Small RNA Sibling Pairs Ryfa and Ryfb in Shigella Dysenteriae and Their Impact On

Small RNA Sibling Pairs RyfA and RyfB in Shigella dysenteriae and their Impact on

Pathogenesis

A dissertation presented to

the faculty of

the College of Arts and Sciences of Ohio University

In partial fulfillment

of the requirements for the degree

Doctor of Philosophy

Megan E. Fris

August 2018

This dissertation titled

Small RNA Sibling Pairs RyfA and RyfB in Shigella dysenteriae and their Impact on

Pathogenesis

MEGAN E. FRIS

has been approved for

the Department of Biological Sciences

and the College of Arts and Sciences by

Erin R. Murphy

Associate Professor of Bacteriology

Joseph Shields

Interim Dean, College of Arts and Sciences 3

ABSTRACT

FRIS, MEGAN E., Ph.D., August 2018, Biological Sciences

Small RNA Sibling Pairs RyfA and RyfB in Shigella dysenteriae and their Impact on

Pathogenesis

Director of Dissertation: Erin R. Murphy

Understanding molecular mechanisms which regulate bacterial virulence is essential for discovering new therapeutics. Alarmingly, some bacterial species are already unresponsive to antibiotic cocktails, and in the case of Shigella dysenteriae, will up- regulate virulence factors under the duress of antibiotic treatment. While molecular mechanisms involving the regulation of S. dysenteriae pathogenesis have been studied, many are still unknown, especially those of which are mediated by RNA. Ribo-regulators that have been characterized in Shigella, and published to date, have all been shown to impact virulence processes. Many other sRNAs have been discovered in Shigella and E. coli, yet, remain uncharacterized. One sRNA of interest, RyfA, was originally discovered in Escherichia and found to have nearly exact sequence identity to that of RyfA in

Shigella flexneri. Interestingly, in S. dysenteriae, 95% identical twin copies of RyfA, designated RyfA1 and RyfA2 exist in tandem. Intriguingly, ryfA1-like copies of the gene are only found in pathogenic species of Escherichia and all species of Shigella. Non- pathogenic species of Escherichia contain a ryfA2-like copy. Upstream of both RyfA1 and RyfA2, additional sRNAs exist, termed RyfB1 and RyfB2. While RyfB1 and RyfB2 are only 60% identical to each other they each share large amounts of complementarity to

RyfA1 and RyfA2 respectively, suggesting that these molecules may interact with each 4 other and thus one sRNA may regulate the abundance and/or activity of the other. Indeed, overexpression of RyfB1 leads to reduced levels of RyfA1 in the bacterial cell but has no effect on the levels of RyfA2. Unexpectedly, it was it was found that RyfB2 alters the abundance of both RyfA1 and RyfA2 when produced in S. dysenteriae. Furthermore, overproduction of RyfA1 slightly inhibits RyfA2 levels, while RyfA2 overproduction strongly inhibits RyfA1 levels. In summary, sRNAs RyfA2, RyfB2, and RyfB1 all function to include tight control over RyfA1. The overproduction of RyfA1 results in a lack of cell to cell spread in S. dysenteriae, suggesting the number of sRNAs regulating

RyfA1 is critical for optimal regulation. Lastly, overproduction of RyfB2 in S. dysenteriae results in increased plaque number. The complex interplay between these sRNAs is likely important for fine-tuned regulation of invasion and virulence in S. dysenteriae. Downstream targets of RyfA1 and RyfA2 may shed some light on genes/proteins impacted resulting in the above described phenotypes.

DEDICATION

Dedicated to Matt, my sister Sarah, and my parents Anna, and Will for encouraging hard

work and not sweating the small stuff.

ACKNOWLEDGMENTS

Special thanks to Michelle Pate, for the technical help, assistance, and unmatched organizational skills. Thank you to Erin Murphy for her invaluable mentorship. I would also like to thank my committee members Ronan Carroll, Peter Coschigano, and Sarah

Wyatt for the useful discussion, feedback, and occasional technical assistance.

Additionally, a thank you to the NIH for financial support and Ohio University for laboratory space and equipment.

TABLE OF CONTENTS

Page

Abstract ...... 3 Dedication ...... 5 Acknowledgments...... 6 List of Tables ...... 9 List of Figures ...... 10 Chapter 1: Riboregulators: Fine-Tuning Virulence in Shigella ...... 12 1.1 Abstract ...... 12 1.2 Introduction ...... 12 1.2.1 CsrB and CsrC ...... 16 1.2.2 RyhB ...... 18 1.2.3 RnaG ...... 20 1.2.4 ShuA ...... 21 1.3 Discussion ...... 23 Chapter 2: Sibling sRNA RyfA1 Influences Shigella dysenteriae Pathogenesis ...... 25 2.1 Introduction ...... 25 2.2 Materials and Methods ...... 28 2.3 Results ...... 37 2.3.1 Twin copies of RyfA are produced by Shigella dysenteriae ...... 37 2.3.2 The presence of a ryfA1-like gene is associated with pathogenicity ...... 42 2.3.3 RyfA1 impacts virulence in Shigella dysenteriae ...... 44 2.3.4 Overproduction of RyfA1 leads to inhibition of cell-to-cell spread by Shigella dysenteriae ...... 47 2.3.5 RyfA1 overproduction results in elimination of ompC, a transcript encoding a major outer membrane protein ...... 49 2.3.6 Regulation of RyfA1 by RyfB1 ...... 50 2.3.7 RyfA1 does not encode a small protein under conditions tested ...... 53 2.3.8 RyfB1 Overproduction does not influence plaque formation or ompC ...... 56 2.4 Discussion ...... 57 Chapter 3: RyfA2 and RyfB2 ...... 62 8

3.1 Abstract ...... 62 3.2 Methods ...... 62 3.3 Introduction ...... 63 3.3 Results ...... 64 3.3.1 RyfA2 overproduction does not affect plaquing of Shigella dysenteriae ...... 64 3.3.2 RyfB2 overproduction increases plaquing ability of Shigella dysenteriae ...... 66 3.3.3 RyfB2 inhibits RyfA2 levels ...... 68 3.3.4 Swapping the RyfB2 putative interaction site ...... 70 3.3.5 RyfB2 influences RyfA1 levels ...... 73 3.3.6 Other sRNA interactions ...... 75 3.4 Discussion ...... 77 Chapter 4: Downstream Targets of RyfA1 and RyfA2 ...... 81 4.1 Abstract ...... 81 4.2 Material and Methods ...... 81 4.3 Introduction ...... 83 4.4 Results ...... 85 4.4.1 Transcriptome of RyfA1 and RyfA2 ...... 85 4.4.2 Proteome of RyfA1 and RyfA2 ...... 89 4.5 Conclusions and Discussion ...... 99 Chapter 5: Conclusions and General Discussion ...... 105 References ...... 107 Appendix: Supplemental Figures and Tables ...... 123

LIST OF TABLES

Page

Table 1: Summary of riboregulators in Shigella ...... 16

Table 2: Downstream targets of RyfA1 overexpression through RNA-seq analysis ...... 87

Table 3: Downstream targets of RyfA2 overexpression through RNA-seq analysis ...... 88

Table 4: Transcriptomic data of RyfA1 and RyfA2 compared ...... 89

Table 5: Proteins significantly impacted by RyfA1 overexpression ...... 91

Table 6: Membrane proteins associated uniquely with RyfA1 overproduction ...... 93

Table 7: Proteins significantly impacted by RyfA2 overexpression ...... 94

Table 8: RyfA2 specific iron/heme regulation proteins ...... 95

Table 9: RyfA1 and RyfA2 overexpression protein level comparison ...... 96

Table 10: Transcript/protein levels significantly impacted by RyfA1 or RyfA2 overproduction ...... 98

LIST OF FIGURES

Page

Figure 1: sRNA network involved with Shigella virulence ...... 15

Figure 2: RyfA1 and RyfA2 are unique to Shigella dysenteriae ...... 38

Figure 3: RyfA1 and RyfA2 structure and sequence ...... 39

Figure 4: Shigella dysenteriae produces both RyfA1 and RyfA2 ...... 41

Figure 5: Bacterial isolates that encode for RyfA1-like molecules are enteropathogens . 43

Figure 6: Overproduction of RyfA1 has no significant impact on the growth of Shigella dysenteriae ...... 45

Figure 7: RyfA1 overproduction inhibits cell to cell spread by Shigella dysenteriae ...... 48

Figure 8: Increased production of RyfA1 results in undetectable levels of ompC transcript ...... 50

Figure 9: RyfB1 specifically inhibits RyfA1 ...... 52

Figure 10: GFP was not produced under a putative ryfA1 translation start site ...... 55

Figure 11: Overproduction of RyfB1 does not impact plaque formation of Shigella dysenteriae ...... 57

Figure 12: RyfA2 overexpression impacts growth ...... 65

Figure 13: RyfA2 overproduction does not impact plaque formation ...... 66

Figure 14: RyfB2 overexpression does not influence growth of Shigella dysenteriae..... 67

Figure 15: Overproduction of RyfB2 increases plaque number ...... 68

Figure 16: RyfB2 overproduction decrease RyfA2 levels ...... 69

Figure 17: RyfB overexpression inhibits RyfA levels ...... 70

Figure 18: Mutant RyfBswap binding ...... 71

Figure 19: RyfB2swap overproduction does not impact the growth of Shigella dysenteriae...... 72

Figure 20: RyfB2swap also retains an increased plaque phenotype...... 73 11

Figure 21: RyfB2 overproduction impacts RyfA2 and RyfA1...... 74

Figure 22: RyfB2 overproduction results in a decrease of RyfA1...... 75

Figure 23: RyfA1, RyfA2, and RyfB2 have no impact on RyfB1...... 76

Figure 24: RyfA1 and RyfA2 overproduction impacts each other...... 77

Figure 25: RyfA and RyfB regulation schematic...... 80

Figure 26: RyfA1 may have a putative binding site near the 3’ end of ompR...... 102 12

CHAPTER 1: RIBOREGULATORS: FINE-TUNING VIRULENCE IN SHIGELLA

Riboregulators: Fine-Tuning Virulence in Shigella was previously published in the journal Frontiers in Cellular Infection and Microbiology on January 27, 2016.

1.1 Abstract

Within the past several years, ribo-regulation has become increasingly recognized for its importance in controlling critical bacterial processes. Regulatory RNA molecules, or riboregulators, are perpetually responsive to changes within the micro-environment of a cell. Notably, several characterized riboregulators have been shown to control virulence in pathogenic bacteria, as is the case for each riboregulator characterized to date in

Shigella. The timing of virulence gene expression and the ability of the pathogen to adapt to rapidly changing environmental conditions is critical to the establishment and progression of infection by Shigella species; ribo-regulators have been shown to mediate each of these important processes. This mini review will present the current state of knowledge regarding RNA-mediated regulation in Shigella by detailing the characterization and function of each identified riboregulator in these pathogens.

1.2 Introduction

Shigella are bacterial pathogens highly adapted for colonizing the human gut, a process that is facilitated by their unique lifestyle. The bacteria are passed from host to host via the fecal oral route of transmission. After surviving the acidic environment of the stomach, they travel the length of the intestinal tract to the site of infection, the colonic epithelium [1]. Once at the site of infection Shigella transverse the colonic epithelium by passage through M-cells and are subsequently presented to, and engulfed by, 13 macrophages. Once inside the macrophage, Shigella species induces lysis of the phagocytic cell, releasing the bacteria to the basal-lateral surface of the epithelium. [2]

Next, Shigella invade human intestinal epithelial cells using a type three secretion system

(T3SS) and begin to replicate within the eukaryotic cytoplasm. [3] Finally, the bacteria utilize host actin to spread from one eukaryotic cell to the next, a process that destroys intestinal epithelial cells thus contributing directly to the development of symptoms of a

Shigella infection, namely bloody diarrhea [1,3].

To establish and progress an efficient infection, Shigella species precisely regulate the expression of virulence-associated genes in response to specific environmental conditions encountered within the host; a collection of complex processes in which regulatory RNA molecules play critical, and increasingly recognized roles

(Figure 1) [4,5]. Their ability to mediate a rapid, specific response makes riboregulators ideal molecules to mediate the regulation of virulence gene expression in response to changes within a pathogen’s micro-environment. Riboregulators characterized thus far in

Shigella include several regulatory small RNAs (sRNAs) and one RNA thermometer [6–

12]. Despite the fact that they function to regulate the expression of different target genes, and that they utilize a variety of molecular mechanisms, all riboregulators described in

Shigella to date share two important features; 1) they each respond, directly or indirectly, to changes in specific environmental conditions, and 2) they all are nestled within large regulatory networks that impact virulence-associated processes (Table 1 and Figure 1).

This review will examine all characterized riboregulators in Shigella, with special 14 emphasis placed on a discussion of how each was discovered, as well as their functions and impact on virulence-associated processes.

Figure 1: sRNA network involved with Shigella virulence. All sRNAs which have been characterized in Shigella thus far are interconnected within the virulence network. RyhB inhibits VirB which can lead to inhibition of invasion. CsrB/C can inhibit CsrA which is needed for invasion. RnaG inhibits icsA transcription and limits spread. ShuA is important for heme uptake and survival of the pathogen during the infection process.

Table 1: Summary of riboregulators in Shigella Riboregulator Type Environmental Influence Target CsrB/CsrC Sequestering sRNAs Carbon Metabolism CsrA

RnaG cis-encoded sRNA Temperature icsA

RyhB trans-encoded sRNA Iron VirB

5’ UTR ShuA RNA Thermometer Iron and Temperature ShuA

Each riboregulator described in Shigella has also contributed to virulence in some way. The sRNAs seem to affect invasion and spread while the RNA thermometer is important with iron acquisition.

1.2.1 CsrB and CsrC

Two sRNAs which are important regulatory molecules for carbon metabolism in

E. coli are the carbon storage regulators CsrB and CsrC [13–15]. Both CsrB and CsrC belong to a unique class of sRNAs which bind and sequester multiple copies of their target protein to inhibit its activity [13–15]. In this case, CsrB and CsrC bind to and inhibit CsrA, a protein that promotes the production of other proteins necessary for glycolysis while inhibiting the production of proteins required for gluconeogenesis

[13,14,16–18]. CsrA also indirectly up-regulates the production of CsrB and CsrC through the UvrY/BarA two-component system, thus regulating its own activity [19,20].

Many genes involved with carbon regulation in E. coli, including those encoding CsrA,

CsrB and CsrC are conserved in Shigella.

Interestingly, researchers have demonstrated that CsrA activity is linked to virulence in S. flexneri [6]. Specifically, a mutation in csrA inhibits the ability of S. flexneri to invade eukaryotic cells and to spread from one eukaryotic cell to the next within a monolayer [6]. Additionally a mutation in csrB or csrC which, in turn, indirectly 17 increases the amount of free CsrA, allows S. flexneri to invade slightly more effectively than the wild-type strain [6]. When CsrB or CsrC are overproduced, CsrA activity is inhibited as expected, and S. flexneri is no longer able to invade host cells [6]. CsrA is hypothesized to impact Shigella virulence by two mechanisms. By the first mechanism,

CsrA facilitates the activity of phosphofructokinase A (PfkA), which in turn upregulates the master virulence regulator in Shigella, virF [6,21]. By the second mechanism, CsrA and PfkA may impact glycosylation of LPS on the surface of Shigella. Changes in LPS structure, mediated by alterations in the degree of glycosylation, have been shown to impact the exposure of the type III secretion system (T3SS) on the surface of Shigella

[22]. Given these observations, changes in the Csr regulatory pathway could influence exposure of the Shigella T3SS [6,22,23]. Regardless of the underlying molecular mechanism, CsrA clearly impacts virulence-associated processes in Shigella [6]. As RNA molecules that function to regulate CsrA activity, CsrB and CsrC are thus implicated in the control of Shigella virulence.

Although different in size, no unique functions for CsrC and CsrB have been found thus far in either E. coli or Shigella. Investigating CsrB and CsrC under a number of different environmental could reveal unique activity and/or production patterns for these regulators [14] Additionally, the biological significance of sibling sRNAs with apparent functional redundancies is still unclear [24]. It has been suggested that the extremely short half-life of CsrB and CsrC contributes to the speed by which the sRNAs can regulate CsrA, and that CsrB and CsrC allow for fine-tuning of gene expression in response to changes of carbon sources [15]. The short half-life of CsrB and CsrC could 18 also contribute to the rapid responses needed for the regulation of virulence-associated processes based on environment specific alterations in carbon source availability.

1.2.2 RyhB

Another sRNA indirectly influenced by an environmental factor is RyhB.

Originally discovered in E. coli, RyhB has been demonstrated to be an important regulator of iron metabolism [25]. Bacteria need iron for survival, but too much iron can kill the organism, thus the production of iron uptake systems and iron storage systems are tightly regulated [26–28]. RyhB plays an important role in maintaining the critical balance between the strict requirement, and potential toxicity, of iron in E. coli [25,26]

The production of RyhB itself, in both E. coli and Shigella is regulated in response to iron availability via the activity of the iron-responsive transcriptional regulator Fur [9,25,29–

31]. Under conditions of high iron, Fur functions to inhibit RyhB production. Fur- dependent repression of RyhB production in turn relieves the RyhB-mediated repression of genes encoding various iron containing enzymes and iron storage proteins in E. coli

[25]. RyhB is conserved between E.coli and Shigella where the regulator has been shown to impact the expression of several targets conserved between the two genera [9].

In addition to the role that RyhB plays in iron metabolism, RyhB has been implicated in the regulation of virulence gene expression in S. dysenteriae [9,10,12].

Specifically, RyhB inhibits the transcription of virB, a gene encoding a protein that functions to promote the expression of several virulence-associated genes in Shigella.

[9,12,21,32,33]. Although the exact molecular mechanism underlying RyhB-dependent inhibition of virB transcription remains unknown, complementarity between the template 19

DNA strand within the virB open reading frame and RyhB exists and is required for the observed regulation; data that suggests a novel regulatory mechanism may be at play

[10].

RyhB allows for iron responsive regulation of the Shigella virulence cascade. In the relatively iron-rich environment of the human gut, Fur is likely active and functioning to repress the production of RyhB. With decreased production of RyhB, VirB production proceeds and the protein functions to promote the expression of several virulence- associated genes including icsP [10,34,35]. IcsP protease limits IcsA (a protein required to polymerize the actin tail used by Shigella to spread from one eukaryotic cell to the next) from being produced prior to invasion into the host cell [12,36–43]. Once Shigella enter the host cell, iron conditions become limiting, and as a result Fur-mediated repression of RyhB production is relieved. Once produced in the low iron environment

RyhB functions to represses virB expression, which in turn limits IcsP production

[12,35]. Decreased IcsP production results in increased activity of IcsA which in turn facilitates host actin polymerization and cell to cell spreading by the bacterium [40–44].

Inhibition of virB transcription is not the only way by which RyhB influences virulence-associated processes in Shigella. In addition to its role in modulating VirB production, RyhB also indirectly regulates the expression of genes encoding factors required for acid resistance, an essential aspect of infection initiation by this pathogen

[45].

