The regulatory network of natural competence and transformation of Vibrio cholerae

Abstract

The human pathogen Vibrio cholerae is an aquatic bacterium frequently encountered in rivers, lakes, estuaries, and coastal regions. Within these environmental reservoirs, the bacterium is often found associated with zooplankton and more specifically with their chitinous exoskeleton. Upon growth on such chitinous surfaces, V. cholerae initiates a developmental program termed "natural competence for genetic transformation." Natural competence for transformation is a mode of horizontal gene transfer in bacteria and contributes to the maintenance and evolution of bacterial genomes. In this study, we investigated competence gene expression within this organism at the single cell level. We provide evidence that under homogeneous inducing conditions the majority of the cells express competence genes. A more heterogeneous expression pattern was observable on chitin surfaces. We hypothesize that this was the case due to the heterogeneity around the chitin surface, which might vary extensively with respect to chitin degradation products and autoinducers; these molecules contribute to competence induction based on carbon catabolite repression and quorum-sensing pathways, respectively. Therefore, we investigated the contribution of these two signaling pathways to natural competence in detail using natural transformation assays, transcriptional reporter fusions, quantitative RT-PCR, and immunological detection of protein levels using Western blot analysis. The results illustrate that all tested competence genes are dependent on the transformation regulator TfoX. Furthermore, intracellular cAMP levels play a major role in natural transformation. Finally, we demonstrate that only a minority of genes involved in natural transformation are regulated in a quorum-sensing-dependent manner and that these genes determine the fate of the surrounding DNA. We conclude with a model of the regulatory circuit of chitin-induced natural competence in V. cholerae.

Figure 1. Visualization of competence gene expression on chitin surfaces.

V. cholerae cells were grown on chitin beads as previously described . The white dashed line notes the edge of the chitin surface. The bacteria carried diverse transcriptional FP reporter fusions. I: vector control containing promoter-less gfp and dsRed; II: plasmid containing gfp driven by the pilA promoter and dsRed downstream of the comEA promoter region; III: swapped reporter genes in comparison to II; IV: plasmid containing gfp driven by the gyrA promoter and dsRed driven by the comEA promoter. Bacteria were grown statically for 48 h before pictures were taken. The order of the images here and in the following figures is (from left to right): green channel (GFP), red channel (dsRed), and merged image composed of a phase contrast image overlaid by both fluorescence channel images. Scale bar = 5 µm.

Figure 2. Competence genes are expressed by the majority of cells under homogeneous competence-inducing conditions.

V. cholerae strains were grown aerobically in defined artificial seawater medium with the addition of either N-acetylglucosamine (GlcNAc) or hexa-N-acetylchitohexaose (GlcNAc)₆ as sole carbon source. The latter is known as a potent inducer of natural competence/transformation . Competence gene expression was quantified for fluorescence intensities using flow cytometry. Reporter fusions are indicated above each panel. Panel A: promoter-less gfp and dsRed reporter; panel B: [P_pilA]-gfp/[P_comEA]-dsRed; and panel C: [P_comEA]-gfp and [P_pilA]-dsRed. Panel D: [P_gyrA]-gfp/[P_comEA]-dsRed. The flow cytometry graphs indicate the number of cell counts on the y-axis and the fluorescence signal intensity (as arbitrary units, AU) on the x-axis.

Figure 3. Artificial induction of natural transformation by expression of the competence regulatory gene tfoX in cis.

V. cholerae cells were grown in rich medium in the absence (−) or presence (+) of the artificial inducer arabinose (0.02%). Cells were either tested for natural transformability (panel A) or competence gene promoter activity based on FP reporters (panels B and C). Panel A: Transformation frequencies are given on the y-axis for competence-uninduced (−) and competence-induced (+) bacteria. <d.l. = below detection limit. Panel B and C: V. cholerae cells harboring the different transcriptional reporter fusions were grown without or with competence induction. Bacteria were either visualized by epifluorescence microscopy (panel B; image arrangements as in Figure 1; Scale bar = 5 µm) or measured with respect to relative fluorescence units (RFU) and optical density at 600 nm (panel C). Panel C: RFU per OD₆₀₀ values are given on the y-axis. All experiments in Figure 3 were repeated at least three independent times. Error bars reflect standard deviations. Statistics were applied using the Student's t test. * P<0.05, ** P<0.01, n.s. = not significant.

