Nucleoid-associated proteins shape the global protein occupancy and transcriptional landscape of a clinical isolate of Vibrio cholerae

ABSTRACT Vibrio cholerae, the causative agent of the diarrheal disease cholera, poses an ongoing health threat due to its wide repertoire of horizontally acquired elements (HAEs) and virulence factors. New clinical isolates of the bacterium with improved fitness abilities, often associated with HAEs, frequently emerge. The appropriate control and expression of such genetic elements is critical for the bacteria to thrive in the different environmental niches they occupy. H-NS, the histone-like nucleoid structuring protein, is the best-studied xenogeneic silencer of HAEs in gamma-proteobacteria. Although H-NS and other highly abundant nucleoid-associated proteins (NAPs) have been shown to play important roles in regulating HAEs and virulence in model bacteria, we still lack a comprehensive understanding of how different NAPs modulate transcription in V. cholerae. By obtaining genome-wide measurements of protein occupancy and active transcription in a clinical isolate of V. cholerae, harboring recently discovered HAEs encoding for phage defense systems, we show that a lack of H-NS causes a robust increase in the expression of genes found in many HAEs. We further found that TsrA, a protein with partial homology to H-NS, regulates virulence genes primarily through modulation of H-NS activity. We also identified few sites that are affected by TsrA independently of H-NS, suggesting TsrA may act with diverse regulatory mechanisms. Our results demonstrate how the combinatorial activity of NAPs is employed by a clinical isolate of an important pathogen to regulate recently discovered HAEs. IMPORTANCE New strains of the bacterial pathogen Vibrio cholerae, bearing novel horizontally acquired elements (HAEs), frequently emerge. HAEs provide beneficial traits to the bacterium, such as antibiotic resistance and defense against invading bacteriophages. Xenogeneic silencers are proteins that help bacteria harness new HAEs and silence those HAEs until they are needed. H-NS is the best-studied xenogeneic silencer; it is one of the nucleoid-associated proteins (NAPs) in gamma-proteobacteria and is responsible for the proper regulation of HAEs within the bacterial transcriptional network. We studied the effects of H-NS and other NAPs on the HAEs of a clinical isolate of V. cholerae. Importantly, we found that H-NS partners with a small and poorly characterized protein, TsrA, to help domesticate new HAEs involved in bacterial survival and in causing disease. A proper understanding of the regulatory state in emerging isolates of V. cholerae will provide improved therapies against new isolates of the pathogen.

For each case, the heatmap shows a division of genes into seven equally-populated bins discretizing the log2 fold change, the heat map shows the enrichment or depletion of members of the indicated GO term in that expression bin.B) AT percentage of differentially expressed genes in ∆hns, ∆tsrA, and ∆hns∆tsrA versus wild type.Significant differentially expressed genes (q-value less than or equal to 0.1) are indicated in black and q-value of greater than 0.1 are indicated in gray.Figure S9: Volcano plot of differential protein abundances obtained from Tandem Mass Tag Mass Spectrometry from the strains lacking ihfA vs. wild type.Fur regulon and iron homeostasis protein groups are color-matched to the RNA-seq volcano plot for consistency (Figure 2B).The horizontal dashed line indicates the -log10(q-value) of 0.1 and the left and right vertical dashed lines represent log2(fold change) of -1 and 1, respectively.
Figure S2: A) Mean differences in protein occupancy of EPODs between the average of indicated genotypes and the wild type.Values are plotted separately for IPOD, IPOD-HR and RNA polymerase ChIP-Seq.B) Mean differences in protein occupancy of negative EPODs (nEPODs) between the average of indicated genotypes and the wild type.Values are plotted separately for IPOD, IPOD-HR and RNA polymerase ChIP-Seq.

Figure
Figure S3: Difference (relative to WT) in IPOD-HR occupancies and difference (relative to WT) in RNA polymerase ChIP-seq rz-scores in indicated genotypes at individual replicate levels.For IPOD-HR, a 50 bp rolling median was used as the fundamental unit of data centered at each genomic location, and the plotted values reflect the pseudomedian of those values across the indicated genomic features.Error bars show 95% confidence intervals calculated using the wilcox.testfunction in.For RNA polymerase ChIP-seq, we followed a similar procedure as for IPOD-HR, except that we used gene-level means of the ChIP occupancies as individual units of data for the pseudomedian calculations.

Figure S4 :
Figure S4: Heatmaps showing the differences of means of A) IPOD, B) IPOD-HR and C) RNA polymerase ChIP-seq rz-scores in the indicated conditions at the biological replicate level, relative to the average for wild type.
Figure S5: A) IPOD-HR occupancy traces in the VPI-1 island for the indicated strains.B) Distribution of scores in the wild type and ∆ihfA Vibrio cholerae KDS1 in the regions obtained from V5-H-NS ChIP-seq study in strain C6706 (data from [71]) in the main text).C) IPOD, IPOD-HR and RNA polymerase ChIP-seq occupancy tracks for all of the genotypes in the CTX region of V. cholerae, with the gray track being the V5-H-NS ChIP-seq from C6706 (data from [71]).D) IPOD, IPOD-HR and RNA polymerase ChIP-seq occupancy tracks for all of the genotypes in the PLE region of V. cholerae.
Figure S6:A) GO term enrichment classification analysis of RNA-sequencing results of indicated genotypes: ∆hns, ∆tsrA, and ∆hns∆tsrA.For each case, the heatmap shows a division of genes into seven equally-populated bins discretizing the log2 fold change, the heat map shows the enrichment or depletion of members of the indicated GO term in that expression bin.B) AT percentage of differentially expressed genes in ∆hns, ∆tsrA, and ∆hns∆tsrA versus wild type.Significant differentially expressed genes (q-value less than or equal to 0.1) are indicated in black and q-value of greater than 0.1 are indicated in gray.

Figure S7 :
Figure S7: GO term enrichment classification analysis of the upregulated genes (defined by a q value <0.1 and a log2 fold change at least three times the standard error of the mean) from the RNA-sequencing results of indicated genotypes: ∆hns, ∆tsrA, and ∆hns∆tsrA.
Figure S8: A) Correlation plot of rz-log ratios of RNA polymerase ChIP-seq vs. Input (y axis) with RNA-seq log10(Transcripts per million (TPM)) (x axis) in the wild type KDS1.The line represents a robust linear model of Log10(TPM) as a function of RNA polymerase ChIP-seq in wild type.The slope is shown in blue.Genes with less than 0.0001 TPM are omitted from the plot.B) Occupancy traces of IPOD, IPOD-HR and RNA polymerase ChIP-seq of the whole SXT-VchInd6 genomic feature in untreated and MMC treated wild type V. cholerae.The shaded rectangles are shown for visual comparison of regions with differences in RNA polymerase binding between the untreated and MMC treated cells.