Evolutionarily, as an sRNA, RyhB allows for differential regulation than a protein. To compensate for small changes in iron availability within the human host, the 20 synthesis of RyhB can quickly be inhibited by Fur. RyhB can also become a fully active regulator after only transcription, giving it an edge over a protein regulator which would require more energy and time to synthesize [46]. Its fundamental features as an iron- regulated ribo-regulator allow for quick, efficient modulation of target gene expression by RyhB in response to the subtle changes of environmental iron availability experienced by the pathogen throughout the course of a natural infection.

1.2.3 RnaG

RnaG is unique among Shigella sRNA in that, unlike the others, it was identified and characterized in Shigella, and it is encoded on the large virulence plasmid. Similar to other Shigella sRNAs however, is the fact that production of RnaG is regulated in response to a specific environmental cue and once produced, it functions to regulate a virulence-associated process. Specifically, RnaG production is indirectly controlled in response to environmental temperature and once produced, functions to regulate the transcription of icsA, a virulence associated gene required for actin-based motility of

Shigella species [7,8,43]. Two coordinated mechanisms allow RnaG to mediate transcriptional control of icsA. First, rnaG and icsA have convergent promoters in close proximity to each other [7]. As such, activity of the rnaG promoter results in decreased activity of the icsA promoter through promoter interference [7]. Second, as a result of their over-lapping arrangement, and thus nucleic acid complementarity, RnaG can bind directly to the icsA transcript via kissing complexes, alter the structure of the growing transcript, and lead to early transcriptional termination [7]. Through these two, non- mutually exclusive molecular mechanisms, transcription of the important virulence factor 21

IcsA is inhibited by RnaG, thus directly implicating this small RNA in controlling

Shigella pathogenesis.

RnaG is likely produced during the time Shigella first enters the host until the pathogen reaches the site of infection. During this time, RnaG would inhibit premature expression of icsA, in turn, inhibit the expression of proteins required for host actin polymerization [7]. At 37 ˚C (the environmental temperature first encountered from initial entry into the host), VirF promotes the transcription of icsA and inhibits that of rnaG [8]. During the initial stages of infection it is possible that VirF levels are not high enough to induce adequate icsA transcription for polymerization and, due to specific environmental factors such as pH and osmolarity, may not be high enough until Shigella reaches the colonic epithelium [33,47,48]. In this case, RnaG production during these initial stages of infection would inhibit aberrant icsA expression between the time that

Shigella first enters the host and when the pathogen reaches the site of infection. Such temporal timing would prevent premature production of IcsA and possibly dampen any immune system alarms which may be set off in the presence of the protein.

1.2.4 ShuA

The final Shigella ribo-regulator is an RNA thermometer located within the 5’ untranslated region (UTR) of S. dysenteriae shuA (Shigella heme uptake), a gene encoding an outer-membrane heme receptor that is essential for the utilization of heme as a source of nutrient iron by the pathogen [49,50]. RNA thermometers function to modulate translation efficiency from the transcript on which they are housed by the formation of an inhibitory structure(s) that physically blocks binding of the ribosome to 22 the transcript at non-permissive (low) temperatures [51]. As the environmental temperature rises the inhibitory structure is destabilized, the ribosomal binding site is exposed and translation of the regulated gene proceeds. The shuA RNA thermometer represents the first RNA thermometer characterized in any Shigella species [11] .

Although identified initially in Shigella, an identical regulator functions to control expression of the orthologous gene chuA in pathogenic E. coli where this gene product is a virulence determinant [52–55]. Transcription of shuA is subject to iron-dependent regulation by Fur, the protein, while translation of the shuA transcript is subject to temperature-dependent regulation by the activity of the cis-encoded RNA thermometer

[11,49,50]. It is important to note, that for bacterial pathogens increased environmental temperature can act as an important signal that the organism has entered the host, the environment where production of virulence-associated factors will provide the most benefit.

Only under particular environmental conditions will ShuA be efficiently produced. Under conditions where iron is abundant, shuA transcription will be repressed by the activity of Fur, regardless of environmental temperature. Conditions where iron is depleted, but the environmental temperature is relatively low, the FourU RNA thermometer will inhibit translation of shuA. Only under iron-limiting conditions, and at temperatures corresponding to that within the human body (37 ˚C), will ShuA be produced [11]. The transcriptional and translation regulation mediated by Fur and the shuA RNA thermometer, respectively, function together to ensure maximal production of

ShuA under conditions of poor iron availability and increased temperature, precisely the 23 condition encountered within the human body; the only environment in which Shigella will encounter heme as a potential source of essential nutrient iron.

1.3 Discussion

The riboregulators in Shigella, described in this review, all respond (directly or indirectly) to environmental changes, and all of them function within larger regulatory networks to influence pathogenesis of these species. CsrB/CsrC are regulated in response to carbon availability, RnaG is regulated in response to temperature, RyhB is regulated in response to iron availability, and finally the activity of the shuA RNA thermometer is regulated in response to temperature. Importantly, every Shigella ribo-regulator characterized to date functions to influence (and are influenced by) the production of factors involved in one or more virulence-associated processes, and thus must themselves be considered virulence determinants. This observation raises the question, why are some virulence processes in Shigella controlled by protein-based regulation while others are controlled, at least in part, by the activity of riboregulators? In all sRNA found in

Shigella thus far, proteins (VirF, H-NS, Fur and UvrY) seem to be the initial regulator controlling a given step of a specific virulence-associated process. In each case, the riboregulator functions to modulate a specific virulence-associated activity for some duration of time, and then due to an environmental trigger, quickly switches off and allows

Shigella to proceed to the next stage of pathogenesis (Figure 1). Shigella experience many changes in environmental conditions during host-to-host transmission and throughout the course of a natural infection. Perhaps Shigella evolved to favor riboregulators over protein regulators in conditions under which rapid specific changes to 24 the production of one or just a few genes would be more beneficial to the organism than turning on/off an entire large regulon. Perhaps, processes required for the initial induction of pathogenesis in Shigella are controlled by protein regulators rather than riboregulators because quick reactions to false positive signals for pathogenesis could be detrimental to the survival of the bacteria, while a lag in protein regulation may temper those signals thus reducing the frequency of such detrimental events. [46].

More research needs to be carried out on riboregulators in Shigella to fully understand their functions and roles in virulence regulation [56]. Additionally, many putative sRNAs should be examined to see if they have small proteins missed by predictor programs [57]. Lastly, once sRNAs and riboregulators are fully understood, it is possible that their function could be targeted as novel treatments for shigellosis. 25

CHAPTER 2: SIBLING SRNA RYFA1 INFLUENCES SHIGELLA DYSENTERIAE

PATHOGENESIS

Sibling sRNA RyfA1 Influences Shigella dysenteriae was previously published in the journal Genes on January 26, 2017.

2.1 Introduction

Species of the bacteria Shigella (S. dysenteriae, S. sonnei, S. boydii and S. flexneri) are the causative agents of shigellosis, a highly infectious diarrheal disease in humans. Each year, an estimated 164 million people are infected by Shigella, resulting in

1.1 million deaths [1]. Devastatingly, children under the age of five in developing nations account for the majority of Shigella associated deaths due to lack of clean water, hydration, nutrition, and access to treatment [1]. Shigella infections, however, are not limited to developing countries; the Centers for Disease Control estimates 500,000 cases of shigellosis in the United States each year [2]. Worldwide prevalence of Shigella is vast, yet universally safe treatment and prevention are lacking [1]. These facts, together with the increasing rates of antibiotic resistance seen in Shigella species across the globe

[3] highlight the relevance of these pathogens as continuing threats to human health, and thus the need for further investigation. Of particular importance are studies that reveal the molecular mechanism underlying the ability of Shigella species to survive within the human host and cause disease. These mechanisms must be fully characterized so that they can be specifically targeted by therapeutics, and by doing so, prevent or lessen the morbidity and mortality associated with shigellosis. 26

The basic pathogenesis pathway of Shigella is well characterized [4]. Infection of a new host occurs upon ingestion of Shigella during the consumption of contaminated food or water [1]. Once ingested, the pathogen transits the length of the gastrointestinal tract before reaching the colonic epithelium, the site where infection is initiated. Shigella cross from the lumen of the colon to the basolateral side of the colonic epithelium via uptake by microfold cells (M-cells) within the Peyer’s patches [5]. Following uptake by

M-cells, the bacteria are presented to, and taken up by macrophages. Shigella escape killing by the macrophages by inducing apoptosis of the phagocytic cells, a process that results in the release of the pathogen to the submucosa [5]. From here, Shigella utilize a

Type III secretion system (T3SS) to orchestrate their own uptake into the nutrient rich cytoplasm of the colonic epithelial cell [4,6]. Once within the eukaryotic cell, Shigella replicates and manipulates host cell actin to facilitate rapid inter- and intra-cellular spread

[1,4,7–9]. The invasion of, replication within, and spread between cells of the colonic epithelium by Shigella provokes severe inflammation and destruction of the intestinal colonic epithelia, processes that result directly in the symptoms associated with shigellosis [1,4,7]. While studies characterizing virulence-associated processes of

Shigella pathogenesis are numerous, a complete understanding of these processes and the complex regulatory networks controlling them has not yet been achieved. Of particular interest is the growing evidence that RNA-mediated regulation plays a significant role in controlling virulence-associated processes in Shigella [10–16], a topic about which much remains to be learned. 27

It is increasingly recognized that regions within bacterial chromosomes previously assumed to be intergenic often encode riboregulators; RNA molecules that respond to a variety of signals and function to regulate the expression or activity of target genes or proteins, respectively [17–19]. Ribo-regulators can be classified into two major groups based on their location in relation to that of their target or targets. Cis-acting riboregulators are those that are located within the regulated transcript itself while trans- acting ribo-regulators are encoded separately from their target. sRNAs that interact with target transcripts generated from gene(s) located distal to where the sRNA is encoded are classic examples of trans-acting ribo-regulators. While the understanding of bacterial sRNAs, their regulatory mechanisms, and their impact on critical cellular processes continues to grow exponentially, less well characterized are a sub-family of sRNAs termed sibling sRNAs [20]. So named due to their extensive similarly to each other, the relative function and significance of sibling sRNAs remain hotly debated [20]. What is known about sibling sRNAs is that many have important roles in regulating virulence- associated processes; however, the significance of having more than one near identical copy of the regulator is often not completely understood.

To date, all riboregulators characterized in Shigella function to influence virulence-associated processes, highlighting the importance of future studies of this class of regulators in Shigella and related pathogens [16]. Here, we characterize RyfA1, one of a sibling pair of sRNAs in S. dysenteriae. While many bacterial species carry a single copy of ryfA, few contain sibling copies. With this study it is demonstrated that RyfA1 influences the virulence-associated process of intracellular spread as well as the level of 28 ompC transcript, a gene encoding a major outer-membrane protein associated with

Shigella virulence. Additionally, it is demonstrated that RyfA1 production is controlled by RyfB1, a second sRNA encoded divergent to ryfA1 that shares 17 nucleotides of complementarity with RyfA1. Together, these sRNAs demonstrate the potential complex interplay between small transcripts in S. dysenteriae, an interaction that may allow for more precise and responsive regulation within virulence-associated networks. Disruption of these regulatory systems could be key in determining targets for therapeutic agents.

2.2 Materials and Methods

Growth conditions

Escherichia coli K12 DH5α was grown in Luria-Bertani (LB) broth, tryptic soy broth

(TSB) or agar at 37 °C. All strains of S. dysenteriae were grown in LB broth or TSB broth in a 200 rpm shaking incubator or cultured on tryptic soy broth agar (TSBA) plates with 0.01% wt/vol Congo red at 37 °C unless noted otherwise. When needed to maintain selection for the presence of a plasmid, ampicillin was used at a concentration of

50μg/mL.

Sequence alignment and structural predictions of RyfA1 and RyfA2

E. coli K12 MG1655 ryfA was used as a reference in a Basic Local Alignment Search from the National Center of Biotechnology Information database

(http://blast.ncbi.nlm.nih.gov/Blast.cgi) in order to find all other ryfA genes in bacteria whose genome are included. The genes ryfA1 and ryfA2 were aligned with each other using Clone Manager ver. 9. Structures of RyfA1 and RyfA2 were predicted using M- fold analyses (http://unafold.rna.albany.edu/?q=mfold). Species containing any ryfA gene 29 were compared within the five nucleotide variable region using a program written in

JavaScript which scored gene likeness to ryfA1 or ryfA2 based on points. The greater the likeness score, the more similar. Supplemental figures comparing the likeness of all ryfA genes were generated using CLC genomics workbench (QIAGEN, Valencia, CA).

Cloning ryfA1 and ryfB1 into expression vectors ryfA1 was amplified from the S. dysenteriae chromosome [15] using specific primers containing MfeI and HindIII restriction sites (Table S1). ryfB1 was also amplified from the S. dysenteriae chromosome using primers containing MfeI and SacI restriction sites.

Amplicons were run on an agarose gel and purified using the QIAquick gel extraction kit

(QIAGEN, Valencia, CA) as per the protocol. Restriction enzymes MfeI and HindIII from (New England Biolabs Inc., Ipswitch, MA) were used to digest the ryfA1 containing amplicons as well as plasmid pQE2, an expression plasmid with an ITPG inducible promoter (QIAGEN, Valencia, CA). Restriction enzymes MfeI and SacI were used to digest the ryfB1 amplicon as well as pQE2. Both ryfA1 and ryfB1 were subsequently ligated into pQE2 using T4 ligase (New England Biolabs Inc., Ipswitch MA) creating plasmids pRyfA1 and pRyfB1 respectively (Table S2). The resulting plasmids were introduced into competent E. coli K12 DH5α using heat shock transformation. Each plasmid was then extracted from DH5α using a QIAmini prep kit (QIAGEN, Valencia,

CA) as per their instructions, sequence verified, and introduced into competent S. dysenteriae using electroporation. 30

Construction of GFP translational reporter plasmid p5’UTR RyfA1

A double-stranded insert containing the putative 5’ UTR of ryfA1 (up to and including the putative translational start site) flanked by short single stranded sequences compatible with overhangs generated by the activity of NsiI and NheI restriction endonucleases was generated by oligo annealing. Once generated, the insert was ligated into plasmids pXG-

10 [21] previously digested with restriction endonucleases NsiI and NheI (New England

Biolabs Inc., Ipswitch, MA). Following ligation and introduction into competent E. coli

DH5α by heat-shock transformation [22] the resulting translational reporter plasmid, termed p5’UTR_RyfA1, was sequence verified using Sanger Sequencing (Ohio

University Genomics Facility, Athens, OH).

Western blot

E. coli DH5α carrying pXG-0, pXG-1, or p5’UTR RyfA1 were grown to stationary phase at 37 °C. Using a ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington,

DE), the optical density at 600 nm was measured and approximately 5 X 108 bacterial cells were pelleted. The pelleted cells were suspended in 200 μl of a Laemmli protein dye

(Bio-Rad) supplemented with 5% 2-mercaptoethanol. Following suspension, the samples were boiled for 10 minutes and stored at -20 °C until further use. 15 μl of each whole cell protein preparation was separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on a 7.5% polyacrylamide gel. A PVDF membrane was prepared by soaking in methanol for 10 minutes and rinsing with water three times.

Proteins present in the polyacrylamide gel were then transferred to the prepared PVDF membrane. After the transfer, a 10% milk solution in PBST was used to block the 31 membrane overnight at 4 °C. Following blocking, anti-Gfp monoclonal IgG stabilized antibody preparation (Roche, Indianapolis, IN) was diluted in PBST and 5% milk by

1:1000 and added to the membrane for incubation at 4 °C for one hour. The membrane was subsequently washed 3 times for 5 minutes each time in PBST, and then incubated in

PBST and a solution of 10% milk in PBST at 4 °C for 10 minutes. The secondary antibody (goat anti-mouse HRP conjugated IgG (Bio-Rad)) was diluted 1:20,000 in

PBST and 5% milk and incubated with the membrane for 1 hour at 4 °C. Following incubation, the membrane was washed as above and the resulting signal visulaized using the ImmuneStar WesternC reagents (Bio-Rad) and a ChemiDoc XRS+ Imaging System

(Bio-Rad).

RNA isolation

Following growth under conditions indicated for each subsequent analysis (qRT-PCR,

Northern Blot) total RNA was isolated as follows. Bacterial cells present in 3 mL of culture were pelleted and lysed with the addition of 10% sodium dodecyl sulfate (SDS), and 3M sodium acetate (pH of 5.2), followed by brief vortexing, then heating the cells and contents for 7 minutes at 90 ˚C. TRIzol (Thermo Fisher Scientific, Waltham, MA) was subsequently added to the lysed cells, and all contents were transferred to a phase- lock tube (5 PRIME, Gaithersburg, MD). Nucleic acid was isolated as per the factory protocol. RNA was precipitated overnight at -80 °C using 100% EtOH. The following day, RNA samples were centrifuged for 15 minutes on high at 4 °C and the supernatant discarded. Next, 1 mL of ice cold 75% EtOH was added to the tube and the RNA pelleted again by centrifugation for 15 minutes on high at 4 °C. The supernatant was once again 32 discarded and the RNA pellet dried, and subsequently rehydrated in 54 μl nuclease free water. Next, Turbo DNase (Ambion, Austin, TX) was used according to kit instructions to remove any DNA remaining in each RNA sample. To ensure that the RNA samples were free from DNA contamination, 1 μl of each was used as template in a screening

PCR using primers for the conserved gene sodB and subsequently the amplicon was visualized on an agarose gel. DNA-free RNA was measured using the Thermo Scientific

Nanodrop 2000c spectrophotometer for concentration and quality based on 260/280 nm and 260/230 nm ratios.

Northern blot

Following growth at 37 °C in a shaking incubator RNA was isolated as described above.

Ten micrograms of RNA were run on a 10% polyacrylamide gel with 7 M urea and 1x

TBE (89 mM boric acid, 89 mM Tris, and 2 mM EDTA). A Low Molecular Weight

DNA Ladder (New England BioLabs, Ipswich, MA) was labeled with [ɣ-32P]-ATP using polynucleotide kinase and run on the same gel as the RNA sample. RNA and ladder were separated at 150 volts in 1x TBE buffer. Nucleic acids were then transferred to a

Hybond™-N+ membrane (GE Healthcare, Piscataway, NJ) at 50 v for 2 hrs. The samples and ladder were then crosslinked using UV to the membrane. Probes were labeled by polynucleotide kinase with [ɣ-32P]-ATP. The membrane was pre-hybridized in

ULTRAhyb®-Oligo Buffer (Ambion, Austin, TX) at 42 °C for at least 2 hrs in a rotating incubator. Radio-labeled probes were then added to the pre-hybridized membrane and allowed to hybridize overnight at 42 °C in a rotating incubator. The membrane was washed with 2x SSC (300 mM NaCl, and 30 mM NaC6H5O7) + 0.1% SDS, 1x SSC + 33

0.1% SDS, and 0.5x SSC + 0.1% SDS for 30 min at 42 °C in a rotating incubator. The membranes were then exposed to film and visualized by autoradiography.

