A novel phase variant of the cholera pathogen shows stress-adaptive cryptic transcriptomic signatures

Abstract

Background: In a process known as phase variation, the marine bacterium and cholera pathogen Vibrio cholerae alternately expresses smooth or rugose colonial phenotypes, the latter being associated with advanced biofilm architecture and greater resistance to ecological stress. To define phase variation at the transcriptomic level in pandemic V. cholerae O1 El Tor strain N16961, we compared the RNA-seq-derived transcriptomes among the smooth parent N16961, its rugose derivative (N16961R) and a smooth form obtained directly from the rugose at high frequencies consistent with phase variation (N16961SD).

Results: Differentially regulated genes which clustered into co-expression groups were identified for specific cellular functions, including acetate metabolism, gluconeogenesis, and anaerobic respiration, suggesting an important link between these processes and biofilm formation in this species. Principal component analysis separated the transcriptome of N16961SD from the other phase variants. Although N16961SD was defective in biofilm formation, transcription of its biofilm-related vps and rbm gene clusters was nevertheless elevated as judged by both RNA-seq and RT-qPCR analyses. This transcriptome signature was shared with N16961R, as were others involving two-component signal transduction, chemotaxis, and c-di-GMP synthesis functions.

Conclusions: Precise turnarounds in gene expression did not accompany reversible phase transitions (i.e., smooth to rugose to smooth) in the cholera pathogen. Transcriptomic signatures consisting of up-regulated genes involved in biofilm formation, environmental sensing and persistence, chemotaxis, and signal transduction, which were shared by N16961R and N16961SD variants, may implicate a stress adaptation in the pathogen that facilitates transition of the N16961SD smooth form back to rugosity should environmental conditions dictate.

Keywords: Bioinformatics; Cholera; Phase variation; RNA sequencing; Transcriptomics; Vibrio cholerae.

Fig. 1

Experimental procedure for isolation of N16961-derived phase variants. a Beginning with isolated colonies, parental strain N16961 was subjected to daily passaging with plating following the indicated approximate number of bacterial generations. Single randomly selected rugose isolates, designated RU1, RU2 and RU3, were then chosen. In the passaging experiment where RU1 was eventually selected, plating was also performed after ~30 generations, but there were no clearly distinguishable rugose isolates present on those plates; b Beginning with isolated colonies of RU1, RU2, and RU3, the passaging and plating procedure was repeated, and single randomly chosen smooth isolates, designated SD1, SD2 and SD3, were selected at the end

Fig. 2

Biofilm formation of N16961 phase variants. a Following 48 h of static incubation at 30 °C, liquid cultures were poured off and the remaining attached biofilm material was stained with 0.1% crystal violet and quantified by measuring the absorbance at 570 nm. Represented in the graph are the average absorbance values of 12 replicates of N16961, 18 replicates (6 per individual variant) of N16961R, 18 replicates (6 per individual variant) of N16961SD, and 6 replicates of the uninoculated control. Error bars show standard deviations. Samples indicated with the same letter were not significantly different according to Tukey’s post-test (P < 0.05). b Following 48 h of static incubation at 30 °C, liquid cultures were carefully poured out and the pellicles that had formed at the air-broth interface during incubation were retained and repositioned at the side of each glass tube to be photographed

Fig. 3

Motility of N16961 phase variants. a Representative plate showing motility zones of variants following overnight incubation. b Following overnight incubation of motility plates, motility zones were measured in mm. Represented in the figure are the averages of 10 plates, with error bars depicting the standard deviations. Samples indicated with same letter were not statistically different according to Tukey’s post-test (P < 0.05)

Fig. 4

Principal Component Analysis. The principal components (PC 1 and PC 2) identified together account for 58.94% of the overall variability of the dataset. Phase variants were grouped by PCA based on transcriptomic similarities

Fig. 5

Transcription analysis of vps and rbm genes among N16961 phase variants. a Normalized RNA expression profiles based on RNA-seq results for biofilm-related genes. Peaks corresponding to the number of transcripts that mapped to regions containing the vpsI, vpsII, and rbm gene clusters are shown in parallel tracks for each phase variant. N16961 samples are depicted in green, N16961R samples are shown in red, and N16961SD samples are shown in blue. Although our analysis reported similar up-regulated expression values of the genes of all three clusters in the N16961R and N16961SD samples as compared to N16961, some genes of the vpsI cluster fell out of the range of statistical significance in the N16961SD samples (Additional file 10: Table S6). These include the vpsI gene, which encodes a glycosyltransferase, the vpsE and vpsH genes, whose products are predicted to be involved VPS export, and the vpsF and vpsJ genes, which encode proteins of unknown function. The increased P _adj values obtained for these genes may be a result of the apparent differences in expression of the vpsI cluster in the SD3 sample versus SD1 and SD2, which can be seen in the figure. b RT-qPCR verification of up-regulation of genes representative of the vpsI, vpsII, and rbm gene clusters in N16961R and N16961SD samples. The graph depicts the log₂ fold changes of gene transcripts calculated for each selected gene relative to their transcript abundances in the N16961 sample group at a confidence interval of 95%. All data were normalized respective to the reference gene, gyrA, and the error bars depict the normalized quantity standard error

Fig. 6

Overview of the differentially regulated genes discussed in this paper. The first symbol next to each gene name represents the qualitative expression change from N16961 to N16961R, the second symbol denotes the change from N16961R to N16961SD and the third represents the difference from N16961 to N16961SD. Symbol types are as follows: a dash indicates no significant change in gene expression, an upward arrow indicates up-regulation of the gene, and a downward arrow indicates down-regulation of the gene. Summary of central metabolic pathways shown within the cell includes the locations of gene functions as labeled with numbers. Pathways or portions of a pathway are colored as follows: red = gluconeogenesis; purple = pyruvate dehydrogenase complex; gold = acetate metabolism; green = glyoxylate bypass; blue = TCA cycle. As identified in this study, genes with the expression pattern ↑-↑ (or ↑↓↑ in the case of ggt) compose the transcriptomic signatures shared between N16961R and N16961SD strains