1006). Characteristics of ligand recognition by a glmS self-cleaving ribozyme, Angew Chem Int Ed Engl. 21(20), 6600-00.

216. Lunse C E, Schmidt M S, Wittmann V, Mayer G (1022). Carba-sugars activate the glmS-riboswitch of Staphylococcus aureus, ACS Chem Biol. 6(1), 611-0.

211. Wang G-N, Lau P S, Li Y, Ye X-S (1021). Synthesis and evaluation of glucosamine-6-phosphate analogues as activators of glmS riboswitch, Tetrahedron. 60(26), 0201-0221.

210. Schwalbe H, Buck J, Fürtig B, Noeske J, Wöhnert J (1001). Structures of RNA switches: insight into molecular recognition and tertiary structure, Angewandte Chemie International Edition. 26(0), 2121-2120.

210. Hickey S F, Hammond M C (1022). Structure-guided design of fluorescent S-adenosylmethionine analogs for a high-throughput screen to target SAM-I riboswitch RNAs, Chem Biol. 12(0), 021-16.

200. Liu J, Zeng C, Zhou S, Means J A, Hines J V (1021). Fluorescence assays for monitoring RNA-ligand interactions and riboswitch-targeted drug discovery screening, Methods Enzymol. 110, 060-00.

202. Yu C H, Olsthoorn R C (1021). Monitoring ribosomal frameshifting as a platform to screen anti-riboswitch drug candidates, Methods Enzymol. 110, 001-00. Chapter 1 62

201. Chen L, Cressina E, Leeper F J, Smith A G, Abell C (1020). A fragment- based approach to identifying ligands for riboswitches, ACS Chem Biol. 1(2), 011-0.

200. Colizzi F, Lamontagne A M, Lafontaine D A, Bussi G (1022). Probing riboswitch binding sites with molecular docking, focused libraries, and in- line probing assays, Methods Mol Biol. 2200, 222-12.

202. Subramaniam S, Mehrotra M, Gupta D (1000). Virtual high throughput screening (vHTS)--a perspective, Bioinformation. 0(2), 22-1.

201. Winkler W C, Cohen-Chalamish S, Breaker R R (1001). An mRNA structure that controls gene expression by binding FMN, Proc Natl Acad Sci U S A. 00(11), 21000-20.

206. Breaker R R, Blount K F, Puskarz I J, Wickiser J K (1000). Methods and Compositions Related to the Modulation of Riboswitches.US PATENT 10000001110 A2. Availabe from https://www.google.com/patents/US10000001110?dq=us+10000001110 +A2&hl=en&sa=X&ved=0ahUKEwij2rvrnY1PAhUCxRQKHRaoDKMQ 6AEIHDAA, Internet, accessed September 22, 1026.

201. Breaker R R, Barrick J, Mandal M (1020). Glycine riboswitches, methods for their use, and compositions for use with glycine riboswitches.US PATENT 10200010021 A2.Available from https://www.google.com/patents/US10200010021?dq=us+10200010021 +A2&hl=en&sa=X&ved=0ahUKEwiL-aCgno1PAhWLORQKHeG- DagQ6AEIHjAA, Internet, accessed September 22, 1026.

200. Batey R, Garst A (1022). Lysine riboswitch and compositions and uses thereof. US PATENT 10220106600 A2. Available from https://www.google.com/patents/US10220106600?dq=us+10220106600 +A2&hl=en&sa=X&ved=0ahUKEwjUvc1tno1PAhWCXRQKHdDKBXk Q6AEIHDAA, Internet, accessed September 22, 1026.

200. Strobel S A, Breaker R R, Smith K E (1021). GEMM riboswitches, structure-based compound design with GEMM riboswitches, and methods and compositions for use of and with gemm riboswitches. US PATENT 10210201002 A2. Available from https://www.google.com/patents/US10210201002?dq=us+10210201002 +A2&hl=en&sa=X&ved=0ahUKEwidgMSvn21PAhUCsxQKHTOaCjgQ 6AEIHDAA, Internet, accessed September 22, 1026.

200. Breaker R R, Baker J, Lee E, Sudarsan N, Weinberg Z (1020). Cyclic di- GMP-II Riboswitches, Motifs, and Compounds, and Methods for Their Use. US PATENT 10200220011 A2. Available from https://www.google.com/patents/US10200220011?dq=us+10200220011 Chapter 1 61

+A2&hl=en&sa=X&ved=0ahUKEwj0- ZrIn21PAhXGORQKHReICOkQ6AEIHDAA , Internet, accessed September 22, 1026.

202. Breaker R, Baker J, Sudarsan N, Weinberg Z, Roth A, Ames T, et al. (1021). Fluoride-responsive riboswitchs, fluoride transporters, and methods of use. US PATENT 10210010000 A2. Available from https://www.google.com/patents/US10210010000?dq=us+10210010000 +A2&hl=en&sa=X&ved=0ahUKEwiLy6jpn21PAhVFzRQKHaVsCM0Q 6AEIHDAA, Internet, accessed September 22, 1026.

201. Breaker R R, Lim J, Strobel S A, Cochrane J C (1020). glms riboswitches, structure-based compound design with glms riboswitches, and methods and compositions for use of and with glms riboswitches. US PATENT 10200012210 A2. Available from https://www.google.com/patents/US10200012210?dq=us+10200012210 +A2&hl=en&sa=X&ved=0ahUKEwig00_0n21PAhXLuBQKHYZuCQIQ 6AEIHDAA, Internet, accessed September 22, 1026.

200. Breaker R R, Barrick J E, Roth A, Winkler W (1020). PreQ2 riboswitches and methods and compositions for use of and with preQ2 riboswitches. US PATENT 10200021111 A2. Available from https://www.google.com/patents/US10200021111?dq=us+10200021111 +A2&hl=en&sa=X&ved=0ahUKEwjh2pCQoI1PAhVCchQKHenPAsIQ6 AEIHDAA, Internet, accessed September 22, 1026.

202. Ke A (1020). Method for identifying SMK box riboswitch modulating compounds. US PATENT 10200120111 A2. Available fromhttps://www.google.com/patents/US10200120111?dq=us+10200120 111+A2&hl=en&sa=X&ved=0ahUKEwikx_acoI1PAhUDVBQKHZi0CG 0Q6AEIHDAA, Internet, accessed September 22, 1026.

201. Blanchard S C, Micura R, Haller A (1021). Methods and reagents for analyzing riboswitches using FRET. US PATENT 10210116626 A2. Available from https://www.google.com/patents/US10210116626?dq=us+10210116626 +A2&hl=en&sa=X&ved=0ahUKEwixptOtoI1PAhUIGhQKHfb2AJUQ6 AEIHDAA, Internet, accessed September 22, 1026.

206. Mulhbacher J, Lafontaine D, MALOUIN F, Allard M, Marsault E (1021). Guanine riboswitch binding compounds and their use as antibiotics. US PATENT 10210210000 A2. Available from https://www.google.com/patents/US10210210000?dq=us+10210210000 +A2&hl=en&sa=X&ved=0ahUKEwjLsIy6oI1PAhWB1RQKHemTACIQ 6AEIHDAA, Internet, accessed September 22, 1026. Chapter 1 66

201. Agris P F (1021). T-box riboswitch-binding anti-bacterial compounds. WO PATENT 1021020202 A2. Available from https://www.google.com/patents/WO1021020202A2?cl=en&dq=wo+102 1020202+A2&hl=en&sa=X&ved=0ahUKEwjoh6rNoI1PAhVEVxQKHfu UCMAQ6AEIHDAA, Internet, accessed September 22, 1026.

200. Huang J D, Jin Y (1020). Two-way, portable riboswitch mediated gene expression control device. US PATENT 10200002000 A2. Available from https://www.google.com/patents/US10200002000?dq=us+10200002000 +A2&hl=en&sa=X&ved=0ahUKEwiZ1fbYoI1PAhWDxxQKHZ26BesQ 6AEIHDAA, Internet, accessed September 22, 1026.

200. Aharoni A, Bocobza S, Shapira M (1020). Thiamine pyrophosphate (TPP) riboswitch mutants producing vitamin B2 enriched food and feed crops. US PATENT 10200200000 A2. Available from https://www.google.com/patents/US10200200000?dq=us+10200200000 +A2&hl=en&sa=X&ved=0ahUKEwj0sOH1oI1PAhWM0RQKHcqqDPU Q6AEIHDAA, Internet, accessed September 22, 1026.

100. Gaur R K, Pillai S G (1022). Artificial riboswitch for controlling pre- mRNA splicing. US PATENT RE20000 E2. Available from https://www.google.com/patents/USRE20000?dq=us+RE20000+E2&hl= en&sa=X&ved=0ahUKEwiN2oGNoY1PAhXCaxQKHTHPBBMQ6AEI HDAA , Internet, accessed September 22, 1026.

102. Koedam J C (2010). The mode of action of pyrithiamine as an inductor of thiamine deficiency, Biochim Biophys Acta. 10(1), 000-22.

101. Di Girolamo M, Di Girolamo A, Cini C, Coccia R, De Marco C (2006). Thialysine utilization for protein synthesis by CHO cells, Physiol Chem Phys Med NMR. 20(0), 210-62. 100. Ban N and Thore S, (1022) Riboswitches. US PATENT 10220200600 A2. Available from https://www.google.com/patents/US10220200600?dq=us+10220200600 +A2&hl=en&sa=X&ved=0ahUKEwi- sbGgoY1PAhUL2xQKHQB1B_0Q6AEIHDAA, Internet, accessed September 22, 1026. 10220200600 A2

102. Batey R and Gilbert S, (1022) Sam-II riboswitch and uses thereof. US PATENT 10210210000 A2. Available from https://www.google.com/patents/US10210210000?dq=us+10210210000 +A2&hl=en&sa=X&ved=0ahUKEwiEnfywoY1PAhVFORQKHSUJBSY Q6AEIHDAA, Internet, accessed September 22, 1026. Chapter 1 61

101. Patel D, Serganov A, and Huang L, (1022) Modulators of RNA riboswitches. US PATENT 10220101120 A2. Available from https://www.google.com/patents/US10220101120?dq=us+10220101120 +A2&hl=en&sa=X&ved=0ahUKEwiomr60oY1PAhVFORQKHSUJBSY Q6AEIHDAA, Internet, accessed September 22, 1026.

106. Bertozzi C R, et al., (1021) Riboswitch based inducible gene expression platform., US PATENT 10210122602 A2. Available from https://www.google.com/patents/US10210122602?dq=us+10210122602 +A2&hl=en&sa=X&ved=0ahUKEwi_2ufLoY1PAhXM6RQKHSXlB6E Q6AEIHDAA, Internet, accessed September 22, 1026.

101. Breaker R R, et al., (1020) Crystalline xpt guanine riboswitch from Bacillus subtilis and structure-based compound identification employing said riboswitch. US PATENT 10000160110. Available from https://www.google.com/patents/US10000160110?dq=us+10000160110 &hl=en&sa=X&ved=0ahUKEwjKhZ0coY1PAhWEyRQKHSP2Az0Q6A EIHDAA, Internet, accessed September 22, 1026.

100. Fitzgerald J J, Alexander R L and Meduri A J, (1020) EMP1: Ethyl- Methyl, di Methyl, tri Methyl Pyruvate Acid Esters: A Tool for Regulating HbA2c and a Riboswitch Activator., US PATENT 10200221011 A2. Available fromhttps://www.google.com/patents/US10200221011?dq=us+10200221 011+A2&hl=en&sa=X&ved=0ahUKEwi1mIfooY1PAhVG0RQKHRmE DYoQ6AEIHDAA, Internet, accessed September 22, 1026.

100. Jung G Y, et al., (1020) Method for screening for high l-tryptophan producing microorganisms using riboswitch. WO PATENT 1020202210 A2. Available from https://www.google.com/patents/WO1020202210A2?cl=en&dq=us+1020 202210+A2&hl=en&sa=X&ved=0ahUKEwjbkuvzoY1PAhXFWxQKHS AeD1IQ6AEIIzAB, Internet, accessed September 22, 1026.

120. Batey R, Edwards A and Reyes F, (1022) S-adenosyl-(l)-homocysteine (SAH) riboswitches and compositions and uses thereof. WO PATENT 1022260216 A0. Available from https://www.google.com/patents/WO1022260216A0?cl=en&dq=wo+102 2260216+A0&hl=en&sa=X&ved=0ahUKEwiZw- 1Hoo1PAhVEkRQKHZKSCvEQ6AEIHDAA, Internet, accessed September 22, 1026.

Chapter 0: Evolutionary Origin and Conserved Structural Building Blocks of Riboswitches and Ribosomal RNAs: Riboswitches as Probable Target site for Aminoglycosides Interaction

Chapter 0 60

3.1 Abstract

Riboswitches, as noncoding RNA sequences, are usually located in 1'UTR region of mRNA, which control gene expression through direct ligand binding. Sporadic reports on structural relation of riboswitches with ribosomal RNAs (rRNA), raises an interest in possible similarity between riboswitches and rRNAs evolutionary origins. Since aminoglycoside antibiotics affect microbial cells through binding to functional sites of the bacterial rRNA, finding any conformational and functional relationship between riboswitches and rRNA is utmost important in both of medicinal and basic research. Analysis of the riboswitches structures using bioinformatics and computational tools revealed that there were highly structural conserved relationship of riboswitches and rRNAs but no particular sequence based similarity between them. The building blocks including "hairpin loop containing UUU", "peptidyl transferase center conserved hairpin A loop"," helix 21" and "S1 (G0) hairpin" were detected as high identical rRNA motifs associated with all kinds of riboswitches. The possible similar functional properties of riboswitches with rRNAs were evaluated based on the affinity of paromomycin antibiotic (targeting “A site” of 26S rRNA) to riboswitches via docking methods. Surprisingly, binding energies of paromomycin with different riboswitches are more considerably favourable than the binding energy of paromomycin with “26S rRNA A site”. Therefore the high affinity of paromomycin to bind riboswitches in comparison with rRNA “A site” suggests new insight about riboswitches as possible targets for aminoglycoside antibiotics. In conclusion, there are definite resemblances between riboswitches and rRNA in terms of 0D structures/functions. These findings are considered as a possible supporting evidence for evolutionary origin of riboswitches/rRNAs and also their role in the exertion of antibiotics effects to design new drugs based on the concomitant effects via rRNA/riboswitches.

Chapter 0 10

3.2 Introduction

Today, it is evident that RNAs are not just intermediates between DNA and proteins. Their catalytic and regulating characteristics have been more verified since more than a decade ago. It has been revealed that there are RNA-based mechanisms which regulate gene expression in response to internal or external signals [2-0]. Accordingly, RNA structures play an essential role in this process and determines the fate of the mRNA [2-1]. As ribosome binds mRNA before transcription is completed, most regulatory regions are located within the 1' untranslated region (UTR) of mRNAs. These regulatory regions contain either cis acting binding sites or trans-acting regulators (non-coding RNAs). Riboswitches, usually found within the 1'UTR of mRNAs, are cis acting RNA elements. They can adopt various conformations in response to environmental signals, including stalled ribosomes, uncharged tRNAs, elevated temperatures or small molecule ligands [reviewed in 0]. These metabolite sensors, which identified a decade ago [0], regulate the genes involved in the uptake and use of related metabolites without proteins interpretation [2, 0]. An ever-increasing number and variety of riboswitches are being identified in bacteria, as well as some eukaryotes. For example, as much as 15 of all Bacillus subtilis genes are regulated by riboswitches that bind to metabolites such as flavin mononucleotide (FMN), thiamin pyrophosphate (TPP), S-adenosylmethionine (SAM), lysine, and purines. Riboswitches generally consist of two parts: the aptamer region, a conserved sequence which binds the ligand, and the so-called expression platform, which regulates gene expression through alternative RNA structures that affect transcription or translation [20, 22]. Upon binding of the ligand, the riboswitch changes conformation which forms or disrupts transcriptional terminators or antiterminators, respectively. Therefore, in order to find out their mechanistic details, 1D and 0D structure of riboswitches’ aptamers [21] and their binding characteristics [20] were extensively analyzed experimentally or computationally [22-26]. On the other hand, other possible Chapter 0 12

interactions are suggested to introduce some molecules as new drugs which exert their effects via riboswitches [21, 20]. RNA structure is basically expressed at the sequence or primary structure level, the secondary and tertiary levels. Initially, RNA motifs were identified at the sequence level as generally existing short sequences in functional RNAs, such as transfer RNA (tRNA) or ribosomal RNA (rRNA) [01]. Base-pairing or secondary structure constitutes both the canonically base-paired regions (helices) and non- paired regions (loops). Structural studies and comparative sequence analyses have suggested that biological RNAs are composed primarily of conserved structural building blocks or motifs [61] of secondary and tertiary structures. Forms and functions of RNAs in biological systems which connected to their three- dimensional (3D) structures lead RNA molecules to perform specific roles. However, there are some similarities between various motifs in RNAs types with diverse functionalities. Barrick and Breaker in 6117 detected some motifs in riboswitches, which are close in relation to rRNA structures [60]. In 6112, an artificial riboswitch for the aminoglycoside antibiotic, neomycin B, was engineered [11] which partially resembles the ribosomal A-site, the natural target for aminoglycoside antibiotics [10]. Based on the previous studies we aimed to investigate relation between rRNAs and riboswitches structures and their subsequent functions such as binding to specific antibiotics. In this context, in current study we attempted to survey structural similarity including primary, 1D, 0D and motifs as well as functional similarity (interaction with paromomycin) among rRNAs and riboswitch elements through bioinformatics and computational tools. Chapter 0 11

3.3 Materials and Methods

3.3.1 Databases and Programs

Information regarding sequences of riboswitches data were collected from Rfam database [12] and 0D structure information was obtained through PDB database (www.pdb.org). Multiple sequence alignment was carried out via ClustalW implemented in Mega5 program. Needle (http://www.ebi.ac.uk/Tools/psa/emboss_needle/nucleotide.html) and Water (http://www.ebi.ac.uk/Tools/psa/emboss_water/nucleotide.html) servers were applied in order to accomplish optimal pairwise global and local alignment, respectively [11]. Also, functional and structural alignments were done by means of SARA server (http://structure.biofold.org/sara/) [16] and R0D Align (http://rna.bgsu.edu/R0DAlign) servers [11], respectively. Eventually, the interactions of an antibiotic with RNA structures were investigated using Autodock version 261 program [10].

3.3.2 Functional and structural riboswitch alignments

Ten riboswitch classes which have not only the most representatives in microorganisms [10], but also have available PDB structures, were selected. Their PDB codes which represent preferably unbound state of riboswitches were extracted first from Rfam (http://rfam.sanger.ac.uk/) and then PDB database (http://www.rcsb.org/pdb/home/home.do). They included TPP (PDB code: 1GDI), FMN (PDB code: 1YIE), SAM-I (PDB code: 0IQN), lysine (PDB code: 0D0X), glycine (PDB code: 0OX0), purine (PDB code: 2FE1), c-di-GMP-I (PDB code: 0IWN), c-di-GMP-II (PDB code: 0Q0Z), preQ2(PDB code: 0FU1), THF (PDB code: 0SUY) riboswitches. Each of them was analyzed via SARA server (http://structure.biofold.org/sara/) to perform structure based function alignment. Default parameters are -1600 for opening gap, -0660 for extension gap and 0 consecutive vectors (2 atoms) for length of the Unit-Vector used to generate the Chapter 0 10

comparison matrix. The results were sorted out based on the PSS (percentages of secondary structure identity) and then top rRNA structures were selected and their sequences were obtained.

3.3.3 Sequences alignment

All rRNA sequences resulted from function alignment and have high similarity with a definite riboswitch sequence were aligned applying ClustalW [31] implemented in Mega5 [30] with the default parameters for multiple alignment stages as 05 and 2.22 for gap opening and gap extension penalties, respectively. If necessary, minor adjustments were manually made to the alignments. Also, global and local pairwise alignments of same rRNAs with the riboswitches sequences were conducted through Needle and Water server programs. All the parameters were set by default as 01 and 1.5 for gap opening and gap extension penalties, respectively.

3.3.4 Motifs categorization

Each type of riboswitches was categorized based on the criteria of more than 10 percent secondary structure identity to find similar rRNA motifs. Afterward, motifs were sorted out based on the number of similar riboswitches types. Also, average of PSS (percentages of secondary structure identity) level for each type of motifs was determined. In addition, according to the results, types of riboswitches are catagorized based on each kind of motif they could possess. Chapter 0 12

3.3.5 Docking

3.3.5.1 Preparation of the macromolecules All crystal structures of 20 riboswitches (PDB IDs: 1GDI, 1YIE, 0IQN, 0D0X, 0OX0, 2FE1, 0IWN, 0Q0Z, 0FU1, 0SUY) and 26S-rRNA A-site (PDB ID: 2J1T) were used. Water and ligand molecules from the structures were eliminated by the software program ViewerPro Version 160. Also non-polar hydrogens and Gasteiger charges were added during the preparation of the macromolecule input file using the AutoDockTools package.

3.3.5.2 Preparation of the ligand Complex of paromomycin and 26S-rRNA A-site (PDB code: 2J1T) was used to retreive the three-dimensional (0D) structure of paromomycin using ViewerPro Version 1. Gasteiger charges were added to the obtained structure of the ligand and the rotatable bonds were set to 0 using AutoDock Tools.

3.3.5.3 Preparation of the grid files The complex crystal structure of paromomycin antibiotic and 26S-rRNA A-site motif (PDB code: 2J1T) was selected as a control for assessments of the results. In order to find out the similar and conserved building blocks of the riboswitches with the rRNA, the 0D structure of 26S-rRNA A-site and 20 different riboswitches were structurally aligned via R0D Align. The aligned part of each riboswitch with “A site” motif of 26s rRNAs was considered to generate Grid maps. The interaction of the antibiotic ligand (paromomycin) and 26S-rRNA A- site as well as the homologue parts in different riboswitches were analyzed. Grid maps were generated by AutoGrid 261 based on the superimposed area of each riboswitch with 26S-rRNA A-site. The points of the grids were 60×16×00; 02 ×06×01; 11×06×06; 16×60×200; 200×200×200; 00×00×00; 200× 00×00; 16×00×200; 200×12×06; 00×12×00; 11×06×06 for 2j1t, 0iwn, Chapter 0 11

0q0z, 0ox0, 0d0x, 0fu1, 2fe1, 0iqn, 0suy, 1gdi and 1yie, respectively, with a grid spacing of 06011A˚.

3.3.5.4 Preparation of the docking files Molecular docking was carried out using AutoDock Version 261 based on the Lamarckian genetic algorithm [01]. For each complex, 200 independent docking runs were conducted containing a population of 210 randomly positioned individuals. The maximum number of energy-evaluation retries and generations were 1100000 and 11000, respectively. Also, crossover rate of 060 and a mutation rate of 0601 were set up. During docking, macromolecules were set rigid, whereas all the torsional bonds of ligands were set free. The docking results were clustered according to a root-mean-square deviation (RMSD) tolerance of 160 A˚. The structure of 26S-rRNA A-site (PDB code: 2J1T) was taken as control for docking. First, the ligand (paromomycin) was removed from the complex. Then, the ligand-free structure of 26S-rRNA A-site was used for docking of ligand into binding site for 200 independent runs. At last, binding energies obtained from docking of riboswitches with paromomycin were compared to that of calculated for paromomycin-26S-rRNA complex.

3.3.5.5 Statistical Analyses Where needed, results were evaluated by excel (version 1001) and SPSS (version 26). Statistical analyses were performed using one-way analysis of variance (ANOVA). Statistical assessment of difference between mean values was performed by least significance difference (LSD) test at p<0601 using SPSS (26 version) software.

Chapter 0 16

3.4 Results and discussion

3.4.1 Structural riboswitch alignments

Recent interest in non-coding RNA transcripts has culminated in a rapid increase of deposited RNA structures in the PDB database. However, functional classification and characterization of the RNA structure have not completely addressed. There are many bioinformatics tools to investigate 1D and 0D structural alignments of DNA and RNA structures [00-01]. SARA (Structure Alignment of Ribonucleic Acids) web server (http://sgu.bioinfo.cipf.es/services/SARA) [16] is a promising program for aligning RNA 0D structures via PDB files based on unit-vector root mean square (URMS). Herein, PDB codes of a total of 20 riboswitch types were analyzed by SARA server and all rRNA PDB codes were collected. For all PDB codes in each group of data achieved from SARA server, primary, secondary and tertiary similarity percentages were calculated. Also they were sorted by average natural logarithm of PID (Percentage of sequence identity), PSS (percentages of secondary structure identity( and PSI (percentages of tertiary structure identity). For instance, among similar rRNA PDB codes for lysine riboswitch, the highest 1D similarity belongs to “helix 21” (PDB code: 2WTS-chain A) which shares 1066, 200 and 0160 % in primary, secondary and tertiary identity with this type of riboswitch. According to Figure 0.2, there is a correlation between tertiary structure identity (based on PSI) and secondary structure identity (based on PSS) with approximate squared correlation coefficient (R1) of 060. As a result, the observed correlation between the similarity of 1D and 0D structures demonstrate that both 1D and 0D structures can be utilized for similarity studies in current study. Besides, those concerned with RNA-ligand interactions, generally give greater weight to secondary structure similarity, as ligand binding sites typically consist of a single type of secondary structure [06]. As a result and due to interaction analysis in the subsequent stages of this study, secondary structure could be more useful and debatable in our study. Hence, in the following steps, sorting in order of PSS (based on secondary structure identity)

was performed. Chapter 0 11

Figure 3.1 Correlation of secondary structure identity and 3D structure of rRNAs with the mentioned riboswitches. Vertical and horizontal axes are percentage of secondary structure similarity (PSS) and tertiary structure identity (PSI) of rRNAs with associated riboswitches, respectively. There is high correlation between PSS and PSI of all selected rRNAs (Table S2-Table S20) with 20 different types of riboswitches (Average R1 ~ 060). Chapter 0 10

3.4.2 Motifs categorization

The Structural Classification of RNA (SCOR) is a database designed to provide a comprehensive perspective and understanding of RNA motif structures, functions, tertiary interactions and their relationships (http://scor.berkeley.edu/). It is an inclusive, manually created source of RNA It is believed that GNRA motif has high selectivity and specificity toward different kinds of compounds [01-00]; a common characteristics for riboswitches. GNRA tetaraloop is a common motif in some riboswitches including glycine, purine, FMN, THF, c-di-GMP I (see Figure 0.1 and Table 062). The structural similarity of this motif with rRNAs is more than 11 percent. Consequently, the findings proposed that the common motifs in riboswitches structures should have common functional properties in similar rRNAs such as binding of ligand molecules. The ligand binding characteristics of 20 riboswitches and RNAs which share similar motifs with them were considered as well (see section 0.0). Considering these findings, common evolutionary aspects of riboswitches and rRNAs are confirmed. As a result the resemblance of rRNA building blocks and riboswitches domains may be resulted either from connecting evolution or the dependent byproducts of historical events such as local segment duplication and recombination mechanisms that cause elevation of structural complexity of natural functional molecules.