Quantitative real-time PCR analysis

Total RNA was isolated from triplicate samples using methods described above following growth of each strain to the mid-logarithmic phase in LB supplemented with

20 μM IPTG, 50 μg/μL ampicillin and 0.01% deoxycholic acid. To generate cDNA for analysis, iScript (Bio-Rad, Hercules, CA) was used following manufacturer’s instructions. cDNA samples were diluted 1 to 10 and used as template in Real-time PCR analyses with TaqMan probe chemistries for analysis of ryfA1, ryfA2, and ryfB2 or

Sybrgreen and iQ Supermix (Bio-rad, Hercules, CA) for analysis of ompC, as per protocol instructions. All samples were amplified and analyzed in a Bio Rad C1000™

Thermal Cycler with a CFX96 Real-Time System under standard reaction conditions optimized for each primer set (Table S1). Fold changes were calculated using the ΔΔCt method [23] with target amounts normalized to those of the housekeeping gene rrsA in each sample and expressed relative to a selected control sample.

Plaque assay

Henle cells were cultured at 37 °C in an atmospheric condition of 5% CO2 to 80% confluency in 6 well polystyrene tissue culture plates (Corning Inc. Costar, Corning, NY) in the presence of Henle Cell Media (composed of Gibco Minimal Essential Media)

(Invitrogen Corp. Carlsbad, CA), supplemented with 10% Fetal Bovine Serum (FBS), 2 mM glutamine, and 1X non-essential amino acids (Lonza, Switzerland). 34

Plaque assays were performed as described previously by Oaks et al. 1985 [24] with minor modifications. S. dysenteriae strains were cultured on TSBA with Congo red and

50 μg/mL ampicillin. Three colonies from each strain were used to inoculate separate 3 mL cultures of TSB and 50 μg/mL ampicillin, that were subsequently grown overnight in a shaking incubator at 30 °C. 100 μL of each overnight culture was used to inoculate 3 mL of fresh TSB supplemented with 50 μg/mL ampicillin, 20 μM IPTG, and 0.01%

DOC. Strains were cultured at 37 °C in a shaking incubator to the mid-logarithmic phase of growth prior to the addition of 1x104 bacteria from each culture into a separate well of a 6 well polystyrene tissue culture plate containing a monolayer of Henle cells at approximately 80% confluency, 2 mL of Henle media, 50 μg/mL ampicillin, and 20 μM

IPTG. Following inoculation with the bacteria, the tissue culture plates were spun for 10 minutes in a Beckman Coulter Allegra 25R centrifuge at 600 x g at room temperature and then incubated for 90 minutes in a 37 °C incubator under 5% CO2 atmospheric conditions. Following incubation, Henle media was suctioned off and to the monolayer a wash containing 2 mL of fresh Henle media supplemented with 50 μg/mL ampicillin, 20

μg/mL gentamicin, 0.3% glucose, and 20 μM IPTG was added. Following incubation for

72 hrs at 37 °C and in 5% CO2 atmospheric conditions the Henle cell monolayers and bacteria were stained with Giemsa-Wright stain (Camco, Ft. Lauderdale, FL) and washed with double distilled water twice. Plaques were analyzed for number and size.

Spread assay

Spread assays were performed as a modified version of the plaque assay. S. dysenteriae strains were grown as described above but to near confluency. Once the bacteria reached 35 mid-logarithmic growth phase, 1x104 Shigella cells were used to infect a near confluent monolayer of Henle cells in Henle media containing 50 μg/mL ampicillin and 20 μM

IPTG. Plates were centrifuged as described above. Following centrifugation, plates were subsequently incubated for 60 minutes at 37 °C at an atmospheric condition of 5% CO2.

Next, Henle cells were washed with 2 mL of Henle media, and then an overlay of Henle

Media containing 0.3% glucose, 50 μg/mL ampicillin, 20 μM IPTG, and 20 μg/ml of gentamicin was added. Plates were incubated for 6 hrs at 37 °C under 5% CO2 atmospheric conditions. After this incubation, Wright-Giemsa stain (Camco, Ft.

Lauderdale, FL) was used to stain the plates. The plates were washed with distilled water following staining and air-dried. 100 Henle cells containing ≥3 bacteria and that were in contact with at least 2 other Henle cells within the monolayer were selected for scoring.

Spread was scored if one or more of the Henle cells in physical contact with the first also contained bacteria.

Invasion assay

Invasion assays were performed as a modification to plaque assays. S. dysenteriae strains were grown as described above. Once reaching mid-logarithmic growth phase, 2x108 bacteria were used to infect a 60% confluent monolayer of Henle cells as detailed above.

Each well contained Henle media, and was supplemented with 50 μg/ml ampicillin and

20 μM IPTG. Plates were centrifuged as described above and incubated for 30 minutes at

37 °C at 5% atmospheric CO2. Following incubation, plates were washed with 2 ml of

Henle media, and then covered in Henle cell media containing 0.3% glucose, 50 μg/mL ampicillin, 20 μM IPTG, and 20 μg/mL gentamicin. Plates were incubated for 90 minutes 36 at 37 °C under 5% CO2 atmospheric conditions. Following incubation, plates were stained with Wright-Giemsa stain (Camco, Ft. Lauderdale, FL). Next, plates were washed with distilled water, air dried, and invaded Henle cells were counted. Henle cells that contained ≥3 bacteria and were physically isolated from other Henle cells, were scored as invaded.

Next generation sequencing

The following RNA isolation protocol was adapted from Carroll et al. 2014 [25] with minor modifications. S. dysenteriae strains carrying either the empty vector (pQE2) or the ryfA1 expression plasmid (pRyfA1) (Table S2) were cultured to the mid-logarithmic phase of growth at 37 °C in 3 mL LB containing 50 μg/mL ampicillin and 0.01% DOC and 20 μM IPTG. Next, 750 μL of RNA preserving solution (95% EtOH, 5% phenol) was added to each tube. Bacterial cells present in each sample were pelleted and the supernatant discarded. To isolate total RNA, QIAGEN RNEasy kit (QIAGEN, Valencia,

CA) was used as per instructions. Immediately following, nucleic acid was treated with

Turbo DNAse (Ambion, Austin, TX). RNA was checked for purity and concentration using Agilent 2100 Bioanalyzer on an RNA Nano 6000 chip. Ribosomal RNA was depleted from the samples using Ribozero (Illumina, San Diego, CA) and MicrobExpress

(Thermo Fisher Scientific, Waltham, MA) as per their respective instructions. RNA was ethanol precipitated in the presence of 3M sodium acetate and 1 μg/μl RNA-grade glycogen (Thermo Fisher Scientific, Waltham, MA) overnight at -80 °C. The following day, samples were centrifuged at 4 °C for 30 minutes. The supernatant was discarded and

1 mL of ice cold 70% EtOH was added. Samples were spun in a microcentrifuge on high 37 for 10 minutes at 4 °C. Again, the supernatant was discarded and the RNA was allowed to air dry for 5 minutes. Each RNA pellet was suspended in 20 μL of nuclease free water and analyzed for purity and concentration as above. RNA-seq was performed using an

Ion Torrent Next PGM and 200 bp read chemistry and a 318 sequencing chip at the Ohio

University Genomics Core Facility. Data was analyzed using CLC Genomics Workbench

Version 8 (QIAGEN, Valencia, CA). The RNA-seq data files have been deposited in

GEO under accession number GSE87727.

2.3 Results

2.3.1 Twin copies of RyfA are produced by Shigella dysenteriae

Originally identified by in silico-based genomic analyses designed to identify genes encoding previously uncharacterized sRNA molecules, the putative sRNA RyfA was predicted to be encoded by S. flexneri and several strains of E. coli [26]. Further in silico analyses revealed that ryfA is present in other species of Shigella (S. boydii, S. sonnei and S. dysenteriae) (Figures S1-2). Of note, two genes predicted to encode nearly identical sibling sRNAs, here designated ryfA1 and ryfA2, were found to exist in tandem on the S. dysenteriae chromosome (Figure 2). ryfA1 and ryfA2 share 95% identity at the nucleic acid level, and each has greater than 90% sequence identity to the singlet ryfA from other species (Figure 2). While several single nucleotide alterations exist between ryfA1 and ryfA2, the largest run of nucleotide dissimilarities occurs in a five-nucleotide stretch, here termed the variable region (Figure 3a). This five nucleotide variable region is not limited to ryfA1 and ryfA2 in S. dysenteriae, indeed, across all isolates containing ryfA, the least conserved region is the variable region (Figure S1). In S. dysenteriae the 38 variable region influences the predicted structure of RyfA1 and RyfA2 in a subtle but potentially important way. Specifically, while otherwise identical, the predicted structure of RyfA1 and RyfA2 varies within a single stem loop structure containing the sequences from within the variable region of each (Figure 3b) [27]. Given the significance of single stranded regions within sRNAs in mediating specific interactions between the sRNA and its regulatory target(s), the variation predicted between RyfA1 and RyfA2 raises the possibility that these two sRNAs may have unique regulatory targets, and thus unique functions, in vivo.

Figure 2: RyfA1 and RyfA2 are unique to Shigella dysenteriae. S. dysenteriae encodes for two nearly identical RyfA molecules, sibling sRNAs here designated RyfA1 and RyfA2. RyfA1 and RyfA2, represent putative sibling sRNAs which are 95% identical to one another and contain greater than 90% nucleic acid identity to the singlet RyfA encoded at the same chromosomal location in all other species of Shigella as well as several isolates of E. coli.

Figure 3: RyfA1 and RyfA2 structure and sequence. (a) The predicted genes ryfA1 and ryfA2 were aligned using Clone Manager version 9. The genes encoding the sibling RyfA molecules are 95% identical in S. dysenteriae. Nucleotide differences between ryfA1 and ryfA2 are highlighted with light gray rectangles and the identified five nucleotide “variable region” of each is boxed in black. (b) M-fold (http://unafold.rna.albany.edu/?q=mfold) [27] was used to predict the structures of RyfA1 and RyfA2. Although the predicted structure between each sibling RyfA is nearly identical, the stem loops outlined in the black boxes vary between the two molecules. The observed structural differences between RyfA1 and RyfA2 result from the varied nucleic acid sequences within the five nucleotide variable region (as indicated by the gray boxes) identified upon a comparison of each ryfA gene.

The first step in characterizing RyfA1 and RyfA2 from S. dysenteriae was to determine if the bacterium produces one or both of these putative molecules. To this end, a northern blot was performed using a probe specific for a region conserved between 40

RyfA1 and RyfA2; sRNAs predicted to be 303nt and 305nt in length, respectively. A single band of approximately 300nt was detected in the northern blot analysis, confirming that S. dysenteriae produces at least one RyfA molecule (Figure 4a). To investigate the presence of both RyfA1 and RyfA2 in wild-type S. dysenteriae, reverse transcriptase

PCR (RT-PCR) analyses were performed. Specifically, cDNA was generated from total

RNA isolated from wild-type S. dysenteriae and used as template in amplification reactions pairing a conserved reverse primer with a forward primer specific to the variable region of either RyfA1 or RyfA2 (Figure 4b). Plasmids containing either ryfA1 or ryfA2 were used as control templates to ensure the ryfA1 specific primers only amplified ryfA1 and not ryfA2, and vice-versa. Results from the RT-PCR analyses demonstrate that amplification by each primer set is specific to the corresponding RyfA molecule and, importantly, that wild-type S. dysenteriae produces both RyfA1 and RyfA2

(Figure 4c).

Figure 4: Shigella dysenteriae produces both RyfA1 and RyfA2 (a) Northern blot analysis using total RNA isolated from wild-type S. dysenteriae following grown to the mid-logarithmic phase at 37 °C and a radio-labeled probe specific to sequences conserved between RyfA1 and RyfA2. The predicted sizes of RyfA1 and RyfA2 are 303 and 305 nt, respectively. (b) Schematic depicting the location and sequence specificity of the amplification primers used in the reverse transcriptase analysis. Each forward primer overlaps the five nucleotide variable region of RyfA1 or RyfA2, thus providing specificity of amplification. (c) Reverse transcriptase PCR demonstrating that both RyfA1 and RyfA2 are produced by wild-type S. dysenteriae under the conditions tested. Using the ryfA1 specific forward primer, amplification occurs when cDNA generated from S. dysenteriae or a plasmid carrying ryfA1 is used as a template (pRyfA1), but not when a plasmid carrying ryfA2 is used as a template (pRyfA2). Similarly, using the ryfA2 specific forward primer, amplification occurs when cDNA generated from S. dysenteriae or a plasmid carrying ryfA2 is used as a template, but not when a plasmid carrying ryfA2 is used as a template. The RNA used to generate the cDNA amplification template is used itself as a template with each primer set to ensure that the sample is free from DNA contamination.

2.3.2 The presence of a ryfA1-like gene is associated with pathogenicity

In silico analysis of all sequenced bacterial isolates available in NCBI

(http://www.ncbi.nlm.nih.gov/) and ShiBASE (http://www.mgc.ac.cn/ShiBASE/) databases was carried out in order to identify those that carry one or more ryfA gene.

Interestingly, ryfA was identified only in Shigella and Escherichia isolates. Next, using an in-house program written in JavaScript, the identified ryfA genes were scored and grouped based on likeness within the five nucleotide variable region. Our analysis revealed that four iterations of the five nucleotide variable region exist in Shigella and

Escherichia, three versions being grouped as “ryfA1-like” by containing higher GC content (CACCC, CCCCC, and CGCGT) and a single ryfA2 version with higher T content (TGTTT) (Figure 5). Strikingly, of the isolates that contain a ryfA gene, all that encode a “RyfA1-like” molecule are pathogenic (Figure 5), and more specifically are enteropathogenic. With few exceptions, bacterial isolates that carry a ryfA2 gene are either non-pathogenic or are uropathogenic. Given the observed link between the presences of a ryfA1-like gene and pathogenicity, S. dysenteriae RyfA1 was selected for further investigation.

Figure 5: Bacterial isolates that encode for RyfA1-like molecules are enteropathogens. All bacterial isolates whose genome is sequenced and available on NCBI were evaluated for the presence of ryfA. Next all identified species were grouped based on the similarity of nucleic acid sequence within the ryfA five nucleotide variable region. Four different iterations of the variable region were found including TGTTT, CGCGT, CCCCC, and CACCC. Three of the four variable region iterations contained higher GC content and were termed “RyfA1-like”. The number of isolates containing each RyfA iteration as well as the number of those with each group that are pathogenic are indicated.

2.3.3 RyfA1 impacts virulence in Shigella dysenteriae

Overproduction of RyfA1 does not significantly impact the growth of S. dysenteriae

To characterize the function of RyfA1 in S. dysenteriae, ryfA1 was amplified and ligated into an expression plasmid such that production of the RyfA1 molecule was placed under the control of an ITPG inducible promoter (pRyfA1). To ensure that RyfA1 is predictably produced from the pRyfA1 plasmid, quantitative Real-time PCR (qRT-

PCR) analyses were carried out using a primer probe set designed to distinguish ryfA1 from ryfA2 by binding of the probe over the five nucleotide variable region. The growth of S. dysenteriae carrying pRyfA1 in the presence of 20 μM IPTG results in a 100-fold increase in the relative levels of RyfA1 as compared to those measured in the strain carrying the empty vector control (pQE2) grown under identical conditions (Figure 6a).

Increased levels of RyfA1 have no significant effect on the growth of S. dysenteriae as determined by growth analysis of the strain carrying pRyfA1 or the vector control in the presence of 20 μM IPTG (Figure 6b).

Figure 6: Overproduction of RyfA1 has no significant impact on the growth of Shigella dysenteriae. An IPTG inducible ryfA1 expression plasmid was created (pRyfA1) and introduced into S. dysenteriae. (a) Quantitative Real-time PCR analysis of the relative levels of RyfA1 present in S. dysenteriae carrying pRyfA1 or the pQE2 vector control following the growth of each strain in the presence of IPTG. RyfA1 levels were calculated using the ΔΔCt method [23] in which they are normalized to the level of rrsA measured in each sample and expressed relative to the level of RyfA measured in a single vector control sample. All data are the average of biological triplicate analyses and error bars represent one standard deviation. *** denotes a statistically significant difference 46 with p < 0.0001. (b) Growth analysis of S. dysenteriae carrying the pQE2 vector control was compared to that of the strain carrying pRyfA1 under both non-inducing (0 μM IPTG) and inducing (20 μM IPTG) conditions. No significant differences were observed between the growth of S. dysenteriae carrying pRyfA1 or the vector control under either condition tested.

2.3.4 Overproduction of RyfA1 leads to inhibition of cell-to-cell spread by Shigella

dysenteriae

The ability of RyfA1 to impact virulence processes in S. dysenteriae was investigated using a series of in vitro tissue culture based analyses. The first assays completed were plaque assays that measure the cumulative ability of the bacterium to invade, replicate within, and spread between eukaryotic cells within a monolayer.

Maintaining the strains under inducing conditions to ensure increased production of

RyfA1 from the pRyfA1 plasmid, the ability of S. dysenteriae carrying pRyfA1 to form plaques within a monolayer of eukaryotic cells was compared to that of the strain carrying the vector control. Interestingly, S. dysenteriae carrying pRyfA1 formed an equivalent number of plaques as compared to those formed by the strain carrying the vector control, yet the plaques formed were dramatically smaller (Figure 7a, 7b). These data indicate that increased production of RyfA1 inhibits plaque formation and specifically, based on the conserved number and decreased size of the plaques, predicts that it is the process of cell-to-cell spread that is inhibited.

To determine which step of plaque formation is specifically influenced by RyfA1, in vitro invasion assays and spread assays were performed. For each of these assays, wild-type S. dysenteriae carrying either the pQE2 vector control or the RyfA1 producing plasmid pRyfA1 were cultured under inducing conditions. While S. dysenteriae expressing RyfA1 from pRyfA1 invaded eukaryotic cells at the same efficiency as the strain carrying the vector control grown (Figure 7c) the efficiency of cell-to-cell spread was significantly less (Figure 7d). Taken together these data clearly demonstrate that 48 increased production of RyfA1 limits plaque formation by inhibiting the ability of S. dysenteriae to efficiently spread from one cell to the next, a process essential for full virulence of this bacterial pathogen.

Figure 7: RyfA1 overproduction inhibits cell to cell spread by Shigella dysenteriae. (a) No significant difference was observed in the number of plaques formed between the vector control and pRyfA1. (b) Representative image of plaques formed by S. dysenteriae carrying either the pQE2 vector control or the ryfA1 producing plasmid pRyfA1 cultured and maintained under inducing conditions. The magnified image is that of Henle cells (larger light purple cells) and bacterial cells (dark purple smaller cells) surrounding the plaque formed by the indicated strain. Quantification of the ability of S. dysenteriae carrying the RyfA1 producing plasmid pRyfA1 or the vector control to invade Henle cells (c) and to spread between Henle cells within a subconfluent monolayer (d). All assays were carried out following growth of the indicated strain in the presence of IPTG, 49 and the inducer was maintained throughout the analyses. Data presented panel (a), (c), and (d) are the average of biological triplicate analyses and error bars represent one standard deviation. ** denotes a statistically significant difference with p < 0.001.

