TsrA Regulates Virulence and Intestinal Colonization in Vibrio cholerae

Cholera is a potentially lethal disease that is endemic in much of the developing world. Vibrio cholerae, the bacterium underlying the disease, infects humans utilizing proteins encoded on horizontally acquired genetic material. Here, we provide evidence that TsrA, a Vibrionaceae-specific protein, plays a critical role in regulating these genetic elements and is essential for V. cholerae virulence in a mouse intestinal model.

ABSTRACT Pathogenic strains of Vibrio cholerae require careful regulation of horizontally acquired virulence factors that are largely located on horizontally acquired genomic islands (HAIs). While TsrA, a Vibrionaceae-specific protein, is known to regulate the critical HAI virulence genes toxT and ctxA, its broader function throughout the genome is unknown. Here, we find that deletion of tsrA results in genomewide expression patterns that heavily correlate with those seen upon deletion of hns, a widely conserved bacterial protein that regulates V. cholerae virulence. This correlation is particularly strong for loci on HAIs, where all differentially expressed loci in the DtsrA mutant are also differentially expressed in the Dhns mutant. Correlation between TsrA and H-NS function extends to in vivo virulence phenotypes where deletion of tsrA compensates for the loss of ToxR activity in V. cholerae and promotes wild-type levels of mouse intestinal colonization. All in all, we find that TsrA broadly controls V. cholerae infectivity via repression of key HAI virulence genes and many other targets in the H-NS regulon. IMPORTANCE Cholera is a potentially lethal disease that is endemic in much of the developing world. Vibrio cholerae, the bacterium underlying the disease, infects humans utilizing proteins encoded on horizontally acquired genetic material. Here, we provide evidence that TsrA, a Vibrionaceae-specific protein, plays a critical role in regulating these genetic elements and is essential for V. cholerae virulence in a mouse intestinal model. KEYWORDS Vibrio cholerae, gene regulation, H-NS, TsrA, computational biology, genetics, virulence regulation V ibrio cholerae is the causative agent of the potentially lethal disease cholera.
TsrA is a Vibrionaceae-specific protein that is by far most common in the genomes of organisms within the Vibrio genus, as determined via BLAST-based (33) protein homology searches (data not shown). TsrA has been shown to regulate type VI secretion system (T6SS) genes (namely, Hcp) in coordination with quorum-sensing pathways and further affects both expression of toxT and the ability of V. cholerae to colonize the small intestine (34). Despite this knowledge, TsrA's larger impacts on V. cholerae gene regulation have not been explored. Here, we elaborate on previous findings and provide more clarity regarding TsrA's impact on gene regulation throughout the V. cholerae genome. Our transcriptomics analyses demonstrate that TsrA mimics the ability of H-NS to repress acquired genetic elements on canonical pathogenicity islands. We further show that this gene plays a critical role in controlling intestinal colonization, with deletion of tsrA completely compensating for the attenuation observed when ToxR, an essential virulence gene regulator, is also deleted in an infant mouse intestinal model. Our findings illustrate a large role in controlling V. cholerae virulence for this small protein.