Chapter 0 10

Figure 3.2 Correlation of ribosomal RNA motifs with 11 different types of riboswitches. Vertical axis demonstrates the number of similar riboswitches types which share at least 105 secondary structure identity (PSS>105) with the shown rRNA motifs (average of PSS for each motif is shown above on the relative column). Please note that 2 motifs including "Hairpin loop containing UUU", "Helix 21", "Peptidyl transferase center conserved hairpin A loop", and "S1 (G0) Hairpin" are conserved with all 20 types of riboswitches with PSS values of 00, 10, 10, 005 PSS, respectively.

Chapter 0 00

3.4.3 Sequences alignment

Multiple sequence alignment is a way of arranging the sequences of DNA, RNA, or protein molecules to similar regions that may be a consequence of functional, structural, or evolutionary origin. Multiple alignments are often used in identifying conserved sequences across a group of sequences hypothesized to be evolutionarily related. Herein, multiple sequence alignment of the riboswitches and similar rRNAs with more than 10 percent secondary structure identity (PSS, discussed above) was carried out. Although, all of the sequences were structurally similar to each other (PSS>105) there were not any particular sequence similarity results using different programs (ClustalW1 and M-Coffee, data not shown). In order to do the alignment more precisely, pairwise alignment of each rRNA with the riboswitches was accomplished by Needle and Water servers as global and local alignment tools, respectively (Table S2-Table S1).

Chapter 0 02

Table 3.1 Types of riboswitches which share more than 515 secondary structure identity with ribosomal RNA motifs. The type of riboswitch which include a motif is checked under its name.

Glyci Lysine Purine THF FMN TPP preQ1 SAM c-di-GMP I c-di-GMP II Motif* Riboswitchne 11S Conserved 196 Hairpin         11 rRNA A site  11 rRNA A site Complex Parmomycin    32S rRNA Sarcin/ricin loop      32S Ribosomal RNA Hairpin 23         31S rRNA Sarcin/ricin loop        A site      Central domain complex with protein          CUCAA Pentaloop          GNRA tetraloop         Hairpin loop containing UUU           Helix 31  Helix 53           Helix III       Loop 12         Loop E         Peptidyl transferase center conserved hairpin           A loop Pre ribosomal RNA        Ribosomal Protein L3/3D rRNA Complex      S3 (G1) Hairpin           UGAA tetraloop        11S Conserved 196 Hairpin         Total number 11 11 11 11 11 11 9 11 11 11 * The name of each motif is extracted from SCOR (Structural Classification of RNA, http://scor.berkeley.edu/).

Figure 0.0 illustrates the results achieved for each pairwise global alignment between every rRNAs and associated riboswitches. Each one of 20 riboswitches sequences was aligned with related rRNAs and an average for every kind of riboswitches was represented in Mean ± SEM (Standard Error of the Mean). As it was shown, global identity percentage is considerably lower than 115 identity. More importantly, based on local alignment there is no unique region in rRNA sequences aligned with the similar riboswitches (Table S2-Table S20). As a result, no common assembly of nucleotides was recognized in the alignments by different programs. Despite this fact that aptamer domains of riboswitches are highly conserved sequences [2, 0], no particular conserved element was observed Chapter 0 01

in their similar rRNAs. In agreement with our findings, it is commonly established that functionally important RNA sequences could be less conserved than their structures [20] where maintaining the structure is more important than maintaining the sequence [22]. For instance, a study on human and mouse genome sequences suggested that there are corresponding non-coding RNA sequences regions between human and mouse with common RNA structures which are not alignable in primary sequence [22]. Therefore, in spite of the structurally similarity of riboswitches and rRNAs, no particular conserved primary sequence element observed in their similar rRNAs using pairwise/multiple alignment approach.

Figure 3.3 Average of global pairwise alignment similarity percentage of rRNA sequences with structurally-based similar riboswitches via Needle program [25]. The sequences of all structurally similar rRNAs (PSS>105.) were aligned with related riboswitch and the similarity percentages were represented in Mean ± SEM for all types of riboswitches (it should be noted that the average number of rRNAs for the analysis were more than 00 strings for each type of Chapter 0 00

riboswitches). Global identity percentages in the range of 225 to 165 denote no sequence correlation between riboswitch and structurally conserved rRNAs (For complete data see Table S2).

3.4.4 Functional properties of riboswitches-rRNAs motifs

Ligand binding is one of the main functions of RNAs for structural stabilization, as well as producing signals. RNA ligand binding is important for ribozymes, riboswitches and splicing functions, along with mediating RNA-protein and RNA- RNA intermolecular interactions [21]. The first ribosomal RNA identified as a small molecule target was 26S RNA component of the prokaryotic ribosome [20]. Various antibiotics, such as aminoglycosides, affect bacterial cells through binding to functional sites of the bacterial rRNA which leads to miscoding during the translation process. Since similarity between riboswitches and different rRNA motifs were revealed in previous parts of current study, a question was raised that “could binding characteristics of these structures be similar too?”. Therefore, the binding affinity of riboswitches for paromomycin as a characteristic of functional properties of these structures was evaluated for a common motif in riboswitches- rRNA structures using molecular docking approach via Autodock 261. Paromomycin is a member of aminoglycosides antibiotics family that has high functional affinity for “A site” motif of 26s rRNAs [22-26]. The aligned part of each riboswitch with “A site” motif of 26s rRNAs (2J1T) was considered to evaluate the possibility of functional antibiotic binding affinity. Table 061 provides the binding energy of docked ligand with riboswitches and 26S rRNA structures. According to the docking results, except for c-di-GMP I riboswitch, there is high functional affinity of paromomycin to riboswitches in comparison with 26S rRNA. Statistical analysis showed that apart from c-di-GMP I, p value for all kinds of riboswitches relative to “26S rRNA A site” is 0600. As a result, there is a remarkable significant increase was occurred in binding energy of different riboswitches types with paromomycin. The range of appropriate binding energies for riboswitches is from -20 to -11 kcal/mol whereas -2261 kcal/mol for Chapter 0 02

“26S rRNA A site” (see Table 061). Seven riboswitches including lysine, THF, SAM, c-di-GMP II, purine, glycine and TPP riboswitches have 261-~1 times higher affinity for paromomycin than “A site” motif. Among them, lysine, THF, SAM and c-di-GMP II have more than 105 of secondary structure similarity with “A site” motif (Table 062). However, despite of less similarity of purine and TPP riboswitches with “A site” motif, they have also considerable low binding energy with paromomycin. But only c-di-GMP I riboswitch showed completely different functional behavior of interaction with the desired ligand. It could be possibly due to nucleotide types in defined binding site which cause weak electrostatic interaction with paromomycin. According to the Table 061, maximum and minimum binding energies of each type of riboswitches demonstrate that the most involved intermolecular energy in the interactions is electrostatic energy.

Figure 3.6 shows RMSD (relative to time zero) against binding energy for all kinds of riboswitches and “26S rRNA A site”. Accordingly, the steadiness of all graphs means that most conformations in studied structures have similar behavior to interact with receptors. Consequently, it verified the docking results and similar condition of docking in all of the riboswitches and 26S rRNA A site. Figure 3.5 illustrates the schematic interaction of paromomycin with “A site” and riboswitches types. In the case of paromomycin-“A site” complex, the ligand binds to minor groove of riboswitches too. However, this kind of binding is not observed in c-di-GMP I riboswitch which may correspond the possible reason for high binding energy relative to other complexes. Apart from c-di-GMP I riboswitch, paromomycin bound to all types of riboswitches which may reduce the access of specific ligands of riboswitches. These findings also support a report which introduced an engineered riboswitch for the aminoglycoside antibiotic neomycin B [11]. The resulting neomycin B responsive RNA-element partially resembles the ribosomal A-site, the natural target for aminoglycoside antibiotics [10]. Furthermore, recently Jia et al. discovered an aminoglycoside-binding riboswitch that is related to the induction of aminoglycosides antibiotic resistance [21]. Chapter 0 01

The targeting of RNA with small molecules is the complementary or even basic of targeting of proteins. Through this phenomenon, riboswitches demonstrate regulatory mechanisms in which proteins do not take part. Furthermore, the importance of RNA-binding small molecules such as antibiotics is undeniable. All clinically approved drugs which exert their effect by binding to RNA are totally recognized as rRNA-targeting molecules [06]. Accordingly, it could be discussed that some of resulted motifs in this study could be alternative binding sites for antibiotics and related small molecules in riboswitches. As though, if further studies on other motifs and antibiotics verify these findings, there is another mechanism for antibiotics effects or resistance apart from rRNA binding signaling in bacteria. It means these kinds of small molecules may bind to same motifs in riboswitches to cause the bacteria’s death or growth repression. However, this suggestion needs more computational and experimental confirming findings. Chapter 0 06

Figure 3.4 RMSD vs. binding energy for 11 types of riboswitches and “18S rRNA A site” based on Autodock results. Vertical and horizontal axes represented RMSD (relative to the time zero) and binding energy of each docked conformations, respectively.

Chapter 0 01

Table 3.2 Binding energy of paromomycin interactions with different types of riboswitches and “18S rRNA A site” as receptors. Binding energy of each interaction is divided to van der Waals energy, hydrogen bonding energy, desolvation energy and electrostatic energy.

Max binding energy(kcal/mol) Min binding energy(kcal/mol) Mean binding Total Total Receptors* energy ± SD vdW + Hbond + Electrostatic vdW + Hbond + Electrostatic binding binding (kcal/mol) desolv Energy Energy desolv Energy Energy energy energy

Lysine (3d1x) -22.252 -8.62 -21.41 -25.1 -5.5 -12.4 -21.22 1.12 -21.24 ± THF (3suy) -5.5 -14.44 -22.6 -3.35 -18.44 -12.15 1.42 -14.15 ± SAM (3iqn) -6.32 -15.8 -21.24 -3.88 -18.88 -12.84 1.22 c-di-GMP II -14.33 ± -5.82 -12.48 -21.4 -3.45 -18.46 -12.25 (3q3z) 1.81 -14.12 ± Purine (4fe5) -8.12 -12.25 -21.54 -4.61 -15.56 -12.2 1.82 -16.42 ± Glycine (3ox1) -8.14 -12.46 -21.63 -2.52 -12.18 -18.64 1.41 -16.52 ± TPP (2gdi) -8.32 -18.42 -21.58 -3.1 -15.66 -18.3 1.44 -15.12 ± preQ (3fu2) -5.82 -15.33 -16.32 -3.34 -11.42 -12.12 1.84 -13.12 ± FMN (2yie) -5.18 -12.24 -15.21 -3.86 -4.85 -11.85 1.43 c-di-GMP I 1.53 ± 1.28 -2.14 2.64 -25.1 -3.42 2.46 2.2 (3iwn) -11.25 ± A site (1j2t)** -5.24 -11.43 -13.54 -3.54 -4.23 -11.16 1.24

* Receptors indicated riboswitches with their PDB codes. ** “A site” refers to “26S rRNA A site” set as a control.

Fig. 1 Chapter 0 00

Figure 3.5 Three dimensional representation of docked paromomycin and corresponding RNAs. The ligand Paromomycin was shown stick-line and all of receptors were presented charge-space including (A) “26S rRNA A site” (2J1T), (B) SAM riboswitch (0IQN), (C) Glycine riboswitch (0OX0), (D) Purine riboswitch (2FE1), (E) Lysine riboswitch (0D0X), (F) THF riboswitch (0SUY), (G) c-di-GMP II riboswitch (0Q0Z), (H) TPP riboswitch (1GDI), (I) FMN riboswitch (1YIE), (J) c-di-GMP I riboswitch (0IWN), (K) preQ riboswitch (0FU1). The conformation with lowest binding energy was selected for presentation of each complex. Chapter 0 00

3.5 Conclusion In this study, the relation between ribosomal RNA and riboswitch structures in terms of structural and functional similarity were evaluated. Our findings indicated these two types of RNAs are structurally similar (secondary and tertiary based level) rather than in primary sequences. Accordingly, similar motifs with high average of secondary structure identity for all types of riboswitches as Hairpin loop containing UUU, Peptidyl transferase center conserved hairpin A loop, Helix 21 and S1 (G0) Hairpin were detected. Consequently, investigation on the realtionship between binding sites of aminoglycosides in rRNA and riboswitches structures via docking method revealed that riboswitch structures may bind more tightly than “26S rRNA A site” to paromomycin. Considering other studies suggesting any kind of structural, functional or evolutionary similarity of ribosomal RNAs and riboswitches, these results could verify that these two apparent diverse types of RNA show strong correspondence to each other. Chapter 0 00

3.8 References

2. Montange R K, Batey R T (1000). Riboswitches: emerging themes in RNA structure and function, Annu Rev Biophys. 01, 221-00. 1. Toledo-Arana A, Repoila F, Cossart P (1001). Small noncoding RNAs controlling pathogenesis, Curr Opin Microbiol. 20(1), 201-0. 0. Waters L S, Storz G (1000). Regulatory RNAs in bacteria, Cell. 206(2), 621-10. 2. Serganov A, Patel D J (1000). Amino acid recognition and gene regulation by riboswitches, Biochim Biophys Acta. 2100(0-20), 101-622. 1. Serganov A (1000). The long and the short of riboswitches, Curr Opin Struct Biol. 20(0), 112-0. 6. Geissmann T, Marzi S, Romby P (1000). The role of mRNA structure in translational control in bacteria, RNA Biol. 6(1), 210-60. 1. Cruz J A, Westhof E (1000). The dynamic landscapes of RNA architecture, Cell. 206(2), 602-0. 0. Grundy F J, Henkin T M (1006). From ribosome to riboswitch: control of gene expression in bacteria by RNA structural rearrangements, Crit Rev Biochem Mol Biol. 22(6), 010-00. 0. Nahvi A, Sudarsan N, Ebert M S, Zou X, Brown K L, Breaker R R (1001). Genetic control by a metabolite binding mRNA, Chem Biol. 0(0), 2020- 2020. 20. Nudler E, Mironov A S (1002). The riboswitch control of bacterial metabolism, Trends Biochem Sci. 10(2), 22-1. 22. Mandal M, Breaker R R (1002). Gene regulation by riboswitches, Nat Rev Mol Cell Biol. 1(6), 212-60. 21. Petrone P M, Dewhurst J, Tommasi R, Whitehead L, Pomerantz A K (1022). Atomic-scale characterization of conformational changes in the preQ(2) riboswitch aptamer upon ligand binding, J Mol Graph Model. 00, 210-01. 20. Ling B, Wang Z, Zhang R, Meng X, Liu Y, Zhang C, et al. (1000). Theoretical studies on the interaction of modified pyrimidines and purines with purine riboswitch, J Mol Graph Model. 10(2), 01-21. 22. Vicens Q, Mondragon E, Batey R T (1022). Molecular sensing by the aptamer domain of the FMN riboswitch: a general model for ligand binding by conformational selection, Nucleic Acids Res. 00(20), 0106-00. Chapter 0 02

21. Kelley J M, Hamelberg D (1020). Atomistic basis for the on-off signaling mechanism in SAM-II riboswitch, Nucleic Acids Res. 00(2), 2001-200. 26. Gong Z, Zhao Y, Chen C, Xiao Y (1021). Computational study of unfolding and regulation mechanism of preQ2 riboswitches, PLoS One. 1(0), 2-0. 21. Mulhbacher J, Brouillette E, Allard M, Fortier L C, Malouin F, Lafontaine D A (1020). Novel riboswitch ligand analogs as selective inhibitors of guanine-related metabolic pathways, PLoS Pathog. 6(2), 2-22. 20. Daldrop P, Reyes F E, Robinson D A, Hammond C M, Lilley D M, Batey R T, et al. (1022). Novel ligands for a purine riboswitch discovered by RNA-ligand docking, Chem Biol. 20(0), 012-01. 20. Woese C R, Winker S, Gutell R R (2000). Architecture of ribosomal RNA: constraints on the sequence of "tetra-loops", Proc Natl Acad Sci U S A. 01(12), 0261-12. 10. Leontis N B, Westhof E (1000). Analysis of RNA motifs, Curr Opin Struct Biol. 20(0), 000-0. 12. Barrick J E, Breaker R R (1001). The distributions, mechanisms, and structures of metabolite-binding riboswitches, Genome Biol. 0(22), 2-20. 11. Weigand J E, Sanchez M, Gunnesch E B, Zeiher S, Schroeder R, Suess B (1000). Screening for engineered neomycin riboswitches that control translation initiation, RNA. 22(2), 00-01. 10. Duchardt-Ferner E, Weigand J E, Ohlenschläger O, Schmidtke S R, Suess B, Wöhnert J (1020). Highly modular structure and ligand binding by conformational capture in a minimalistic riboswitch, Angewandte Chemie International Edition. 20(01), 6126-6120. 12. Burge S W, Daub J, Eberhardt R, Tate J, Barquist L, Nawrocki E P, et al. (1020). Rfam 2260: 20 years of RNA families, Nucleic Acids Res. 22, D116-D101. 11. Rice P, Longden I, Bleasby A (1000). EMBOSS: the European Molecular Biology Open Software Suite, Trends Genet. 26(6), 116-1. 16. Capriotti E, Marti-Renom M A (1000). SARA: a server for function annotation of RNA structures, Nucleic Acids Res. 01, W160-W161. 11. Rahrig R R, Leontis N B, Zirbel C L (1020). R0D Align: global pairwise alignment of RNA 0D structures using local superpositions, Bioinformatics. 16(12), 1600-01. 10. Morris G M, Huey R, Lindstrom W, Sanner M F, Belew R K, Goodsell D S, et al. (1000). AutoDock2 and AutoDockTools2: Automated docking with selective receptor flexibility, J Comput Chem. 00(26), 1101-02. Chapter 0 01

10. Breaker R R (1021). Riboswitches and the RNA world, Cold Spring Harb Perspect Biol. 2(1). 2-21. 00. Thompson J D, Higgins D G, Gibson T J (2002). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice, Nucleic Acids Res. 11(11), 2610-00. 02. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (1022). MEGA1: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods, Mol Biol Evol. 10(20), 1102-0. 01. Morris G M, Goodsell D S, Halliday R S, Huey R, Hart W E, Belew R K, et al. (2000). Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function, Journal of computational chemistry. 20(22), 2600-2661. 00. Chang Y F, Huang Y L, Lu C L (1000). SARSA: a web tool for structural alignment of RNA using a structural alphabet, Nucleic Acids Res. 06, W20-W12. 02. Dror O, Nussinov R, Wolfson H (1001). ARTS: alignment of RNA tertiary structures, Bioinformatics. 12(Suppl 1), 21-10. 01. Ferre F, Ponty Y, Lorenz W A, Clote P (1001). DIAL: a web server for the pairwise alignment of two RNA three-dimensional structures using nucleotide, dihedral angle and base-pairing similarities, Nucleic Acids Res. 01, W610-W660. 06. Thomas J R, Hergenrother P J (1000). Targeting RNA with small molecules, Chem Rev. 200(2), 2212-112. 01. Yan Z, Baranger A M (1002). Binding of an aminoacridine derivative to a GAAA RNA tetraloop, Bioorg Med Chem Lett. 22(10), 1000-00. 00. Yan Z, Rao Ramisetty S, Bolton P H, Baranger A M (1001). Selective recognition of RNA helices containing dangling ends by a quinoline derivative, Chembiochem. 0(22), 2610-62. 00. Yan Z, Sikri S, Beveridge D L, Baranger A M (1001). Identification of an aminoacridine derivative that binds to RNA tetraloops, J Med Chem. 10(21), 2006-202. 20. Chursov A, Walter M C, Schmidt T, Mironov A, Shneider A, Frishman D (1021). Sequence-structure relationships in yeast mRNAs, Nucleic Acids Res. 20(0), 016-61. 22. Torarinsson E, Sawera M, Havgaard J H, Fredholm M, Gorodkin J (1006). Thousands of corresponding human and mouse genomic regions Chapter 0 00

unalignable in primary sequence contain common RNA structure, Genome Res. 26(1), 001-0. 21. Hendrix D K, Brenner S E, Holbrook S R (1001). RNA structural motifs: building blocks of a modular biomolecule, Q Rev Biophys. 00(0), 112-20. 20. Moazed D, Noller H F (2001). Interaction of antibiotics with functional sites in 26S ribosomal RNA, Nature. 011(6212), 000-02. 22. Francois B, Szychowski J, Adhikari S S, Pachamuthu K, Swayze E E, Griffey R H, et al. (1002). Antibacterial aminoglycosides with a modified mode of binding to the ribosomal-RNA decoding site, Angew Chem Int Ed Engl. 20(20), 6101-0. 21. Vicens Q, Westhof E (1002). Crystal structure of paromomycin docked into the eubacterial ribosomal decoding A site, Structure. 0(0), 621-10. 26. Fourmy D, Recht M I, Blanchard S C, Puglisi J D (2006). Structure of the A site of Escherichia coli 26S ribosomal RNA complexed with an aminoglycoside antibiotic, Science. 112(1102), 2061-12. 21. Jia X, Zhang J, Sun W, He W, Jiang H, Chen D, et al. (1020). Riboswitch control of aminoglycoside antibiotic resistance, Cell. 211(2-1), 60-02.

Chapter 2: Riboswitches as potential targets for aminoglycosides compared with rRNA molecules: in Silico study

Chapter 5 01

4.1 Abstract

Riboswitches are cis acting riboregulaters in non-coding region of the mRNAs. Their possible contribution in antibiotic targeting especially for FMN, TPP and lysine riboswitches in bacteria has been revealed since a decade ago. Regarding some studies on the possibility of the interaction between aminoglycosides and the artificial riboswitches, in this study we attempted to evaluate the binding potential of different types of aminglycosides including gentamicin, amikacin, kanamaycin, neomycin, tobramaycin, sisomicin and paromomycin with various classes of riboswitches using computational methods. Applying AutoDock vina, it was shown that the binding energy of each kind of riboswitches with different types of aminoglycosides (riboswitches-aminoglycosides) is almost similar or sometimes more than the binding energy of the aminoglycoside with the corresponding binding site of “26S rRNA A site” (26S rRNA A site- aminoglycosides complex) as aminoglycosides’ target site. The affinity between riboswitches and aminoglycosides is almost the same or higher than the affinity of riboswitches toward natural ligands. In this study ampicillin was used as the negative control drug and 1S rRNA was employed as the negative control RNA. Results showed that the binding energies of riboswitches-ampicillin and 1S rRNA-aminoglycosides complexes are usually greater than the energy of riboswitches-aminoglycosides. Accordingly, lysine, glycine and SAM-I riboswitches were recognized as the best RNA targets for all of the aminoglycosides because of their higher binding energy. In the next step, docking results were further validated by rDock program. Furthermore, it was shown that hydrogen binding makes a key role in the interaction between aminoglycosides and riboswitches. Moreover, MD simulation studies on lysine riboswitch- paromomycin complex confirmed the stability of the docked structure in the solvent containing magnesium and chloride ions. In conclusion, these computational findings support the hypothesis of possible role of riboswitches as aminoglycoside targets.

Chapter 5 06

4.2 Introduction

Riboswitches as non-coding sequences usually located in 1'UTR of mRNAs, are involved in gene regulation through binding to various small molecules without protein interpretation. Ligand binding to the conserved structure of riboswitches causes the folding shift of RNA molecule and halting of transcription and/or translation of downstream related genes [reviewed in 2]. Their unique characteristics on specific and selective binding to various molecules make them interesting biological devices. Since early 1000s, when the riboswitches were primarily introduced [1], the possibility of their application as antibiotic targets has been raised dramatically. At the beginning, the riboswitch-targeting mechanism of four well-known antibiotics including roseoflavin [0, 2], pyrithiamine [1], L-aminoethylcysteine (AEC) and DL-2-oxalysine [6] were approved as ligands for FMN, TPP and lysine riboswitches. Afterwards, some efforts have been taken to find drug-like compounds through drug discovery methods in order to achieve some reliable effective antibacterials.