2.3.5 RyfA1 overproduction results in elimination of ompC, a transcript encoding a

major outer membrane protein

Inhibition of cell-to-cell spread in S. dysenteriae can result from misregulation or malfunction of virulence factors. RNA-seq analyses were performed as a means to identify those genes for which transcript levels change significantly as a result of increased RyfA1 production. When the transcriptome of wild-type S. dysenteriae producing RyfA1 from pRyfA1 was compared to that of the strain carrying the empty vector, the single largest fold change in transcript level (~7,000 fold) was observed for ompC, a transcript encoding the major outer membrane protein C (OmpC). The negative impact of RyfA1 production on ompC levels was confirmed using qRT-PCR (Figure 8).

Interestingly, OmpC has been shown to influence Shigella virulence. Specifically, inactivation of ompC in S. flexneri has been shown to inhibit cell-to-cell spread by the pathogen, as measured using in vitro tissue-culture based analyses [28]. Together, these data demonstrate that increased production of RyfA1 in wild-type S. dysenteriae results in undetectable levels of ompC and a phenotype that phenocopies that of S. flexneri lacking ompC, an inability to spread from one cell to the next within a monolayer of human epithelial cells. 50

Figure 8: Increased production of RyfA1 results in undetectable levels of ompC transcript. Quantitative Real-time PCR measuring the relative abundance of ompC transcript levels in S. dysenteriae carrying the RyfA1 producing plasmid pRyfA1 or the vector control following growth in the presence of IPTG. Using the ΔΔCt method [23] ompC levels are normalized to that of rrsA in each sample and are expressed relative to a single vector control sample. Data presented are the average of analyses completed in biological triplicate with error bars representing one standard deviation. B.L.D indicates that the target transcript was below the level of reliable detection.

2.3.6 Regulation of RyfA1 by RyfB1

The RNA-seq analyses provided unique insight into the ryfA1/ryfA2 locus on the

S. dysenteriae chromosome. Specifically, a small transcript (~100nt in length) was detected upstream of, and encoded divergently to, each ryfA gene (Figure 9a). These small transcripts are likely to encode putative sRNAs, due to a lack of an ATG start site or a discernable ribosomal binding site. Designated ryfB1 and ryfB2, these putative sRNAs share 60% nucleic acid homology to each other. Despite the fact that the coding regions for ryfA1 and ryfB1 do not overlap, in silico analysis demonstrated that the 51 transcripts share greater than 10 nucleotides of complementarity with each other.

Specifically, within an 18 nucleotide long region, 16 nucleotides demonstrate complementarity between the ryfA1 and ryfB1 transcripts (Figure 9b). It is noteworthy that the region of complementarity between RyfA1 and RyfB1 overlapped the variable region of RyfA1. Together these observations lead to the testable hypothesis that ryfB1 encodes an sRNA that functions to specifically regulate RyfA1.

To experimentally determine if RyfB1 influences RyfA1 levels, and if this regulation is specific, an expression plasmid was generated in which ryfB1 was cloned under the control of an IPTG inducible plasmid (pRyfB1) and introduced into wild-type

S. dysenteriae. Following growth under inducing conditions, quantitative RT-PCR analysis was used to measure the relative levels of RyfA1, RyfA2, and RyfB1 in S. dysenteriae carrying either pRyfB1 or the vector control. The obtained data demonstrate that increased levels of RyfB1 result in a significant decrease in the relative amounts of

RyfA1 (Figure 9c). The specificity of RyfB1 regulation on RyfA1 is demonstrated by the fact that increased production of RyfB1 has no significant effect on the relative abundance of RyfA2 (Figure 9c).

Figure 9: RyfB1 specifically inhibits RyfA1. (a) Next generation sequencing revealed ~100nt long transcripts immediately preceding, and divergent to, ryfA1 and ryfA2. The gene preceding ryfA1 has been termed ryfB1, while that preceding ryfA2 has been termed ryfB2. (b) In silico analyses of RyfB1 revealed a region, in which 17 of 18 nucleotides share complementarity to sequences within RyfA1. Of note, this region of complementarity between RyfA1 and RyfB1 overlaps the five-nucleotide variable region of the latter, indicated by the black box. Numbers indicate the base location relative to the 5’ end of each molecule (+1). (c) Quantitative Real-time PCR analyses of the relative levels of RyfB1, RyfA1 and RyfA2 in S. dysenteriae carrying the RyfB1 producing plasmid pRyfB1 or the vector control following growth of both strains under inducing conditions. The relative abundance of each target was calculated using the ΔΔCt method [23] in which target transcript levels are normalized to the level of rrsA present in each sample and are expressed relative to the level in that target in a single vector control sample. Data are the average of biological triplicates and errors bars indicate one standard 53 deviation. * denotes a significant difference with p<0.01 while ** denotes a significant difference with p<0.001.

2.3.7 RyfA1 does not encode a small protein under conditions tested

The genetic arrangement, relative size, and nucleic acid complementarity between the ryfA1 and ryfB1, as well as the ability of RyfB1 to influence steady-state levels of

RyfA1 (Figure 9b) is highly reminiscent of a Type I toxin-antitoxin locus, especially that of the zorO-orzO systems of E. coli O157:H7 [29,30]. In the case of the E. coli zorO- orzO toxin-antitoxin system, the larger zorO transcript encodes a small peptide which is toxic to E. coli O157:H7, and the sRNA OrzO inhibits production of that toxin [29,31].

The regulation of zorO by OrzO is dependent on 18 nucleotides of complementarity between the two transcripts [29]. In order to test if the ryfB1-ryfA1 pair encode a classical Type I toxin-antitoxin system, we first searched for putative ribosomal binding sites and potential start codons within each transcript by in silico analyses using Clone

Manager Version 9. Within ryfA1, a putative ribosomal binding site and corresponding

ATG start codon were located that, if functional, would be predicted to result in the synthesis of a small protein composed of 30 amino acids. To test whether or not a protein could be made from the ryfA1 transcript, the putative 5’untranslated region up to and including the identified predicted translational start of the gene was ligated into plasmid pXG-10 (Table S2) between a constitutive promoter and in frame with the gfp reporter gene lacking its own ribosomal binding and translational start sites. If the cloned sequence from ryfA1 is capable of mediating ribosomal binding and translation initiation under the conditions tested, GFP would be produced; a prediction that was tested using 54

Western Blot analysis. Specifically, total protein was extracted from triplicate strains of

E. coli DH5α containing either p5’UTR RyfA1, or the positive/negative control plasmids.

No GFP protein was detected in the strains carrying p5’UTR RyfA1 (Figure 10a). To ensure gfp transcript was being produced, qRT-PCR was performed on triplicate samples

(Figure 10b). Significantly more gfp transcript was detected in the positive control and p5’UTR RyfA1, than the negative control. Together, these data suggest that, under the conditions tested, no protein is generated from the ryfA1 transcript.

Figure 10: GFP was not produced under a putative ryfA1 translation start site. Sequences within the ryfA1 transcript do not mediate translation initiation. (a) Western blot analysis was used to detect GFP production in E. coli carrying a negative control plasmid (pXG- 0), the experimental plasmid p5’UTR RyfA1, or a positive control plasmid (pXG-10). Under the conditions tested, no GFP was produced in the strain carrying p5’UTR RyfA1, indicating that the cloned sequences do not support translation initiation, and thus that the ryfA1 transcript likely does not encode a small peptide. (b) Quantitative RT-PCR was used to ensure that the same triplicate samples used to detect GFP protein were producing gfp transcript. Strains carrying either the positive control and or the p5’UTR RyfA1 plasmid produce significantly more transcript than the negative control. The small differences in the level of gfp transcript between the positive control and p5’UTR RyfA1 are not significant. Error bars represent one standard deviation. * indicates a significant difference (where p <0.05). Relative levels were calculated using the ΔΔCt method [23] where gfp levels are normalized to the amount of rrsA present in each sample.

2.3.8 RyfB1 Overproduction does not influence plaque formation or ompC

Increased production of RyfA1 results in both a decrease in the steady-state level of ompC and an inhibition of the ability of S. dysenteriae to form plaques in a monolayer of human epithelium cells (Figure 7 and Figure 8). The finding that RyfB1 production decreased the steady-state levels of RyfA1 (Figure 9) allows for the evaluation of the impact of reduced RyfA1 levels on the relative abundance of ompC and on the ability of S. dysenteriae to form plaques in a monolayer of human epithelium cells. To this end, first, the ability of S. dysenteriae overexpressing RyfB1 from the pQE plasmid to form plaques in a Henle cell monolayer was compared to that of the strain carrying the vector control. No significant differences in plaque formation were observed between S. dysenteriae containing the vector control and the strain overexpressing pRyfB1 (Figure 11a,b). Next, using the same set of strains, the impact of RyfB1 overproduction on the steady-state levels of ompC was directly investigated using qRT-PCR. An increase in RyfB1 levels had no significant effect, either positive or negative, on the relative amount of ompC as compared to levels measured in the strain carrying the vector control (Figure 11c). Possibly, decreasing the levels of

RyfA1 through overproduction of RyfB1, may not lead to the opposite phenotypic effects.

Figure 11: Overproduction of RyfB1 does not impact plaque formation of Shigella dysenteriae. (a) Overproduction of RyfB1 by S. dysenteriae did not alter the number of plaques as compared to a vector control. (b) A monolayer of Henle cells was infected with S. dysenteriae containing a vector control or overproducing RyfB1 (pRyfB1). Overproduction of RyfB1 did not alter the plaque phenotype as from the vector control. (c) Quantitative RT-PCR was used to measure expression of RyfB1 and ompC. Significant overproduction of RyfB1 did not impact the relative levels of ompC transcript. * denotes a significant difference where p<0.001. The relative levels of RyfB1 and ompC were calculated using the ΔΔCt method[23]. The housekeeping gene, rrsA, was used as an internal control.

2.4 Discussion

The presented studies demonstrate that, unlike related species, S. dysenteriae encodes and produces two RyfA molecules, RyfA1 and RyfA2. Consistent with the positive correlation between the presence of a RyfA1-like molecule and 58 virulence, S. dysenteriae RyfA1 has now been shown to influence virulence-associated processes in this pathogen. Specifically, increased production of RyfA1 inhibits the ability of S. dysenteriae to spread from cell to cell within a monolayer of human epithelial cells, and results in near-elimination of ompC transcript, a porin protein important for cell-to-cell spread in S. flexneri [58]. Furthermore, RNA-seq analysis revealed the presence of a short divergently encoded transcript immediately upstream of ryfA1. The transcribed product of this newly identified gene, designated RyfB1, shares 17 nucleotides of complementarity to RyfA1 and when overproduced results in a significant decrease in RyfA1 levels. Taken together, these data demonstrate that

RyfA1 impacts cell-to-cell spread by S. dysenteriae, a process that is crucial for the pathogenesis of the pathogen. Furthermore, RyfA1 is now implicated in a complex regulatory network involving ompC and an additional sRNA RyfB1. As such, both

RyfA1 and RyfB1 are now implicated in the regulation of virulence-associated processes in S. dysenteriae, and thus represent two newly identified virulence factors.

The specific regulatory mechanisms and action of RyfB1 and RyfA1 have yet to be elucidated. Based on the data presented above, some predictions can be made.

RyfA1 and RyfB1 share 17 nucleotides of complementarity, a finding that leads to the model that RyfB1 influences RyfA1 levels via a direct interaction between these two molecules. Of the 17 nucleotides with complementarity between RyfA1 and RyfB1, four are contained within the RyfA1 variable region, providing a potential mechanism for specificity of regulation. Complementary nucleic acids surrounding those included in the variable region are conserved between RyfA1 and RyfA2, a finding that might 59 suggest that RyfB1 could downregulate both RyfA1 and RyfA2. This prediction is not supported by the experimental data presented above, which demonstrate that RyfB1 overproduction inhibits only RyfA1 levels (Figure 9). While the observed specificity may simply be due to the number of complementary nucleotides (RyfB1 has a greater number of complementary nucleotides to RyfA1 than RyfA2), the secondary structure of the RyfA molecules may also influence binding activity. Indeed, the variable region of RyfA1 is located almost entirely in a single-stranded region, indicating that if

RyfA1 and RyfB1 interact directly, initial pairing between these two molecules may occur at the single-stranded variable region [56], thereby mediating the specificity of

RyfB1 to modulate the levels of RyfA1 and not that of RyfA2. Further studies will be needed to confirm and understand this predicted interaction, as well as the specific molecular mechanism by which RyfA1 modulates the steady-state level of ompC in S. dysenteriae.

While OmpC is necessary for cell-to-cell spread in S. flexneri, the protein also plays a role in the survival of gastrointestinal bacteria [58,59]. The relatively small porin size of OmpC (as compared to that of OmpF) protects the bacteria by slowing the diffusion of biosalts and toxins [60–62]. While ompC is regulated by numerous factors and regulatory networks [59], the additional control of ompC by RyfA1, whether it be through impacting timing of expression or the total amount of OmpC, may endow additional benefits to enteric pathogens containing this version of the

RyfA molecule when faced with the harsh gut environment. 60

Many species of Escherichia and all species of Shigella contain a ryfA gene; however, few species contain both sibling copies, provoking interesting questions such as: what gave rise to either a single ryfA1-like iteration, a single ryfA2 iteration or the sibling pair? Why is the presence of an ryfA1-like gene associated with enteropathogens while the presence of a ryfA2 gene is associated with non-pathogenic or uropathogenic isolates? Based on the conservation of all identified ryfA genes, it is interesting to speculate that at some time RyfA played a role in a conserved process. In this model, it is proposed that the most evolutionarily ancient form of RyfA is that of

RyfA2, the form of the regulator most highly represented in non-pathogenic enterobacteria, species from which pathogenic enterobacteria evolved. For pathogenic species, the harsh environment of the gut provides a powerful selection pressure, one not experienced by non-pathogens or uropathogens, and one that over time may have selected for a gene duplication event of ryfA [63]. For this scenario to be plausible, the gene duplication event must have provided a selective advantage to the organisms having experienced it. Based on the presented studies, it is possible that the presence of a second RyfA (RyfA1-like) molecule afforded the bacteria more precise control of

OmpC production, an advantage that likely increased survival rates of the enteropathogens [64]. Based on the phylogenetic tree of Shigella and E. coli species, the gain of ryfA1-like genes is likely to have occurred concurrently, yet independently, in different species [65]. Loss of the original ryfA gene in some pathogenic species, an event leaving just the ryfA1-like gene, may have occurred later when acquired mutations afforded the RyfA1-like molecule the ability to regulate both the conserved 61 process and ompC. Continued investigations will shed light on the evolutionary history of RyfA1 and RyfA2, as well as the relative functions of each.

This study is the first in what is likely to be a series of studies to fully elucidate the molecular mechanisms and physiological consequences of the complex regulatory network revealed here. While it is clear that RyfB1 influences RyfA1 levels and that

RyfA1, in turn, influences ompC levels, many interesting questions remain. The exact molecular interplay between RyfA1/RyfB1 and ompC is the subject of current investigation. While at this point in time we cannot rule out the possibility that ryfA1 encodes a small protein under the conditions we tested [57,66,67], our data thus far support the hypothesis that both RyfA1 and RyfB1 function as sRNA regulators. Specifically, the data presented are consistent with the model of a regulatory network in which RyfB1 functions to modulate the levels of RyfA1, which in turn functions to regulate ompC levels. One sRNA regulating another sRNA is not unprecedented; however, this is the first incidence of such regulation described in Shigella [68]. An additional level of regulation by sRNA/sRNA interactions could be advantageous to a pathogen that must survive fluctuating hostile conditions by allowing for rapid and receptive changes in transcript levels.

CHAPTER 3: RYFA2 AND RYFB2

3.1 Abstract

Previously, sRNAs have been demonstrated to impact the pathogenesis of

Shigella. One sRNA, RyfA is found in all species of Shigella. Interestingly, sibling copies, termed RyfA1 and RyfA2, were discovered in S. dysenteriae. Upstream of each sibling RyfA, another sRNA exists, termed RyfB1 and RyfB2 respectively. RyfA1 has previously been demonstrated to impact the pathogenesis of Shigella and RyfB1 influences RyfA1. Here we start characterization of RyfA2 and RyfB2. Overproduction of RyfA2 did not influence the pathogenesis of Shigella. However, overproduction of

RyfB2 increased the number of plaques S. dysenteriae could form. Additionally, increasing the levels of RyfB2 resulted in a decrease of RyfA2, and unexpectedly RyfA1 suggesting a mechanism by which these sRNA siblings communicate.

3.2 Methods

Growth conditions

Strains of S. dysenteriae were grown in Luria-Bertani broth (LB) or tryptic soy broth (TSB) or agar at 37 °C. Escherichia coli K12 DH5α was grown under the same conditions when necessary. Liquid cultures of either S. dysenteriae or E. coli were grown at 200 rpm in a shaking incubator set at 37 °C. S. dysenteriae was plated on TSBA plates with 0.01% wt/vol Congo Red. Maintaining the presence of plasmids was achieved using ampicillin at a concentration of 50 μg/ml. 63

Cloning RyfB2 into an expression vector

To clone RyfB2 into an expression vector, ryfB2 was amplified from the S. dysenteriae chromosome using primers with SacI and MfeI sites (Table S1 and Table S2).

Next, the resulting amplicon was inserted into a TOPO cloning vector and transformed by heat-shocking into DH5α. The resulting insert was sequenced. Next, TOPO containing ryfB2 was digested with SacI-HF and MfeI-HF and ligated into vector pQE creating pRyfB2. The new plasmid was transformed into S. dysenteriae through heat shock.

Colonies were screened in pQE with primers pQE for and pQE rev.

Other experimental methods

Plaque assays and growth curves and qRT-PCR were performed as detailed in

Chapter 2.2 Methods.

3.3 Introduction

Shigella utilize small RNAs (sRNAs) for survival and efficient pathogenesis in the human host [69]. One set of sRNAs described previously (RyfA1 (A1) and RyfB1

(B1)) are linked to pathogenesis. Specifically, A1 overproduction limits the ability of S. dysenteriae to spread from cell to cell [70]. The A1 sibling copy, RyfA2 (A2), remains uncharacterized. Species of E. coli and Shigella that contain only the ryfA2-like gene are not associated with pathogenesis, yet, S. dysenteriae and E. coli O157:H7 retain both ryfA1 and ryfA2, indicating evolutionary importance.

As is the case with A1 and B1, upstream to ryfA2 is another putative sRNA molecule termed ryfB2. Like A1/B1, A2 and RyfB2 (B2) share a large putative binding region (17 of 19 nucleotides are complementarity) that spans the five-nucleotide variable 64 region. Binding within the five-nucleotide variable region suggests a mechanism for specificity in which B2 binds A2 and not A1. Here, we characterize the impact of A2 and

B2 overproduction on S. dysenteriae pathogenesis. Additionally, we investigate the role each sRNA may have in influencing the other RyfAs/RyfBs.

3.3 Results

3.3.1 RyfA2 overproduction does not affect plaquing of Shigella dysenteriae

While in silico analysis shows ryfA2-like genes are not linked to pathogenesis

[70], ryfA2 is conserved in S. dysenteriae and E. coli O157:H7. To explore the role of A2, a copy of ryfA2 was inserted into an IPTG-inducible plasmid (termed pRyfA2) and transformed into S. dysenteriae. IPTG inducible over-production of A2 was confirmed by qRT-PCR (Figure 12a). A significant 66- fold increase of A2 was detected. Next, the growth of S. dysenteriae overproducing A2 was monitored over 12 hours. As a control, S. dysenteriae containing pRyfA2 or the vector control were grown with or without IPTG.