RESULTS
TsrA deletion promotes expression of H-NS regulon. Previous work showed that TsrA regulates V. cholerae gene expression of ctxA and toxT and that TsrA is structurally similar to the oligomerization domain of H-NS (34). These observations suggested TsrA might have a similar function to H-NS. To investigate this hypothesis, we compared the global transcriptome profiles of the parental C6706 V. cholerae strain to isogenic Dhns and DtsrA strains (see Table S1 in the supplemental material). All strains were grown exponentially at 37°C in Luria-Bertani (LB) medium, since H-NS is known to repress virulence-associated genes under this growth condition (21). As an initial control, we verified that any effects observed upon tsrA deletion are not a function of decreased H-NS protein concentration by Western blotting. Using RpoB as a loading control, we saw no difference in H-NS protein levels between an H-NS-V5 strain and a tsrA mutant derivative of said strain, as detected with anti-V5 antibody (Fig. 1A).
In line with previous estimates (24), our data show that the H-NS regulon encompasses nearly 600 genes ( Fig. 2A; see also Table S1 in the supplemental material). These include, as expected, genes associated with virulence and T6SS (Table 1). Although generally less extreme, RNA expression changes upon deletion of tsrA heavily mirror those observed in the Dhns mutant for a large subset of genes, especially genes on HAIs (Table 1; see also Table S1). When looking at all significantly differentially expressed loci in both strains regardless of fold change, the expression levels of HAI genes were more strongly correlated (adjusted R 2 = 0.644) than their progenitor genome counterparts (adjusted R 2 = 0.582) (Fig. 2B). With regard to effect size across all genes that significantly changed expression by 2-fold or more in the DtsrA strain versus the wild type, 181 loci (roughly 86%) exhibited similar behavior in the Dhns strain ( Fig. 2A). These 181 overlapping loci include all 37 HAI genes that are differentially expressed in the DtsrA strain.
TsrA demonstrates GC and HAI independent effects on both V. cholerae chromosomes as well ( Table 1). Expression of genes associated with T6SS, such as vipAB (35,36), was increased in both mutants despite exhibiting GC content comparable to background levels. These findings agree with and expand upon previously observed links between TsrA and HCP levels (34). In addition, tricarboxylic acid (TCA) cycle enzyme genes, such as oadB and citG, are downregulated in both knockout strains. Since TCA cycle products are known to repress ToxT expression in V. cholerae (37), transcriptional regulation of these genes by TsrA provides a link between cellular response to environmental cues and regulation of virulence genes. A few metabolism-related genes also appeared to be regulated by TsrA but not H-NS, most notably loci involved in chitin utilization (VC0616-VC0619) (38)(39)(40) (Table 1). In sum, TsrA, like H-NS, functions in and regulates key pathways controlling the broader V. cholerae life cycle (19).
TsrA plays a critical role in mouse intestinal colonization. Deletion of tsrA has been shown to increase V. cholerae colonization in an infant mouse model and affect expression of a few genes dually regulated by H-NS and ToxR (34). We previously showed the importance of ToxR in V. cholerae host colonization could be abrogated by deleting H-NS (12). Given the ability of TsrA to regulate virulence gene expression and its parallel effects with H-NS, we hypothesized that TsrA may likewise play a critical role in host colonization. We used an infant mouse model of intestinal colonization to test this hypothesis. As found previously (34), deletion of tsrA leads to a modest hypercolonization of the infant mouse with the DtsrA mutant out-colonizing a wild-type C6706 strain by ;4-fold (Fig. 3). Remarkably, we show that deletion of tsrA completely negates V. cholerae's dependence on ToxR to colonize the infant mouse intestine (Fig. 3A). The near wild-type infection levels of the DtsrA DtoxRS strain are in stark contrast to the drastically reduced infectivity seen in the DtoxRS single mutant, providing a clear testament to the potency of TsrA-mediated virulence repression. This phenotype was complemented by exogenous expression of tsrA in the DtsrA DtoxRS mutant, which showed an extreme colonization defect in line with the DtoxRS single mutant (Fig. 3B). These results implicate TsrA as a high level regulator of critical V. cholerae virulence systems.

DISCUSSION
Our results support previous TsrA research and suggest an expanded role for this protein in fine-tuning expression of the complex virulence cascade of V. cholerae. As a testament to TsrA's importance, deletion of tsrA wholly overcomes the infant mouse intestinal colonization defects seen in a DtoxRS strain. We further show that TsrA stands as a potent coregulator of HAI genes and other portions of the H-NS regulon most responsible for virulence. TsrA's regulatory activities mirror and supplement those of H-NS.
Since virulent strains of V. cholerae rely on ToxR and its regulon to facilitate intestinal colonization (4, 41), it is little surprise that H-NS and TsrA regulons overlap so heavily at sites, such as VPI-1 and the CTX prophage, that are also controlled by ToxR (21, 23). TsrA-mediated repression at these loci likely explains how a DtsrA DtoxRS strain maintains wild-type levels of intestinal colonization, a phenotype previously observed in a Dhns DtoxRS mutant (12).
Since TsrA lacks a DNA binding domain but has some weak homology to the H-NS oligomerization domain (34), it may act through interactions with H-NS that affect the latter's ability to influence gene expression, as HHA is known to do in Escherichia coli (28,42,43). Unfortunately, we were unable to purify TsrA after multiple attempts to confirm an interaction with H-NS in vitro. It is clear that future genetic and biochemical studies will be needed to fully determine how TsrA functions and influences H-NS activity.
TsrA's low relative protein abundance compared to H-NS clarifies the smaller effect size of most transcriptomic changes in the DtsrA strain compared to more intense changes in the Dhns strain. These data, as well as TsrA's known role in coordinating T6SS expression in coordination with quorum-sensing pathways, generally support a a The indicated genes showed significant differences in expression between one or both mutant strains and a wild-type C6706 Vibrio cholerae strain. DtsrA L2FC = log 2 (DtsrA gene abundance/wild-type gene abundance); Dhns L2FC = log 2 (Dhns gene abundance/wild-type gene abundance); q value = FDR adjusted P value; normalized GC = GC content/average chromosomal GC content.
model of TsrA having a more specialized function than H-NS. In this model, if H-NS is the master regulator of virulence gene expression, then TsrA is a master modulator, fine-tuning expression levels in response to some unknown environmental cues. Our results suggest that TsrA serves an important role in V. cholerae gene regulation by controlling the expression of key virulence genes and other H-NS targets. Since V. cholerae's survival in diverse environments depends on precise control of varied genomic elements at specific times, a factor such as TsrA that can modulate and target expression of specific genes helps potentiate V. cholerae's impressive proclivity to adapt and thrive.