The aminoglycoside antibiotics are important therapeutic compounds in the treatment of severe bacterial infections. They exert their effects via binding to A site of 26S rRNA in 00S ribosomal subunit and cause interference in mRNA translation [1-0]. The presence of the cationic amine groups of the aminoglycosides causes binding to the negatively charged pockets in folded RNA [20]. However, in addition to “26S rRNA A site”, it has been shown that aminoglycoside neomycin could bind to several other RNAs including the trans- activating response (TAR) [22], Rev response element (RRE) RNAs of HIV- 2[21] and catalytic RNAs, such as group I introns, RNase P, and the hammerhead and hepatitis delta virus ribozymes [20-26]. It is established that conformational changes in the RNA molecules can be occurred by drug binding at such sites [21, 20]. The binding capability of the aminoglycosides to the folded RNA structures has been applied to produce RNA aptamers [20]. In addition, designing artificial riboswitches for aminoglycosides have been conducted in the past decade [10- 10]. For instance, a neomycin B binding artificial riboswitch was designed [10]

Chapter 5 01

which has later been studied structurally for the binding of ribostamycin and tobramycin [11]. Moreover, an in vitro interaction evaluation has been shown that some aminoglycosides inhibit the glms riboswitch [12]. According to the literature, there have been some reports about the structural similarity between riboswitches and rRNAs which raises the possibility of functional connection between these two types of RNA molecules. In 1001 Some structurally similar motifs to rRNA structures have been detected in riboswitches [65]. In 6112, an artificial riboswitch for neomycin B (an aminoglycoside antibiotic), was engineered [10] which partially resembles the ribosomal A site, the natural target for aminoglycoside antibiotics [16]. A comprehensive study on this similarity was carried out by our group [11].

Molecular docking is an important tool in structural biology and computational drug design. It is commonly used to calculate the binding modes of ligands and drug candidates to their protein-nucleic acid targets to predict the affinity and activity of the small molecule drugs [10]. Consequently, there is a wide range of uses and applications for molecular docking, including drug discovery and affinity prediction [10]. The most cited docking tool, namely AutoDock [00] was used to predict the binding orientation of aminoglycosides to 0 types of riboswitches. The last version of AutoDock namely AutoDock Vina significantly enhances the average accuracy and speed of the binding mode predictions compared to AutoDock 2 for molecular docking [02].

Regarding the structural similarity between rRNAs and riboswitches, the possibility of binding of paromomycin, as a representative of aminoglycosides, to different types of riboswitches were predicted [11]. In this study we attempted to evaluate and validate the binding potential of seven aminoglycosides including paromomycin, gentamicin, amikacin, kanamaycin, neomycin, tobramaycin, sisomicin against nine types of riboswitches through computational methods.

Chapter 5 00

4.3 Methods

4.3.1 Molecular docking

4.3.1.1 Preparation of Macromolecules for AutoDock Vina Nine riboswitch classes which have not only the most representatives in microorganisms [01], but also have available PDB structures, were selected. Their PDB codes which represent preferably unbound state of riboswitches were extracted first from Rfam (http://rfam.sanger.ac.uk/) and then PDB database (http://www.rcsb.org/pdb/home/home.do). These riboswitch classes included TPP (PDB code: 1GDI), FMN (PDB code: 1YIE), SAM-I (PDB code: 0IQN), lysine (PDB code: 0D0X), glycine (PDB code: 0OX0), purine (PDB code: 2FE1), c-di- GMP-II (PDB code: 0Q0Z), preQ2 (PDB code: 0FU1), THF (PDB code: 0SUY) riboswitches. In addition, the PDB structures of “26S rRNA A site” were obtained from 2J1T, 1ET0, 1G1Q, 1ESI, 1ETS, 2LC2, 2F0U codes for paromomycin, gentamicin, amikacin, kanamaycin, neomycin, tobramaycin and sisomicin, respectively. 1S Ribosomal RNA structure was extracted from 2C1X PDB code and was used as a negative RNA control. Water and ligand molecules were removed using ViewerPro program Version 160. In addition, non-polar hydrogens and Gasteiger charges were added during the preparation of the macromolecule input file using the AutoDockTools package 26166 [00].

4.3.1.2 Preparation of the ligand for AutoDock Vina The structures of seven aminoglycosides including paromomycin, gentamicin, amikacin, kanamaycin, neomycin, tobramaycin and sisomicin were obtained from corresponding crystal structures using ViewerPro program. In addition, the specific ligand of each riboswitch was retrieved from the complex structure of the riboswitch from PDB codes of 0d0u, 0owi, 1yie, 0iqr, 0suh, 0fu1, 2fe1, 0q0z, 1gdi for lysine, glycine, FMN, SAM, THF, preQ, purine, c-di-GMP-II and TPP molecules, respectively. The structure of ampicillin as a negative control antibiotic was obtained from 0D structure complex with ompF porin (i.e. PDB code: 2KR2). All rotatable bonds within the ligands were allowed to rotate freely

Chapter 5 00

and Gasteiger charges were added to the obtained structure of the ligand using the AutoDockTools package 26166.

4.3.1.3 Docking procedure All dockings were performed using AutoDock Vina 26261 [02] which is a new generation of docking software from the Molecular Graphics Lab (http://vina.scripps.edu). The grid box of each riboswitch was set according to the similar part of the riboswitches with “26S rRNA A site” based on our previous study [11]. In addition, the grid boxes for various A sites were obtained based on the binding site of the aminoglycosides. The number of 10 modes was set for each docking run. Other parameters were kept to their default values. First, docking was carried out for every aminoglycoside and riboswitches. The pdbqt file of each docking procedure was created and the best binding energy was acquired. The binding energies of the aminoglycosides with related “26S rRNA A sites” and also the binding energies of the riboswitches with their own specific ligands were considered as positive controls. The interaction between riboswitches and ampicillin as well as the interaction of 1S rRNA with aminoglycosides were considered as negative controls. The interaction features of selected conformations were analyzed using AutoDockTools package 26166.

4.3.2 Docking validation rDock program is a fast and versatile docking tool for docking small molecules against for nucleic acids [00]. Docked conformations of paromomycin and gentamicin (as sample aminoglycosides) with highest binding energy in AutoDock Vina were selected to be validated and re-scored via rDock.

4.3.2.1 Docking preparation for rDock At first the system definition parameter file was developed to define the receptor file, ligand and scoring functions. Cavity mapping was carried out based on "Reference ligand method" and the radius of cavity mapping region, radius of small probe, minimum cavity volume to accept, `maximum number of cavities to accept, receptor atom radius increment for excluded volume and grid resolution

Chapter 5 200

for mapping, were considered 260 Å, 260 Å, 200 Å0, 20, 060 Å and 061 Å, respectively. Receptor and ligand files were prepared in Mol1 and SD formats, respectively, using Open Babel 16061 program. Finally cavity mapping was performed using rbcavity module (one of the executable programs of rDock). Besides, where needed rbrmsd was used to calculate the RMSD between pose predictions and the PDB structure.

4.3.2.2 Docking processing for rDock Having done the cavity mapping, docking process was performed using rbdock with definition of input files including ligand, system definition parameter file, docking protocol file and output SD format file. The file dock_solv.prm was considered as the docking protocol file based on SF1 scoring function which is compatible with nucleic acids according to the reference guide [00].

4.3.2.3 Post-processing and analysis of results In order to sort out the docking conformations based on the total score, sdsort program was utilized. The output SD file of rbdock was considered as input file for sdsort. Afterwards, the summary of scores (total, inter, intra and vdw) was obtained through sdreport program. According to the program guide, total score is a weighed sum of intermolecular, ligand intramolecular, site intramolecular and external restraint terms. Inter score is the most important term as it represents the receptor-ligand interaction score. Intra score shows the energy difference between the ligand and the input ligand conformation.

4.3.3 Molecular dynamics simulation

In order to confirm the data from docking results we performed additional computational molecular dynamics (MD) simulation approach via GROMACS program [02]. To this end, the complex (lysine riboswitch-paromomycin) was placed in a cubic box center of 216262 nm, 216262 nm, 216262 nm with periodic boundary condition and solvated by TIP0P water molecules [01]. Mg1+ and Cl- counterions were added to maintain overall system electroneutrality. The simulation was performed with GROMACS 160 suite program using

Chapter 5 202

CHARMM11 force field [06]. The Berendsen temperature coupling was used to keep the system at 000 K, and the constant of coupling was 062 ps. The particle mesh Ewald (PME) algorithm was applied to calculate long-range electrostatics interactions with a cutoff of 060 nm and a cutoff of 262 nm was set for van der Waals interactions. The Verlet leapfrog integrator with an integration time step of 1 fs was used and LINCS algorithm was employed to keep all bonds involving hydrogen atoms rigid.

The system was subjected to 100000 steps of steep descent energy minimization. Then, position restrained molecular dynamics was subsequently carried out for 100 ps. Finally, MD simulation was run for 1000 ps to the whole system.

4.4 Results

4.4.1 Docking of riboswitches with various aminoglycosides

AutoDock Vina [01] is an open-source program for molecular docking. In comparison to AutoDock 2, AutoDock Vina significantly enhances the average accuracy of the binding predictions. Docking was performed on nine types of riboswitches (lysine, glycine, purine, FMN, SAM, TPP, PreQ, c-di-GMP-II, THF) as receptors interacting with 1 types of aminoglycosides (paromomycin, gentamicin, amikacin, kanamaycin, neomycin, tobramaycin, sisomicin) as ligands using AutoDock Vina. The rRNA molecule was used as a control for all of aminoglycosides because of having “26S rRNA A site” as a known binding site for aminoglycosides. Therefore, the docked results for the complexes of riboswitches-aminoglycosides were compared with the binding energy of docked antibiotics with “26S rRNA A site”. Moreover, the corresponding natural metabolites of riboswitches were docked and used as a positive control (named own ligand) for investigation of interactions with riboswitches (see Figures 2.2- 2.1). The RMSD calculated by rDock between predicted poses of natural ligands and their PDB structure were almost 2~1 A°. The interactions of some complexes including the 1S rRNA-aminoglycosides and riboswitches-ampicillin complexes

Chapter 5 201

were considered as negative controls. The binding energy of such complexes was almost 0-2 kcal/mol (Figure 2.2-2.1).

Figure 4.1 Docking of paromomycin (A) the chemical structure of paromomycin. (B) Dark black columns demonstrate the binding energies of paromomycin with different riboswitches. Dark grey columns show the binding energy of each riboswitch with its own natural ligands. Light grey columns show the binding energy of each riboswitch with ampicillin (C) The interaction between the riboswitches and paromomycin. Hydrogen bindings were shown as green dot lines.

As seen in Figure 2.2, the binding energy of all of the riboswitches-paromomycin is approximately close to the binding energy of the interaction between rRNA A site-paromomycin. It was revealed that the binding energy of most of riboswitches-paromomycin complexes is more (negative) than the binding energy

Chapter 5 200

of riboswitches-natural ligands complexes. In addition, the affinity of aminoglycosides-riboswitches is almost 1-fold of the affinity of riboswitches-- ampicillin and 1S rRNA-paromomycin (Figure 2.2-2.1).

Figure 4.2 Docking of gentamicin (A) the chemical structure of gentamicin. (B) Dark black columns demonstrate the binding energies of gentamicin with different riboswitches, dark grey columns show the binding energy of each riboswitch with its own natural ligands and light grey columns show the binding energy of each riboswitch with ampicillin based on AutoDock Vina results. (C) The interaction between the riboswitches and gentamicin. Hydrogen bindings were shown as green dot lines.

Chapter 5 202

Figure 4.3 Docking of neomycin (A) the chemical structure of neomycin. (B) Dark black columns demonstrate the binding energies of neomycin with different riboswitches, dark grey columns show the binding energy of each riboswitch with its own natural ligands and light grey columns show the binding energy of each riboswitch with ampicillin based on AutoDock Vina results. (C) The interaction between the riboswitches and neomycin. Hydrogen bindings were shown as green dot lines.

Besides, the binding energies of the riboswitches-aminoglycosides are comparable to those of riboswitches-natural ligands. Even the binding energies of the glycine, lysine and purine riboswitches-aminoglycosides are higher than those of riboswitches-natural ligands. In addition, according to Figure 2.2B, lysine, glycine and SAM-I riboswitches have the best binding energies to paromomycin in comparison to “26S rRNA A site”. Figure 2.2C illustrates the 0D structure of the interactions between paromomycin and lysine, glycine and SAM-I riboswitches

Chapter 5 201

and also “26S rRNA A site”. The green dot lines illustrate the hydrogen binding between the ligand and the RNA molecule. It was shown that hydrogen bindings exist mostly between paromomycin and guanine and adenine nucleotides.

Figure 4.4 Docking of kanamycin (A) the chemical structure of kanamycin. (B) Dark black columns demonstrate the binding energies of kanamycin with different riboswitches, dark grey columns show the binding energy of each riboswitch with its own natural ligands and light grey columns show the binding energy of each riboswitch with ampicillin based on AutoDock Vina results. (C) The interaction between the riboswitches and kanamycin. Hydrogen bindings were shown as green dot lines.

Chapter 5 206

Figure 4.5 Docking of amikacin (A) the chemical structure of amikacin. (B) Dark black columns demonstrate the binding energies of amikacin with different riboswitches, dark grey columns show the binding energy of each riboswitch with its own natural ligands and light grey columns show the binding energy of each riboswitch with ampicillin based on AutoDock Vina results. (C) The interaction between the riboswitches and amikacin. Hydrogen bindings were shown as green dot lines.

Chapter 5 201

Figure 4.8 Docking of sisomicin (A) the chemical structure of sisomicin. (B) Dark black columns demonstrate the binding energies of sisomicin with different riboswitches, dark grey columns show the binding energy of each riboswitch with its own natural ligands and light grey columns show the binding energy of each riboswitch with ampicillin based on AutoDock Vina results. (C) The interaction between the riboswitches and sisomicin. Hydrogen bindings were shown as green dot lines.

Chapter 5 200

Figure 4.2 Docking of tobramycin (A) the chemical structure of tobramycin. (B) Dark black columns demonstrate the binding energies of tobramycin with different riboswitches, dark grey columns show the binding energy of each riboswitch with its own natural ligands and light grey columns show the binding energy of each riboswitch with ampicillin based on AutoDock Vina results. (C) The interaction between the riboswitches and tobramycin. Hydrogen bindings were shown as green dot lines.

Approximately the same pattern was observed for other types of aminoglycosides (Figures 2.1-2.1). As seen, of 0 classes of riboswitches, lysine, glycine and SAM- I riboswitches showed higher binding energy to interact with different aminoglycosides. For some riboswitches, the natural metabolites binding energies are less than aminoglycosides’ binding energies and vice versa for others. For instance, c-di-GMP-II riboswitch showed higher affinity to its own metabolite

Chapter 5 200

ligand in comparison to the affinity of this riboswitch to tested aminoglycosides (Figure 262B-Figure 261B). In addition, binding energies of SAM-I riboswitch with kanamycin (Figure 2.2B), amikacin (Figure 2.1B), sisomicin (Figure 2.6B) and tobramycin (Figure 261B) were higher than the binding energy of SAM-I riboswitch-SAM complex. Regarding interaction between antibiotics with lysine, glycine, SAM-I riboswitches and “26s rRNA A site”, most hydrogen bindings make interaction between the ligands and guanine/adenine nucleotides.

4.4.2 Docking validation rDock program which can be used for docking against nucleic acids very efficiently, was applied to re-score docked conformations. The conformations with higher affinity were conducted as ligands to be docked once again with rDock program. The main jobs were performed by the programs rbcavity (cavity generation) and rbdock (docking) [00].

As shown in Figure 2.0A, according to total scores, all of the studied riboswitches showed considerable affinity to paromomycin in the range of -26600 – -11626 for c-d-GMP-II and lysine riboswitches, respectively. Then, the best total score belongs to lysine and THF riboswitches with the total score of -11626 and -20600. Regarding total score of -26620 for “A site rRNA”, it can be inferred that all studied riboswitches showed more affinity to paromomycin than its known target. Also, according to Figure 2.0B, gentamicin showed even higher affinity to different riboswitches in the range of -20600 – -10601 for lysine and SAM-I riboswitches, respectively. Then, lysine riboswitch may be also the best target for gentamicin in comparison to other riboswitches. Moreover, the difference between the total scores of riboswitches and “A site rRNA” (-11600) is more than the case of paromomycin. In addition, the intermolecular scores are higher than intramolecular scores in paromomycin. However, in the case of gentamicin intermolecular scores are lower which reflects the some differences in the binding mode of interaction. On the other hand, van der waals forces are approximately equal between two aminoglycosides-riboswitches bindings.

Chapter 5 220

Figure 4.6 rDock results of paromomycin and gentamicin docking (A) Total, intra, inter and van der waals scores of docking paromomycin against different riboswitches acquired by rDock program.(B) Total, intra, inter and van der waals scores of docking gentamicin against different riboswitches acquired by rDock program.

4.4.3 Molecular dynamics simulation

GROMACS 160 suite program was used to confirm the docking results and evaluate the interaction in an environment filled with water, Mg1+ and Cl- ions. The PDB structure of docked lysine riboswitch and paromomycin was applied as the starting structure. In addition, separate ligand and riboswitch structure (with approximate distance of 6 Å) was simulated as negative control.

Figure 2.0A demonstrates the Root-mean-square deviation (RMSD) of the whole system relative to the initial system during the simulation time. As shown, the system reaches equilibrium after 2100 ps and RMSD value of 0 nm.

Chapter 5 222

Figure 2.0B and Figure 260C show the distances fluctuation between C2 atom of paromomycin and P atom of A200 in lysine riboswitch. The distances fluctuation is very high in negative control (Figure 2.0B), whereas the distances variation is very low in docked form of complex (Figure 2.0C). In other words, the interaction between the aminoglycoside and the riboswitch is strong enough to keep the ligand in the complex form during the simulation.

Chapter 5 221

Figure 4.4 Molecular dynamic simulation results of lysine riboswitch- paromomycin interaction (A) Time dependences of RMSDs (nm) of system relative to the initial system in MD simulation. (B) Time courses of length between C2 of paromomycin and P of A200 in lysine riboswitch in negative control and (C) docked structure during MD simulation.

Chapter 5 220

4.5 Discussion

Non-coding RNAs have been considered as important elements for gene regulations in the cell more than a decade ago. A sort of cis acting riboregulators, so-called riboswitches, has attracted so many attentions in this regard since their discovery in 1001 [1]. Their structures consist of two essential parts including aptamer domain and expression platform. Structurally conserved aptamer domain binds to specific ligand and makes alteration in three dimensional structures of expression platform allosterically, followed by down-regulation or up-regulation of the corresponding genes [00]. Several types of riboswitches with specificity to particular ligands have been identified up to now [reviewed in 2]. However, lots of them are looking forwards to be discovered in the near future. The possibility of antibiotic targeting by riboswitches has been studied considerably. Lysine [6] and guanine [00] riboswitches are most studied targets for antibiotic designing. Although, the success to design completely effective antibiotics is rare [20], full efforts have been taken to improve drug discovery in this regard.

Aminoglycosides are important therapeutic agents for battling against severe infections. Their known mechanism is related to halting the translation through binding to 26S rRNA A site structure [1]. In addition, the binding of some aminoglycosides to artificial riboswitches has been investigated in the number of studies [10-10]. On the other hand, a comprehensive study on the similarity of rRNA and different riboswitches structures have been conducted and possible targeting of riboswitches by paromomycin has been suggested based on docking results [11]. Accordingly in this study, the affinity of other aminoglycosides was investigated and verified via AutoDock Vina and rDock programs, respectively. Moreover, a sample MD simulation was conducted to evaluate the interaction in water and ions environment. In this study, first the PDB structures of the representatives of nine kinds of riboswitches were extracted. It should be noted that based on different studies on bound and free-state of the riboswitches, the RMSD of the atoms are not high (approximately lower than 1 A°) [reviewed in 22]. It means that the cell environment itself could fold the riboswitches properly

Chapter 5 222

[21]. Therefore, in spite of the rare PDB codes for free-state riboswitches, docking on their bound structure could be acceptable as well. However, selecting the unbound state was preferable. As shown in Figures 2.2-2.1, the binding energies of different riboswitches with related aminoglycosides are approximately similar or more than the affinity of the 26S rRNA A site with the aminoglycoside. However, the binding energy of the lysine, glycine and SAM-I riboswitches are significantly favorable than those of other riboswitches. In the case of paromomycin, lysine, THF and SAM-I riboswitches were best receptors according to our previous study [11]. This could be due to differences in the scoring function of AutoDock 2 (mostly based on electrostatic forces) and AutoDock Vina (mostly based on hydrophobic and hydrogen binding). However, lysine and SAM-I riboswitches have been considered as better targets according to both methods. In addition, the affinity of riboswitches toward aminoglycosides is more than the binding energy of negative controls including riboswitches- ampicillin and 1S rRNA-aminoglycosides complexes (Figure 2.2-2.1). The exception of c-di-GMP-II riboswitch for all aminoglycosides and SAM-I riboswitch for kanamycin (Figure 2.2B), amikacin (Figure 2.1B), sisomicin (Figure 2.6B) and tobramycin (Figure 2.1B) demonstrates that these riboswitches may fail to bind aminoglycosides in competition with their own natural ligands. Binding characteristics of different aminoglycosides showed that strong hydrogen bindings exist between the aminoglycoside molecules and guanine-adenine nucleotides. This characteristic is similar between the iteraction of aminoglycoside-A site and aminoglycoside-riboswitch interactions. rDock program as a valid and strong newly developed docking method for ribonucleic acids [00] have been utilized to re-score the results of AutoDock Vina. Based on the observations (Figure 2.0) almost all riboswitches- aminoglycoside complexes demonstrate higher total scores in comparison to rRNA-aminoglycosides in the case of paromomycin and gentamicin. Showing high total score, lysine riboswitch was predicted as the most possible target for aminoglycosides in this study which is in accordance with the results of AutoDock Vina and Autodock 2 [11]. However, all riboswitches illustrated better

Chapter 5 221

affinity to aminoglycosides in comparison to rRNAs. According to the results, although the binding modes of paromomycin and gentamicin may be different with each other, vdw forces are almost equal among riboswitches and aminoglycosides.

Taken the whole, it seems that though different riboswitches show affinity to aminoglycosides, some of them are particularly better than others in terms of higher binding affinity such as lysine riboswitch verified by rDock scoring function.

It should be noted that molecular docking has a number of limitations beside its strength. For example, its dependence on the search algorithm and scoring function makes the performance of this approach quite different. Besides, the rigidity of binding site is not quite similar with what happens in reality. In addition, calculation of the binding energy in the vacuum condition without considering the effect of environmental molecules is an important limitation of docking method. Particularly, the presence of cationic ions such as Mg1+ is so important for the behavior of the riboswitches in the cell. As a result, in the next step, the complex of lysine riboswitch-paromomycin was studied through MD simulation. MD simulation enabled us to evaluate the interaction in the mixture of water and ions, especially Mg1+. Regarding the RMSD value the system is equilibrated. The result of the simulation showed that the distance of the ligand and the receptor was not varied considerably during the simulation, while the distance variation was considerably high in the negative control simulation (Figure 2.0). These findings approved the possibility of interaction between lysine riboswitch and paromomycin even in the ionic environment.

However, computational methods cannot replace experimental validation of the results and further in vitro and in vivo studies are needed to confirm the results.

4.8 Conclusion

In this study the binding affinity of 1 aminoglycosides including paromomycin, gentamicin, amikacin, kanamaycin, neomycin, tobramaycin and sisomicin to

Chapter 5 226

different types of riboswitches was evaluated and validated through docking methods and MD simulation studies. In conclusion, according to AutoDock Vina results, almost all kinds of aminoglycosides have considerable affinities to the riboswitches. Among them lysine, glycine and SAM-I riboswitches showed higher binding energies. In addition, the binding energies of riboswitches- aminoglycosides are lower than the binding energies of negative control complexes including riboswitches-ampicillin and 1S rRNA-aminoglycosides. Additionally, most hydrogen bindings are formed by guanine and adenine nucleotides. Docking validation through rDock program confirmed data as it was shown that the total score is higher in riboswitches-paromomycin and riboswitches-gentamicin complexes in comparison with “26S rRNA A site”- aminoglycosides interactions. MD simulation study on lysine riboswitch- paromomycin complex approved the docking results even within the solvent in the presence of magnesium and chloride ions. Refereeing all results, this study has strengthened the idea in support of the potential of the riboswitches’ as drug targets for aminoglycosides. Further computational and experimental studies are suggested in this context.

Chapter 5 221

4.2 References

2. Winkler W C, Breaker R R (1001). Regulation of bacterial gene expression by riboswitches, Annu Rev Microbiol. 10, 201-121. 1. Nahvi A, Sudarsan N, Ebert M S, Zou X, Brown K L, Breaker R R (1001). Genetic control by a metabolite binding mRNA, Chem Biol. 0(0), 2020-2020. 0. Serganov A, Huang L, Patel D J (1000). Coenzyme recognition and gene regulation by a flavin mononucleotide riboswitch, Nature. 210(1101), 100-1. 2. Lee E R, Blount K F, Breaker R R (1000). Roseoflavin is a natural antibacterial compound that binds to FMN riboswitches and regulates gene expression, RNA Biol. 6(1), 201-02. 1. Sudarsan N, Cohen-Chalamish S, Nakamura S, Emilsson G M, Breaker R R (1001). Thiamine pyrophosphate riboswitches are targets for the antimicrobial compound pyrithiamine, Chem Biol. 21(21), 2011-01. 6. Blount K F, Wang J X, Lim J, Sudarsan N, Breaker R R (1001). Antibacterial lysine analogs that target lysine riboswitches, Nat Chem Biol. 0(2), 22-0. 1. Carter A P, Clemons W M, Brodersen D E, Morgan-Warren R J, Wimberly B T, Ramakrishnan V (1000). Functional insights from the structure of the 00S ribosomal subunit and its interactions with antibiotics, Nature. 201(6001), 020-0. 0. Fourmy D, Recht M I, Blanchard S C, Puglisi J D (2006). Structure of the A site of Escherichia coli 26S ribosomal RNA complexed with an aminoglycoside antibiotic, Science. 112(1102), 2061-12. 0. Davies J, Davis B D (2060). Misreading of ribonucleic acid code words induced by aminoglycoside antibiotics. The effect of drug concentration, J Biol Chem. 120(21), 0021-6. 20. Hermann T, Westhof E (2000). Docking of cationic antibiotics to negatively charged pockets in RNA folds, Journal of Medicinal Chemistry. 21(1), 2110-62. 22. Faber C, Sticht H, Schweimer K, Rosch P (1000). Structural rearrangements of HIV-2 Tat-responsive RNA upon binding of neomycin B, J Biol Chem. 111(11), 10660-6. 21. Kirk S R, Luedtke N W, Tor Y (1000). Neomycin-acridine conjugate: a potent inhibitor of Rev-RRE binding, Journal of the American Chemical Society. 211(1), 000-002.