While a significant growth defect was observed in stationary phase in S. dysenteriae overproducing A2, no growth defect was apparent in lag, mid-logarithmic, or logarithmic growth phases (Figure 12b).

Figure 12: RyfA2 overexpression impacts growth. (a) A plasmid containing an IPTG- inducible ryfA2 was created and termed pRyfA2. Levels of RyfA2 were measured in S. dysenteriae containing pRyfA2 or the vector control after induction with 20 mM IPTG and grown to mid-logarithmic phase. Significantly higher levels of RyfA2 was detected in pRyfA2 than the vector control. Levels were normalized to the housekeeping gene rrsA. (b) A growth curve analysis was performed with S. dysenteriae with pRyfA2 or the vector control with and without 20 mM IPTG. A decrease in growth at stationary phase was observed in S. dysenteriae containing pRyfA2 with IPTG.

While a significant growth defect was noticed in S. dysenteriae overproducing A2 at stationary phase (Figure 12b), Shigella are optimal for infection at mid-logarithmic growth phase. Plaque assays were performed with S. dysenteriae overproducing A2 to 66 examine the impact of the sRNA on pathogenesis. Overproduction of A2 had no significant effect on plaque formation under the conditions tested as compared to the vector control (Figure 13). The lack of impact on A2 overproduction on the pathogenesis of S. dysenteriae was expected, given that ryfA2-like genes are not associated with pathogenic species of Shigella and E. coli.

Figure 13: RyfA2 overproduction does not impact plaque formation. S. dysenteriae overproducing RyfA2 was used to infect a monolayer of Henle cells and compared to S. dysenteriae containing a vector control. No significant differences were observed in the number or size of plaques. (b) Plaque assays were repeated in triplicate and no statistically significant differences were noted.

3.3.2 RyfB2 overproduction increases plaquing ability of Shigella dysenteriae

Upstream of ryfA2 another putative sRNA molecule termed ryfB2 exists [70]. To understand how B2 may impact S. dysenteriae pathogenesis, a strain was built in which ryfB2 was inserted into an ITPG-inducible plasmid (termed pRyfB2) and transformed into S. dysenteriae. Next, a growth curve was performed to investigate the impact of B2 overproduction on growth (Figure 14). The growth of S. dysenteriae either containing pRyfB2 or the vector control was monitored over 12 hours. Strains were exposed to 67 either 20 mM IPTG or no IPTG. No significant growth effects were observed between any strains (Figure 14). To further investigate how B2 impacts S. dysenteriae, a plaque assay was performed in which S. dysenteriae overproduced B2. Interestingly, overproduction of B2 lead to a significant increase in the number of plaques as compared to the strain carrying the vector control (Figure 15).

1 nm Vector Control 600

OD OD pRyfB2

Vector Control + IPTG

pRyfB2 + IPTG

0.1 2 3 4 5 6 7 8 9 10 11 12 Hours Post Innoculation

Figure 14: RyfB2 overexpression does not influence growth of Shigella dysenteriae. A strain of S. dysenteriae containing pRyfB2 under an IPTG-inducible promoter was grown for 12 hours at 37 °C. Each hour the OD600 value was measured. No significant growth defects were observed in S. dysenteriae containing pRyfB2 as compared to the vector control.

Figure 15: Overproduction of RyfB2 increases plaque number. (a) A plaque assay was performed in which S. dysenteriae overproducing RyfB2 (pRyfB2) was compared to a vector control. Significantly more plaques of similar size were produced in pRyfB2. (b) Plaque assays were performed in triplicate. A Student’s T-Test was used to measure differences between the average number of plaques in the vector control compared to S. dysenteriae overproducing pRyfB2 (*p<0.05).

3.3.3 RyfB2 inhibits RyfA2 levels

Because overproduction of B2 led to more efficient plaquing, RyfB2 was evaluated in silico. Between B2 and A2, a region where 17 out of 19 nucleotides are complementary was discovered (Figure 16a). Like B1/A1 pair, the complementarity region spans the five-nucleotide variable region, suggesting a method for specific interaction between B2 and A2 (Figure 16a). To start characterizing the impact of B2 on

A2, S. dysenteriae overproducing B2 were grown to the mid-logarithmic growth phase.

Quantitative Real-Time PCR was used to measure expression levels of A2. Half of the levels of A2 were found in S. dysenteriae overproducing B2 as compared to those in the strain carrying the vector control (Figure 16b).

Figure 16: RyfB2 overproduction decrease RyfA2 levels. (a) RyfB2 and RyfA2 share a large region of complementarity at the 5’ end of RyfB2. Out of 19 nucleotides, 17 are complementary between RyfA2 and RyfB2. The box indicates the five-nucleotide variable region of RyfA2. RyfA1 and RyfA2 are mostly distinguished by this five nucleotide stretch. (b) Using qRT-PCR, RyfA2 levels in S. dysenteriae were measured and compared between a strain overproducing RyfB2 and a vector control. Calculations were performed using the ΔΔCt method, where levels of RyfB2 were normalized to the internal housekeeping gene rrsA. Student’s T-test was used to compare expression levels of the vector control to that of S. dysenteriae overproducing RyfB2. Significantly less RyfA2 was detected in the strain overproducing RyfB2 (**p<0.01).

Figure 17: RyfB overexpression inhibits RyfA levels. Overproduction of B2 resulted in a decrease of A2. Additionally, overproduction of B1 resulted in a decrease in the levels of A1 [70].

3.3.4 Swapping the RyfB2 putative interaction site

B2 overproduction not only increased plaque number but decreased A2 as well

(Figure 16b and Figure 17), it may be possible that a decrease in A2 is the cause of increased plaque number. To test our hypothesis, a RyfB2swap (B2swap) mutant was created in which the variable region interaction site was switched to B1-like interaction site predicted to reduce the interactions between B2 and A2 (Figure 18). Next, a growth curve was performed to test the impact of B2swap on the growth of S. dysenteriae. No significant growth phenotypes were observed between the strains (Figure 19). 71

Figure 18: Mutant RyfBswap binding. Five nucleotides in the putative interaction site between A2 and B2 were mutated to match the B1/A1 interaction site. The mutated bases are indicated in red. The five-nucleotide variable region of A2 is indicated by a black box. The mutations decreased the number of complementary nucleotides between B2 and A2 from 17 to 13.

1 nm

600 Vector Control Vector Control + IPTG OD at OD at pRyfB2swap pRyfB2swap + IPTG 0.1 2 3 4 5 6 7 8 Hours Post Innoculation

Figure 19: RyfB2swap overproduction does not impact the growth of Shigella dysenteriae. Growth of S. dysenteriae overproducing B2swap was compared to that of a vector control over eight hours by measuring optical density at 600 nm with and without ITPG induction. No significant growth phenotypes were observed.

To analyze the impact of B2swap on virulence, S. dysenteriae overproducing

B2swap were used in plaque assay and compared to a vector control. Unexpectedly, overproduction of B2swap resulted in a significant increase in the number of plaques. No phenotypic changes were observed between pRyfB2swap and pRyfB2 (Figure 20).

Figure 20: RyfB2swap also retains an increased plaque phenotype. S. dysenteriae overproducing RyfB2 swap was compared to that of a vector control in a plaque assay. Compared to the vector control, S. dysenteriae overproducing pRyfB2swap produced significantly more plaques. Plaque assays were performed in triplicate. Differences in the number of plaques were measured using a Student’s T-test. Significantly more plaques were found when S. dysenteriae overproduced RyfB2swap(*p<0.05).

3.3.5 RyfB2 influences RyfA1 levels

Further examination of the in silico analyses between B2/A2 and B2/A1 revealed a single nucleotide of complementarity within the five-nucleotide variable region of A1 and B2 (Figure 21a). Within the A1 and A2 molecules, the five-nucleotide variable region corresponds to a single stranded stem loop region and is predicted to be the initial seed region between A1/B1[70]. However, within the seed region of A1, a single nucleotide of complementarity exists between B2, it may be possible that A1 and B2 interact. To test if B2 impacts A1 levels qRT-PCR was performed on S. dysenteriae strains overproducing B2 and an empty vector as a control. Indeed, overproduction of B2 led to a significant decrease in the levels of A1 (Figure 21b and Figure 22).

Figure 21: RyfB2 overproduction impacts RyfA2 and RyfA1. (a) RyfB2 and RyfA2 share a single nucleotide of complementarity within the five-nucleotide variable region as indicated by the bold bar. The black box shows the variable region. (b) S. dysenteriae overproducing RyfB2 was compared to that of S. dysenteriae containing a vector control. Relative expression of RyfA1 and RyfA2 was measured using the ΔΔCt method where rrsA, a housekeeping gene, was used to normalize expression. RyfB2 overexpression resulted in a 3-fold significantly decreased levels of RyfA1 using Student’s T-test (**p<0.01) and a 2-fold significantly reduced levels of RyfA2 (*p<0.05). Relative levels of expression were determined using the housekeeping gene rrsA as an internal control.

Figure 22: RyfB2 overproduction results in a decrease of RyfA1. While B1 overproduction did not influence A2 levels, B2 overproduction results in a decrease of A1 and A2 levels. B1 and B2 are 60% identical, suggesting independent functions for each RyfB as well.

3.3.6 Other sRNA interactions

Due to crosstalk between A1 and B2/A2 (Figure 22), the many possible relationships between the sRNAs A1, A2, B1, and B2 were investigated. First, S. dysenteriae overproducing A1 or A2 were examined for their impact on B1 levels.

Neither A1 or A2 significantly altered the levels of B1 (Figure 23a). Additionally, qRT-

PCR was performed on S. dysenteriae overproducing B2 measuring B1 levels (Figure

23b). Again, no significant changes were found in the levels of B1 indicating that within this sRNA regulatory network, B1 only decreases levels of A1[70]. 76

Figure 23: RyfA1, RyfA2, and RyfB2 have no impact on RyfB1.(a) S. dysenteriae overexpressing RyfA1 or RyfA2 were measured for levels of RyfB1. No significant alterations in the levels of RyfB1 were found as compared to the vector control. (b) S. dysenteriae overproducing RyfB2 was measured for levels of RyfB1. No significant differences were found in levels of RyfB1. Levels were normalized to the housekeeping gene rrsA.

Although overexpression of A1/A2/B2 did not impact levels of B1 (Figure 23), other interactions may be occurring with the prior mentioned sRNAs. To examine the relationship between A1 and A2, each sRNA was overproduced in S. dysenteriae and the levels of A1 and A2 were measured respectively. Interestingly, overproduction of A1 77 resulted in a significant 2-fold decrease of A2 (Figure 24). Furthermore, overproduction of A2 resulted in a significant 3-fold decrease in the levels of A1 (Figure 24).

Figure 24: RyfA1 and RyfA2 overproduction impacts each other. S. dysenteriae overproducing RyfA1, RyfA2, or containing an empty vector control were compared for levels of RyfA1 and RyfA2 with qRT-PCR. Levels of RyfA1 and RyfA2 were normalized to the internal housekeeping gene rrsA and relative expression values were calculated using the ΔΔCt method. Overproduction of RyfA1 resulted in a significant 1.5 fold decrease of RyfA2 (*p<0.05) and overproduction of RyfA2 resulted in a 500 fold decrease of RyfA1. Significance was calculated using the Student’s T-Test (**p<0.01, ***p<0.001).

3.4 Discussion

The RyfAs and RyfBs are dynamically involved with one another through unexpected means. While A1 overproduction decreases the ability of S. dysenteriae to 78 spread from cell to cell[70], B2 overproduction increases the ability of S. dysenteriae to form plaques. Additionally, overexpression of B2 decreases both A2 and A1, while overproduction of B1 impacts only levels of A1 [70] (Figure 25). Lastly, overproduction of A2 decreases A1 and vice versa.

Unexpectedly, the data shows B2 overproduction lowers levels of both A1 and

A2. Nucleotides surrounding the variable region are complementarity between B2 and

A1 (much like B1 and A2) but in addition, the central base of the variable region in A1 compliments B2 (ACGAA) while no base pairs in the A2 variable region compliment B1.

Perhaps this lone base pair increases the complementarity just enough within the single- stranded stem loop to allow the interaction of A1 and B2. Additionally, swapping the interaction site of the B2 variable region (ACGAA  GGGGG) did not affect plaque phenotype of B2swap overproduction likely because the swap may have retained the ability to decrease A1. However, merely reducing A1 levels is not enough to explain the increased plaque production as B1 overproduction, which decreases A1 explicitly, did not cause the same phenotype[70]. It is possible that regions outside the variable region are essential for regulatory activities of RyfAs and RyfBs. Perhaps B2 affects targets outside of A1 and A2 leading to increased plaque production. B1 and B2 are around 60% identical, suggesting that B1 and B2 are likely changing a range of different targets from one another.

In addition to the interplay between B2 and the RyfAs, A1 and A2 impact the levels of one another. Overproduction of A2 leads to a sharp decrease in A1 and overproduction of A1 leads to a smaller, but still significant, decrease in A2. Neither A1 79 nor A2 overproduction leads to changes in the levels of B1. Overall B1, B2, and A2 all control production of A1(Figure 25). Overproduction of A1 decreases cell to cell spread and OmpC (outer membrane protein C) production, it is likely necessary to tightly control regulation of OmpC through control of A1 [70]. The concentrations of the different outer membrane porin proteins, OmpC and OmpF, are central in responding to changes in the environment [59]. OmpC is a slightly smaller porin protein (1.1 nm) that is activated in response to high osmolarity conditions while OmpF is a marginally larger porin protein

(1.2nm) that is activated under low osmolarity conditions [59]. Between these two porin proteins, a balance must be achieved in which nutrients can be acquired by Shigella but at the same time, adequate protection remains to stop bile-salts and other factors within the gut destroying the Shigella. To this end, a complicated series of sRNAs functioning to regulate the production of porins could give an advantage in tempering small changes in

OmpC.

In silico analysis demonstrates whenever a ryfA-like gene is present, a ryfB-like gene is also present. In species with a single copy of ryfA, a single copy of ryfB is present. It is interesting that S. dysenteriae and E. coli O157: H7 retain both ryfA1 and ryfB1 as well as ryfA2 and ryfB2, especially since B2 can decrease levels of A2 and A1.

Perhaps having an additional control mechanism over A1 besides B1 is advantageous or even necessary in pathogens such as S. dysenteriae. When the ryfA or ryfB genes are expressed is the subject of further investigations and will likely reveal the importance of maintaining two sets of ryfA/ryfB.

Figure 25: RyfA and RyfB regulation schematic. RyfB1, RyfB2, and RyfA2 all regulate RyfA1 levels. RyfA1 overproduction results in a decrease in outer membrane protein C, the smaller of the two main outer membrane proteins in Shigella. RyfA1 overproduction decreases RyfA2 and vice versa. Interestingly, RyfB2 regulates not only RyfA2 but RyfA1 suggesting importance for maintaining both sets of ryfA/ryfB within the chromosome.

CHAPTER 4: DOWNSTREAM TARGETS OF RYFA1 AND RYFA2

4.1 Abstract

The characterized molecular factors associated with Shigella pathogenesis are mostly proteins. Riboregulators involved with shigellosis are less understood. RyfA1 and

RyfA2 are two sibling small RNAs (sRNAs) in S. dysenteriae that are implicated in pathogenesis. Overexpression of RyfA1 limits the ability of S. dysenteriae to spread from cell to cell in a plaque assay. Overproduction of RyfA2 does not impact S. dysenteriae in plaque assay, however RyfA2 overproduction decreases levels of RyfA1. Elucidating the regulatory targets of RyfA1 and RyfA2 may aid in understanding how each sRNA affects

Shigella pathogenesis. The transcriptome and proteome of S. dysenteriae with and without overproducing RyfA1 or RyfA2 were established and analyzed to identify those virulence factors that are influence by RyfA1 and/or RyfA2.

4.2 Material and Methods

Next generation RNA-sequencing

The relative abundance of all transcripts present in S. dysenteriae carrying pRyfA1, pRyfA2 or the vector control were measured by RNA-sequencing analysis performed as detailed in Chapter 2.

TMT-10 plex

S. dysenteriae containing pRyfA1, pRyfA2, or the vector control were streaked on tryptic soy broth agar (TSBA) plates with 0.01% wt/vol Congo Red and 50 μg/ml ampicillin. Plates were incubated overnight at 37 °C. Following overnight growth, three colonies from each strain were used to inoculate tubes with Luria-Bertani (LB) broth and 82

50 μl/mL ampicillin. Culture tubes were incubated at 30 °C and shaken at 200 rpm overnight. For each culture tube 25 ml of LB, 50 μl/ml ampicillin, and 0.01% deoxycholate (DOC) were inoculated with 500 μl of overnight culture. Cells were grown at 37 °C and shaken at 200 rpm for approximately four hours until the cultures reached the mid-logarithmic growth phase. After reaching the mid-logarithmic growth phase, 24 ml of culture was pelleted by centrifugation at 12,000g for 5 minutes at 4 °C. The supernatant was discarded and the pellet was resuspended in 1 ml of standard wash buffer composed of 10 mM Tris, and 5 mM magnesium acetate. Spinning and wash steps were repeated three times. After the final spin, all standard cell wash buffer was removed via pipette. The pellet was subsequently resuspended in standard cell lysis buffer composed of 30 mM Tris, 2 M thiourea, 7 M urea, 4% w/v of CHAPS, and 5 mM DTT, at a pH of

8.5. Each resuspended was put on ice for 10 minutes and then incubated overnight at -80

°C. The following day six freeze-thaw cycles were performed in which the tubes were thawed at 37 °C for 30 minutes and then frozen at -80 °C for an additional 30 minutes.

After the sixth and final repeat, scrapings from each sample were used to streak on TSBA with 0.01% w/v % Congo Red and 50 μl/ml ampicillin plates. Plates were incubated overnight at 37 °C. After no growth was observed on the plates, samples were sent on dry ice to the Proteomics and Metabolomics Facility at the Center for

Biotechnology/University of Nebraska-Lincoln. At the facility, samples were alkylated and reduced, followed by a trypsin digestion. Next, samples were desalted and dissolved in 0.1 M TEAB for Tandem Mass Tag (TMT) labeling. Then, triplicate samples were mixed and using high-pH reverse phase chromatography the samples were fractionated. 83

Fractionations were run by nanoLC-MS/MS using a 2 hour gradient on a 0.075 mm x 250 mm C18 Waters CSH column feeding into a Q-Exactive HF mass spectrometer. Samples were subsequently analyzed with Mascot (Matrix Science London, UK, version 2.6.1).