MATERIALS AND METHODS
Bacterial strains, plasmids, and media. Strains and plasmids used in this study are listed in Table S1 in the supplemental material. Strains were grown in lysogeny broth/agar at 37°C. The antibiotics streptomycin (100 mg/ml) and carbenicillin (75 mg/ml) were used for selection as needed. X-Gal (5bromo-4-chloro-3-indolyl-b-D-galactopyranoside) was used at 40 mg/ml.
Plasmid and strain construction. All cloning products were sequence verified. For in-frame deletion constructs, surrounding genomic DNA was amplified by crossover PCR and cloned into pWM91 for subsequent sacB-mediated allelic exchange (44). To add the V5 epitope tag to H-NS, hns was amplified from the genome using primers, including the epitope sequences, to add the appropriate tag to the resulting PCR product. For complementation constructs, the original genes with their native promoters were PCR amplified off chromosomal DNA and cloned into plasmid pWKS30 (45).
Western blot analysis. Equal amounts of cells grown at 37°C in LB medium were harvested during exponential phase. Cells were pelleted, resuspended in loading buffer, and separated on a NuPAGE Bis- Tris gel (Thermo Fisher). After transfer, membranes were blotted with monoclonal anti-V5 antibody (Sigma-Aldrich) or anti-RpoB antibody (BioLegend). RpoB was blotted as a loading control. Pierce ECL Western blotting substrate (Thermo Scientific) was added before exposing the X-ray film. Experiments were carried out in at least biological triplicates.
RNA sequencing. RNA sequencing (RNA-seq) was performed essentially as previously described (46). Total RNA was extracted from cells in exponential phase growing at 37°C in LB medium using a Direct-zol RNA MiniPrep kit with TRI-Reagent (Zymo Research). DNase treatment was carried out using a Turbo DNA-free kit (Ambion, Inc.). Ribosomal RNA was depleted using a Ribo-Zero rRNA removal kit for Gram-negative bacteria (Illumina). Sequencing libraries were then prepared for the Illumina sequencing platform. Experiments were repeated in biological triplicates.
Data analysis and visualization. RNA-seq data were aligned to a transcriptome (47) derived from the El Tor N16961 reference genome (ASM674v1) (48). RNA abundances were quantified using Kallisto version 0.43.1 (49), and differential expression was calculated using DESeq2 version 1.18.1 (50). All other data were analyzed using R version 3.6 (51) with the Tidyverse family of packages (52). All visualizations were generated with ggplot2 version 3.2.1 (53).
Infant mouse colonization assays. Assays were performed as previously described (12). At least five mice were tested for each mutant. P values were calculated using Tukey's honest significant difference test following one-way analysis of variance.
Ethics approval. The mouse experiment was reviewed and approved by the UT Austin IACUC (approval AUP-2018-00354).
Data availability. Raw sequence reads for the RNA-seq data were uploaded to the Sequence Read Archive (SRA) (https://www.ncbi.nlm.nih.gov/sra) under accession number SRP242320. Processed RNAseq results are provided in Table S1 in the supplemental material.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.