Chapter 5 220

20. Rogers J, Chang A H, von Ahsen U, Schroeder R, Davies J (2006). Inhibition of the self-cleavage reaction of the human hepatitis delta virus ribozyme by antibiotics, J Mol Biol. 110(1), 026-11. 22. Mikkelsen N E, Brannvall M, Virtanen A, Kirsebom L A (2000). Inhibition of RNase P RNA cleavage by aminoglycosides, Proc Natl Acad Sci U S A. 06(22), 6211-60. 21. Stage T K, Hertel K J, Uhlenbeck O C (2001). Inhibition of the hammerhead ribozyme by neomycin, RNA. 2(2), 01-202. 26. von Ahsen U, Davies J, Schroeder R (2002). Antibiotic inhibition of group I ribozyme function, Nature. 010(6021), 060-10. 21. Murchie A I, Davis B, Isel C, Afshar M, Drysdale M J, Bower J, et al. (1002). Structure-based drug design targeting an inactive RNA conformation: exploiting the flexibility of HIV-2 TAR RNA, J Mol Biol. 006(0), 611-00. 20. Davis B, Afshar M, Varani G, Murchie A I, Karn J, Lentzen G, et al. (1002). Rational design of inhibitors of HIV-2 TAR RNA through the stabilisation of electrostatic "hot spots", J Mol Biol. 006(1), 020-16. 20. Famulok M, Huttenhofer A (2006). In vitro selection analysis of neomycin binding RNAs with a mutagenized pool of variants of the 26S rRNA decoding region, Biochemistry. 01(22), 2161-10. 10. Weigand J E, Sanchez M, Gunnesch E B, Zeiher S, Schroeder R, Suess B (1000). Screening for engineered neomycin riboswitches that control translation initiation, RNA. 22(2), 00-01. 12. Morse D P (1001). Direct selection of RNA beacon aptamers, Biochem Biophys Res Commun. 010(2), 02-202. 11. Schmidtke S R, Duchardt-Ferner E, Weigand J E, Suess B, Wohnert J (1020). NMR resonance assignments of an engineered neomycin-sensing riboswitch RNA bound to ribostamycin and tobramycin, Biomol NMR Assign. 2(2), 221-0. 10. Vandenengel J E, Morse D P (1000). Mutational analysis of a signaling aptamer suggests a mechanism for ligand-triggered structure-switching, Biochem Biophys Res Commun. 010(2), 12-6. 12. Lünse C E. (1021). Investigation of riboswitches as new antibacterial targets: Identification and characterization of novel synthetic and natural riboswitch modulators with effect on bacterial cell growth. PhD Dissertation, University of Bonn. 11. Barrick J E, Breaker R R (1001). The distributions, mechanisms, and structures of metabolite-binding riboswitches, Genome Biol. 0(22), 2-20.

Chapter 5 220

16. Duchardt-Ferner E, Weigand J E, Ohlenschlager O, Schmidtke S R, Suess B, Wohnert J (1020). Highly modular structure and ligand binding by conformational capture in a minimalistic riboswitch, Angew Chem Int Ed Engl. 20(01), 6126-0. 11. Mehdizadeh Aghdam E, Barzegar A, Hejazi M (1022). Evolutionary origin and conserved structural building blocks of riboswitches and ribosomal RNAs: riboswitches as probable target site for aminoglycosides interaction, Adv Pharm Bull. 2(0), 111-101. 10. Barzegar A, Moosavi-Movahedi A A, Mahnam K, Ashtiani S H (1020). Chaperone-like activity of alpha-cyclodextrin via hydrophobic nanocavity to protect native structure of ADH, Carbohydrate Research. 021(1), 120- 0. 10. Morris G M, Lim-Wilby M (1000). Molecular docking, Methods Mol Biol. 220, 061-01. 00. Morris G M, Huey R, Lindstrom W, Sanner M F, Belew R K, Goodsell D S, et al. (1000). AutoDock2 and AutoDockTools2: Automated docking with selective receptor flexibility, Journal of Computational Chemistry. 00(26), 1101-02. 02. Trott O, Olson A J (1020). AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading, Journal of Computational Chemistry. 02(1), 211-62. 01. Breaker R R (1021). Riboswitches and the RNA world, Cold Spring Harb Perspect Biol. 2(1). 2-21. 00. Ruiz-Carmona S, Alvarez-Garcia D, Foloppe N, Garmendia-Doval A B, Juhos S, Schmidtke P, et al. (1022). rDock: a fast, versatile and open source program for docking ligands to proteins and nucleic acids, PLoS Computational Biology. 20(2), 2-1. 02. Pronk S, Pall S, Schulz R, Larsson P, Bjelkmar P, Apostolov R, et al. (1020). GROMACS 261: a high-throughput and highly parallel open source molecular simulation toolkit, Bioinformatics. 10(1), 021-12. 01. Jorgensen W L, Chandrasekhar J, Madura J D, Impey R W, Klein M L (2000). Comparison of simple potential functions for simulating liquid water, J Chem Phys. 10(1), 016-001. 06. Bjelkmar P r, Larsson P, Cuendet M A, Hess B, Lindahl E (1020). Implementation of the CHARMM force field in GROMACS: Analysis of protein stability effects from correction maps, virtual interaction sites, and water models, J Chem Theory Comput. 6(1), 210-266. 01. Trott O, Olson A J (1020). AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading, J Comput Chem. 02(1), 211-62.

Chapter 5 210

00. Nudler E, Mironov A S (1002). The riboswitch control of bacterial metabolism, Trends Biochem Sci. 10(2), 22-1. 00. Mulhbacher J, Brouillette E, Allard M, Fortier L C, Malouin F, Lafontaine D A (1020). Novel riboswitch ligand analogs as selective inhibitors of guanine-related metabolic pathways, PLoS Pathog. 6(2), 2-22. 20. Ster C, Allard M, Boulanger S, Lamontagne Boulet M, Mulhbacher J, Lafontaine D A, et al. (1020). Experimental treatment of Staphylococcus aureus bovine intramammary infection using a guanine riboswitch ligand analog, Journal of Dairy Science. 06(1), 2000-0. 22. Liberman J A, Wedekind J E (1021). Riboswitch structure in the ligand- free state, Wiley Interdiscip Rev RNA. 0(0), 060-02. 21. Tyrrell J, McGinnis J L, Weeks K M, Pielak G J (1020). The cellular environment stabilizes adenine riboswitch RNA structure, Biochemistry. 11(20), 0111-01.

Chapter 1: TPP riboswitch characterization in Alishewanella tabrizica and Alishewanella aestuarii and comparison with other TPP riboswitches

Chapter 3 211

5.1 Abstract

Riboswitches are located in non-coding areas of mRNAs and act as sensors of cellular small molecules, regulating gene expression in response to ligand binding. The TPP riboswitch is the most widespread riboswitch occurring in all three domains of life. However, it has rarely been characterized in environmental bacteria other than Escherichia coli and Bacillus subtilis. In this study, TPP riboswitches located in the 1’-UTR of the thiC operon from Alishewanella tabrizica and Alishewanella aestuarii were identified and characterized. Moreover, Affinity analysis of TPP binding to the TPP aptamer domains originated from A. tabrizica, A. aestuarii, E.coli, and B. subtilis were studied and compared using In-line probing and Surface Plasmon Resonance (SPR). TPP binding to the studied RNAs from A. tabrizica and A. aestuarii caused distinctive changes of the In-line cleavage pattern, demonstrating them as functional TPP riboswitches. With a dissociation constant of 1 – 2 nM (depending on the method utilized), the affinity of TPP binding was highest in A. tabrizica , followed by the motifs sourced from A. aestuarii, E. coli, and B. subtilis. The observed variation in their TPP-binding affinity might be associated with adaptation to the different environments of the studied bacteria.

Chapter 3 210

5.2 Introduction

Several gene regulation mechanisms have been established to adapt the cells to various changing conditions. However, RNA has been shown to play a significant role in controlling gene expression. RNA regulatory elements are usually divided into trans-acting riboregulators such as microRNAs (miRNA) and short interfering RNAs (siRNA) and cis-acting regulating RNAs within non-coding parts of RNA [reviewed in 2]. Riboswitches, as one of the cis-acting type elements, are usually located in 1’- UTRs of mRNAs where they regulate the transcription and/or translation of the respective gene through specific binding to small molecular ligands [1]. The organization of riboswitches usually consists of an aptamer domain and an expression platform. Specific binding of the corresponding ligand to the aptamer domain causes a conformational change in the expression platform leading to the creation of a transcription terminator structure and/or masking of the ribosomal binding site (RBS) to prevent translation initiation or vice versa[0]. Until now about 00 different classes of riboswitches have been identified. However, many candidate riboswitches identified by bioinformatics still await ligand-assignment [2, 1].

Thiamine pyrophosphate (TPP) is the active form of vitamin B2 and an important factor in several metabolic pathways [6]. The TPP riboswitch is the most widespread riboswitch and was also one of the first riboswitches introduced by Breaker and co- workers in E. coli [1] and Mironov et al. in B. subtilis [0]. In E. coli the UTR upstream of the thiM and thiC open reading frames were studied and conserved parts of the aptamer domain were observed. The mechanism of gene regulation in thiM RNA was suggested to be Shine Dalgarno-masking. However, the riboswitch in thiC is responsible for controlling gene expression by both SD-masking and transcription attenuation. In Bacillus subtilis, the mRNA upstream of the tenA ORF was studied and a thi-box was also identified and transcription termination was shown to be controlled by TPP binding. Later, it was revealed that some representatives of TPP

Chapter 3 212

riboswitch also exist in eukaryotes including fungi (in the Aspergillus oryzae thiA gene [0]), as well as plants (Oryza sativa 0’-UTR of a putative thiC gene [20], the tomato LeTHIC transcript [22], A. thaliana [21], and the algae Chlamydomonas reinhardtii thi2 and thiC transcripts [20]. The gene regulation mechanisms were studied for eukaryotic TPP riboswitches that controls intron splicing [0] involving long distance base pairing in Neurospora crassa [22]. Structural studies were carried out on the crystallographic structure of TPP riboswitch in thiM mRNA [21, 26] and Arabidopsis thaliana [21]. In addition, the folding mechanism upon ligand binding was studied in at the single-molecule level using both force spectroscopy and single molecule-Förster resonance energy transfer recently (smFRET) [21]. Moreover, the TPP aptamer has been used to construct synthetic riboswitches utilizing ribozymes as expression platform [20]. TPP riboswitches in E. coli and B. subtilis were the first examples of riboswitches as targets of antibiotics. In this case, the ligand analog pyrithiamine was demonstrated to bind and trigger the riboswitch, thereby shutting off the TPP biosynthesis [20]. This discovery led to the general idea of designing novel antibiotics based on structure of riboswitches [10]. Consequently, some drug discovery approaches were utilized to find TPP riboswitches targeting antibiotics [26, 12, 11]. Owing to the potential of riboswitches in novel antibiotic design and discovery, pathogenic bacteria are interesting area to the study of riboswitches. However, environmental microorganisms other than B. subtilis have rarely been used for studying riboswitches. These bacteria due to their different conditions as well as possibly diverse metabolic pathways could show different patterns of gene regulation such as riboswitch-mediated mechanisms. Here we studied the TPP riboswitches of two Alishewanella species. The genus Alishewanella from the phylum of -Proteobacteria was first introduced in 1000 and the first bacterium in this genus, considering the close proximity to genus Shewanella, was named Alishewanella fetalis [10]. About a decade later, other

Chapter 3 211

species in this genus were identified including Alishewanella jeotgali [12], Alishewanella aestuarii [11] and Alishewanella agri [16] isolated from traditional fermented food, tidal flat sediment, and landfill soil, respectively. Alishewanella tabrizica, a Gram-negative, aerobic, motile and rod-shaped bacterium was identified by our group in 1021 isolated from a Qurugöl Lake located in the north-west of Iran [11]. Alishewanella solinquinati is the last species introduced in this genus isolated from soil contaminated with textile dyes [10]. Apart from some annotations regarding the prediction of cobalamin and c-di-GMP riboswitches (GeneBank accession number of LCWL11111111.1 and LCWL11111143.1, respectively) no experimental identification or confirmation of riboswitches have been carried out in this genus. In this study, we identify and characterize the TPP riboswitch of thiC operon in A. aestuarii and A. tabrizica and compare these with TPP riboswitches in E. coli and B. subtilis. For this purpose, TPP binding was examined using In-line probing and surface plasmon resonance (SPR) methods to determine the affinity of TPP to these motifs.

5.3 Materials and Methods

5.3.1 Preparation of RNA sequences from different bacteria

5.3.1.1 Providing TPP riboswitch template for in vitro transcription In order to prepare TPP riboswitches RNAs, first PCR amplifications with corresponding primers (Table 162) were carried out. T1 RNA polymerase promoter sequence was added upstream of the TPP riboswitch aptamer domain. For RNAs to be used in SPR, forward primerswere designed to insert a linker of 12 bp between T1 RNA polymerase promoter sequence and TPP riboswitch. The procedure to provide templates for each bacterium and PCR details are explained below.

Chapter 3 216

Table 5.1 Oligos used in order to obtain target TPP riboswitches in studied bacteria.

Oligo Sequence (1'-0') Purpose GGCTAGTCTAGAGCTTGTCGGAGTGCATTAGGGCGATATTGCCCTGAAGCTGAGACCGCATTGCG GGATCCGTTGAACCTGATCAGGCTAATACCTGCGAAGGGAACAAGTAGTGATTATGTACCAGCTG 2 ACCAGCTACGAATACCGAGGCTTACAAATCTCTATTCTGAGTTGTGTCATCTGATAAGCAACGCA Synthesized sequence of 1’UTR of thiC operon from A. aestuarii AGACCTGCTGAATAGCAGTTTGATTTTGCTGCTGACACTTAATCCTATCCAACACTCCTGACAAG CATTTTCCCTTAACTTTTTGGTGAGATTGCTCCATGGTAGTCG 1 TAATACGACTCACTATAGGAATAGTTACTGGGGGTGCC F primera to amplify TPP riboswitch from B. subtilis (adding TPPSb)-In-line probing 0 TAATACGACTCACTATAGGAAGACACGATCTGCGACGTTTAATAGTTACTGGGGGTGCC F primer to amplify TPP riboswitch from B. subtilis(adding TPPS+reverse complement TM60 + TTT)-SPR 2 AAAACAATTCCTCTACTTCCCCACG R primerc to amplify TPP riboswitch from B. subtilis-In-line probing and SPR 1 TAATACGACTCACTATAGGCTTGTCGGAGTGCCTTAACTG F primer to amplify TPP riboswitch from E. coli (adding TPPS)-In-line probing 6 TAATACGACTCACTATAGGAAGACACGATCTGCGACGTTTCTTGTCGGAGTGCCTTAACTG F primer to amplify TPP riboswitch from E. coli (adding TPPS+reverse complement TM60 + TTT)-SPR 1 CTTGTTCCCTTCGCAGGTATTAG R primer to amplify TPP riboswitch from E. coli -In-line probing and SPR 0 ACAAACAGCCCCAACTCAG F primer to amplify thic gene (part) in A. tabrizica 0 GCAATTGCGTAACCCTTTTAC R primer to amplify thic gene (part) in A. tabrizica 20 GACCAAGATACAAACTTGTCGGAGTG F primer to amplify 1'UTR + 1' part of of thic gene in A. tabrizica R primer to amplify 1'UTR + 1' part of of thic gene in A. tabrizica 22 CATCGGCTCGGCTTCAGGATG based on the product of primers 6 and 1) F primer to amplify TPP riboswitch from A. Tabrizica (adding TPPS) -In-line probing 21 TAATACGACTCACTATAGGCTTGTCGGAGTGCCTTAGG template was PTZ/AtabTPP F primer to amplify TPP riboswitch from A. Tabrizica (adding TPPS+reverse complement TM60 + TTT)-SPR 20 TAATACGACTCACTATAGGAAGACACGATCTGCGACGTTTCTTGTCGGAGTGCCTTAGG template was PTZ/AtabTPP R primer to amplify TPP riboswitch from A. Tabrizica - In-line probing and SPR 22 TTAACCACTACTTGTTCCCTTCG template was PTZ/AtabTPP F primer to amplify TPP riboswitch from A. aestuarii (adding TPPS)- In-line probing 21 TAATACGACTCACTATAGGCTTGTCGGAGTGCATTAG template was PTZ/AaesTPP F primer to amplify TPP riboswitch from A. aestuarii (adding TPPS+reverse complement TM60 + TTT)-SPR 26 TAATACGACTCACTATAGAAGACACGATCTGCGACGTTTGCTTGTCGGAGTGCATTAG template wasPTZ/AaesTPP R primer to amplify TPP riboswitch from A. aestuarii - In-line probing and SPR 21 TAGCTGGTCAGCTGGTACAT template was PTZ/AaesTPP a Forward primer b T1 RNA polymerase promoter sequence c Reverse primer

Chapter 3 211

160626262 A. aestuarii TPP riboswitch The genome of A. aestuarii is available under the accession number of RefSeq: NZ_ALAB11111111.1 [10]. The sequence of thiC operon containing 1’UTR was extracted from the genome and then was confirmed as a TPP riboswitch by RNAReg 160 scanning. The whole sequence of 1’UTR was synthesized by GeneArt Company (Table 162-oligo 2) and was cloned in pTZ15R/T plasmid using TA cloning kit (Theremo scientific) (pTZ/AaesTPP). Constructed pTZ/AaesTPP was used as the template for PCR reaction. Besides, to obtain desired TPP riboswitch aptamer domain for In-line probing and SPR, primers 21/21 and 26/21 (Table 162) were employed, respectively.

160626261 A. tabrizica TPP riboswitch Not being available the complete genome of A. tabrizica, multiple alignment using Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/) was done on 1’UTR and thiC gene of A. aestuarii, A. jeotgali and A. agri to identify conserved sequences for primer design. In the first round of PCR, part of thiC gene was amplified using primers 0/0 (Table 162) and PCR product was sequenced. Then, reverse primer 22 was designed based on the sequenced area and forward primer 20 was designed based on the alignment results of the predicted aptamer domains of TPP riboswitches of the bacteria (Table 162). The PCR reaction was then conducted and the product was cloned into pTZ15R/T plasmid (pTZ/AtabTPP), sequenced and the obtained sequence was confirmed as TPP riboswitch using RNAReg 160 scanning (http://regrna1.mbc.nctu.edu.tw/). The whole sequence of 1’UTR and thiC gene partial sequence was submitted to NCBI under the accession number of KU668141. PCR reaction was carried out to obtain desired TPP riboswitch aptamer domain using pTZ/AtabTPP as the template as well as primers 21/22 and 20/22 for In-line probing and SPR experiments, respectively.

Chapter 3 210

160626260 E. coli and B. subtilis TPP riboswitch In order to obtain TPP riboswitch of E. coli and B. subtilis, the annotated TPP riboswitches in 1’UTR of thiC gene of accession numbers CP114225.1 and CP114188.1 were amplified, respectively, using colony PCR. For E. coli, primers 1/1 and 6/1 (Table 162) and for B. subtilis primers 1/2 and 0/2 (Table 162) were employed to obtain TPP riboswitch aptamer domain for In-line probing and SPR experiments, respectively.

160.26262 Secondary structures Secondary structures of TPP riboswitches were predicted with Mfold (http://unafold.rna.albany.edu/?q=mfold) [00].

5.3.1.2 In vitro transcription Ethanol precipitated PCR products were employed as templates for in vitro transcription using a reaction containing T1 RNA polymerase, Pyrophosphatase (ppase), RNase inhibitor and incubated for 0 h at 01 ºC. DNA was removed by DNaseI treatment. RNA samples were run in 05 denaturing PAGE (urea 0 M). Then, the desired bands were cut out and RNAs were extracted from the gel using crush soak buffer (20 mM HEPES, 100 mM NaCl, and 2 mM EDTA, pH 161). The eluted RNAs were precipitated with ethanol and resuspended in Mili-Q water.

5.3.2 In-line probing

In-line probing was carried out based on the protocol described by Regulski and Breaker [02] as follows. The experiment was conducted to observe any possible cleavage difference in ligand bound and unbound riboswitches.

5.3.2.1 RNA labeling In order to label the RNAs, 10 pmol of RNA was dephosphorylated using antarctic phosphatase. After removing the enzyme by phenol extraction, dephosphorylated RNA was labeled subsequently with α-01P-ATP using T2 polynucleotide kinase.

Chapter 3 210

RNAs were passed through Illustra MicroSpin G-11 Column (GE Healthcare) and purified over 05 denaturing PAGE (urea 0 M) and eluted from gel pieces in crush soak buffer, with subsequent precipitation with ethanol and resuspension in Mili-Q water.

5.3.2.2 In-line probing In line probing reaction was made containing 2µl 1’ 01P-labeled RNA (10000- 200000 cpm), 0 µl water, 1 µl 1x In-line reaction buffer (200 mM Tris-HCl, 20 mM

MgCl1, 100 mM KCl, pH 060) and 2 µl of TPP solution in the desired concentration. For making a dilution row, different concentrations of TPP were prepared from 20 mM stock solution in In-line reaction buffer. The reactions were incubated for approximately 20 h at 11ºC. The control reactions including RNase T2 RNA digestion (T2) and partial alkaline hydrolysis (–OH) ladders were prepared as 20 μl reactions and quenched with 20 μL loading buffer. Dried gels were exposed on a phosphorimager screen overnight and scanned using a Typhoon laserscanning system (GE Healthcare). Band volumes were determined using Quantity One software and accordingly Kd (dissociation equilibrium constant) was calculated as described by Regulski and Breaker [02] using following equilibrium: sample value − min value Fraction RNA cleaved = max value − min value Max and min are the highest and lowest cleavage values for each altered band in comparison to control, respectively. The concentration of TPP required to produce half maximal variation of cleavage provides an estimate of the Kd for TPP-TPP riboswitch complex.

5.3.3 SPR assay

5.3.3.1 Immobilization of biotinylated single stranded DNA on SA chip All SPR measurements were carried out at 112C using a Biacore T200 instrument. DNA immobilization and RNA capturing were done based on the method described

Chapter 3 200

by Liu et al [01]. Streptavidin coated sensor-chip (SA-chip) were purchased from Biacore and a biotinylated single stranded DNA (1‘ Biotin- CGTCGCAGATCGTGTCTTCC) was immobilized on all flow cells. HSB-EP buffer (20 mM HEPES pH 162, 210 mM NaCl, 0 mM EDTA, 060015, v/v polysorbate 10) was used as running buffer. First, in order to remove any unbound streptavidine from the chip, activation buffer (2 M NaCl, 10 mM NaOH) was injected for 2 min (with 10 µl/min flow rate) seven times. Afterwards, prepared solution containing 11 nM DNA in HSB-EP buffer was injected with 1 µl/min flow rate over the each cell. The injection was stopped once proper Response Unit (RU) was achieved. The steps were repeated for all of the flow cells.

5.3.3.2 Riboswitch loading on SA-chips RNA preparation and purification was carried out as previously described in In-line probing section. The difference was forward primers which add a complementary region to the biotinylated single stranded DNA and TTT linker beside to the T1 polymerase promoter sequence. The solutions of prepared RNA (aptamer domain of TPP riboswitch) with the concentration of 2 µM were prepared for injection. Running buffer containing 20 mM HEPES pH 162, 200 mM NaCl, 200 mM KCl, 1mM MgCl1 was used with starting 10 µl/min flow rate and each type of RNA solution was injected on each flow cell to get a suitable immobilization amount and flow cell 2 was left without RNA loading. For renewal and changing the type of riboswitch loading, regeneration buffer (11 mM NaOH) was injected over flow cells (for 1 min with 20 µl/min flow rate) to remove the last RNAs properly. The injection needle was rinsed with running buffer before injection of new RNA solution.

5.3.3.3 Real-time binding experiment for TPP Riboswitch-TPP Binding assays were carried out with a constant flow rate of 10 µl/min using running buffer containing 20 mM HEPES pH 162, 200 mM NaCl, 200 mM KCl, 1mM

MgCl1. Flow cell 2 immobilized with biotinylated DNA was as reference flow cell and other flow cells were loaded with aptamer domains of TPP riboswitch in A.

Chapter 3 202

aestuarii, A. tabrizica, E.coli and B. subtilis. TPP solution was injected at least with six different concentrations over all channels. The range of concentrations was considered between 2 nM - 2 µM prepared in the running buffer by serial dilutions of a 20 µM TPP stock solution based on Kd reported in the literature and In-line probing results. TPP was injected for 1 min at a flow rate of 00 µl/min. TPP was allowed to dissociate by flushing the cells with running buffer for 1 min at the same flow rate. Data points were recorded every second. The binding sensorgrams were obtained from the subtraction F1-F2, F0-F2 and F2-F2.