The software searched the cRAP_20150130 databased and the protein FASTA Shigella

SD_197 database (http://www.uniprot.org/proteomes/UP000002716; 2017013, 5728 entries) and the S. dysenteriae virulence plasmid database

(http://www.ncbi.nlm.nih.gove/nuccore/CP000035.1?report=fasta) assuming the digestion enzyme trypsin. A fragment ion mass tolerance of 0.06 Da and a parent ion tolerance of 10.0 PPM was used to search Mascot. As a fixed modification TMT and carbamidomethyl were specified. Variable modifications used were oxidation of methionine and cysteine. A decoy database of random peptides was used to create a false discovery rate of 1% and 5% respectively. Proteome Discoverer (Thermo, version 2.2) was used to quantify proteins. The average S/N was set to 10 and the co-isolation threshold was set at 50%. Ratio and log2 were also recorded.

4.3 Introduction

Proteomic and transcriptomic methods can be used to identify a variety of targets and aid in outlining different pathways essential to bacterial virulence and survival. sRNAs are a class of riboregulators which contribute to virulence of Shigella, yet many of their specific regulatory targets are unknown [69]. sRNAs can impact targets in a variety of ways. First, sRNAs can bind with the target mRNA directly and result in degradation, or in rarer cases stabilization, of the transcript [46,56,71]. Secondly, sRNAs can bind with the target mRNA and induce structural changes in the molecule that 84 influence translational efficiency [46,56,71]. Although sRNAs are known to impact their targets through other mechanisms, mediating targets post-transcriptionally via stability or degradation are the two main pathways by which sRNAs impact the production of regulated products. By elucidating the transcriptome (all transcript levels present in the cell) and the proteome (all protein levels present in the cell) regulatory network patterns may emerge to aid in identification of the direct and indirect targets of RyfA1 and RyfA2.

Identification of sRNA targets not only helps in understanding the function of the sRNA but also in providing potential information for the development of novel therapeutics in treating bacterial diseases.

Two sibling sRNAs, RyfA1 and RyfA2 have previously been characterized by overexpressing each sRNA in S. dysenteriae and observing the impact on the ability of the pathogen to complete several virulence-associated processes using in vitro tissue culture-based assays [70]. Overproduction of RyfA1 results in a decrease in the ability of

S. dysenteriae to spread from cell to cell [70]. On the other hand, abundant RyfA2 levels do not impact S. dysenteriae in a plaque assay. However, increasing the levels of RyfA2 does significantly decrease RyfA1, suggesting an auxiliary function for RyfA2 on

Shigella virulence.

To find targets of RyfA1 and RyfA2, the proteome and transcriptome of S. dysenteriae overproducing either respective sRNA was analyzied. Analysis of the data revealed overlapping and specific targets for each sRNA including a number of known virulence determinants. While the genes impacted by RyfA1 and RyfA2 will need further 85 investigation, the influence that these sRNAs appear to have on S. dysenteriae is substantial.

4.4 Results

4.4.1 Transcriptome of RyfA1 and RyfA2

One mechanism by which sRNAs influence the outcome of gene expression is through targetting mRNAs post transcriptionally to degrade or stabilize the target transcript. S. dysenteriae overexpressing either RyfA1 or RyfA2 was analyzed through

RNA-sequencing analysis and the relative abundance of each transcript compared to that in the wild-type strain to identify those transcript influencec by RyfA1 and/or RyfA2

(Table 2 and Table 3). Triplicate samples were combined to one tube before RNA- sequencing analysis. To add stringency for our RNA-seq data set, genes which were upregulated or down-regulated by at least 3-fold as compared to the vector control were considered significant (rather than a 2-fold cut off from uncombined samples). Overall, the transcriptomes of RyfA1 and RyfA2 were distinct from one another. Only 18 of the

424 targets were shared by these sRNAs (Table 4). Of these shared genes, two were differentially regulated, one of which is a known virulence factor in S. flexneri, csrB[6].

Twelve known virulence factors were impacted by RyfA1 overproduction (Table

2) including mxiK, mxiL, mxiH, ipgB1, ipgF, ipgA, spa-orf10, csrC, ospB, ospD3, spa29, and ipaH1.4[6,72–83]. All genes listed (except csrC) are encoded within the entry region of the large virulence plasmid. The entry region of Shigella contains genes encoding components of the type three secretion system and effectors/chaperones necessary for invading into a host cell (see review [3]). sRNAs often interact with their target via 86 complementary binding. However, in silico analysis revealed no significant nucleic acid complementarity between RyfA1 and any of the prior mentioned mRNA transcripts.

Possibly, RyfA1 is modulating a regulator post-transcriptionally.

Compared to RyfA1, RyfA2 overexpression in S. dysenteriae influenced few virulence genes (Table 3). One gene altered by RyfA2 overexpression was ospC4.

Located in the entry region, ospC4 is 96% identical to ospC2 and ospC3 [84]. However, a frameshift deletion within ospC4 is thought to make this gene product inactive and no further experimental investigations have been performed with this gene.

Table 2: Downstream targets of RyfA1 overexpression through RNA-seq analysis

RNA isolation was performed on a strain of S. dysenteriae overproducing RyfA1 and RNA-seq analysis was performed. Genes either upregulated (green) or downregulated (red) by at least 3-fold were considered significant. Genes involved with pathogenesis are bolded. Of note, ompC was downregulated ~7,000 fold.

Table 3: Downstream targets of RyfA2 overexpression through RNA-seq analysis

RNA-seq analysis was performed on S. dysenteriae overproducing RyfA2. Genes upregulated (green) and downregulated (red) by at least 3-fold were considered significant. Genes in bold are virulence factors.

Table 4: Transcriptomic data of RyfA1 and RyfA2 compared

Only 18 genes were impacted by both RyfA1 and RyfA2 overproduction. Out of the 18 genes impacted, two were differentially expressed. The virulence factor csrB increased with RyfA2 overproduction and decreased in the presence of RyfA1 overproduction. The second gene differentially impacted by both RyfA1 and RyfA2 was yeiE, which is not a known virulence factor.

4.4.2 Proteome of RyfA1 and RyfA2

To compare proteins impacted by either RyfA1 or RyfA2 overproduction, triplicate cultures of S. dysenteriae overproducing either sRNA or containing the vector control were grown to mid-logarithmic growth phase in liquid media. Total protein was extracted and submitted for proteomic analysis. Initially, the ratio of proteins was compared between each strain. Proteins which had ratios with a significant difference

(p<0.05, Student’s T-Test) between samples and also had an increase of at least 1.2 fold or a decrease of -0.45 fold decrease (also known as a -1.2 fold decrease) were examined further. 90

Between RyfA1 and RyfA2 overproduction, 387 proteins were significantly impacted. RyfA1 overproduction resulted in a total of 159 up-regulated proteins and 25 down-regulated proteins (Table 5) and many were involved with the outer membrane

(Table 6). Overproduction of RyfA2 in S. dysenteriae resulted in 184 up-regulated proteins and 19 down-regulated proteins (Table 7) of which some are involved with heme and iron regulation (Table 8). Interestingly, 144 proteins were influenced by both RyfA1 and RyfA2, indicating that, unlike the RNA-seq data, RyfA1 and RyfA2 share more targets at the protein level (Table 9).

Table 5: Proteins significantly impacted by RyfA1 overexpression

Proteins extracted from S. dysenteriae overproducing RyfA1 were compared to that of a control vector. First, the ratio by which the protein levels of RyfA1 to the vector control were compared and ratios which a significant difference (p<0.05) were gathered. Next, proteins were further analyzed by their level of fold change as compared to the vector control. Proteins which were significantly downregulated by 0.5-fold are dark red. Proteins significantly downregulated >0.59-fold are light red. Significantly upregulated proteins by <1.2 fold are light green. Proteins significantly upregulated between 1.5 and 2.0 fold are medium green. Proteins upregulated significantly more than 2.0 fold are dark green. Genes in bold are considered virulence factors of Shigella. To determine the p- value a Student’s T-Test was used.

Uniquely to RyfA1, proteins which were impacted were found to be involved with membrane integrity. These proteins included OmpA, PspC, PspB, PspA, CysW,

YegH, GlpF, SDY_3076, OmpF, OmpC, and YgiM. Of the membrane associated proteins exclusively impacted by RyfA1, two distinguishable operons are observable. The first being the phage-shock-protein (Psp) system named for the initial observed upregulation when bacteria are in the presence of phage [85]. The Psp system is involved with stabilizing the cytosolic membrane under stress conditions (such as filamentous 92 phage infection, extreme temperatures, osmolarity changes, the mislocalization of secretin proteins) by maintaining the proton motive force [85,86]. Furthermore, the Psp system is induced when S. flexneri are infecting a macrophage [87], understandably considering the hostile environment of the macrophage. Three of the six members of the

Psp operon were upregulated at the protein level when RyfA1 was overproduced, indicating S. dysenteriae may be experiencing membrane associated stress, or is picking up a false signal of membrane associated stress.

The second noticeable membrane associated operon impacted uniquely to overproduction of RyfA1 were the porin proteins OmpC and OmpF. The OmpC porin protein is has a slightly smaller porin diameter (1.1 nm) compared to the OmpF porin protein (1.2 nm) [59]. Both porin proteins are regulated by the two component regulatory system EnvZ/OmpR where under high osmolarity conditions, ompF transcription is repressed and ompC transcription is turned on. In low osmolarity conditions, ompF transcription is turned on and ompC transcription is repressed. Overproduction of RyfA1 uniquely resulted in a decrease in OmpC at the protein (0.1

(7,000 fold decrease) and an increase in OmpF protein (>1.2 fold) while OmpR and EnvZ remained unaffected [58,59,88,89]. Again, these results suggest overproduction of RyfA1 impacts membrane associated stress of the bacterial cell.

Table 6: Membrane proteins associated uniquely with RyfA1 overproduction Protein Function Regulation Reference Porin protein, “a molecular Swiss army knife”, OmpA known to be involved with adhesion and invasion 1.5-2 fold [90] of EHEC PspC May sense conditions to activate PspA >2 fold [85] PspB May sense conditions to activate PspA >2 fold [85] Involved with membrane integrity and proton PspA >2 fold [85] motive force. CysW Sulfate transport permease protein >1.2 fold YegH Putative transport protein >1.2 fold GlpF Glycerol transport protein >1.2 fold [91] Putative hydrogenase subunit (predicted to be SDY_3076 >1.2 fold transmembrane) OmpF Porin protein 1.2 nm >1.2 fold [59] OmpC Porin protein 1.1 nm <0.1 fold [59] YgiM Uncharacterized transmembrane protein <0.5 fold

Table 7: Proteins significantly impacted by RyfA2 overexpression

Protein levels were compared between S. dysenteriae overexpressing RyfA2 and an empty control vector. Proteins which were significantly downregulated (p<0.01) are dark red. Proteins significantly downregulated (p<0.05) are light red. Significantly upregulated proteins (p<0.01) are dark green. Proteins significantly upregulated (p<0.05) are light green. Genes in bold are considered virulence factors of Shigella. Significance was determined using Student’s T-Test.

Of note, some proteins associated with pathogenesis of Shigella were uniquely impacted by RyfA2 overproduction. These proteins include MxiD, OspE1, IcsA, IpaH_5,

IpaH4.5, IpaH7.8, IpaH1.4, and IpaH9.8. These virulence factors were significantly upregulated by at least 1.2-fold. Interestingly, most of these virulence factors specifically impacted by RyfA2 overproduction are secreted secondary effectors from the T3SS during an infection[3]. OspE1, IpaH_5, IpaH4.5, IpaH7.8, and IpaH9.8 and are mainly associated with dampening the effects of the immune system (excluding OspE1). The two exceptions are MxiD, which a component of the T3SS necessary for the secretion of all

Ipa invasions [78,92] and IcsA. While a RyfA2 overproduction plaque phenotype is not observed, possible misregulation of these secondary effectors may impact the survival of

S. dysenteriae in a macrophage, as compared to wild-type S. dysenteriae. 95

Furthermore, in addition to the acid resistance genes impacted by both RyfA1 and

RyfA2 overproduction, RyfA2 overproduction resulted in specifically influencing some proteins associated with acid resistance including PhoN1 (which is thought to have homology to an acid prophosphotase) and ClcA. In E. coli ClcA has been characterized as a H+/Cl- exchange transporter [93], which may be important for maintaining cell homeostasis under low pH conditions.

Lastly, overproduction of RyfA1 and RyfA2 influence iron and heme associated genes. Unique to RyfA2 overproduction and heme/iron regulation the following proteins were impacted; IscR, FeoC, CcmA, NapG, and NapC. The proteins and their functions are listed below (Table 8).

Table 8: RyfA2 specific iron/heme regulation proteins Protein Function Regulation Reference IcsR Represses the icsRSUA operon. This operon is needed to <0.59 [94] synthesize Fe-S clusters. FeoC Characterized in Salmonella FeoC has oxygen <0.59 [95] responsive Fe-S clusters. Mediates ferrous iron transport. CcmA Part of an ABC transporter complex, at one point >1.2 thought to export heme NapG Fe-S protein involved with electron transfer >1.5 [96] NapC Ferric reductase activity >2.0 [97]

Table 9: RyfA1 and RyfA2 overexpression protein level comparison

Proteins influenced by both RyfA1 and RyfA2 were compared for differences in regulation. Darker red indicates down-regulation (fold change <0.2) while lighter red indicates downregulation (fold change >0.59). Dark green indicates up-regulation (fold change >2) and light green also indicates up-regulation (fold change >1.2). Genes in bold are virulence factors of Shigella.

Of interest, both RyfA1 and RyfA2 overproduction increase by at least 1.5-fold the protein VirB, one of the master regulators of Shigella pathogenesis [21,32,98,99].

VirB is responsible for upregulating many genes involved with the entry region on the virulence plasmid. Many downstream targets of VirB at the protein level were upregulated by at least 1.2 fold though RyfA1 and RyfA2 overproduction in S. dysenteriae including; IcsB, IcsP, IpaA, IpaC, IpaD, IpgA, IpgB1, IpgB2, IpgD, IpgE,

MxiA, MxiC, MxiE, MxiG, MxiH, MxiI, MxiK, MxiL, MxiM, MxiN, OppF, OspB,

OspC2, OspC3, OspD1, OspD2, OspF, Spa13, Spa15, Spa24, Spa32, Spa33, Spa40,

Spa47, Spa-orf10, and VirA [72,80,81,100–110]. The numerous downstream targets of

VirB are involved with production of the T3SS, creation of secreted virulence factors, 97 and entry/spread once Shigella are in the host cell (see [1] for review). Slight increases in the transcripts of VirB targets (less than 3-fold) may have not been detected as significant in the next generation sequencing analysis yet may have been enough for an increase in translation to occur, and thus, result in approximately 1.5-fold more protein.

Together RyfA1 and RyfA2 overproduction also result in the impact of some genes involved with iron homeostasis regulation. These proteins include Ftn, DppA,

FrdB, YecL, AcnB, and CcmE. All proteins are upregulated, especially Ftn and DppA of which increase by >2 and >1.5-fold respectively. Both Ftn and DppA are involved with heme binding. Ftn is a bacterial ferritin which is capable of forming a spherical shape to contain iron and DppA has been described in E. coli as a heme binding dipeptide permease protein [111,112]. Likewise, YecL is also a bacterial ferritin, however, compared to Ftn, it was only upregulated 1.2-fold. FrdB, AcnB, and CcmE are all involved with heme or binding iron as well [113–117]. Interestingly, CcmE is the only protein with opposing regulation. For the majority, however, the upregulation of proteins involved with binding heme/iron and utilizing these two molecules may give us a clue as to why RyfA1 and RyfA2 are important in Shigella as potential sensors of the microbial environment.

RyfA1 and RyfA2 overexpression also impacts a few proteins involved with acid resistance including HdeA and HdeB. Both HdeA and HdeB are characterized in E. coli as acid chaperone proteins which become active during an acid stress response and are thought to be important for enteric pathogens [118,119]. Each of these chaperones is downregulated when either RyfA1 or RyfA2 is overexpressed.

Table 10: Transcript/protein levels significantly impacted by RyfA1 or RyfA2 overproduction

Protein and transcript levels were compared between the RNA-sequencing data and the TMT-10plex data for both RyfA1 and RyfA2. Red indicates downregulated genes or decreased levels of protein while green indicates upregulated genes or increased levels of protein. For the protein data (dark red < 2 fold 1.2 fold dark green < 2 fold). For the transcriptomic data (dark red < 5 fold, light red < 3 fold, light green > 3 fold, dark green > 5 fold). More stringent parameters were used to select for genes of interest in the transcriptomic data because combined triplicate samples were used in processing the RNA.

4.5 Conclusions and Discussion

Analysis of the proteomic and transcriptomic profiles of S. dysenteriae overproducing either RyfA1 or RyfA2 revealed proteins and genes involved with virulence. Of interest are CsrB, OmpC/OmpF and other membrane associated proteins,

VirB and targets of VirB, as well as proteins involved with iron homeostasis and acid resistance genes. [6,33,38,44,58,89,99,120,121]. While the direct and indirect interactions between the sRNAs and their targets remain unclear, in silico data may start to reveal putative regulation of virulence associated pathways.

One virulence factor, CsrB, was differentially expressed in the transcriptomic data upon overproduction of RyfA1 or RyfA2. RyfA1 overproduction downregulates csrB while RyfA2 upregulates csrB. CsrB, or carbon storage regulator B, is involved with

Shigella pathogenesis [6,13,15]. Specifically, CsrB overproduction results in S. flexneri losing the ability to invade efficiently in a plaque assay [6]. On the other hand, decreasing

CsrB results in a hypervirulent strain of S. flexneri [6]. RyfA1 overproduction decreases

CsrB, yet, S. dysenteriae becomes less virulent. Likely, the loss of the ability to spread from cell to cell is not a result of the influence of RyfA1 on the carbon storage regulatory system. Additionally, in silico analysis did not reveal any putative binding sites between

RyfA1 and genes involved with carbon storage regulation.

Another protein influenced by both RyfA1 and RyfA2 was VirB, one of the master regulators of virulence gene expression in Shigella [32,33,99]. VirB levels increased as a result of both RyfA1 and RyfA2 overproduction, and many of the targets of VirB were also positively impacted at the protein level. Alterations in VirB expression 100 may result in poorly timed expression of virulence factors and could impact the efficacy of Shigella infection. However, both RyfA1 and RyfA2 overproduction similarly increased VirB, but only RyfA1 impacted pathogenesis in plaque assay [70], indicating the influence RyfA1 and RyfA2 have on VirB may not be responsible for the defect in cell to cell spread. At the same time, and as is true for all dis-regulated genes, aberrant expression of VirB may be critical for the survival of Shigella in the human host and not observable in plaque assay conditions, and thus, still have impact on virulence.

Virulence regulation of Shigella is not only controlled by VirB, but also controlled by the environmental responses to acid and iron availability[11,27,45,122]. In general, genes which are involved with heme/iron regulation and homeostasis seem to be upregulated. Acid resistance proteins are a little less clear in their directional regulation.

Environmental responses are crucial for the survival and deployment of virulence factors for Shigella[9,26,27,30,45,123]. Perhaps, RyfA1 and RyfA2 may be important in sensing changes in iron/heme availability or pH during an infection. Further experimentation will help elucidate this potential function.

