Transcriptional Silencing by TsrA in the Evolution of Pathogenic Vibrio cholerae Biotypes

Abstract

Vibrio cholerae is a globally important pathogen responsible for the severe epidemic diarrheal disease called cholera. The current and ongoing seventh pandemic of cholera is caused by El Tor strains, which have completely replaced the sixth-pandemic classical strains of V. cholerae To successfully establish infection and disseminate to new victims, V. cholerae relies on key virulence factors encoded on horizontally acquired genetic elements. The expression of these factors relies on the regulatory architecture that coordinates the timely expression of virulence determinants during host infection. Here, we apply transcriptomics and structural modeling to understand how type VI secretion system regulator A (TsrA) affects gene expression in both the classical and El Tor biotypes of V. cholerae We find that TsrA acts as a negative regulator of V. cholerae virulence genes encoded on horizontally acquired genetic elements. The TsrA regulon comprises genes encoding cholera toxin (CT), the toxin-coregulated pilus (TCP), and the type VI secretion system (T6SS), as well as genes involved in biofilm formation. The majority of the TsrA regulon is carried on horizontally acquired AT-rich genetic islands whose loss or acquisition could be directly ascribed to the differences between the classical and El Tor strains studied. Our modeling predicts that the TsrA protein is a structural homolog of the histone-like nucleoid structuring protein (H-NS) oligomerization domain and is likely capable of forming higher-order superhelical structures, potentially with DNA. These findings describe how TsrA can integrate into the intricate V. cholerae virulence gene expression program, controlling gene expression through transcriptional silencing.IMPORTANCE Pathogenic Vibrio cholerae strains express multiple virulence factors that are encoded by bacteriophage and chromosomal islands. These include cholera toxin and the intestinal colonization pilus called the toxin-coregulated pilus, which are essential for causing severe disease in humans. However, it is presently unclear how the expression of these horizontally acquired accessory virulence genes can be efficiently integrated with preexisting transcriptional programs that are presumably fine-tuned for optimal expression in V. cholerae before its conversion to a human pathogen. Here, we report the role of a transcriptional regulator (TsrA) in silencing horizontally acquired genes encoding important virulence factors. We propose that this factor could be critical to the efficient acquisition of accessory virulence genes by silencing their expression until other signals trigger their transcriptional activation within the host.

Keywords: TsrA; Vibrio cholerae; horizontal gene transfer; structural modeling; transcriptional regulation.

FIG 1

TsrA is a transcriptional repressor of a relatively small set of V. cholerae genes. (A to C) Volcano plot representations of differential expression analysis of the wild type versus the isogenic ΔtsrA mutant for V. cholerae O395-N1 (A), C6706 (B), and Haiti H1 (C) RNA-seq data sets. The x axis shows the log₂ fold change in expression, and the y axis shows the log odds of a gene being differentially expressed (ac, –log₁₀ P value). Genes are shown in blue if they pass the absolute confidence (ac) threshold of ≥1.3 (equivalent to a P value of ≤0.05, pink dotted line). The identities of the top three genes with the highest ac values, and the few genes with negative fold changes, are given. (D) Venn diagram of the number of genes that are members of the TsrA regulon in each of the strains.

FIG 2

The majority of genes in the TsrA regulon are encoded on genetic islands that are most highly activated in the Haiti H1 ΔtsrA strain. (A) Heat map of mean-centered log₂-transformed expression values of members of the TsrA regulon in each of the three biological replicates of each wild-type and mutant strain. Each row in the heat map corresponds to a gene, and each column in the heat map corresponds to a sample (identified above the heat map); if no ortholog of the gene is found in the strain, the field is shaded gray. Genetic islands are indicated on the left of the heat map, and clusters of biofilm formation and T6SS genes are indicated on the right. (B, D, and E) Fold changes from the wild type in the expression of genes encoding signature virulence factors (B), the T6SS (D), and biofilm formation functions (E) in the ΔtsrA mutant. Filled circles indicate genes considered significantly differentially expressed and part of the TsrA regulon in the corresponding strain (ac, ≥1.3). Open circles indicate genes not regulated by TsrA. Horizontal bars indicate mean values. (C) tcpA and toxT mRNA abundances expressed in reads per kilobase per million (rpkM) (y axis on a log₂ scale) in either the WT (black circles) or the ΔtsrA mutant (red circles) of each of the three strains analyzed. Numbers above the C6706 and Haiti H1 data sets indicate the fold rpkM change between the WT and the ΔtsrA mutant. Error bars indicate standard deviations of rpkM measurements for the three independent biological replicates. (F) qRT-PCR measuring the fold change in hcp1 (blue) and ctxA (red) mRNA abundances between the WT and the ΔtsrA mutant in each of the strains analyzed. Error bars represent the standard deviations of measurements obtained from three independent biological replicates.