5.3.3.4 Kd determination The responses were determined from sensorgrams by using BIAevaluation software version 26062. The sensorgrams were fitted with a simplest model for 2:2 interaction between the riboswitch aptamer and TPP. Direct curve fitting of sensograms was determined the association (ka) and dissociation rate (Kd) constants. Dissociation equilibrium constant, Kd, was calculated.

5.4 Results

5.4.1 Sequence and secondary structure analysis of TPP riboswitch in A. aestuarii and A. tabrizica

Among 2 studied microorganisms, only A. tabrizica whole genome was not available. To get the TPP riboswitch aptamer domain from this bacterium, PCR amplification primers were designed based on the alignment of TPP riboswitches of Alishewanella species. Having amplified the desired sequence, it was predicted as an aptamer domain of TPP riboswitch in A. tabrizica, using the RegRNA program. Besides, using same program, the sequence of aptamer domains of TPP riboswitches in A. aestuarii was also identified. The alignment between TPP riboswitches of Alishewanella species and between A. tabrizica, E.coli and B. subtilis is shown in

Chapter 3 201

Figure 162A and Figure 162B, respectively. As shown in the alignment, the first difference among TPP riboswitches of Alishewanella sp. is their length; while, the aptamer domain is predicted as about 200 nt in all of them, the expression platform is about 10 nt longer in A. tabrizica. Furthermore, results showed that the similarity in aptamer domain (starting from position 21 to 221) between A. aestuarii, A. jeotgali and A. agri is remarkably high. However, A. tabrizica contain slightly different nucleotides in positions 10 (C/A), position 01-01 (AT/CGA), 20 (C/G), 21 (A/G),221 (A/G),212 (A/T), 212 (G/A), 216-211 (AT/CC), 210 (C/G), 202 (T/G) which are not co-variations. Comparing expression platform (starting from position 220 to 010), even with the high difference among Alishewanella species, there is high similarity in the area close to start codon (position 022-010). On the other hand the alignment between TPP riboswitch of A. tabrizica, E. coli and B. subtilis (Figure 162B) showed less similarity even in aptamer domain sequences. Nevertheless, the identity percentage between E. coli and A. tabrizica in aptamer domain (105) is more than B. subtilis and A. tabrizica (125) using pairwise alignment which could be the result of taxonomical proximity between E. coli and A. tabrizica.

Chapter 3 200

Figure 5.1 Multiple alignment of TPP riboswitches (A) Multiple alignment of 1’UTR of thiCEFSGH operons of Alishewanella sp. (B) Multiple alignment of 1’UTR of thiC gene in A. tabrizica, E. coli and B. subtilis and conserved base paired stem-loops (P2, P1, P0, P0a, P2 and P1) are identified. The alignment was carried out using CLUSTAL omega with default parameters.

The general secondary structure of the TPP riboswitch extracted from Rfam and the predicted structure of the TPP riboswitch aptamer domain (using RNAfold) from A. tabrizica and A. aestuarii are illustrated in Figure 1.1A, Figure 1.1B and Figure 1.1C, respectively. Considering the secondary structures of the TPP aptamer domain (Figure 1.2B and Figure 1.1A) conserved nucleotides form stem P2 and P1, as well as the bridges between P0 and P0a, P2 and P1. As shown in Figure 1.2B, Figure 1.1B and Figure 1.1C, P0 and P0a is longer in Alishewanella sp. On the other hand, the thi box or P2-P1 and P2-P1 segments are similar in size and structure. However, the sequence and secondary structure of the aptamer domain from B. subtilis has some differences relative to Alishewanella sp. and E. coli.

Chapter 3 201

Figure 5.2 Secondary structure of TPP riboswitch (A) General secondary structure of TPP riboswitch in Rfam (http://rfam.xfam.org/family/TPP) and secondary structure of TPP riboswitch aptamer domain in (B) A. tabrizica and (C) A. aestuarii using RNAfold (http://rna.tbi.univie.ac.at/cgi-bin/RNAfold.cgi). Conserved stem-loops in the structure are assigned.

Chapter 3 206

5.4.2 In-line probing To get insights in the secondary structure of the aptamer domain of the TPP riboswitch of A. tabrizica, A. aestuarii and B. subtilis In-line experiments were conducted. Additionally the apparent Kd of the tested motif against TPP was determined. Figure 1.0, Figure 1.2 and Figure 1.1 illustrate the In-line probing results of aptamer domains of TPP riboswitch in A. tabrizica, A. aestuarii and B. subtilis, respectively. For A. tabrizica the most prominent changes in the cleavage pattern due to TPP binding are at nucleotides G01 (P2-P2 junction), A10, U11 (L1), G60 (P2), U61 (P1-P2 junction), A20, U00 (P0-P0a junction) (Figure 1.0A). For the motif of A. aestuarii changes are at nucleotides A10, U10, G10, A21, U20 (Figure 1.2A). According to structural studies [21], nucleotides in P2, P2-P1 junction and P1-P2 junction are essentials for pyrophosphate recognition with aiding of a pair of hexa-coordinated Mg1+ ions. In addition, nucleotides located in P0-P0a junction are necessary for the recognition of 2-amino-1- hydroxymethylpyrimidine (HMP). Differences in the cleavage pattern upon the addition of TPP for the B. subtilis TPP riboswitch were observed for nucleotides G01, C00, U00, G12, C11, U60, U60, A11, A26, C11 (Figure 1.1A). These nucleotides are mostly located in thi box (P2-P1). However, nucleotides located in P1, P0 and P0a are also involved in binding as explained above. In addition most visible intensity changing nucleotides in TPP incubated RNAs were determined (Figure 1.0A, Figure 1.2A and Figure 1.1A) and normalized cleavage fractions were plotted with different concentration of TPP in order to calculate Kd (Figure 1.0C, Figure 1.2C and Figure 1.1C). For B. subtilis the intensity changes of nucleotide U00 (Figure 1.1A) was employed for Kd calculation. In A. tabrizica and A. aestuarii the most drastic change is related to G60 and G10 (P2-P1 junction), respectively. Those were used for the calculation of the Kd and for the normalization the intensity of the whole corresponding lane was used. Calculated Kd values are presented in Table 161. Comparing the Kd values of the TPP riboswitches derived from the studied bacteria (Table 161), the order of TPP affinity to TPP riboswitch aptamer domain is A. tabrizica > A. aestuarii > B. subtilis.

Chapter 3 201

Table 5.2: Affinity and kinetic parameters of TPP binding to TPP riboswitch aptamer domain using In-line probing and SPR

KD (nM) KD(nM) TPP riboswitch aptamer domain In line SPR Probing A. tabrizica 262 1

E.coli 2660 200 [1]

B. subtilis 12060 210

A. aestuarii 660 16

Chapter 3 200

Figure 5.3 In-line probing of TPP riboswitch A. tabrizica (A) In-line probing of 200 nt TPP riboswitch aptamer domain from A. tabrizica. In the right gel, lanes include precursor RNA, T2 digestion and alkaline hydrolysis of RNA (OH-) and RNA incubated without TPP (-TPP) and with TPP (+TPP), in the left gel and with different concentrations of TPP (2 mM, 200 µM, 20 µM, 2 µM, 200 nM, 20 nM, 2 nM) included. Nucleotides with different intensity between presence and absence of TPP are shown with flags. Conserved structured stem-loops are also assigned in the gel. (B) the sequence used for In-line probing and secondary structure based on RNAfold prediction with highlighting of flagged nucleotides. (C) The extent of RNA cleavage at site G60 normalized to G11 intensity is plotted for different concentrations of TPP to estimate concentration of TPP needed to achieve half maximal conformational change of RNA (apparent Kd).

Chapter 3 200

Figure 5.4 In-line probing of TPP riboswitch A. aestuarii (A) In-line probing of 211 nt TPP riboswitch aptamer domain from A. aestuarii. Lanes include precursor RNA, T2 digestion and Alkalyin hydrolysis of RNA (OH-) and RNA incubated without TPP (-TPP) and with different concentrations of TPP (2 mM, 200 µM, 20 µM). Nucleotides with different intensity between presence and absence of TPP are shown with flags. (B) the sequence used for In-line probing and secondary structure based on RNAfold prediction with highlighting of flagged nucleotides. (C) The extent of RNA cleavage at site G10 normalized to U12 intensity is plotted for different concentrations of TPP to estimate concentration of TPP needed to achieve half maximal conformational change of RNA (apparent Kd).

Chapter 3 220

Figure 5.5 In-line probing of TPP riboswitch B. subtilis (A) In-line probing of 220 nt TPP riboswitch aptamer domain from B. subtilis. In the gel, lanes include precursor RNA, T2 digestion and alkalyin hydrolysis of RNA (OH-) and RNA incubated without TPP (-TPP) and with 2 µM (+TPP) are included. Nucleotides with visible different intensity between presence and absence of TPP are shown with flags. Conserved structured stem-loops are also assigned in the gel. (B) the sequence used for In-line probing and secondary structure based on Mfold prediction with highlighting of flagged nucleotides. (C) The extent of RNA cleavage at site U00 normalized to G21 intensity is plotted for different concentrations of TPP to estimate concentration of TPP needed to achieve half maximal conformational change of RNA (Figure S2) (apparent Kd).

Chapter 3 222

5.4.3 Surface Plasmon Resonance Spectroscopy

To evaluate the affinity parameters of TPP binding to TPP riboswitch aptamer domains of different bacteria, rate of binding and dissociation and apparent equilibrium dissociation constants were measured using surface plasmon resonance (Fig 1.6 and Table 161). Overall, the kinetic profile of the interaction between TPP and TPP riboswitches demonstrated very fast association and dissociation rates (Figure 1.6). As a result, affinity determination based on kinetics was impossible to calculate by the instrument. Instead, affinity determination was carried out by analyzing concentration- dependent steady state responses.

The values of steady state Kd for TPP binding were 262 nM, 660 nM, 2660 nM, and 12060 nM for A. tabrizica , A. aestuarii, E. coli, and B. subtilis, respectively. As a result, the order of TPP affinity to TPP aptamer domains of examined bacteria was shown as A. tabrizica > A. aestuarii > E. coli > B. subtilis which makes the TPP riboswitch aptamer domain of A. tabrizica a significantly better than either of the other tested TPP aptamer domains in terms of binding to TPP. The determined Kd of the SPR measurements are in good accordance with the once obtained by the In-line probing experiments.

Chapter 3 221

Figure 5.8 SPR Affinity analysis of the TPP riboswitch aptamer from B. subtilis, E. coli, A. tabrizica and A. aestuarii. Affinity analysis of the TPP riboswitch aptamer from B. subtilis, E. coli, A. tabrizica and A. aestuarii. Plots are the global fit obtained from a 2:2 binding model based on the experimental data. The apparent equilibrium Kd values are reported in Table 161.

Chapter 3 220

5.5 Discussion

Riboswitches are one of the most important RNA regulatory elements known for their unique characteristics in specific and selective binding to cellular metabolites such as amino acids and derivatives [00], carbohydrates [02], nucleobases and their derivatives [01], ions [06] and coenzymes such as thiamine pyrophosphate (TPP) [1, 0]. TPP as an active form of thiamine is a cofactor playing an important role in several enzymatic reactions including glycolysis, the citric acid cycle and the pentose phosphate pathway. As a result, TPP level maintenance is essential for the cells. The TPP riboswitch was one of the first riboswitches determined in the 1’-UTR of genes involved in biosynthesis and transport of thiamine in bacteria. Later on, this riboswitch was also discovered in fungi, algae, and higher plants. The thiC gene is involved in the production of phosphomethyl pyrimidine synthase enzyme which is the first gene in the operon of ThiCEFSGH in E. coli [6] and a separate gene in B. subtilis. This enzyme starts the thiamine biosynthesis pathway with catalyzing the synthesis of the hydroxymethylpyrimidine phosphate (HMP-P) moiety of thiamine from aminoimidazole ribotide (AIR) [01]. The TPP riboswitch located upstream of the thiC gene was identified in several bacteria. However, the only published kinetically studied TPP riboswitch in the upstream of thiC was in E. coli [1]. The TPP riboswitch is the only riboswitch identified in eukaryotes so far. For instance, it is located in the 0’-UTR of putative thiC in plants (i.e Poa secunda and Oryza sativa [00] , tomato LethiC gene [22]) and thiC of a green algae C. reinhardtii [20]. In the present study, we described the identification of TPP riboswitches located in 1’UTR of thiC operons in A. tabrizica and A. aestuarii, and their comparison with E. coli and B. subtilis TPP riboswitches. The thiC operon of A. aestuarii is annotated in NCBI (accession number: ALAB11111142.1) to code the thiamine biosynthesis protein ThiC, thiamine monophosphate synthase (thiE like protein in E. coli), UBA/THIF-type NAD/FAD binding protein, sulfur carrier protein ThiS, thiazole synthase (thiG like protein in E. coli) and thiamine biosynthesis protein

Chapter 3 222

ThiH. As a result, the operon is obviously similar to thiC operon in E. coli. Accordingly, we predicted the presence of TPP riboswitch in the upstream of this operon in A. aestuarii using RegRNA 160 [00] riboswitch scanning. The predicted 1’UTR of thiC operon in A. aestuarii was synthesized in order to study in vitro the TPP binding characteristics of this element. However, because of non- availability of A. tabrizica’s whole genome, some primers were designed based on the alignments of thiC gene among Alishewanella species. Subsequently, amplified product was sequenced and predicted as TPP riboswitch as well. As shown in Figure 1.2A, the aptamer domains of TPP riboswitch in A. tabrizica and A. aestuarii (position 21-221) are 01 nt and 00 nt long, respectively. In addition, despite some slightly differences located in positions 00-10 in A. tabrizica, the aptamer domain is mostly identical in four species of Alishewanella. Nevertheless, considering the alignment of 1’UTR of thiC gene among A. tabrizica, E. coli and B. subtilis (Figure 1.2B), the similarity between aptamer domains is much lower. However, the identity percentage between E. coli and A. tabrizica is more than the similarity between B. subtilis and A. tabrizica. This observation is in agreement with the phylogenetic relation of riboswitches sequences’ conservation [20, 22] as Alishewanella belongs to the same phylum (proteobacteria) of E. coli’s.

Considering the mechanism of gene regulation, despite sequence variations in the expression platforms, there is a high similarity (2005) close to start codon (positions 022-010). As position 011-060 (GGTG) and 011-010 (CACC) are predicted as Shine Dalgarno sequence (SD) and anti-SD, respectively, this hints for a SD masking mechanism. However, the first assigned TPP riboswitch upstream of thiC operon in E. coli was proposed to control gene regulation by both transcription attenuation and translation termination [1]. Secondary structure prediction results in the same well-known structure for TPP riboswitches including conserved stem loops P2, P1, P2 and P1 for the newly assigned TPP riboswitches of A. tabrizica and A. aestuarii. P0 and P0a as less conserved parts are much longer in A. tabrizica and A. aestuarii relative to E. coli and B. subtilis.

Chapter 3 221

TPP riboswitch is a type II class as it goes to long distance rearrangements following ligand binding, in contrast to type I riboswitches with local distance rearrangements [21]. Anthoney and colleagues characterized folding and energy of TPP aptamer originated from Arabidopsis thaliana thiC gene of using a single- molecule optical-trapping method. The study confirmed the type II behavior of the riboswitch as well as a hierarchical folding sequence. Accordingly, secondary structure forms without the presence of TPP. However, tertiary structure forms slowly and in concert with ligand binding. Moreover, it was proposed that TPP binding progress, from weak to strong binding, may be a final step in the folding process to guarantee selectivity of riboswitch activation [20].

According to the In-line probing results, TPP riboswitch functionality of studied sequences was approved. The cleavage sites in A. tabrizica and A. aestuarii are located in Thi-box (P2 and p1) and P0-P0a junction which is in accordance with previously studied TPP riboswitches in bacteria and eukaryotes. However, in B. subtilis, two cleavage sites following TPP binding (A26 and C11 with reduced and increased intensity) are located in P0 and P0a, respectively. According to previous studies, the TPP-bound aptamer adopts a special folded structure in which one sensor helix arm (P1-P0) makes an intercalation pocket for the pyrimidine moiety of TPP, and the other sensor helix arm (P2-P1) forms a binding pocket for the pyrophosphate moiety of TPP that also involves in the binding to bivalent metal ions [21, 21, 22]. Using single-molecule fluorescence resonance energy transfer (smFRET), Haller and colleagues elucidate more the relationship between TPP recognition and aptamer folding. Their results on E. coli thiM TPP aptamer indicated that the P2 switch helix forms a primarily folded structure in the presence of Mg1+ alone. Also, P2 and other regions around TPP- binding site exhibit an unexpected degree of plasticity which probably results in facilitating the entry and exit of the TPP ligand and influence gene regulation [21].

The kinetic and affinity parameters were measured using In-line probing and Surface Plasmon Resonance methods (Table 2). Kd of TPP aptamer in E. coli (In- line probing) is extracted from Winkler et al study [1]. According to both In-line

Chapter 3 226

probing and SPR-based Kds, the affinity of TPP binding to TPP riboswitch aptamer domains (1’UTR of thiC gene) is the highest in A. tabrizica and lowest in B. subtilis. However Kd values obtained from SPR experiments are a little higher which is because of slightly different assay conditions and the difference of the method for affinity determination. It could be suggested that differences between the affinities of TPP riboswitches from different bacterial species is the result of a possible adaptation dependent on the available thiamine in the environment. Especially for A. tabrizica as a marine bacterium, lower concentration of TPP is needed to shut down TPP biosynthesis in the cell. This could be an advantage in order to save the bacterium in the aqueous environment with variant sources of essentials.

5.8 Conclusion

In this study, TPP riboswitches upstream of the thiC operon in A. tabrizica and A. aestuarii were identified and characterized. TPP binding affinity to different TPP riboswitch aptamer domains from 1’-UTRs of thiC genes sourced from A. tabrizica, A. aestuarii, E. coli and B. subtilis were compared. In-line probing results of RNAs from A. tabrizica and A. aestuarii incubating with and without TPP showed distinct changes in the cleavage patterns, demonstrating that the RNAs are functional TPP riboswitches. Despite the longer P0 and P0a elements in the riboswitch originated from A. tabrizica, the structure is similar to the general secondary structure of TPP riboswitches known basically in E. coli. The high degree of similarity of the 1’-UTRs of thiC close to the start codon among Alishewanella sp. suggests SD masking as the possible mechanism of this type of TPP riboswitch. According to the Kd values determined by In-line probing and SPR methods, the affinity of TPP binding was highest in A. tabrizica (1 nM determined by In-line probing and 2 nM by SPR) and it was in the order of TPP aptamer domains sourced from A. tabrizica > A. aestuarii > E. coli > B. subtilis.

Chapter 3 221

5.2 References

2. Waters L S, Storz G (1000). Regulatory RNAs in bacteria, Cell. 206(2), 621-10. 1. Nahvi A, Sudarsan N, Ebert M S, Zou X, Brown K L, Breaker R R (1001). Genetic control by a metabolite binding mRNA, Chem Biol. 0(0), 2020. 0. Serganov A, Patel D J (1021). Metabolite recognition principles and molecular mechanisms underlying riboswitch function, Annu Rev Biophys. 22, 020-10. 2. Weinberg Z, Wang J X, Bogue J, Yang J, Corbino K, Moy R H, et al. (1020). Comparative genomics reveals 202 candidate structured RNAs from bacteria, archaea, and their metagenomes, Genome Biol. 22(0), R02- R20. 1. Weinberg Z, Barrick J E, Yao Z, Roth A, Kim J N, Gore J, et al. (1001). Identification of 11 candidate structured RNAs in bacteria using the CMfinder comparative genomics pipeline, Nucleic Acids Res. 01(22), 2000-20. 6. Vander Horn P B, Backstrom A D, Stewart V, Begley T P (2000). Structural genes for thiamine biosynthetic enzymes (thiCEFGH) in Escherichia coli K-21, J Bacteriol. 211(2), 001-01. 1. Winkler W, Nahvi A, Breaker R R (1001). Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression, Nature. 220(6020), 011-6. 0. Mironov A S, Gusarov I, Rafikov R, Lopez L E, Shatalin K, Kreneva R A, et al. (1001). Sensing small molecules by nascent RNA: a mechanism to control transcription in bacteria, Cell. 222(1), 121-16. 0. Kubodera T, Watanabe M, Yoshiuchi K, Yamashita N, Nishimura A, Nakai S, et al. (1000). Thiamine-regulated gene expression of Aspergillus oryzae thiA requires splicing of the intron containing a riboswitch-like domain in the 1'-UTR, FEBS Lett. 111(0), 126-10. 20. Sudarsan N, Barrick J E, Breaker R R (1000). Metabolite-binding RNA domains are present in the genes of eukaryotes, RNA. 0(6), 622-1. 22. Zhao W, Cheng X, Huang Z, Fan H, Wu H, Ling H Q (1022). Tomato LeTHIC is an Fe-requiring HMP-P synthase involved in thiamine synthesis and regulated by multiple factors, Plant Cell Physiol. 11(6), 061-01.

Chapter 3 220

21. Thore S, Frick C, Ban N (1000). Structural basis of thiamine pyrophosphate analogues binding to the eukaryotic riboswitch, J Am Chem Soc. 200(16), 0226-1. 20. Croft M T, Moulin M, Webb M E, Smith A G (1001). Thiamine biosynthesis in algae is regulated by riboswitches, Proc Natl Acad Sci U S A. 202(11), 10110-1. 22. Li S, Breaker R R (1020). Eukaryotic TPP riboswitch regulation of alternative splicing involving long-distance base pairing, Nucleic Acids Res. 22(1), 0011-02. 21. Serganov A, Polonskaia A, Phan A T, Breaker R R, Patel D J (1006). Structural basis for gene regulation by a thiamine pyrophosphate-sensing riboswitch, Nature. 222(1001), 2261-12. 26. Warner K D, Ferre-D'Amare A R (1022). Crystallographic analysis of TPP riboswitch binding by small-molecule ligands discovered through fragment-based drug discovery approaches, Methods Enzymol. 120, 112- 00. 21. Duesterberg V K, Fischer-Hwang I T, Perez C F, Hogan D W, Block S M (1021). Observation of long-range tertiary interactions during ligand binding by the TPP riboswitch aptamer, Elife. 2, 2-21. 20. Wieland M, Benz A, Klauser B, Hartig J S (1000). Artificial ribozyme switches containing natural riboswitch aptamer domains, Angew Chem Int Ed Engl. 20(21), 1121-0. 20. Sudarsan N, Cohen-Chalamish S, Nakamura S, Emilsson G M, Breaker R R (1001). Thiamine pyrophosphate riboswitches are targets for the antimicrobial compound pyrithiamine, Chem Biol. 21(21), 2011-01. 10. Blount K F, Breaker R R (1006). Riboswitches as antibacterial drug targets, Nat Biotechnol. 12(21), 2110-62. 12. Cressina E, Chen L, Abell C, Leeper F J, Smith A G (1022). Fragment screening against the thiamine pyrophosphate riboswitch thiM, Chem Sci. 1(2), 211-261. 11. Warner K D, Homan P, Weeks K M, Smith A G, Abell C, Ferre-D'Amare A R (1022). Validating fragment-based drug discovery for biological RNAs: lead fragments bind and remodel the TPP riboswitch specifically, Chem Biol. 12(1), 102-1. 10. Vogel B F, Venkateswaran K, Christensen H, Falsen E, Christiansen G, Gram L (1000). Polyphasic taxonomic approach in the description of Alishewanella fetalis gen. nov., sp. nov., isolated from a human foetus, Int J Syst Evol Microbiol. 10 Pt 0, 2200-21.

Chapter 3 220

12. Kim M S, Roh S W, Nam Y D, Chang H W, Kim K H, Jung M J, et al. (1000). Alishewanella jeotgali sp. nov., isolated from traditional fermented food, and emended description of the genus Alishewanella, Int J Syst Evol Microbiol. 10(Pt 0), 1020-6. 11. Roh S W, Nam Y D, Chang H W, Kim K H, Kim M S, Oh H M, et al. (1000). Alishewanella aestuarii sp. nov., isolated from tidal flat sediment, and emended description of the genus Alishewanella, Int J Syst Evol Microbiol. 10(Pt 1), 212-2. 16. Kim M S, Jo S K, Roh S W, Bae J W (1020). Alishewanella agri sp. nov., isolated from landfill soil, Int J Syst Evol Microbiol. 60(Pt 0), 1200-100. 11. Tarhriz V, Nematzadeh G, Zununi Vahed S, Hejazi M A, Hejazi M S (1021). Alishewanella tabrizica sp. nov., isolated from Qurugol Lake, Int J Syst Evol Microbiol. 61(Pt 0), 2006-02. 10. Kolekar Y M, Pawar S P, Adav S S, Zheng L Q, Li W J, Shouche Y S, et al. (1020). Alishewanella solinquinati sp. nov., isolated from soil contaminated with textile dyes, Curr Microbiol. 61(2), 212-0. 10. Jung J, Choi S, Chun J, Park W (1021). Genome sequence of pectin- degrading Alishewanella aestuarii strain B22(T), isolated from tidal flat sediment, J Bacteriol. 202(20), 1216. 00. Zuker M (1000). Mfold web server for nucleic acid folding and hybridization prediction, Nucleic Acids Res. 02(20), 0206-21. 02. Regulski E E, Breaker R R (1000). In-line probing analysis of riboswitches, Methods Mol Biol. 220, 10-61. 01. Liu Y, Wilson W D (1020). Quantitative analysis of small molecule- nucleic acid interactions with a biosensor surface and surface plasmon resonance detection, Methods Mol Biol. 620, 2-10. 00. Rodionov D A, Vitreschak A G, Mironov A A, Gelfand M S (1000). Regulation of lysine biosynthesis and transport genes in bacteria: yet another RNA riboswitch?, Nucleic Acids Res. 02(10), 6120-11. 02. Winkler W C, Nahvi A, Roth A, Collins J A, Breaker R R (1002). Control of gene expression by a natural metabolite-responsive ribozyme, Nature. 210(6000), 102-6. 01. Nelson J W, Sudarsan N, Phillips G E, Stav S, Lunse C E, McCown P J, et al. (1021). Control of bacterial exoelectrogenesis by c-AMP-GMP, Proc Natl Acad Sci U S A. 221(21), 1000-02. 06. Furukawa K, Ramesh A, Zhou Z, Weinberg Z, Vallery T, Winkler W C, et al. (1021). Bacterial riboswitches cooperatively bind Ni(1+) or Co(1+) ions and control expression of heavy metal transporters, Mol Cell. 11(6), 2000-00.