In addition to sensing iron availability and pH, RyfA1 may be a part of a regulatory network influencing the EnvZ/OmpR two component regulatory system

[59,124]. OmpR protein is a highly conserved transcriptional regulator which is best known for response to high or low osmotic conditions [59,124]. OmpR can mitigate the response to osmolarity through alteration of ompF and ompC transcription. OmpF is a smaller porin protein and OmpC is a slightly larger one [59]. In the promoter region of ompF and ompC, OmpR binding sites are present, termed F1, F2, F3, F4 and C1, C2, C3, 101 respectively[89]. In conditions of low osmolarity, EnvZ phosphorylates OmpR and then

(OmpR~P) binds to F1, F2, F3, and C1[89,125]. The binding of OmpR~P to these sites stimulates the production of OmpF and inhibits OmpC. Under conditions of high osmolarity research models and experimental data demonstrate OmpR~P levels increase in the cell [125]. OmpR~P then binds to F1, F2, F3, and F4 causing a bend in the promoter region of ompF and preventing transcription[125]. In the promoter region of ompC, all sites become occupied by OmpR~P leading to an increase in transcription

[89,121,125].

RyfA1 overproduction decreases ompC and increases OmpF, two targets of

OmpR. Between RyfA1 and ompR mRNA, a putative binding site exists at the 3’ end of

OmpR, about 20 base pairs upstream of envZ (Figure 26). Additionally, this putative binding site overlaps the five nucleotide variable region. RyfA1 may be influencing the

OmpR~P activities through controlling envZ translation (Figure 26). 102

Figure 26: RyfA1 may have a putative binding site near the 3’ end of ompR. (a) In silico analysis revealed a potential binding site between the sRNA RyfA1 and ompR mRNA. Furthermore, the putative binding site overlaps the five nucleotide variable region of RyfA1. 12 of 15 nucleotides are complimentary. (b) The putative binding site of RyfA1 to ompR is located at the 3’ end and approximately 20 bases before the stop codon/start codon of ompR/envZ. (c) Although the ribosomal binding site of envZ is already likely weak, RyfA1 may be inhibiting translation of envZ further by altering the mRNA structure.

OmpR and EnvZ are controlled by the same ompB promoter and are translated from a single polycistronic mRNA [126]. Researchers have demonstrated that the levels 103 of OmpR in E. coli are significantly higher than the levels of EnvZ [126]. Specifically, a single cell was determined to have approximately 3,500 molecules of OmpR and 100 molecules of EnvZ. In Salmonella other researchers have demonstrated that translation is approximately 10 times more efficient of OmpR than EnvZ [127]. Additionally, it is thought that because envZ/ompR are transcribed at the same rate, post-transcriptional regulation is key [126].

While altered levels of expression of envZ or ompR are not observed when RyfA1 is overexpressed, protein levels seem to be unaffected as well. However, the alterations in the low levels of EnvZ may not be detectable from our proteomic data. In our model,

RyfA1 binds to the ompR/envZ transcript near the translation start codon/translation stop codon of EnvZ and OmpR respectively (Figure 26c). Potentially, that interaction between

RyfA1 and the polycistronic mRNA may alter the structure in such a way that translation of EnvZ further decreases due to hiding the already weak ribosomal binding site. With less EnvZ present in the cell, less OmpR is phosphorylated leading to a decrease in

OmpR~P. Less OmpR~P results in less ompC transcription. However, OmpR~P must not be lacking entirely since OmpF translation and expression is still observed. The slight changes in envZ translation encompass what is known about sRNAs, in that they impact expression changes gradually.

Overexpression of RyfA1 leads to a severe decrease in ompC as well as an increase in OmpF, mimicking conditions of low osmolarity. It could be that RyfA1 overproduction is signaling low osmolarity conditions even though Shigella are 104 experiencing high osmolarity conditions in plaque assay leading to a decrease in cell to cell spread. Further experimentation would be needed to prove this model.

The possible interactions of RyfA1 and envZ translation could shed light on a previously unknown mechanism, one which RyfA1 induces a low osmolarity response in the cell. Previous research has shown mutations in the ompB locus (EnvZ/OmpR) resulted in a decrease in the virulence gene regulon [128]. However, exactly how

EnvZ/OmpR is interacting with the virulence plasmid remains unknown [58,128,129]. If

RyfA1 is signaling a low osmolarity environment to the cell, inhibition of virulence factors may inadvertently occur. Additional regulation of virulence factors via EnvZ may be advantageous for a pathogen such as Shigella in that, the bacteria response to changes in osmolarity may be quicker and better suited for pathogenesis in the gut.

The proteome and transcriptome of S. dysenteriae overproducing either RyfA1 or

RyfA2 may give some clue as to the role of these sRNAs in virulence of the pathogen.

The importance of both molecules in S. dysenteriae as compared to other Shigella species or pathogenic E. coli species is still not understood. RyfA1 and RyfA2 overexpression significantly impact several virulence factors in S. dysenteriae. Exploration of the possible relationships between the sRNAs and their putative targets may solidify the models described above, or, reveal unexpected regulatory patterns. Additionally, understanding the molecular mechanisms by which RyfA1 and RyfA2 regulate target genes may suggest reasons for evolutionary conservation of both sRNAs in S. dysenteriae.

105

CHAPTER 5: CONCLUSIONS AND GENERAL DISCUSSION

The great success of antibiotics in treating bacterial disease resulted in an equally successful complacency of studying how bacteria cause disease and latter, how bacteria develop resistance to antibiotics. While our understanding of the molecular causes of disease and resistance has substantially grown over the past few decades, new discoveries regarding molecular regulation by bacterial pathogens and emerging pathogens will be essential for keeping up with ever-evolving bacterial populations.

The greater our molecular toolkit and understanding of the bacteria which inhabit humans and the environment that we live in, the better equipped our physicians will be in preventing and combating bacterial infection and disease. Ribo-regulators are regarded as a small part of bacterial pathogenesis regulation, however, further analysis of infectious organisms through next generation sequencing will continue to reveal a surprisingly large role ribo-regulators play in all aspects of bacterial functions[130].

The sRNAs described here (RyfA1, RyfA2, RyfB1, RyfB2) have further expanded the frontiers of small RNA research due to several unique attributes. First,

RyfA1/RyfB1 and RyfA2/RyfB2 are two of just a few instances of RNA regulating another sRNA. Remarkably, RyfB2 regulates two sRNAs – this is significant as it represents the first identified instance of a single sRNA regulating more than one other sRNA target. Second, RyfA1/RyfA2 are a part of class of sRNAs known as sibling sRNAs. While this class of sRNAs is not entirely understood, in many cases, sibling sRNAs are only known to have redundant functions based on test conditions [24].

However, RyfA1 and RyfA2 present an opportunity to study two sRNAs, which are 95% 106 identical, yet have different functions. Such studies will advance the foundation of the field as precious few examples of sibling sRNAs with unique function have been identified [24].

The exact molecular interaction between the RyfAs and RyfBs has only been analyzed in silico and remains to be investigated further. Are the interaction sites between

RyfA and RyfB as predicted? Does RyfB binding to RyfA induce immediate degradation or does RyfB binding cause sequestration, and thus alter function of the corresponding

RyfA molecule(s)? Does the single base pair within the five nucleotide variable region of

RyfA1/RyfB2 allow for interactions to occur between the two?

In addition to the questions about the exact molecular interactions of these sRNAs, the environmental conditions under which the RyfAs and RyfBs are expressed are still unknown. Understanding the environmental conditions in which RyfA1 downregulates RyfA2 and vice versa may better explain why each sRNA was conserved in S. dysenteriae. Perhaps, timing (influenced by the environment Shigella encounters) affects the expression of each sRNA (RyfB1 and RyfB2 included) and modulates the infection processes.

The four sRNAs are just a small part of the molecular regulation which S. dysenteriae utilizes to cause infection within the human host. Perhaps better characterization of these molecules will result in a greater understanding of Shigella pathogenesis, specifically S. dysenteriae, and importantly inform the generation of novel therapeutics to treat or prevent infections with this and related bacterial pathogens. 107

REFERENCES

1 Jennison, A.V. and Verma, N.K. (2003) Shigella flexneri infection: pathogenesis

and vaccine development. FEMS Microbiol. Rev. 28, 43–58

2 Wassef, J.S. et al. (1989) Role of M-cells in initial antigen uptake and in ulcer

formation in the rabbit intestinal loop model of shigellosis. Infect. Immun. 57, 858–

3 Schroeder, G.N. and Hilbi, H. (2008) Molecular Pathogenesis of Shigella spp.:

Controlling Host Cell Signaling, Invasion, and Death by Type III Secretion. Clin.

Microbiol. Rev. 21, 134–56

4 Skippington, E. and Ragan, M.A. (2012) Evolutionary Dynamics of Small RNAs

in 27 Escherichia coli and Shigella Genomes. Genome Biol. Evol. 4, 330–345

5 Peng, J. et al. (2011) An Integrated Approach for Finding Overlooked Genes in

Shigella. PLoS One 6, e18509

6 Gore, A.L. and Payne, S.M. (2010) CsrA and Cra influence Shigella flexneri

pathogenesis. Infect. Immun. 78, 4674–82

7 Giangrossi, M. et al. (2010) A novel antisense RNA regulates at transcriptional

level the virulence gene icsA of Shigella flexneri. Nucleic Acids Res. 38, 3362–

3375

8 Tran, C.N. et al. (2011) A multifactor regulatory circuit involving H-NS, VirF and

an antisense RNA modulates transcription of the virulence gene icsA of Shigella

flexneri. Nucleic Acids Res. 39, 8122–34

9 Murphy, E.R. and Payne, S.M. (2007) RyhB, an Iron-Responsive Small RNA 108

Molecule, Regulates Shigella dysenteriae Virulence. Infect. Immun. 75, 3470–7

10 Broach, W.H. et al. (2012) VirF-Independent Regulation of Shigella virB

Transcription is Mediated by the Small RNA RyhB. PLoS One 7, e38592

11 Kouse, A.B. et al. (2013) RNA-mediated thermoregulation of iron-acquisition

genes in Shigella dysenteriae and pathogenic Escherichia coli. PLoS One 8,

e63781

12 Africa, L.A. et al. (2011) The Iron-Responsive Fur/RyhB Regulatory Cascade

Modulates the Shigella Outer Membrane Protease IcsP. Infect. Immun. 79, 4543–

4549

13 Liu, M.Y. et al. (1997) The RNA Molecule CsrB Binds to the Global Reguatory

Protein CsrA and Antagonizes Its Activity in Escherichia coli. J. Biol. Chem. 272,

17502–17510

14 Weilbacher, T. et al. (2003) A novel sRNA component of the carbon storage

regulatory system of Escherichia coli. Mol. Microbiol. 48, 657–670

15 Romeo, T. (1998) Global regulation by the small RNA-binding protein CsrA and

the non- coding RNA molecule CsrB. Mol. Microbiol. 29, 1321–1330

16 Romeo, T. et al. (1993) Identification and Molecular Characterization of csrA, a

Pleiotropic Gene from Escherichia coli That Affects Glycogen Biosynthesis,

Gluconeogenesis, Cell Size, and Surface Properties. J. Bacteriol. 175, 4744–55

17 Yang, H. et al. (1996) Coordinate Genetic Regulation of Glycogen Catabolism and

Biosynthesis in Escherichia coli via the CsrA Gene Product. J. Bacteriol. 178,

1012–1017 109

18 Sabnis, N.A. et al. (1995) Pleiotropic Regulation of Central Carbohydrate

Metabolism in Escherichia coli via the Gene csrA. J. Biol. Chem. 270, 29096–

29104

19 Gudapaty, S. et al. (2001) Regulatory Interactions of Csr Components : the RNA

Binding Protein CsrA Activates csrB Transcription in Escherichia coli. J.

Bacteriol. 183, 6017–6027

20 Suzuki, K. et al. (2002) Regulatory Circuitry of the CsrA / CsrB and BarA / UvrY

Systems of Escherichia coli. J. Bacteriol. 184, 5130–5140

21 Adler, B. et al. (1989) A dual transcriptional activation system for the 230 kb

plasmid genes coding for virulence-associated antigens of Shigella flexneri . Mol.

Microbiol. 3, 627–35

22 West, N.P. et al. (2005) Optimization of Virulence Functions Through

Glucosylation of Shigella LPS. Science 307, 1313–1317

23 Hong, M. and Payne, S.M. (1997) Effect of mutations in Shigella flexneri

chromosomal and plasmid-encoded lipopolysaccharide genes on invasion and

serum resistance. Mol. Microbiol. 24, 779–791

24 Caswell, C.C. et al. (2014) Sibling rivalry: related bacterial small RNAs and their

redundant and non-redundant roles. Front. Cell. Infect. Microbiol. 4, 1–13

25 Massé, E. and Gottesman, S. (2002) A small RNA regulates the expression of

genes involved in iron metabolism in Escherichia coli. Proc. Natl. Acad. Sci. U. S.

A. 99, 4620–4625

26 Andrews, S.C. et al. (2003) Bacterial iron homeostasis. FEMS Microbiol. Rev. 27, 110

215–237

27 Payne, S.M. et al. (2006) Iron and pathogenesis of Shigella: Iron acquisition in the

intracellular environment. BioMetals 19, 173–180

28 Fisher, C.R. et al. (2009) Genetics and Virulence Association of the Shigella

flexneri Sit Iron Transport System. Infect. Immun. 77, 1992–9

29 Bagg, A. and Neilands, J.B. (1987) Ferric Uptake Regulation Protein Acts as a

Repressor, Employing Iron (II) as a Cofactor To Bind the Operator of an Iron

Transport Operon in Escherichia coli . Biochemistry 26, 5471–5477

30 Hantke, K. (2001) Iron and metal regulation in bacteria. Curr. Opin. Microbiol. 4,

172–177

31 de Lorenzo, V. et al. (1988) Metal Ion Regulation of Gene Expression Fur

Repressor-Operator Interaction at the Promoter Region of the Aerobactin System

of pColV-K30. J. Mol. Biol. 203, 875–884

32 Beloin, C. et al. (2002) Molecular Dissection of VirB, a Key Regulator of the

Virulence Cascade of Shigella flexneri. J. Biol. Chem. 277, 15333–15344

33 Kane, K.A. and Dorman, C.J. (2012) VirB-mediated positive feedback control of

the virulence gene regulatory cascade of Shigella flexneri. J. Bacteriol. 194, 5264–

34 Castellanos, M.I. et al. (2009) VirB Alleviates H-NS Repression of the icsP

Promoter in Shigella flexneri from Sites More Than One Kilobase Upstream of the

Transcription Start Site. J. Bacteriol. 191, 4047–4050

35 Wing, H.J. et al. (2004) Regulation of IcsP, the Outer Membrane Protease of the 111

Shigella Actin Tail Assembly Protein IcsA, by Virulence Plasmid Regulators VirF

and VirB. J. Bacteriol. 186, 699–705

36 Wing, H.J. et al. (2005) Modulation of an Outer Membrane Protease Contributes

to the Virulence Defect of Shigella flexneri Strains Carrying a Mutation in the virK

Locus. Infect. Immun. 73, 1217–1220

37 Egile, C. et al. (1997) SopA, the outer membrane protease responsible for polar

localization of IcsA in Shigella flexneri. Mol. Microbiol. 23, 1063–1073

38 Shere, K.D. et al. (1997) Disruption of IcsP, the major Shigella protease that

cleaves IcsA, accelerates actin-based motility. Mol. Microbiol. 25, 451–62

39 Steinhauer, J. et al. (1999) The unipolar Shigella surface protein IcsA is targeted

directly to the bacterial old pole: IcsP cleavage of IcsA occurs over the entire

bacterial surface. Mol. Microbiol. 32, 367–77

40 Goldberg, M.B. and Theriot, J.A. (1995) Shigella flexneri surface protein IcsA is

sufficient to direct actin-based motility. Proc. Natl. Acad. Sci. U. S. A. 92, 6572–

6576

41 Lett, M.C. et al. (1989) virG, a Plasmid-Coded Virulence Gene of Shigella

flexneri: Identification of the virG Protein and Determination of the Complete

Coding Sequence. J. Bacteriol. 171, 353–359

42 Makino, S. et al. (1986) A Genetic Determinant Required for Continuous

Reinfection of Adjacent Cells on Large Plasmid in S. flexneri 2a. Cell 46, 551–555

43 Bernardini, M.L. et al. (1989) Identification of icsA, a plasmid locus of Shigella

flexerni that governs bacterial intra- and intracellular spread through interaction 112

with F-actin. Proc. Natl. Acad. Sci. U. S. A. 86, 3867–3871

44 Goldberg, M.B. et al. (1993) Unipolar localization and ATPase activity of IcsA, a

Shigella flexneri protein involved in intracellular movement. J. Bacteriol. 175,

2189–2196

45 Oglesby, A.G. et al. (2005) Fur regulates acid resistance in Shigella flexneri via

RyhB and ydeP. Mol. Microbiol. 58, 1354–1367

46 Beisel, C.L. and Storz, G. (2010) Base pairing small RNAs and their roles in

global regulatory networks. FEMS Microbiol. Rev. 34, 866–882

47 Nakayama, S. and Watanabe, H. (1995) Involvement of cpxA, a Sensor of a Two-

Component Regulatory System , in the pH-Dependent Regulation of Expression of

Shigella sonnei virF gene. J. Bacteriol. 177, 5062–5069

48 Porter, M.E. and Dorman, C.J. (1994) A Role for H-NS in the Thermo-Osmotic

Regulation of Virulence Gene Expression in Shigella flexneri. J. Bacteriol. 176,

4187–4191

49 Mills, M. and Payne, S.M. (1997) Identification of shuA, the Gene Encoding the

Heme Receptor of Shigella dysenteriae, and Analysis of Invasion and Intracellular

Multiplication of a shuA Mutant. Infect. Immun. 65, 5358–63

50 Mills, M. and Payne, S.M. (1995) Genetics and regulation of heme iron transport

in Shigella dysenteriae and detection of an analogous system in Escherichia coli

O157: H7. J. Bacteriol. 177, 3004–3009

51 Kortmann, J. and Narberhaus, F. (2012) Bacterial RNA thermometers: molecular

zippers and switches. Nat. Rev. Microbiol. 10, 255–265 113

52 Wyckoff, E.E. et al. (1998) Structure of the Shigella dysenteriae haem transport

locus and its phylogenetic distribution in enteric bacteria. Mol. Microbiol. 28,

1139–52

53 Okeke, I.N. et al. (2004) Molecular Epidemiology of the Iron Utilization Genes of

Enteroaggregative Escherichia coli. J. Clin. Microbiol. 42, 36–44

54 Torres, A.G. et al. (2001) TonB-Dependent Systems of Uropathogenic Escherichia

coli: Aerobactin and Heme Transport and TonB Are Required for Virulence in the

Mouse. Infect. Immun. 69, 6179–6185

55 Hoffmann, H. et al. (2001) Distribution of the outer membrane haem receptor

protein ChuA in environmental and human isolates of Escherichia coli. Int. J.