FIG 3

TsrA structural homology modeling. (A) The V. cholerae N terminus of TsrA (Arg 38 to Lys 93) (magenta) is structurally homologous to the N-terminal S. Typhimurium H-NS oligomerization domain (Ser 2 to Ala 82) (blue) but contains a gap (dotted line) that predicts a shorter protein fold than that of H-NS (by 13 amino acids [aa] or ∼18 Å). The TsrA N terminus (residues 1 to 37) is excluded from the modeled structure, thus omitting alpha helix H1. The known alpha helices of S. Typhimurium, H1 to H4, are indicated. (B) TsrA oligomerization sites 1 and 2 were modeled after the structure of sites 1 and 2 in S. Typhimurium H-NS to build a TsrA trimer and a superhelix, with a ∼220 Å rise and a ∼70 Å diameter. Site 1 is always positioned on the exterior surface of the helix, while site 2 always points inward to the center of the helix. (C) Detailed view of the TsrA site 2 model structure, where hydrophobic interactions of conserved residues are analogous to those observed in S. Typhimurium H-NS oligomerization site 2. One protomer is displayed as a secondary structure and stick model; the other, as molecular surface. The surface is colored to highlight different residue properties: blue, nitrogen atoms; red, oxygen atoms; green, hydrophobic atoms; yellow, sulfur atoms. (D) Amino acid sequence alignments of the modeled domains of V. cholerae TsrA and S. Typhimurium H-NS. Predicted and known secondary structures are shown above and below for TsrA and H-NS, respectively, and TsrA residue numbers are shown for reference. Residues that compose sites 1 and 2 are marked with a black line. Alpha helices H2, H3, and H4, derived from the S. Typhimurium H-NS structure, are underlined in green. The TsrA site 1 proline box motif HPXXHH that caps the alpha helix, where H is hydrophobic, and X is any amino acid, is boxed in red. Conserved hydrophobic residues that support the site 2 structure in panel C are marked with asterisks.

FIG 4

The V. cholerae TsrA and H-NS regulons largely overlap. (A) Volcano plot representation of the differential expression analysis of the wild type versus the isogenic Δh-ns mutant of V. cholerae C7258 RNA-seq data sets. The x axis shows the log₂ fold change in expression, and the y axis shows the log odds of a gene being differentially expressed (ac, –log₁₀ P value). Genes are shown in blue if they pass the absolute confidence threshold (ac, ≥1.3; equivalent to a P value of ≤0.05). (B) Venn diagram of the overlap between the TsrA and H-NS regulon members of V. cholerae strains C6706 and C7258, respectively. (C) Fold changes between the ΔtsrA and Δh-ns mutants and wild-type strains of C6706 and C7258 in the expression of all differentially regulated genes in either regulon. The gray horizontal bar represents the mean fold change in each regulon, 2.2- and 16-fold, for the TsrA and H-NS regulons, respectively. (D) Heat map of mean-centered, log₂-transformed expression values of members of the H-NS and TsrA regulons. Two and three biological replicates are depicted for the H-NS (47) and TsrA (this study) data sets, respectively. Each row in the heat map represents a gene, and each column in the heat map represents a sample (identified above the heat map); if no ortholog of the gene is found in the strain, the field is shaded gray. Genetic islands are indicated on the left of the heat map, and clusters of biofilm formation and T6SS genes are indicated on the right.

FIG 5

The DNA sequences of TsrA regulon members are AT rich. Shown are the GC percentages of either the coding sequences (CDS) or the 200-nucleotide upstream regions of members of the C7258 H-NS regulon (including only those genes negatively regulated by H-NS), the C6706 TsrA regulon, or a randomly selected set of 300 C6706 genes. Pink horizontal lines represent the mean of each distribution.