Chapter 3 210

01. Lawhorn B G, Gerdes S Y, Begley T P (1002). A genetic screen for the identification of thiamin metabolic genes, J Biol Chem. 110(21), 20111- 0. 00. Wachter A, Tunc-Ozdemir M, Grove B C, Green P J, Shintani D K, Breaker R R (1001). Riboswitch control of gene expression in plants by splicing and alternative 0' end processing of mRNAs, Plant Cell. 20(22), 0201-10. 00. Chang T H, Huang H Y, Hsu J B, Weng S L, Horng J T, Huang H D (1020). An enhanced computational platform for investigating the roles of regulatory RNA and for identifying functional RNA motifs, BMC Bioinformatics. 22 Suppl 1, S2. 20. Barrick J E, Breaker R R (1001). The distributions, mechanisms, and structures of metabolite-binding riboswitches, Genome Biol. 0(22), 2-20. 22. Singh P, Sengupta S (1021). Phylogenetic analysis and comparative genomics of purine riboswitch distribution in prokaryotes, Evol Bioinform Online. 0, 100-600. 21. Montange R K, Batey R T (1000). Riboswitches: emerging themes in RNA structure and function, Annu Rev Biophys. 01, 221-00. 20. Anthony P C, Perez C F, Garcia-Garcia C, Block S M (1021). Folding energy landscape of the thiamine pyrophosphate riboswitch aptamer, Proc Natl Acad Sci U S A. 200(1), 2201-0. 22. Edwards T E, Ferre-D'Amare A R (1006). Crystal structures of the thi-box riboswitch bound to thiamine pyrophosphate analogs reveal adaptive RNA-small molecule recognition, Structure. 22(0), 2210-60. 21. Haller A, Altman R B, Souliere M F, Blanchard S C, Micura R (1020). Folding and ligand recognition of the TPP riboswitch aptamer at single- molecule resolution, Proc Natl Acad Sci U S A. 220(22), 2200-00.

Chapter 6: Conclusion, Perspectives and Recommendations

Chapter 1 211

8.1 Conclusion

In the last decades, it is confirmed that RNA is not just an intermediate molecule for making the proteins; as different types of RNA molecules play important roles in the cell surviving. Riboswitches are cis-acting RNAs usually located in the 1’ untranslated region of mRNAs that regulates the transcription and/or translation of adjacent genes. The genes are usually important in virulence or cell survival of bacteria and making them down-regulated could possibly result in the inactivation or death of harmful bacteria. On the other hand, the unique characteristics of riboswitches in selective and specific bindings to small molecules without the interpretation of proteins, make them interesting targets to manipulate the gene regulation. In this way, computational approaches as well as experimental methods are utilized to design new compounds as novel antimicrobial agents as well as studying the affinity of desired compounds to the riboswitches.

In this thesis, tertiary structures of various types of riboswitches were used for computational studies. Since “26S rRNA A site” is one of the approved targets for aminoglycosides, finding the similarity between riboswitches and ribosomal RNA motifs may raise the hypothesis that they can be targeted by aminoglycosides. Docking and molecular dynamic simulations were applied to study the binding energy and stability of these interactions. Finally, results revealed that the affinity of aminoglycosides to riboswitches is approximately comparable with the affinity of aminoglycosides to “26S A site”. Beside to our study some important experimental studies have been carried out which showed the possibility of aminoglycoside targeting of riboswitches. For example, Baird and colleagues showed that out of expectations, kanamycin/riboswitch interaction is specific and even modulating the conformation of the riboswitch in a way that inhibit the binding of cognate ligand [2]. On the other hand there are some reports regarding the presence of specific aminoglycoside-responsive riboswitches in aminoglycoside resistance bacteria [1-2]. In a very recent study, Dar and his colleagues discovered new ribo-regulators in antibiotic resistance bacteria using developed conditional screening method, term-seq. it was indicated that ribo-

Chapter 1 210

regulation of antibiotic resistance genes is very common in Gram-positive bacteria. Additionally, new riboswitch-like antibiotic-dependent ribo-regulators were discovered through this study. Although, this discovery could not exactly identify the regulators as direct binders to the antibiotics, it was another step to show the possibility of riboswitches’ interaction with antibiotics such as aminoglycosides [1]. Taking together, it can be proposed that riboswitches have this potential to be the target of aminoglycosides. However this can be just claimed in theoretical calculations here and hence, experimental validations is still an open problem.

On the other hand, because of riboswitches applications on drug design as well as designing natural sensors for essential metabolites, there is still a need to find riboswitches from new sources. In this regard, Alishewanella tabrizica, newly discovered specie by Tarhriz et al, have been applied to detect a TPP riboswitch from. Having confirmed extracted sequence as a TPP riboswitch, comparison of TPP affinity to TPP riboswitches of A. tabrizica, A. aestuarii, E. coli and B. subtilis have been carried out using in-line probing and SPR experiments. According to the KD values, the affinity of TPP binding was the highest in A. tabrizica (1 nM resulted from in-line probing and 2 nM resulted from SPR) and it was in the order of TPP aptamer domains sourced from A. tabrizica > A. aestuarii > E. coli > B. subtilis. It gives the impression for A. tabrizica as a marine bacterium that fast regulation of TPP maintenance in the cell is essential in order to save the bacterium from fast transport of molecules in a aqueous environment.

8.2 Perspectives

Regarding the possible interaction between aminoglycosides and riboswitches (based on the computational results and after experimental validation), it is possible to introduce riboswitches as a novel target for this type of antibiotics. Therefore, riboswitches can be considered as a target in order to design new lead compounds based on aminoglycosides. In addition, antibiotic resistance to aminoglycosides can be concerned in the level of riboswitches mistargeting.

Chapter 1 212

A new TPP riboswitch discovery in A. tabrizica can lead to knowing a new pattern for TPP riboswitch structures. This riboswitch high affinity to TPP can also be regarded to design any artificial riboswitch aptamer for fast detection of TPP level.

8.3 Recommendations

2. Experimental validation of aminoglycoside-riboswitch interaction is the first priority. 1. Other known antibiotics-riboswitch interactions can also be studied computationally and experimentally. 0. Riboswitch mutations in different antibiotic resistance could be evaluated in order to find any correlations between this possible interactions and MDR. 2. Employing new TPP riboswitch from A. tabrizica as a template to design a fast biosensor for TPP. 1. TPP aptamer of A. tabrizica can be used for designing new engineered RNA switches for different applications in biotechnology or gene therapy.

Chapter 1 211

8.4 References

2. Baird N J, Inglese J, Ferre-D'Amare A R (1021). Rapid RNA-ligand interaction analysis through high-information content conformational and stability landscapes, Nat Commun. 6(0000), 2-20. 1. Chen D, Murchie A I (1022). An aminoglycoside sensing riboswitch controls the expression of aminoglycoside resistance acetyltransferase and adenyltransferases, Biochim Biophys Acta. 2000(20), 012-0. 0. He W, Zhang X, Zhang J, Jia X, Sun W, Jiang H, et al. (1020). Riboswitch control of induction of aminoglycoside resistance acetyl and adenyl- transferases, RNA Biol. 20(0), 2166-10. 2. Jia X, Zhang J, Sun W, He W, Jiang H, Chen D, et al. (1020). Riboswitch control of aminoglycoside antibiotic resistance, Cell. 211(2-1), 60-02. 1. Dar D, Shamir M, Mellin J R, Koutero M, Stern-Ginossar N, Cossart P, et al. (1026). Term-seq reveals abundant ribo-regulation of antibiotics resistance in bacteria, Science. 011(6101), aad0021-aa0011.

Appendix

Appendix B

Table S1 Pairwise sequence alignment between lysine riboswitch (182 bp)with below sequence and similar rRNA PDB structures. GGACGGAGGCGCGCCCGAGAUGAGUAGGCUGUCCCAUCAGGGGAGGAAUCGGGGACGGCUGAAAGGCGAGGGCGCCGAAGCGA GCAGAGUUCCUCCCGCUCUGCUUGGCUGGGGGUGAGGGGAAUACCCUUACCACUGUCGCGAAAGCGGAGAGCCGUCCA

Identity (%) local alignment Number PDB code Length (bp) Local alignment local global region AGCGGAGAGCCGUCC |.|||| || |||| 1 1wtsA 5737 137 111-111 11 ACCGGA-AG--GUCC CGCUCUGCUU |||| |||.| 3 1rngA 1131 131 95-111 12 CGCU-UGCGU GGAC---GGCUGAAAGGCGAGGG |||| ||||.|| |..||| 2 1k3iA 1139 139 17-52 21 GGACCCGGGCUCAA---CCUGGG GGCG------CGCC |||| |||| 5 1uuuA 1231 135 1-11 19 GGCGUACGUUUCGUACGCC GACGGAGGCG |||.|.|||| 3 1vopA 1131 131 2-11 11 GACUGGGGCG GGGACGGCUGAAAGGCGAGGGC-----GCCGAAGCGAGCAGAGUUCCUCCC ||.|| |||| | ||||||...||.||.| ||| 1 1g1xI 1137 1135 17-121 11 GGCAC--CUGA------CCCCAUGCCGAACUCAGAAGUG-----CCC GGC------UGAAAGGCGAGGGCGCC ||| |||||.|..| |||| 7 1mt5A 1139 137 15-51 21 GGCGUAACGUUGAAAAGUUA---CGCC GGCUGGGGGUGAGGGGAAUACCCUUACCACUG-UCGCGAAAGCGGAGAGCC |||||| ||| |||| .|||| 1 1i2yA 7737 1131 115-111 19 GGCUGG------CUGUUCGC------CAGCC CACUGUCGC ||.| |||| 9 1hs1A 5531 131 171-112 12 CAAU-UCGC GCGCC-----CGAGAUGAGUAGGCUGUCCCAUCAGGGGAG ||.|| |||||.|| || |||| 16 512dA 1231 1231 11-11 21 GCUCCUAGUACGAGAGGA------CC------GGAG GGGACGGCUGAAAGGCGAGGGC-----GCCGAAGCGAGCAGAGUUCCUCCC ||.|| |||| | ||||||...||.||.| ||| 11 1mjiC 7131 1739 12-91 71 GGCAC--CUGA------CCCCAUGCCGAACUCAGAAGUG-----CCC GCUGUCCCAUCAGGGGAGGAAUCG ||| ||..|||.||.|||.|| 13 1q92C 1231 931 21-11 27 GCU----CAGUAGGAGACGAACCG GGAGGCGCGCCCGAGAUGAGUAGGCUGUCCCAUC |||.|| |.|||||| || |||| 12 1qwaA 1239 1132 1-71 21 GGAUGC-CUCCCGAG------UG---CAUC GGCUGUCCCAUCAGGGGAGGAAUCGGGGACGGCUGAAAGGCGAGGGCGCC ||| |||.||.| .||||| |||| |.|||| 15 1pbrA 1131 1137 25-51 25 GGC-GUCACACC------UUCGGG------UGAA------GUCGCC AAGGCGAGGGCGCCGAAGCGAGCAG-----AGUUCCUCCCGCUCUGCUU |.|.|||.|| ||||| |||| .|.|||.||||....|||| 13 1g1xJ 1531 1139 17-111 17 ACGCCGAUGG---CGAAG---GCAGCCACCUGGUCCACCCGUGACGCUU GCGCCCGAGAUGAGUAGGCUGUCCCAUCAGGGGAGGAAUCGGGGACGGCUGAAAGGC |.||||| ||| .|||.|.|| |||| || ||.||.| ||| 11 1i1uD 1139 1131 11-15 71 GGGCCCG------GUA---AGUCUCUUC---GGAG--AU-----ACUGCCG---GGC GGCUGUCCCAUCAGGGGAGGAAUCGGGGACGGCUGAAAGGCGAGGGCGCC ||| |||.||.| .||||| |||| |.|||| 17 1fyoA 1131 1137 25-51 25 GGC-GUCGCACC------UUCGGG------UGAA------GUCGCC GGCUGAAAGGCGAGGGCGCCGAAGC |.||| |||.|| |.|.||||| 11 1jurA 1131 1131 15-11 21 GCCUG--AGGAGA---CUCAGAAGC GGGAAUACCCUUACCACUGUCGCGAAAGCGGAGAGCCGUCC 19 1 1739 1132 121 111 22 |||| |||..|| |||..|| |||| bgzA - GGGA------UACUGCU-UCGGUAA------GUCC GGCUGUCCCAUCAGGGGAGGAAUCGGGGACGGCUGAAAGGCGAGGGCGCC ||| |||.||.| .||||| |||| |.|||| 36 1fypA 1131 1137 25-51 25 GGC-GUCGCACC------UUCGGG------UGAA------GUCGCC GGGAGG-AAUCGGG---GACGGCUGAAAGGCGAGGGCGCCGAAGCGAGCA |||.|| ..||||| ||||| || |.| 31 GGGCGGCCUUCGGGCUAGACGG-UG------GGA GAGUUCCUCCCGCUCUGCUU-GGCUGGGGGUGAGGGGAAUACCCUUACCA ||| |||| |||||| .|| |||. 1kuqB 7937 2239 11-112 11 GAG------GCUUCGGCUGG------UCC--ACCC CUGUCGC .||.||| GUGACGC GCCGAAGCGAGCAGAGUUCCUCC--CGCUCUGCUUGG---CUGGGGGUGA |||| || ||.| |||.| .|| |||.||.| |||| || 33 GCCG--GC---CACA---CCUACGGGGC-CUGGUUAGUACCUGG----GA 1un1F 1137 2139 51-119 11 ----GGGGAAUACCCUUACCACUGUCGCGAAAGCGGAGAGCCGUC .||||||||| || |.||||.| AACCUGGGAAUACC------AG----GUGCCGGC GGAGGCGCGCCCGAGAUGAGUAGGCUGUCCCAUCAGGGGAGGAAUCGGGG ||| || |||| |||| |||| .|||| 32 GGA------CC--GAUG-GUAG--UGUC------UUCGG-- 1d1kB 1132 2131 1-52 71 ACGGCUGAAAGGCGAGGG |.|||||.| ------AUGCGAGAG GAGAUGAGUAGGCUGUCCCAUCAGGGGAGGAAUCGGGGACGGCUGAAAGG |.||||.|||||.| ||| .|||| ||.|.|||| 35 GGGAUGCGUAGGAU------AGG----UGGGA--GCCGCAAGG CGAGGGCGCCGAAGCGAGCAGAGUUCCUCCCGCUCUGCUUGGCUGGGGGU || || ||| 1mzpB 7532 2531 15-171 11 CG------CC------GGU GAGGGGAAU---ACCCUUACC || ||| ||||||.|| GA----AAUACCACCCUUCCC GCCCGAGAUGAGUAGGCU------GUCCCAUCAGGGGAGGAAUCGGGG |||.| ||| ||||||| | |||||| 33 GCCAG-GAU--GUAGGCUUAGAAGCAG--CCAUCA------1qa1D 1131 2231 17-11 11 ACGGCUGAAAGGCGAGGGCGCCGAAGCG----AGC .|.|||| .||||| ||| ----UUUAAAG------AAAGCGUAAUAGC CGGAGGCG----CGCCCG------AGAUGAGUAGGCU |||.||.| |.|||| ||||.|| 31 CGGUGGGGUUGCCUCCCGUACCCAUCCCGAACACGGAAGAUAAG------GUCCCAUCAGGGGAGGAAU--CGGGGACGGCUGAAAGGCGAGGGCGCCGA ||||.|||.| | ||||||...||| |||.|| --CCCACCAGCG------UUCCGGGGAGUACUG------GAGUGC----- 1q11B 7137 7131 1-171 121 AGCGAGCAGAGUUCCUCCCGCUCUGCUUGGCUGGGGGUGAGGGGAAUACC |||||| |||| ||||| ||| -GCGAGC------CUCU------GGGAA-ACC C--UU----ACCAC | || .|||| CGGUUCGCCGCCAC GGUGAGGGGAAUACC ||| |.|||.||| 37 1afxA 1135 132 111-121 12 GGU---GUGAACACC

Appendix C

Table S2 Pairwise sequence alignment between purine riboswitch with below sequence and similar rRNA PDB structures. GGACAUAUAAUCGCGUGGAUAUGGCACGCAAGUUUCUACCGGGCACCGUAAAUGUCCGACUAUGUCC

Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) UCGCGU |.|||| 1 1rngA 1737 1131 11-11 12 UUGCGU GCGUGGAUAUGGCAC |.|||.| ||| 3 1afxA 1737 132 17-25 12 GUGUGAA-----CAC CGGGCACCGUAAAUGUCCGACUAUG ||| ||| |||.|...|||..|| 2 1g1xI 1131 1131 11-11 11 CGG---CCG-AAAGGCUAGACGGUG CGGGC-ACCG ||||| |||| 5 211dB 9131 1131 11-11 21 CGGGCGACCG CGCAAGUUUC-UAC ||.|.||||| ||| 3 1uuuA 5131 2231 25-79 19 CGUACGUUUCGUAC ACCGGGCACCGUAAAUGUCC ||||| ||.|||| 1 1wtsA 1131 1131 71-15 11 ACCGG------AAGGUCC GCGUGGAUAUGGC |||..||.|||.| 7 1fhkA 1932 1132 17-21 11 GCGGUGAAAUGCC GCGUGGAUAUGGCA ||||..|| |.||| 1 1hs1A 5131 1139 17-21 17 GCGUCAAU-UCGCA ACCGGG ||.||| 9 1vopA 1737 1139 71-17 17 ACUGGG CGCAAG |||||| 16 1zihA 11131 1139 25-72 12 CGCAAG UGGC---ACGCAAG |||| .|||.|| 11 1i2yA 1137 1131 22-72 19 UGGCUGUUCGCCAG CCGGGC---ACCGUAAAU--GUCC |||||| ||| | |||| 13 1k3iA 1137 2939 79-15 27 CCGGGCUCAACC-----UGGGUCC UAAUCGCGUGGAUAUGGCACGC ||| |||.||.|.|..|||| 12 1mt5A 1731 2231 1-29 21 UAA---CGUUGAAAAGUUACGC GACU |||| 15 1jurA 11131 1137 21-91 22 GACU AUGGCACGCAAGUUUCUACCGGGCACC |||.|.|.|.||| |||.| 13 1qwaA 1131 2131 21-15 21 AUGCCUCCCGAGU------GCAUC UGGCACGCAAGUUUCUACCGG |.|.|||..|| .||||| 11 512dA 1531 2139 22-12 25 UAGUACGAGAG----GACCGG GGCACCGUAAAUGUCCGACUAUGUC |||||| ||.|| |.|||.| 17 1mjiC 1131 2537 12-11 71 GGCACC-----UGACC--CCAUGCC UAUGGCACGCAAGUUUCUACCGGGCACC ||.|..||| .||| ||||| 11 1q92C 1131 1131 21-15 25 UAGGAGACG------AACC--GCACC ACAUAUAAUCGCGUGGAUAUGGCACGCAAGU |||..| |||.||| |||| 19 1pbrA 1132 2131 7-77 25 ACACCU--UCGGGUG------AAGU GCGUGGAUAUGGCACGCAAGUUUCUACCGGGCACCGUAAAUGUC |||| ||||..|| ||| ||.|| ||| 36 1fyoA 1131 7231 17-11 25 GCGU------CGCACCUU-----CGG-----GUGAA-GUC GGAUAUGGCACGCAAGUUUCUACCGGGCACCGUAAAUGUCC |||||..|| || |.|||| |||| 31 1bgzA 1739 2139 15-15 27 GGAUACUGC------UU------CGGUAA--GUCC GGCAC-GCAAGU------UUCUACCGGGCAC |||.| |.|||| |.||.||||||.| 33 1i1uD 1231 2131 27-11 75 GGCCCGGUAAGUCUCUUCGGAGAUACUGCCGGGCCC GGACAUAUAAUCGCGUGGAUAUGGCACGCAAGUUUCUACCGGGCACCGUA |||| || |||| .|||.|||.|.| 32 GGAC------CG-----AUGG-----UAGUGUCUUCGG------1d1kB 1231 1131 1-11 75 AAUGUCCGACUAUGUC |||...||.||.||| -AUGCGAGAGUAGGUC GCGU-GGAUA--UGGCA--CGCAAGUUUCUACCGGGCACCG--UAAAU |||| ||||| |||.| ||||| |||.||| .|||| 35 1mzpB 1137 1135 17-17 11 GCGUAGGAUAGGUGGGAGCCGCAA------GGCGCCGGUGAAAU GCGUGGAUAUGGCACGCAAGUUUCUACCGGGCACCGUAAAUGUC |||| ||||..|| ||| ||.|| ||| 33 1fypA 1131 7231 17-11 25 GCGU------CGCACCUU-----CGG-----GUGAA-GUC GCGUGGAUAU-GGC----ACGCAAG------UUUCUACCGGGCACCGUAA ||..||||.| ||| |.|| || |||..| |..|.||||| 31 GCCAGGAUGUAGGCUUAGAAGC-AGCCAUCAUUUAAA---GAAAGCGUAA 1qa1D 1231 1731 17-11 11 -AUGUCCGACU |..|| ||| UAGCUC--ACU GGCACGCAAGUUUCUACCGGGC------ACC--GUAAAUGUCCGACUA |.|||.| ||||.|||| ||| |.|||..|..||.|| 37 1un1F 1931 7139 27-12 11 GCCACAC------CUACGGGGCCUGGUUAGUACCUGGGAAACCUGGGAAUA CGGGC---ACCGU---AAAUG-UCCGACUAUGUCC ||||| ||.|| |.|.| |.||.|| .|||| 31 1f7yB 1131 7531 11-15 15 CGGGCUAGACGGUGGGAGAGGCUUCGGCU-GGUCC AUGGC--ACGCAAGUUUCUACCGGG--CA-CCGUAAAUGUCCGAC ||||| |.||| .|.|||.|| || |||| ||| 39 1g1xJ 1737 1231 21-11 11 AUGGCGAAGGCA----GCCACCUGGUCCACCCGU------GAC

Appendix D

Table S3 Pairwise sequence alignment between glycine riboswitch with below sequence and similar rRNA PDB structures. GGCUCUGGAGAGAACCGUUUAAUCGGUCGCCGAAGGAGCAAGCUCUGCGGAAACGCAGAGUGAAACUCUCAGGCAAAAGG ACAGAGUC