Med. Microbiol. 291, 227–30

56 Storz, G. et al. (2011) Regulation by small RNAs in bacteria: expanding frontiers.

Mol. Cell 43, 880–91

57 Storz, G. et al. (2014) Small Proteins Can No Longer Be Ignored. Annu. Rev

Biochem 83, 753–777

58 Bernardini, M.L. et al. (1993) OmpC is involved in invasion of epithelial cells by

Shigella flexneri. Infect. Immun. 61, 3625–3635

59 Pratt, L.A. et al. (1996) From acids to osmZ: Multiple factors influence synthesis

of the OmpF and OmpC porins in Escherichia coli. Mol. Microbiol. 20, 911–917

60 Forst, S. and Masayori, I. (1988) Environmentally Regulated Gene Expression for

Membrane Proteins in Escherichia coli . Annu. Rev. Cell Biol. 4, 21–42

61 Nikaido, H; Varra, M. (1985) Molecular basis of bacterial outer membrane 114

permeability. Microbiol Mol Biol Rev 67, 593–656

62 Mizuno, T. and Mizushima, S. (1990) Signal transduction and gene regulation

through the phosphorylation of two regulatory components: the molecular basis for

the osmotic regulation of the porin genes. Mol. Microbiol. 4, 1077–1082

63 Updegrove, T.B. et al. (2015) How do base-pairing small RNAs evolve? FEMS

Microbiol. Rev. 39, 379–391

64 Lynch, M. and Katju, V. (2004) The altered evolutionary trajectories of gene

duplicates. Trends Genet. 20, 544–549

65 Baeriswyl, S. et al. (2009) Organised Genome Dynamics in the Escherichia coli

Species Results in Highly Diverse Adaptive Paths. PLoS Genet. 5, 1–25

66 Wen, J. et al. (2014) The ZorO-OrzO type I toxin-Antitoxin locus: Repression by

the OrzO antitoxin. Nucleic Acids Res. 42, 1930–1946

67 Wen, J. and Fozo, E. (2014) sRNA Antitoxins: More than One Way to Repress a

Toxin. Toxins (Basel). 6, 2310–2335

68 Tree, J.J. et al. (2014) Identification of Bacteriophage-Encoded Anti-sRNAs in

Pathogenic Escherichia coli. Mol. Cell 55, 199–213

69 Fris, M.E. and Murphy, E.R. (2016) Riboregulators: Fine-Tuning Virulence in

Shigella. Front. Cell. Infect. Microbiol. 6, 2

70 Fris, M.E. et al. (2017) Sibling sRNA RyfA1 Influences Shigella dysenteriae

Pathogenesis. Genes. 50 (8)

71 Waters, L.S. and Storz, G. (2009) Regulatory RNAs in Bacteria. Cell 136, 615–

628 115

72 Jouihri, N. et al. (2003) MxiK and MxiN interact with the Spa47 ATPase and are

required for transit of the needle components MxiH and MxiI, but not of Ipa

proteins, through the type III secretion apparatus of Shigella flexneri. Mol.

Microbiol. 49, 755–767

73 Suzuki, S. et al. (2014) Shigella Type III Secretion Protein MxiI Is Recognized by

Naip2 to Induce Nlrc4 Inflammasome Activation Independently of Pkcδ. PLoS

Pathog. 10, e1003926

74 Blocker, A. et al. (2001) Structure and composition of the Shigella flexneri “needle

complex”, a part of its type III secreton. Mol. Microbiol. 39, 652–663

75 Hachani, A. et al. (2008) IpgB1 and IpgB2, two homologous effectors secreted via

the Mxi-Spa type III secretion apparatus, cooperate to mediate polarized cell

invasion and inflammatory potential of Shigella flexenri. Microbes Infect. 10, 260–

268

76 Costa, S.C.P. and Lesser, C.F. (2014) A multifunctional region of the Shigella type

3 effector IpgB1 is important for secretion from bacteria and membrane targeting

in eukaryotic cells. PLoS One 9,

77 Ohya, K. et al. (2005) IpgB1 Is a Novel Shigella Effector Protein Involved in

Bacterial Invasion of Host Cells. J. Biol. Chem. 280, 24022–24034

78 Allaoui, A. et al. (1993) Characterization of the Shigella flexneri ipgD and ipgF

genes, which are located in the proximal part of the mxi locus. Infect. Immun. 61,

1707–1714

79 Ogawa, M. et al. (2003) IcsB, secreted via the type III secretion system, is 116

chaperoned by IpgA and required at the post-invasion stage of Shigella

pathogenicity. Mol. Microbiol. 48, 913–931

80 Ambrosi, C. et al. (2015) The Shigella flexneri OspB effector: An early

immunomodulator. Int. J. Med. Microbiol. 305, 75–84

81 Zurawski, D. V. et al. (2009) Shigella flexneri type III secretion system effectors

OspB and OspF target the nucleus to downregulate the host inflammatory response

via interactions with retinoblastoma protein. Mol. Microbiol. 71, 350–368

82 Farfán, M.J. et al. (2011) Shigella enterotoxin-2 is a type III effector that

participates in Shigella-induced interleukin 8 secretion by epithelial cells. FEMS

Immunol. Med. Microbiol. 61, 332–339

83 Bongrand, C. et al. (2012) Characterization of the promoter, MxiE box and 5′ UTR

of genes controlled by the activity of the type III secretion apparatus in Shigella

flexneri. PLoS One 7, 19–21

84 Buchrieser, C. et al. (2000) The virulence plasmid pWR100 and the repertoire of

proteins secreted by the type III secretion apparatus of Shigella flexneri. Mol.

Microbiol. 38, 760–771

85 Darwin, A.J. (2005) The phage-shock-protein response. Mol. Microbiol. 57, 621–

628

86 Maxson, M.E. and Darwin, A.J. (2004) Identification of Inducers of the Yersinia

enterocolitica Phage Shock Protein System and Comparison to the Regulation of

the RpoE and Cpx Extracytoplasmic Stress Responses. 186, 4199–4208

87 Lucchini, S. et al. (2005) Transcriptional Adaptation of Shigella flexneri during 117

Infection of Macrophages and Epithelial Cells: Insights into the Strategies of a

Cytosolic Bacterial Pathogen. Infection and Immunity. 73, 88–102

88 Raja, S.B. et al. (2008) Differential expression of ompC and ompF in multidrug-

resistant Shigella dysenteriae and Shigella flexneri by aqueous extract of Aegle

marmelos, altering its susceptibility toward β-lactam antibiotics. Diagn. Microbiol.

Infect. Dis. 61, 321–328

89 Feng, X. et al. (2003) OmpR Phosphorylation and Its Role in Signaling and

Pathogenesis. ASM News 69, 390–395

90 Smith, S.G.J. et al. (2007) A molecular Swiss army knife: OmpA structure,

function and expression. FEMS Microbiol. Lett. 273, 1–11

91 Borgnia, M.J. and Agre, P. (2001) Reconstitution and functional comparison of

purified GlpF and AqpZ, the glycerol and water channels from Escherichia coli.

Proc. Natl. Acad. Sci. 98, 2888–2893

92 Schuch, R. and Maurelli, A.T. (2001) MxiM and MxiJ, base elements of the Mxi-

Spa type III secretion system of Shigella, interact with and stabilize the MxiD

secretin in the cell envelope. J. Bacteriol. 183, 6991–8

93 Jentsch, T.J. et al. (1999) The CLC chloride channel family. Pflügers Arch. Eur. J.

Physiol. 437, 783–795

94 Schwartz, C.J. et al. IscR, an Fe-S cluster-containing transcription factor, represses

expression of Escherichia coli genes encoding Fe-S cluster assembly proteins.

PNAS. 98, 14895-14900

95 Kim, H. et al. Lon-Mediated Proteolysis of the FeoC Protein Prevents Salmonella 118

enterica from Accumulating the Fe(II) Transporter FeoB under High- Oxygen

Conditions. DOI: 10.1128/JB.01826-14

96 Brondijk, T.H.. et al. (2002) Roles of NapF, NapG, and NapH, subunits of the

Escherichia coli periplasmic nitrate reductase in ubiquinol oxidation. Mol.

Microbiol. 44, 245–255

97 Gescher, J.S. et al. (2008) Dissimilatory iron reduction in Escherichia coli:

identification of CymA of Shewanella oneidensis and NapC of E. coli as ferric

reductases. Mol. Microbiol. 68, 706–719

98 Tobe, T. et al. (1993) Transcriptional Control of the Invasion Regularoty Gene

virB of Shigella flexneri - Activation by VirF and Repression by H-NS. J.

Bacteriol. 175, 6142–6149

99 Tobe, T. et al. (1991) Temperature-regulated expression of invasion genes in

Shigella flexneri is controlled through the transcriptional activation of the virB

gene on the large plasmid. Mol. Microbiol. 5, 887–893

100 Burgess, J.L. et al. (2016) Spa47 is an oligomerization-activated type three

secretion system (T3SS) ATPase from Shigella flexneri. Protein Sci. 25, 1037–

1048

101 Morita-Ishihara, T. et al. (2006) Shigella Spa33 is an essential C-ring component

of type III secretion machinery. J. Biol. Chem. 281, 599–607

102 McDowell, M.A. et al. (2016) Characterisation of Shigella Spa33 and Thermotoga

FliM/N reveals a new model for C-ring assembly in T3SS. Mol. Microbiol. 99,

749–766 119

103 Schuch, R. and Maurelli, A.T. (2001) Spa33, a cell surface-associated subunit of

the Mxi-Spa type III secretory pathway of Shigella flexneri, regulates Ipa protein

traffic. Infect. Immun. 69, 2180–2189

104 Tamano, K. et al. (2002) Shigella Spa32 is an essential secretory protein for

functional type III secretion machinery and uniformity of its needle length. J.

Bacteriol. 184, 1244–52

105 Botteaux, A. et al. (2008) Spa32 interaction with the inner-membrane Spa40

component of the type III secretion system of Shigella flexneri is required for the

control of the needle length by a molecular tape measure mechanism. Mol.

Microbiol. 70, 1515–1528

106 Zenk, S.F. et al. (2007) Identification of minor inner-membrane components of the

Shigella type III secretion system “needle complex.” Microbiology 153, 2405–

2415

107 Page, A.-L. et al. (2002) Spa15 of Shigella flexneri, a third type of chaperone in

the type III secretion pathway. Mol. Microbiol. 43, 1533–1542

108 Parsot, C. et al. (2005) A secreted anti-activator, OspD1, and its chaperone, Spa15,

are involved in the control of transcription by the type III secretion apparatus

activity in Shigella flexneri. Mol. Microbiol. 56, 1627–1635

109 Bongrand, C. et al. (2012) Characterization of the Promoter, MxiE Box and 5′

UTR of Genes Controlled by the Activity of the Type III Secretion Apparatus in

Shigella flexneri. PLoS One 7, e32862

110 Kobayashi, T. et al. (2013) The Shigella OspC3 effector inhibits caspase-4, 120

antagonizes inflammatory cell death, and promotes epithelial infection. Cell Host

Microbe 13, 570–83

111 Le Brun, N.E. et al. (2010) Iron core mineralisation in prokaryotic ferritins.

Biochim. Biophys. Acta - Gen. Subj. 1800, 732–744

112 Létoffé, S. et al. (2006) The housekeeping dipeptide permease is the Escherichia

coli heme transporter and functions with two optional peptide binding proteins.

Proc. Natl. Acad. Sci. U. S. A. 103, 12891–6

113 Cole, S.T. et al. (1982) Location and nucleotide sequence of frdB, the gene coding

for the iron-sulphur protein subunit of the fumarate reductase of Escherichia coli.

Eur. J. Biochem. 126, 211–6

114 Gruer, M.J. et al. (1997) Construction and properties of aconitase mutants of

Escherichia coli. Microbiology 143, 1837–1846

115 Gruer, M.J. and Guest, J.R. (1994) Two genetically-distinct and differentially-

regulated aconitases (AcnA and AcnB) in Escherichia coli. Microbiology 140,

2531–2541

116 Cunningham, L. et al. (1997) Transcriptional regulation of the aconitase genes

(acnA and acnB) of Escherichia coli. Microbiology 143, 3795–3805

117 Thöny-Meyer*, L. (2003) A Heme Chaperone for Cytochrome c Biosynthesis†.

DOI: 10.1021/BI035598C

118 Gajiwala, K.S. and Burley, S.K. (2000) HDEA, a periplasmic protein that supports

acid resistance in pathogenic enteric bacteria. J. Mol. Biol. 295, 605–612

119 Kern, R. et al. (2007) Escherichia coli HdeB is an acid stress chaperone. J. 121

Bacteriol. 189, 603–10

120 Hensley, C.T. et al. (2011) Two promoters and two translation start sites control

the expression of the Shigella flexneri outer membrane protease IcsP. Arch.

Microbiol. 193, 263–274

121 Ames, S.K. et al. (1999) C-terminal DNA binding stimulates N-terminal

phosphorylation of the outer membrane protein regulator OmpR from Escherichia

coli. Proc. Natl. Acad. Sci. U. S. A. 96, 11792–7

122 Djapgne, L. et al. (2018) The Pseudomonas aeruginosa PrrF1 and PrrF2 small

regulatory RNAs (sRNAs) promote 2-alkyl-4-quinolone production through

redundant regulation of the antR mRNA. J. Bacteriol. DOI: 10.1128/JB.00704-17

123 Wang, L. et al. (2016) A Novel Small RNA Regulates Tolerance and Virulence in

Shigella flexneri by Responding to Acidic Environmental Changes. 6, 1–12

124 Alphen, W. V and Lugtenberg, B. (1977) Influence of osmolarity of the growth

medium on the outer membrane protein pattern of Escherichia coli. J. Bacteriol.

131, 623–30

125 Mattison, K. et al. (2002) A phosphorylation site mutant of OmpR reveals

different binding conformations at ompF and ompC. J. Mol. Biol. 315, 497–511

126 Cai, S.J. and Inouye, M. (2002) EnvZ-OmpR interaction and osmoregulation in

Escherichia coli. J. Biol. Chem. 277, 24155–61

127 Berman, M.L. and Jackson, D.E. (1984) Selection of lac gene fusions in vivo:

ompR-lacZ fusions that define a functional domain of the ompR gene product. J.

Bacteriol. 159, 750–6 122

128 Bernardini, M.L. et al. (1990) The two-component regulatory system ompR-envZ

controls the virulence of Shigella flexneri. J. Bacteriol. 172, 6274–81

129 Dorman, C.J. et al. (2001) Regulation of virulence gene expression in Shigella

flexneri, a facultative intracellular pathogen. Int. J. Med. Microbiol. 291, 89–96

130 Liu, J.M. and Camilli, A. (2010) A broadening world of bacterial small RNAs.

Curr. Opin. Microbiol. 13, 18–23

131 Broach, W.H. (2014) Small RNAs of Shigella dysenteriae. at

132 Urban, J.H. and Jorg, V. (2007) Translational control and target recognition by

Escherichia coli small RNAs in vivo. Nucleic Acids Res. 35, 1018–1037

123

APPENDIX: SUPPLEMENTAL FIGURES AND TABLES

Supplemental Table 1: Primers and Probes Used for RyfA1 and RyfB1.

Name Sequence Use in this Study Primer RT ryfA1 for CGTAAATCGAGGCCCCC RT-PCR RT ryfA2 for GGCCGTAAATGCAGGTGTTT RT-PCR RT ryfA conserved CTTACCTGCTGTTCTGAAGGTTG RT-PCR rev RyfA1-for TAGCCAATTGCTTAATGGCCCTTTTCGC Construction of pRyfA1 RyfA1-rev CCAAGCTTCCTGATTTTGGCGG Construction of pRyfA1 ryfA TM for AAACGGGTGCTGGCTTTA TaqMan primer qRT-PCR ryfA TM rev CAGCACCACTGGCGATATT TaqMan primer qRT-PCR rrsA TM for GAGTTAGCCGGTGCTTCTT TaqMan primer qRT-PCR rrsA TM rev GGCCTTCGGGTTGTAAAGTA TaqMan primer qRT-PCR RyfB1 for w_MfeI TAGCCAATTGCACTGTGGGGTGCCTG Construction of pRyfB1 RyfB1 rev w_SacI GAGCTCATGGAGTGAATGGGGCG Construction of pRyfB1 RyfB1 Left TM GGGGTGCCTGCGTTGCTC TaqMan primer qRT-PCR RyfB1 Right TM ATTTACCGGTTGAGCCATTG TaqMan primer qRT-PCR ompC-RT-f TTT GCT GTT CAG TAC CAG GG Sybrgreen primer qRT-PCR ompC-RT-r ATAATGGATAGATCCGCCAACG Sybrgreen primer qRT-PCR TCTTAATGGCCCTTTTCGCCGTCTCGCA Construction of p5’UTR RyfA1-prom-1 AACGGGCGCTGGCTTTAGGAAAGGATG RyfA1 TTCCATGG CTAGCCATGGAACATCCTTTCCTAAAG Construction of p5’UTR RyfA1-prom-2 CCAGCGCCCGTTTGCGAGACGGCGAAA RyfA1 AGGGCCATTAAGATGCA RT-RyfB1-L GGGGTGCCTGCGTTGCTC Sybrgreen primer qRT-PCR RT-RyfB1-R ATTTACCGGTTGAGCCATTG Sybrgreen primer qRT-PCR RyfB2 for w_MfeI CAATTGGCTGTGAAGCACCTGC Construction of pRyfB2 new RyfB2 rev_w SacI GAGCTCACAAAAAACCGCCAAAATCA Construction of pRyfB2 new G Probe GCACCACTGGCGATATTGCCGCGATAC ryfA conserved probe Northern Blot GAAGC 6FAM- ryfA1 Fam probe TAAATGCAGGCC/ZEN/CCCCACAGTGC TaqMan qRT-PCR Probe TT-MGB-NFQ 124

6FAM- ryfA2 Fam probe TAAATGCAGGTG/ZEN/TTTCACAGCGC TaqMan qRT-PCR Probe TT-MGB-NFQ 6FAM- rrsA Fam probe ACTCCCTTCC/ZEN/TCCCCGCTGAA- TaqMan qRT-PCR Probe MGB-NFQ 6FAM- RyfB1 probe CACAAGTCA/ZEN/ACCTGCTGGAA- TaqMan qRT-PCR Probe MGB-NFQ

125

Supplemental Table 2: Strains and Plasmids Used for RyfA1 and RyfB1 Investigations.

Strain or Plasmid Description Source

Escherichia coli Strains

DH5α Life Technologies

Shigella dysenteriae Strains

Murphy and Payne, O-4576S1-G Wild Type S. dysenteriae 2007 [15]

Plasmids

pQE2 Expression Vector Qiagen

pRyfA1 ryfA1 expression vector This study

pRyfB1 ryfB1 expression vector This study

pRyfA2 ryfA2 expression vector [131]

pRyfB2 ryfB2 expression vector This study

Urban and Vogel, pXG-10 Translational gfp reporter 2007 [132]

Urban and Vogel, pXG-0 Negative control for gfp reporter 2007 [132]

Urban and Vogel, pXG-1 Positive control for gfp reporter 2007 [132]

p5’UTR RyfA1 putative ryfA1 5’ untranslated region This study 126

Supplemental Figures 1 and 2

Can be found online at http://www.mdpi.com/2073-4425/8/2/50#supplementary [70] ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !

Thesis and Dissertation Services ! !