Number PDB code Identity (%) Similar region in the riboswitch Length Local alignment local global (local alignment) (bp) GCUCUGCG ||| |||| 1 1rngA 1531 1132 12-19 12 GCU-UGCG GGAGAGAAC ||.|.|||| 3 1afxA 5531 139 5-11 12 GGUGUGAAC AAGGAGCAAGCUCUGCGGAAACGC-AGA--GUGAAACUCUCAGGCAAAAGGACAGAGU |||| |.||.||||.|| ||| ||| |||.||.|| 2 1g1xI 1131 2131 77-15 11 AAGG------CGGCCGAAAGGCUAGACGGUG------GGAGAGGGU CGUUUAAUCGGUCGCC |||| |||..|||| 5 1uuuA 1131 1132 11-71 19 CGUU---UCGUACGCC GAACCGUUUAAUCGGUCGCCGAAGG |.|||| ||||| 3 1wtsA 1131 1231 12-11 11 GGACCG------GAAGG GCG--GAAACGC ||| ||||.|| 1 1fhkA 5131 1131 15-11 11 GCGGUGAAAUGC GCGGAAA--CGCA |||..|| |||| 7 1hs1A 1932 1131 15-15 17 GCGUCAAUUCGCA CUGGAGAGAACCGUUUAAUCGGUC ||||.| ||||| 1 1vopA 1135 1131 1-21 17 CUGGGG------CGGUC GGAGCAAGC ||.|||||| 9 1zihA 1139 1131 71-17 12 GGCGCAAGC CUGGAGAGAACCGUUUAAUCGGUCGCCGAAGGAGC |||| |.|| ||||| ||| 16 1i2yA 1239 1937 1-79 19 CUGG------CUGU------UCGCC-----AGC GGCUCUGGAGAGAACCGUUUAAUCGGUC ||||| |||| |.|||| 11 1k3iA 1131 1135 1-21 27 GGCUC------AACC------UGGGUC GGCUCUGGAGAGAACCGUUUAAUCGGU-CGCC ||| |.|.||||.||..|.| |||| 13 1mt5A 1731 1931 1-71 21 GGC------GUAACGUUGAAAAGUUACGCC CUGCGGAAACGCAGA |||.|||.||.|||| 12 1jurA 1131 1132 11-19 22 CUGAGGAGACUCAGA GGA-GCAAGCUCUGCGGAAACGCAGAGUGAAAC ||| || ||| |.|||||.|.| 15 1qwaA 1131 1131 71-11 21 GGAUGC---CUC------CCGAGUGCAUC GAGAGAACCG |||||.|||| 13 512dA 9131 1931 1-15 25 GAGAGGACCG GCA----AGCUC-UGCGGAAACGCAG-AGUG ||| |.|.| |||.| |||.||| |||| 11 1mjiC 1131 2731 71-12 71 GCACCUGACCCCAUGCCG-AACUCAGAAGUG GCUC---UGGAGA-GAACCG |||| .||||| |||||| 17 1q92C 5131 1931 2-15 25 GCUCAGUAGGAGACGAACCG GGC-UC------UGGAGAGAACCGUUUAAUCGGUCGCC ||| || |.|.|.||| |||||| 11 1pbrA 1135 1539 1-71 25 GGCGUCACACCUUCGGGUGAA------GUCGCC GGC-UCUGGAGAGAACC----GUUUAAUCGGUCGCC ||| || |.||| |.|.|| |||||| 19 1fyoA 1231 2131 1-71 25 GGCGUC------GCACCUUCGGGUGAA---GUCGCC GUGAAACU-CUCAGGCAAAAGGACAGAGUC |.||.||| ||..||.| |||| 36 1bgzA 1131 1131 11-11 27 GGGAUACUGCUUCGGUA------AGUC GGUCGCCGAAGGAGCAAG-CUCUGCGGAAACGCAGAGUGAAACUCUCAGGC ||.| ||| |.||| ||||.|||| ||.|||..|.||| 31 1i1uD 1239 7137 21-57 75 GGGC-CCG-----GUAAGUCUCUUCGGA------GAUACUGCCGGGC CCGAAGG-AGCAAGCUCUGCGGAAACGCAGAGUGAAACUCUCAGG ||||.|| || ..|||.||||..|| ||||| ||| 33 1d1kB 1131 7231 71-57 75 CCGAUGGUAG---UGUCUUCGGAUGCG-AGAGU------AGG GAGAACCGUUUAAUCGGUCGCCGAAGGAGCAAGCUCUGCGGAA-ACGCAGAGUGAAA |.||..|||...||.||| .||||| ||.|| .|||.| |||||| 32 1mzpB 1231 7231 11-11 11 GGGAUGCGUAGGAUAGGU------GGGAGC------CGCAAGGCGCCG-GUGAAA GGC-UCUGGAGAGAACC----GUUUAAUCGGUCGCC ||| || |.||| |.|.|| |||||| 35 1fypA 1231 2131 1-71 25 GGCGUC------GCACCUUCGGGUGAA---GUCGCC UGGAGAGAACCGUUUAAUCGGUCGCCGAAGGAGCAAGC-UCUGCGGAAAC ||.|| |.||| ||||.|||.|.| |.| |||. 33 1qa1D 1131 1137 1-57 11 UGUAG------GCUUA------GAAGCAGCCAUCAUUU----AAAG GCAGAGUGA-AACUCUCAGG ..||.||.| |.|||.|.|| AAAGCGUAAUAGCUCACUGG GGCUCUGGAGAGAACCGUUUAAUCGGUCGCCGAAGGAGCAAGCUCUGCGG ||| ||||..||.||| |.|| ||.| ||| || 31 1un1F 1937 7139 1-51 11 GGC-CUGGUUAGUACC------UGGG------AAAC-CUG-GG AA-ACGCAGAGUGAAACUCUCAGGC || || ||| ||| |.||| AAUAC-CAG-GUG------CCGGC GGC--UCUGGAGAGAACCGUUUAAUCGGUCGCCGAAGGAGCAAGCU---- ||| ||.||..||| |||| .|||| |.||| 37 1f7yB 1131 7135 1-11 15 GGCCUUCGGGCUAGA------CGGU------GGGAG-AGGCUUCGG CUGCGGAAACGCAGAGUGAAACUC ||| |...||.| ||||..||| CUG-GUCCACCC---GUGACGCUC CGCCGAAGGAGCAAGCUCUGCGGAAACGCAG------AGUGA ||||||.||.| ||.|||| .|||| 31 1g1xJ 7937 2131 21-19 11 CGCCGAUGGCG------AAGGCAGCCACCUGGUCCACCCGUGA AACUCU ..||.| CGCUUU GGC--UCUGGAGAGAACCGUUUAAUCGGUCGCCGAAGGAGCAAGCU---- ||| ||.||..||| |||| .|||| |.||| 39 1kuqB 1131 7135 1-11 15 GGCCUUCGGGCUAGA------CGGU------GGGAG-AGGCUUCGG CUGCGGAAACGCAGAGUGAAACUC ||| |...||.| ||||..||| CUG-GUCCACCC---GUGACGCUC

Appendix E

Table S4 Pairwise sequence alignment between TPP riboswitch with below sequence and similar rRNA PDB structures. GGACUCGGGGUGCCCUUCUGCGUGAAGGCUGAGAAAUACCCGUAUCACCUGAUCUGGAUAAUGCCAGCGUAGGGAAGUUC

Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) GCCCUUCUGCGU ||.|| ||||| 1 1rngA 5131 1132 12-27 12 GCGCU--UGCGU GCUGAGAAAUACCCGUAUCACC |.||.||| |||| 3 1afxA 1131 1231 21-19 12 GGUGUGAA------CACC AAGGCUG-AGAAA---UACCCGUAUCACCUGAUCUGGAUAAUGCCAGCGUAG-GGAA 2 |||||.| .|||| ||..|| .||| |..||.||.| |||| 1g1xI 1531 7131 21-51 11 AAGGCGGCCGAAAGGCUAGACG------GUGG-----GAGAGGGUGGUGGAA

CUGCGUGAAGGCUGAGAAAUACCCGUAUCACCUG |||.|.| ||| || ||| |||| 5 211dB 1131 2132 11-11 21 CUGGGCG--GGC---GA-----CCG-----CCUG GCGUAGG |||||.| 3 1uuuA 1135 1137 15-57 19 GCGUACG GGAAGUUC |||||.|| 1 1wtsA 1531 1131 57-11 11 GGAAGGUC GGCUGAGAAAUACC |||.|.|||||.|| 7 1fhkA 5131 1731 25-11 11 GGCGGUGAAAUGCC GCGUGAA ||||.|| 1 1hs1A 1135 931 21-21 17 GCGUCAA GACUCGGG ||||.||| 9 1vopA 1531 1132 2-9 17 GACUGGGG GGGUGCCCUUCUGCGUGAAGGCU |||.|| |||.|| 16 1zihA 1731 1231 1-71 12 GGGCGC------AAGCCU GGCUGAGAAAUACCCGUAUCACCUGAUCUGGAUAAUGCCAGC ||||| .|||.|| |||||| 11 1i2yA 7131 2131 25-11 19 GGCUG------GCUGUUC------GCCAGC GGACUCGGGGUGCCCUUCUGCGU ||||.||||.| |...|||.|| 13 1k3iA 1132 1731 1-27 27 GGACCCGGGCU--CAACCUGGGU GCGUGAAGGCUGAGAAA--UACCC |||| ||.|.||| ||| |||.| 12 1mt5A 5131 2135 21-11 21 GCGU-AACGUUGA-AAAGUUACGC GCGUGA--AGGCUGAGAA ||.||| ||.||.|||| 15 1jurA 5232 1137 21-71 22 GCCUGAGGAGACUCAGAA GGGUGCCCUUCUGCGUGAA ||.|| |||.|.|.|||.| 13 1qwaA 1131 1531 1-21 21 GGAUG-CCUCCCGAGUGCA GGACUCGGGGUG |||| |||.||| 11 512dA 1737 1131 1-12 25 GGAC-CGGAGUG CACCUGAUCUGGAUAAUGCC-AGCGUAGGGAAGUUC |||||||.| ..||||| |.|..| |||||.| 17 1mjiC 1131 2132 11-11 71 CACCUGACC----CCAUGCCGAACUCA--GAAGUGC GGUGCCCUUCUGCGUGAAGGCUGAGA--AAUACCCGUAUCACC |||||.| ||...|||| || ||| |||| 11 1q92C 1131 2131 9-19 25 GGUGCUC------AGUAGGAGACGAA---CCG---CACC CCUUCUGCGUGAAGGC ||||| |.||||||.| 19 1pbrA 1132 2137 11-29 25 CCUUC-GGGUGAAGUC GGGGU-GC-CCUUCUGCGUGAAGGC ||.|| || ||||| |.||||||.| 36 1fyoA 5131 2131 5-29 25 GGCGUCGCACCUUC-GGGUGAAGUC GGAUAAUGCCAGCGUAGGGAAGUUC ||||| |.||.|.||.||||.| 31 1bgzA 1131 1931 11-11 27 GGAUA----CUGCUUCGGUAAGUCC GGCUGAGAAAUACCCGUA----UCACCUGAUCUGGAUAAUGCCAGCGUAGGG ||| ||.||| ||..|.|| ||||.|||| ||| 33 1i1uD 1132 7735 25-52 75 GGC------CCGGUAAGUCUCUUCGGA----GAUACUGCC------GGG ACCUGAUCUGGAUA------AUGCCAGCGUAGG ||| ||| |.|| ||||.||.||||| 32 1d1kB 1131 7131 15-57 75 ACC-GAU---GGUAGUGUCUUCGGAUGCGAGAGUAGG GGACUCG------GGGUGCCCUUCUGCGUGAAGG---CUGAGAAAU |||..|| |||.||| || |||| |.|.||||| 35 GGAUGCGUAGGAUAGGUGGGAGCC-----GC---AAGGCGCCGGUGAAAU 1mzpB 1131 7532 1-19 11 ACCCGUAUCACC || |||| AC------CACC UGAAGGCUGAGAAAUACCCGUAUCACCUGAUCUGGAUAAUGCCAGCGUA ||.|||||.||||..|.|| |||| ..|||.|..|||||| 31 1qa1D 1531 1231 27-51 11 UGUAGGCUUAGAAGCAGCC--AUCA------UUUAAAGAAAGCGUA CGGGGUGCCCUUCUGCGUGAAGG------CUGAGAAAUACCCGUAUCAC ||||| ||| || |||.||| || 37 CGGGG---CCU------GGUUAGUACCUGGGAA------AC 1un1F 1731 7131 1-15 11 CUGAUCUGGAUAAUGCCAG ||| || |||.|||| CUG----GG--AAUACCAG GGGUGCCCUUC------UGCGUGAAGGCUGAGAAAUACCCGUAUC |||.|.||||| ||.|.| ||||| || 31 GGGCGGCCUUCGGGCUAGACGGUGGGAG-AGGCU------UC 1f7yB 1132 7131 1-11 15 ACCUGAUCUGGAUAAUGCCA-GCGUAGGGAAGUUC ..|||.| ||| .||| ||.|.|| GGCUGGU------CCACCCGU---GACGCUC AAGGCUG-AGAAA---UACCCGUAUCACCUGAUCUGGAUAAUGCCAGCGU |||||.| .|||| ||..|| .||| |..||.|| 39 AAGGCGGCCGAAAGGCUAGACG------GUGG-----GAGAGGGU 1g1xJ 1531 1531 21-51 11 AG-GGAA .| |||| GGUGGAA

Appendix F

Table S5 Pairwise sequence alignment between SAM riboswitch with below sequence and similar rRNA PDB structures. GGCUUAUCAAGAGAGGUGGAGGGACUGGCCCGAUGAAACCCGGCAGCCAGAAAUGGUGCCAAUUCCUGCAGCGGAAACGU

UGAAAGAUGAGCCG

Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) GCUU |||| 1 1rngA 11131 131 2-1 12 GCUU GGUGGAGGGACUGGCCCGAUGAA-ACC |||| |||| ||| 3 1afxA 1135 1131 11-11 12 GGUG------UGAACACC GGCUUAUCAAGAGA-GGUG---GAGGGACUGGCCCGAUGAAA |||| ||| |||| ||||| ||| ||.|| 2 1g1xI 1131 2135 1-71 11 GGCU------AGACGGUGGGAGAGGG--UGG-----UGGAA GGACUGGC--CCGAUGAAACCCGG ||.|.||| ||| ||.|| 5 211dB 1132 1131 22-17 21 GGGCGGGCGACCG------CCUGG GCGGAAACGUU ||| .||||| 3 1uuuA 5235 1737 51-11 19 GCG--UACGUU GCAGCGGAA |.|.||||| 1 1wtsA 5531 131 11-51 11 GGACCGGAA GGCUUAUCAAGAGA-GGUG---GAGG----GACUGGCCCGAUGAAACCCG |||| ||| |||| |||| |.||||.|| ||||| 7 1kuqB 1131 2139 1-12 15 GGCU------AGACGGUGGGAGAGGCUUCGGCUGGUCC------ACCCG GUGCCAAUUCCUGCA |.|.|||||| ||| 1 1hs1A 5737 1135 11-51 17 GCGUCAAUUC--GCA GACUGGCCCGAU ||||||..||.| 9 1vopA 5131 1135 27-71 17 GACUGGGGCGGU GGC---AGCC ||| |||| 16 1zihA 5131 931 12-11 12 GGCGCAAGCC GACUGGCCCGAUGAAACCCGGCAGCC |.||||| || ..||.||||| 11 1i2yA 1535 1131 27-11 19 GGCUGGC----UG---UUCGCCAGCC ACCCGG---CAGCCAGAAAUGGUGC |||||| ||.||.| |||.| 13 1k3iA 1131 1531 71-19 27 ACCCGGGCUCAACCUG----GGUCC GAAACGUUG-AAAGAUGAGCC |.||||||| ||||.|..||| 12 1mt5A 5132 1539 51-97 21 GUAACGUUGAAAAGUUACGCC GGCCCGAUGAAACCCGGCAGCC ||||.||.||.||.|.|.|||| 15 1jurA 5235 1531 25-11 22 GGCCUGAGGAGACUCAGAAGCC GAUGAAACCCG---GCAGCC ||||...|||| |||.|| 13 1qwaA 1131 1137 72-11 21 GAUGCCUCCCGAGUGCAUCC GCU--UAUCAAGAGAGG--UGGAGGG ||| ||..|.|||||| .||||.| 11 512dA 1131 1537 2-27 25 GCUCCUAGUACGAGAGGACCGGAGUG GCUUAUCAAGAGAGGUGGAGGGACUGGCCCGAUGAAACCCGGCAGC |||.|..|.||||.| |||| |||.| 17 1q92C 7931 1132 2-15 25 GCUCAGUAGGAGACG------AACC--GCACC GGUGGAGGGACUGGCC ||||.|| |.||| 11 1pbrA 1231 1537 11-71 25 GGUGAAG----UCGCC UCCUGCAGCGGAAA |.||||..|||.|| 19 1bgzA 5131 1131 11-55 27 UACUGCUUCGGUAA GGCCCGAU------GAAAC--CCGGCAGCC ||||||.| ||.|| |||| ||| 36 1i1uD 1131 1131 25-11 75 GGCCCGGUAAGUCUCUUCGGAGAUACUGCCGG--GCC GGCCCGAUGAAACCCGGCAGCCAGAAAUGGUGCCAAUUCCUGCAGCGGAA ||.|||||| .|.||| ||.| |||| 31 GGACCGAUG------GUAGUG-----UCUU----CGGA- 1d1kB 1131 2132 25-91 75 ACGUUGAAAGAUGAG ||..|||..|| ----UGCGAGAGUAG AAGAGAGGUGGAGGGAC---UGGC-CCGAUGAAACCCGGCAGCCAGAAAU |.||.|||||| |.|.| .||| |||.|||||..| ||.|| 33 AGGAUAGGUGG-GAGCCGCAAGGCGCCGGUGAAAUAC--CACCC------1mzpB 1131 7137 9-11 11 GGUGCCAAUUCC |||| ------UUCC AGG-UGGAGGGACUGGCCCGAU-GAAACCCGGCAGCC------AGAA ||| ||.||| || | ||| |||||| |||| 32 AGGAUGUAGG--CU------UAGAA-----GCAGCCAUCAUUUAAAGAA 1qa1D 1131 2532 11-11 11 AUGGUGCCAAU---UC-CUG |..|| ||| || ||| AGCGU---AAUAGCUCACUG CUGGCCCGAUGAAACC--CGGCAGCCAGAAAUGGUGCCAAUUCCUGCAGC |.|||| |.||| ||| .||| |||| .|.|.|||| 35 CCGGCC-----ACACCUACGG-GGCC-----UGGU--UAGUACCUG---- 1un1F 1135 1131 21-52 11 GGAAACGUUGAAA---GAUGAGCCG ||||||.|.|.|| .|.|.|||| GGAAACCUGGGAAUACCAGGUGCCG GGCUUAUCAAGAGA-GGUG---GAGG----GACUGGCCCGAUGAAACCCG |||| ||| |||| |||| |.||||.|| ||||| 33 1f7yB 1131 2139 1-12 15 GGCU------AGACGGUGGGAGAGGCUUCGGCUGGUCC------ACCCG CCGAU---GAAACCCGGCAGCCAGAAAUGGUGCCAAUUCCUG---CAGCG ||||| ||| |||||||| |||| ||.|. 31 CCGAUGGCGAA----GGCAGCCA------CCUGGUCCACCC 1g1xJ 1137 7131 71-11 11 GAAACGUU |..|||.| GUGACGCU

Appendix G

Table S8 Pairwise sequence alignment between FMN riboswitch with below sequence and similar rRNA PDB structures. GGAUCUUCGGGGCAGGGUGAAAUUCCCGACCGGUGGUAUAGUCCACGAAUCCAU

Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) GGAUCUU ||..||| 1 1rngA 5131 1731 1-5 12 GGCGCUU GGGUGAA |.||||| 3 1afxA 1135 1132 11-21 12 GUGUGAA GGCAGGGUGAAAUUCCCGACCGGUGGUAUAG ||| ||..||||..|..|| ||||||.|.|| 2 1g1xI 1535 1131 11-11 11 GGC-GGCCGAAAGGCUAGA-CGGUGGGAGAG CGGGGCAGGGUGAAAUUCCCGACCG-GUGG |.||||.|| |||||| .||| 5 211dB 1737 2931 1-71 21 CUGGGCGGG------CGACCGCCUGG GUAUAGUCCACG ||.|.|| ||| 3 1uuuA 1135 2231 71-15 19 GUUUCGU--ACG GACCGGUGGUAUAGUCC |||||| |..|||| 1 1wtsA 1135 2131 21-11 11 GACCGG----AAGGUCC GGCAGGGUGAAAUUCC ||| ||||||||.|| 7 1fhkA 1132 2131 11-21 11 GGC--GGUGAAAUGCC AAUUCCC |||||.| 1 1hs1A 1135 1135 21-25 17 AAUUCGC GGGGCAGGGU ||||| ||| 9 1vopA 1131 2232 9-11 17 GGGGC--GGU GGGGCAGG ||.|||.| 16 1zihA 5131 1237 9-11 12 GGCGCAAG GGCAGGGUGAAAUUCCCGACC |||.||.|| |||.|.|.| 11 1i2yA 1139 2131 11-71 19 GGCUGGCUG---UUCGCCAGC GGAUCUUCGGGGCAGGGUGAAAUUCCCGACCGGUGGUAUAGUCC |||.| |||| |.|.||| ||| |||| 13 1k3iA 1131 7531 1-11 27 GGACC--CGGG------CUCAACC--UGG----GUCC GGC---AGGGUGAAA--UUCC ||| |.|.||||| ||.| 12 1mt5A 1139 2231 11-21 21 GGCGUAACGUUGAAAAGUUAC GACCGGUGGUAUAGUC-CACGAAUCC |.||.|.|| ||.| || |||.|| 15 1jurA 1131 2931 21-12 22 GGCCUGAGG---AGACUCA-GAAGCC UCCCGACCGGUGGUAUAGUCCACGAAUCC ||||| ||| |.|||| 13 1qwaA 1131 7132 21-12 21 UCCCG------AGU----GCAUCC GACCGGUGGUAUAGU |||||| ||| 11 512dA 1131 1935 21-12 25 GACCGG------AGU GGCAGGGUGAAAUUCCCGACCGGUGGUAUAGUCCA---CGAAUCCA |||| ||.|||| ||| ||||..|| 17 1mjiC 1137 2932 11-17 71 GGCA------CCUGACC------CCAUGCCGAACUCA GGUGGU--AUAGUCCACGAAUC ||||.| .|||...|||||.| 11 1q92C 1731 2531 72-11 25 GGUGCUCAGUAGGAGACGAACC AUCUUCGGGGCAGGGUGAAAUUCCC |.|||| |||||||.|..|| 19 1pbrA 1131 2137 7-25 25 ACCUUC------GGGUGAAGUCGCC GGAUCUUCGGGGCAGGGUGAAAUUCCC |.|.|||| |||||||.|..|| 36 1fyoA 1937 2135 1-25 25 GCACCUUC------GGGUGAAGUCGCC GGAU----CUUCGGGGCAGGGUGAAAUUCCC |||| ||||| |.||.|||| 31 1bgzA 1131 2531 1-25 27 GGAUACUGCUUCG------GUAAGUCCC UCUUCGG------GGCAGGGUGAAAUUCCC ||||||| .||.||| ||| 33 1i1uD 1131 2231 1-25 75 UCUUCGGAGAUACUGCCGGG------CCC GACCGGUGGUAUAGUCCACGAAUCC |||||.|||||..|||..||.||.| 32 1d1kB 5231 2131 21-12 75 GACCGAUGGUAGUGUCUUCGGAUGC GGAUCUUCGGG----GCAGGGUGAAAUUCCCGACCGGUGGUAUAGUCCAC ||||....||| |||.||.| ||||||..||| |||| 35 GGAUAGGUGGGAGCCGCAAGGCG------CCGGUGAAAUA--CCAC 1mzpB 1131 1131 1-12 11 GAAUCC ...||| CCUUCC GGAUCUUCGGGGCAGGGUGAAAUUCCC |.|.|||| |||||||.|..|| 33 1fypA 1937 2135 1-25 25 GCACCUUC------GGGUGAAGUCGCC GGAU-----CUUCGGGGCAG------GGUGAAAUUCCCGACCGGU |||| |||.|..|||| .|.|.|||..|..||.||| 31 1qa1D 1231 2931 1-71 11 GGAUGUAGGCUUAGAAGCAGCCAUCAUUUAAAGAAAGCGUAAUAGCUCACUGGU AUCUUCGGGGCAGGGUGAAAUUCCCG----ACCGGUGGUAUAGUCCACG |.||.||||||..||| .|.|.||.| |||.| ||.||| |||.| 37 1un1F 1132 7531 7-15 11 ACCUACGGGGCCUGGU-UAGUACCUGGGAAACCUG-GGAAUA--CCAGG CUUCGGGGCAGGGUGAAAUUCCCGACCGGUGGUA------UAGUCCAC |||||||..|| .||||||.| |.|||||| 31 1f7yB 1731 7131 1-11 15 CUUCGGGCUAG------ACGGUGGGAGAGGCUUCGGCUGGUCCAC GAUCUUCGG----GGCAGGGUGAAAUUCCCGACCGGUGGUAUAGUCCAC ||| || ||||| |.||| |.|||||| 39 1g1xJ 1239 7137 2-11 11 GAU----GGCGAAGGCAG------CCACC------UGGUCCAC CUUCGGGGCAGGGUGAAAUUCCCGACCGGUGGUA------UAGUCCAC |||||||..|| .||||||.| |.|||||| 26 1kuqB 1731 7135 1-11 15 CUUCGGGCUAG------ACGGUGGGAGAGGCUUCGGCUGGUCCAC

Appendix H

Table S2 Pairwise sequence alignment between c-di-GMP I riboswitch with below sequence and similar rRNA PDB structures CACGCACAGGGCAAACCAUUCGAAAGAGUGGGACGCAAAGCCUCCGGCCUAAACGGCAUUGCACUCCGCCGUAGGUAGCG GGGUUACCGAUGG

Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) GGCCUAAACGGCAUUGC ||| || |||| 1 1rngA 1239 1131 11-12 12 GGC------GC-UUGC UGCACUCC ||.||.|| 3 1afxA 5131 931 11-15 12 UGAACACC CGGCCUAAACGGC-AUUGCACUCCGCCGUAGGUAGCGGGGUUACCGAUGG |||||.||| ||| | |.||..||.||.||| ||| 2 1g1xI 1131 2137 11-97 11 CGGCCGAAA-GGCUA------GACGGUGGGAGAGGG------UGG GCGGGGUUACCGA-----UGG ||||| ||| ||| 5 211dB 1231 1137 51-97 21 GCGGG-----CGACCGCCUGG GCCGUAGGUAGCGGGGUUAC |.||||.||..|| ||| 3 1uuuA 1131 1131 11-15 19 GGCGUACGUUUCG----UAC GGA-CGCAAAGCCUCC ||| ||.||.| ||| 1 1wtsA 1131 1135 71-11 11 GGACCGGAAGG--UCC GGCAUUGCACUCCGCC |||..||.|.| ||| 7 1fhkA 1231 1131 11-51 11 GGCGGUGAAAU--GCC CCAUUCGAA |.|||||.| 1 1hs1A 5531 1131 11-21 17 CAAUUCGCA GAGUGGGACG ||.||||.|| 9 1vopA 1131 931 21-71 17 GACUGGGGCG GGACGCAAAGCCU ||.||| |||||| 16 1zihA 1131 1131 71-17 12 GGGCGC-AAGCCU GGCAUUGCACUCCGCCGUAGGUAGC ||| |.||..|.|||| ||| 11 1i2yA 1131 1131 11-59 19 GGC-UGGCUGUUCGCC------AGC GGACGCAAAGCCUCCGGCCUAAAC---GGCAUUGCACUCC |||| ||||.||.||| || ||| 13 1k3iA 1131 1131 71-15 27 GGAC------CCGGGCUCAACCUGGG------UCC GGCAAACCAUUCGAAAGAGUGGGACGC |||..|.|.|| |||| ||| .|||| 12 1mt5A 1135 1135 11-71 21 GGCGUAACGUU-GAAA-AGU--UACGC GAGUGGGACGCA-AAGCC ||| |.|||.|| ||||| 15 1jurA 5531 1131 21-12 22 GAG-GAGACUCAGAAGCC AAGCCUCCGGCCUAAACGGCAUUGCACUCC |.||||||.| |.|||| ||| 13 1qwaA 1737 1931 71-15 21 AUGCCUCCCG------AGUGCA-UCC CUCCGCCGUAGGUA-GCGGGGUUACCG |||| || ||| |.|.|| |||| 11 512dA 1731 2139 11-19 25 CUCC----UA-GUACGAGAGG--ACCG GCACAGGGCAAACCAU-UCGAA---AG-AGUG ||||..| |..|||| .|||| || |||| 17 1mjiC 1231 2131 1-71 71 GCACCUG--ACCCCAUGCCGAACUCAGAAGUG AGU--GGGACGCAAAGCCUCCGGCCUAAACGGCA ||| |.|||| |||.||| 11 1q92C 1132 1531 25-11 25 AGUAGGAGACG------AACCGCA GGCAAACCAUUCGAAAGAGUGGGACGCAAAGCCUCC |.||.||| ||| |.||| |||.|.|| 19 1pbrA 1231 2131 11-11 25 GUCACACC-UUC----GGGUG------AAGUCGCC GGGACGCAAAGCCUCCGGCCUAAACGGCAUUGCACUCCGCC |.|.|||| |||.||| .||.|.| |||| 36 1fyoA 1131 2237 71-51 25 GCGUCGCA---CCUUCGG------GUGAAGU-CGCC GGCAU--UGCACUCCGCCGUAGGU ||.|| || ||.|| |||.|| 31 1bgzA 1231 1131 11-51 27 GGGAUACUG--CUUCG--GUAAGU GGCCUAAACGGCAUUGCACUC-CGCCGUAGGUA--GCGGGG |||| |||.| |.|| |..||.||.|| ||.||| 33 1i1uD 1131 2531 11-17 75 GGCC----CGGUA----AGUCUCUUCGGAGAUACUGCCGGG UUCGAA----AGAGUGGGAC ||||.| |||||.||.| 32 1d1kB 1131 137 19-71 75 UUCGGAUGCGAGAGUAGGUC GGGCAAACCAUUCGAAAGA---GUGGGA--CGCAAAGCCUCCGGCCUAAA ||| ||.||.|.|| |||||| ||| |||.|.|||| |.|| 35 GGG------AUGCGUAGGAUAGGUGGGAGCCGC-AAGGCGCCGG--UGAA 1mzpB 1135 7531 9-15 11 CGGCAUUGCAC---UCC ||..||| ||| ----AUACCACCCUUCC CAGG------GCAAACCAU--UCGAAAGAGUGGGACGCAAAG |||| || |.|||| |..|||| |||| 31 CAGGAUGUAGGCUUAGAAGC-AGCCAUCAUUUAAAG------AAAG 1qa1D 7131 7231 5-51 11 CCUCCGGCCUAAACGGCAUUGCACUCCGCCGUAGGU |.| ||..|| .|||| ||| CGU------AAUAGC---UCACU------GGU CCGGCCUAA---ACGGCAUUGCACUCCGCC--GUAGGUAGCGGGGUUACC ||||||..| |||| .||| ||..|||.|.|||..||| 37 CCGGCCACACCUACGG------GGCCUGGUUAGUACCUGGGAAACC 1un1F 1235 2131 11-97 11 GAUGG ||| --UGG GGACGCAAAGCCUCCGGCCUAAACGGCAUUGCA----CUCCGCCGUAGGU ||.|| ||||.|||.|||.|||| |.|.| ||.|| 31 GGGCG----GCCUUCGGGCUAGACGG--UGGGAGAGGCUUCG------1f7yB 1131 2531 71-92 15 AGCGGGGUUACCGAUG ||.||...|||..|| -GCUGGUCCACCCGUG ACGCACAGGGCAAACCAUUCGAAAGAGUGGGACGCAAAGCCUCC-GGCCU ||||..|.|||.|| || .||||.|| || 39 ACGCCGAUGGCGAA------GG-----CAGCCACCUGG--- 1g1xJ 7139 2131 2-52 11 AAACGGCAUUGCACUCCGCCGU |.||| |||| ------UCCAC----CCGU

Appendix I

Table S6 Pairwise sequence alignment between c-di-GMP II riboswitch with below sequence and similar rRNA PDB structures GCGCGGAAACAAUGAUGAAUGGGUUUAAAUUGGGCACUUGACUCAUUUUGAGUUAGUAGUGCAACCGACCGUGCU

Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) GGCACUUGACUC |||.||||..|| 1 1rngA 5131 1231 77-11 12 GGCGCUUGCGUC GGUUUAAAUUGGGCAC |||.|.|| ||| 3 1afxA 1132 1737 22-75 12 GGUGUGAA-----CAC GAAUGGGUUUAAAUUGGGCACUUGACUCAUUUUGAGUUAGUAGUGCAA |||.||.|..|...||| |||...||.|||.|| 2 1g1xI 1135 7131 15-11 11 GAAAGGCUAGACGGUGG------GAGAGGGUGGUGGAA UGGGCACUUGACUCAUUUUGAGUUAGUAGUGCAACCGACCG ||||| |.||.||||.|.| 5 211dB 7131 1135 71-51 21 UGGGC------GGGCGACCGCCUG UUAGUA ||.||| 3 1uuuA 1737 1132 17-11 19 UUCGUA CGGAA ||||| 1 1wtsA 11131 1135 1-1 11 CGGAA GCGGAAACAAUGAUGAAUG |||| ||| |||| 7 1fhkA 1539 1131 7-21 11 GCGG------UGA--AAUG UAAAUUGG |.||||.| 1 1hs1A 5131 1231 21-77 17 UCAAUUCG GAAUGGG ||.|||| 9 1vopA 1135 1231 15-27 17 GACUGGG GGGCACUUGACU ||||.|..|.|| 16 1zihA 1135 1135 72-17 12 GGGCGCAAGCCU GGCACUUGACUCAUUUUGAGUUAGUAGUGCAACC ||| ||.|| |||.| .||.|| 11 1i2yA 1131 2131 77-11 19 GGC---UGGCU------GUUCG----CCAGCC GGGUUUAAAUUGGG |||.|.||..|||| 13 1k3iA 5131 2131 21-71 27 GGGCUCAACCUGGG AAUGAUGAAUGGGUUUAAAUUGGGC ||.|.|||| ||.||..|| 12 1mt5A 1231 2137 11-71 21 AACGUUGAA------AAGUUACGC UGGGCACUUGACUCA ||.|.| |||||| 15 1jurA 1135 1931 71-11 22 UGAGGA---GACUCA AGUGCAACC ||||||.|| 13 1qwaA 1139 2131 11-11 21 AGUGCAUCC UAGU---AGUGCAACCGACCG |||| ||.| ||||| 11 512dA 1531 2135 11-51 25 UAGUACGAGAG-----GACCG GGCACUUGA-CUCAUUUUG-AGUUAGUAGUGC |||||.||| |.|||...| |.|.||.||||| 17 1mjiC 1131 7132 77-12 71 GGCACCUGACCCCAUGCCGAACUCAGAAGUGC CUCAUUUUGAGUUAGUAGUGCAACCG-ACC |||| .|||.||...||||| ||| 11 1q92C 1135 2531 12-51 25 CUCA------GUAGGAGACGAACCGCACC CUUGACUCAUU---UUGAGUUAGUAG |.|.||.|.|| .||| |||.| 19 1pbrA 1731 2131 75-19 25 CGUCACACCUUCGGGUGA---AGUCG GCACUUGACUCAUUUUGAGUUAGUAGU |||| .||.|.|| |.||| 36 1fyoA 1131 2131 71-11 25 GCAC------CUUCGGGU--GAAGU GGAAACAAUGAUGAAUGGGUUUAAAU |||.|| ||..|.|| |||.| 31 1bgzA 1731 2137 1-71 27 GGAUAC-----UGCUUCGG--UAAGU GCGCGGAAACAAUGAUGAAUGGGUUUAAAUUGGGCACUUGACUCAUUUUG ||.|||.|| |.||| ||..| 33 GCCCGGUAA------GUCUC-UUCGG 1i1uD 7231 7131 1-51 75 AGUUAGUAGUGCAACCGACCGUGC ||.|| |..|||.|| AGAUA------CUGCCGGGC AUGG--GUUUAAAUUGGGCACUU--GACUCAUUUUGAGUUAGUAG |||| ||.| ||| || ||.|..||||| 32 1d1kB 1131 7737 19-19 75 AUGGUAGUGU------CUUCGGA------UGCGAGAGUAG GAUGAAUGGGUUUAAAUUGGGCACUUGACUCAUUUUGAGUUAGUAG-UGC ||||..|.||.| ||.|.|.|| .|| 35 GAUGCGUAGGAU------AGGUGGGAGCCGC 1mzpB 7135 7131 11-57 11 AA-CCGACCGUG || .||.|.||| AAGGCGCCGGUG GCACUUGACUCAUUUUGAGUUAGUAGU |||| .||.|.|| |.||| 33 1fypA 1131 2131 71-11 25 GCAC------CUUCGGGU--GAAGU CAAUGAUGAAUGGGUUUA-AAUUGGGCACUUGAC-UCAUUUUGAG----- |.|.|||| |.||.||| || ||| |.| ||||||..|| 31 CCAGGAUG--UAGGCUUAGAA----GCA---GCCAUCAUUUAAAGAAAGC 1qa1D 1135 7931 11-17 11 UUAGUAGUGCA .||.|||..|| GUAAUAGCUCA GGGCACUUGACUCAUUUUGAGUUAGUA---GUGCAACC------GACC--GUGC |||| || .||||||| |.|.|||| .||| |||| 37 1un1F 1131 1131 72-51 11 GGGC------CU------GGUUAGUACCUGGGAAACCUGGGAAUACCAGGUGC UUGGGCACUUGACUCAUUUUGAG----UUAG--UAGUGCAACC--GACCGUGCU |.||| ||.|| |..|..||| ||.| |.||.||.|| ||| ||| 31 1f7yB 1735 7131 71-51 15 UCGGG--CUAGA--CGGUGGGAGAGGCUUCGGCUGGUCCACCCGUGAC---GCU CGCGGAAACAAUGAUGAAUGGGUUUAAAUUGGG----CACUUGACUCAUU ||| |.|||..||| || |||.|| 39 CGC-----CGAUGGCGAA------GGCAGCCACCUG------1g1xJ 7132 7139 2-12 11 UUGAGUUAGUAGUGCAACC--GACCGUGCU ||.||.|| ||| ||| ------GUCCACCCGUGAC---GCU

Appendix J

Table S4 Pairwise sequence alignment between THF riboswitch with below sequence and similar rRNA PDB structures GGACAGAGUAGGUAAACGUGCGUAAAGUGCCUGAGGGACGGGGAGUUGUCCUCAGGACGAACACCGAAAGGUGGCGGUAC Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) CGUGCGU |.||||| 1 1rngA 1135 939 15-27 12 CUUGCGU GUUGUCCUCAGGACGAACACC |.||| ||||||| 3 1afxA 1231 1139 11-11 12 GGUGU------GAACACC AGGUAAACGUGCGUAAAGUGCCUGAGGGACGGGGAGUUGUCCUCAGGACG ||| || ||..|||| ||..||.||..||.||| 2 AGG----CG-GCCGAAAG-GCUAGACGGUGGGAGAG------1g1xI 1239 2131 11-59 11 AACACCGAAAGGUGGCGGUA |||||.||.| ------GGUGGUGGAA GGCGGUACGUUUACCGCAUCUCGCUG ||||| .|| ||||| ||| 5 211dB 1535 1131 57-91 21 GGCGG-GCG---ACCGC------CUG GGCGGUACGUUUACCGCAUCUCGC ||| |||||||| ||.| ||| 3 1uuuA 5131 1131 57-91 19 GGC-GUACGUUU--CGUA---CGC GGACGGGGAGUUGUCC ||||.||.|| |||| 1 1wtsA 5131 1139 71-11 11 GGACCGGAAG--GUCC GUAAAGUGCC ||.||.|||| 7 1fhkA 1131 1139 22-71 11 GUGAAAUGCC GCGUAAAGU ||||.||.| 1 1hs1A 5531 131 21-21 17 GCGUCAAUU GAAAGGUGGCGGU ||..|| |||||| 9 1vopA 5139 1131 11-51 17 GACUGG-GGCGGU GUGCGUAAAGUGCCU |.|||.|| |||| 16 1zihA 1135 939 11-72 12 GGGCGCAA---GCCU CUCGCUGUUC ||.||||||| 11 1i2yA 9131 1132 92-111 19 CUGGCUGUUC GGA--CGGGGAGUUGUCCUCAGGACGAACACCGAAAGGUGGCGGUAC ||| |||| |||| ||| || |||.| 13 1k3iA 1231 1931 71-11 27 GGACCCGGG------CUCA------ACC------UG--GGUCC AACGUGCGUAAAGU---GCC ||||| .|.||||| ||| 12 1mt5A 5131 1531 11-71 21 AACGU-UGAAAAGUUACGCC GCCUGAGG-GACGGGGAGUUGUCCUCAGGA |||||||| || |||||.| 15 1jurA 1737 1137 29-15 22 GCCUGAGGAGA------CUCAGAA GGAGUUGUCCUCAGGACGAACACCGAAAGGUGGCGGUACGUUUACCGCAUC ||| || |||| |||| || ||||| 13 1qwaA 7932 1931 12-92 21 GGA--UG-CCUC------CCGA------GU-----GCAUC UCCUCAGGAC--GAACACCGAAAGGUG |||| ||.|| ||..||||.| ||| 11 512dA 1135 1932 19-57 25 UCCU-AGUACGAGAGGACCGGA--GUG GGGAGUUGUCCUCAGGACGAACACCGAAAGGUGGC ||.|..||.||.||.|.|||||.|.||| |||.| 17 1mjiC 1135 2231 11-51 71 GGCACCUGACCCCAUGCCGAACUCAGAA--GUGCC GUCCUCAG-----GACGAA---CACC ||.||||| |||||| |||| 11 1q92C 1131 1131 11-11 25 GUGCUCAGUAGGAGACGAACCGCACC AAACGU--GCGUAAAGU-GCC |.||.| |.||.|||| ||| 19 1pbrA 1135 1731 11-71 25 ACACCUUCGGGUGAAGUCGCC CGU------GCGUAAAGU-GCC ||| |.||.|||| ||| 36 1fyoA 1231 1731 15-71 25 CGUCGCACCUUCGGGUGAAGUCGCC GGUAAACGUGC----GUAAAGUGCC ||.|.|| ||| || ||||.|| 31 1bgzA 1131 1132 11-71 27 GGGAUAC-UGCUUCGGU-AAGUCCC GGGACGGGGAGUUGUCCUCAGGACGA-ACACCGAAAGGUGGCGG |||.|.||.|..|.||.|| ||| || || ||.||| 33 1i1uD 1131 2131 71-55 75 GGGCCCGGUAAGUCUCUUC-GGA-GAUAC------UGCCGG GGAC---GGGGAGUUGUCCUCAGGACGAACACCGAAAGGUGGCGGUACGU |||| .||.|| ||||.|| |||.| |||.| |||.|| 32 1d1kB 1131 2131 71-12 75 GGACCGAUGGUAG-UGUCUUC-GGAUG-----CGAGA------GUAGGU UGCGUA----AAGUGCCUGAGGGACGGGGAGUUGUCCUCAGGACGAACAC |||||| |.|| ||||| | 35 UGCGUAGGAUAGGU------GGGAG------C 1mzpB 7931 7131 19-15 11 CGAAAGGUGGCGGU------ACGUUUACC ||.||||.|.|||| ||..||.|| CGCAAGGCGCCGGUGAAAUACCACCCUUCCC CGU------GCGUAAAGU-GCC ||| |.||.|||| ||| 33 1fypA 1231 1731 15-71 25 CGUCGCACCUUCGGGUGAAGUCGCC GACAG--AGUAGGUAAACGUGCGUA-AAGUGCCUGAGGGACGGGGAGUUG |.||| .|||| ||.|| ||| 31 GCCAGGAUGUAG------GCUUAGAAG------1qa1D 7131 7535 2-111 11 UCCUCAGGACGAACACCGAAAGGUGGCGGUACGUUUACCGCAU--CUCGCUGUU ||| |.|.||...||||...||| || || |||.|||.| ----CAG--CCAUCAUUUAAAGAAAGCG------UA----AUAGCUCACUGGU CCUGAGGGACGGGGAGUUGU-----CCU------CAGGACGAACACCGA ||| ||||||..|.|| ||| |.|| |||.||| 37 CCU-----ACGGGGCCUGGUUAGUACCUGGGAAACCUGG--GAAUACC-- 1un1F 1737 2131 71-55 11 AAGGUGGCGG |||||.||| -AGGUGCCGG GUGCGUAAAGUGCCUGAGG----GACG--GGGAGUUG--UCCUCAGGACG |.||| ||||..|| |||| |||||..| ||..|.||.| 31 GGGCG------GCCUUCGGGCUAGACGGUGGGAGAGGCUUCGGCUGGUC- 1f7yB 1232 7139 11-51 15 AACACCGAAAGGUGGCGGU |||| .|||.||.| --CACC----CGUGACGCU ACACCGA-----AAGGUGGC-----GGUACGUUUACCGCAUCUCGCUGU ||.|||| ||||..|| |||.| ||| |.|..||||.| 39 1g1xJ 1731 2731 11-99 11 ACGCCGAUGGCGAAGGCAGCCACCUGGUCC----ACC-CGUGACGCUUU GUUUACCGCAUCUCGCUGUUC

Appendix K

Table S11 Pairwise sequence alignment between preQ1 riboswitch with below sequence and similar rRNA PDB structures AGAGGUUCUAGCUACACCCUCUAUAAAAAACUAA

Identity (%) Similar region in the riboswitch Number PDB code Length (bp) Local alignment local global (local alignment) GGUUCUAGC ||..||.|| 1 1rngA 1135 2131 1-12 12 GGCGCUUGC ACACC ||||| 3 1afxA 11131 2131 11-11 12 ACACC AGAGGUU |||||.| 2 1g1xI 1135 1731 1-5 11 AGAGGGU GCUACACCCU ||.||..||| 5 211dB 5131 2531 11-21 21 GCGACCGCCU AGGUUCUAGCUACACC |.||| |.| |||.|| 3 1uuuA 1131 2131 7-11 19 ACGUU-UCG-UACGCC AGGUUC ||||.| 1 1wtsA 1737 1131 7-1 11 AGGUCC GAGGU |.||| 7 1fhkA 1131 2731 2-1 11 GCGGU CCUCUAU |.||.|| 1 1hs1A 5131 1131 11-21 17 CGUCAAU GAGGU |.||| 9 1vopA 1131 1131 2-1 17 GCGGU GGUUCUAGC ||..|.||| 16 1zihA 1135 2731 1-12 12 GGCGCAAGC GUUCUAGCUACACCC |||| || ||.|| 11 1i2yA 1135 2132 1-19 19 GUUC--GC--CAGCC GCUACACCCU ||| ||.||| 13 1k3iA 1131 7131 11-21 27 GCU-CAACCU GGCGUAACGUUG |||..|||.|.| 12 1mt5A 1135 1931 1-12 21 GGCUCAACCUGG GGCGUAACG---UUGAAAAG |||.|.|.| .|.|.||| 15 1jurA 1131 1232 1-15 22 GGCCUGAGGAGACUCAGAAG GCCUCCCGAG |||| ||| 13 1qwaA 5131 2531 1-11 21 GCCU---GAG GGAUGCCUCCCGAGUG ||| ||.||||| 11 512dA 1231 1131 1-11 25 GGA-----CCGGAGUG AUGC----CUCCCGAGUGC |||| |||...||||| 17 1mjiC 1732 7231 7-15 71 AUGCCGAACUCAGAAGUGC GAUGCCUCCCGAGU------GCAUC |.|| ||| ||| |||.| 11 1q92C 1139 2535 2-21 25 GGUG-CUC---AGUAGGAGACGAACCGCACC CCUCCCGAGUGCA ||| .||.|||.| 19 1pbrA 1932 7131 1-11 25 CCU-UCGGGUGAA CCUCCCGAGUGCA ||| .||.|||.| 36 1fyoA 1932 1735 1-11 25 CCU-UCGGGUGAA GGAUGCCUCCCGAGUGCAUC |||| |.|||.|| 31 1bgzA 1131 7131 1-21 27 GGAU------ACUGCUUC GGAUGCCUC--CCGAG----UGC |.|.|.||| |.||| ||| 33 1i1uD 1131 2732 1-15 75 GUAAGUCUCUUCGGAGAUACUGC GGAUGCCUCCCGAGUGCA ||||| ||||.|.| 32 1d1kB 1131 1131 1-11 75 GGAUG-----CGAGAGUA GGAU------GCCUC------CCGAGUG------CAUCC |||| |||.| ||| ||| ||.|| 35 1mzpB 1739 2131 1-21 11 GGAUAGGUGGGAGCCGCAAGGCGCCG-GUGAAAUACCACCC CCUCCCGAGUGCA ||| .||.|||.| 33 1fypA 1932 1735 1-11 25 CCU-UCGGGUGAA GGAUG ||||| 31 1qa1D 11131 7737 1-1 11 GGAUG GGAUGCCU------CCCGAGUGC |||..||| ||.| |||| 37 1un1F 1132 2131 1-15 11 GGAAACCUGGGAAUACCAG-GUGC GGAUG--CCUCCCGAG ||.|| ||.||||.| 31 1f7yB 1131 2739 1-11 15 GGCUGGUCCACCCGUG GGAUGCCUCCCGAGUGCAUCC ||..|||.||.| ||.||.|| 39 1g1xJ 1135 2531 1-21 11 GGCAGCCACCUG-GUCCACCC

Abstract in Farsi L

چکیده فارسی مقدمه:در تعداد زيادي از mRNA هاي پروکاريوتي قطعاتي در قسمت غير کد کننده به اها رايبوسوييچها وجود داراد ک با اتصال ب متابوليت خاصي بيان ژن پايين دسهت را تنظهي مهي کنند. رايبوسوييچها مهي تواانهد ههد هاي جديهدي بهراي طراحهي آاتهي بيوتيهه هها باشهند . هدف:بنابراين در اين مطالعه بهرآن شهدي تها سهاختار ضهايي برخهي رايبوسهوييا هها را بها RNAهاي ريبوزومي مقايس کرده و اتصال احتمالي رايبوسوييا ها را با آمينوگليکوزيدها مورد مطالع قرار دهي . از طرف ديگر توالي هاي جديد رايبوسوييا ب شکل بالقوه مي توااند ب عنهوان عناصهر کنتهرل بيان در بيوتکنولوژي مورد استفاده قهرار بگيراهد. از ايهن رو در بخه ديگهري از ايهن مطالعه رايبوسوييا TPP در باکتري به تهازگي شهناخت شهده Alishewanella tabrizica مهورد مطالع قرار گر ت. روش ها: مقايس ساختاري بر روي ساختارهاي RNA از طريق ه رديفهي ههاي سهاختاري و عملکردي ااجا گر ت و سپس احتمال اتصال آمينوگليکوزيهدها به رايبوسهوييا هها از طريهق مطالعات داکينگ و شبي سازي ديناميه مولکولي مورد بررسي قرار گر ت. در بخه مربهوب به شناسهايي رايبوسهوييا TPP از روش ههاي In-line probing و SPR استفاده گرديد. نتایج: در مطالعات ساختاري تشاب ضايي بين رايبوسوييا ها و RNA هاي ريبوزومي مشاهده شد. سپس اشان داده شد ک تمايل رايبوسوييا ها ب آمينوگليکوزيدها ب طور قابهل مححظه اي با تمايل rRNA A site ب آمينوگليکوزيدها برابري يا ا زواي مي کند. پس از تاييد عملکرد رايبوسوييا TPP مورد مطالع ، تمايل TPP ب اين رايبوسوييا بها تمايهل آن ب ديگر رايبوسوييا هاي TPP مستخرج از باکتري هاي ديگر مهورد مطالعه قهرار گر هت و رايبوسوييا TPP از A. tabrizica باالترين تمايل را با KD=2-4nM اشان داد. اتيجهه گيههري در مجمههور رايبوسههوييا ههها احتمههاال بهه عنههوان اهههداف جديههد ترکيبههات آمينوگليکوزيدي مي توااند مورد مطالعات بيشتري قرار بگيراد . همچنين رايبوسوييا TPP مستخرج از A. Tabrizica مي توااد ب عنوان يه قطع تنظيمهي قدرتمند مورد استفاده قرار گيرد. واژه های اختصاصی رايبوسوييا، A. tabrizica، آمينوگليکوزيد ها، اارژي اتصال، داکينگ

دااشگاه علو پزشکي تبريز دااشکده داروسازي

پايان اام جهت دريا ت دکتراي تخصصي بيوتکنولوژي دارويي

عنوان کلونینگ و تعیین توالی برخی رایبوسوییچ های Alishewanella tabrizica strain RCRI4 و مطالعه مقایسه ای آن ها با رایبوسوییچ های متناظر در سایر باکتری ها

اگارش الناز مهدي زاده اقد

اساتيد راهنما دکتر محمد سعيد حجازي دکتر ابولفضل برزگر

استاد مشاور دکتر يورگ هارتيگ

شهريور 5931 شماره پايان اام 08