Figure 4. Quorum sensing only regulates a subset of competence genes.

The tfoX-expression construct was transferred onto the chromosome of mutant V. cholerae strains, which were defective in the quorum-sensing circuit. Strains were grown in LB medium with or without 0.02% arabinose and tested for natural transformability (panel A), comEA (panel B), and pilA promoter-driven FP expression (panel C) as described for Figure 3. Experiments were repeated at least three times. Statistically significant differences were calculated using the Student's t test. * P<0.05, ** P<0.01, n.s. = not significant. <d.l. = below detection limit.

Figure 5. Correlation among HapR protein levels, hapA gene expression, and the nuclease Dns.

Panel A and C: Proteins of the indicated strains, each containing artificially inducible tfoX on the chromosome, were separated by SDS-PAGE. After blotting, the relative abundance of proteins HapR (panel A) or Dns (panel C) were determined by detection with protein-specific antibodies. For each sample 6 µg (panel A) and 12 µg total protein (panel C), respectively, were applied per lane. Strains were tested under non-competence-inducing and competence-inducing conditions as indicated above each image. Panel B: HapR-dependent expression of hapA promoter-driven gene expression was quantified as described in Figure 3 for comEA/pilA. The growth conditions were as described for Figure 4. Averages of three independent experiments are indicated. Error bars indicate standard deviations. Statistically significant differences were calculated using the Student's t test. * P<0.05, ** P<0.01, n.s. = not significant.

Figure 6. TfoX drives expression of QS–dependent and QS–independent competence genes.

V. cholerae wild type strain containing artificially inducible tfoX in cis was tested for the expression of different competence genes. Panel A: Different transcriptional FP reporter fusions were tested for TfoX-dependent induction. These fusions were composed of the potential promoter region of the respective competence gene(s) (x-axis) fused to gfp. The housekeeping gene gyrA was tested as control. Relative fluorescence per OD₆₀₀ unit is given on the y-axis. Panel B and C: qRT-PCR data comparing the relative expression of the indicated genes in a wild type strain under competence non-inducing and competence-inducing conditions (panel B). In panel C both the wild type strain and the hapR mutant were tested for competence gene expression under tfoX-expressing conditions. All panels depict averages of at least three independent experiments and error bars indicate standard deviations. Statistically significant differences were determined using Student's t tests. * P<0.05, ** P<0.01, n.s. = not significant.

Figure 7. Model of the regulatory network of natural competence and transformation of V. cholerae.

At least three extracellular and intracellular signaling molecules must be present to allow natural transformation to occur in V. cholerae. 1) Chitin degradation products such as chitin oligomers, which lead to the induction of the sRNA TfoR and the main regulator of transformation TfoX (chitin pathway shown in brown). 2) The secondary messenger cAMP, which has to accumulate within cells (CCR pathway shown in blue). 3) Extracellular autoinducers, with an emphasis on the stronger autoinducer CAI-1, which feed into the quorum-sensing circuit (shown in green). Whereas chitin- and TfoX-dependent induction and the requirement for cAMP and CRP are universal for all, so far investigated, competence genes, the QS-dependent circuit regulates only a subset of those, such as comEA and comEC. Therefore, QS acts as a switch in gene expression and is responsible for the final fate of the surrounding DNA (boxed areas). At a low cell density (LCD), the DNA (shown in purple) is degraded by the nuclease Dns. As a consequence, the cells are non-transformable. At a high cell density (HCD) and, therefore, high abundance of the autoinducer CAI-1, the nuclease gene dns is transcriptionally repressed, whereas comEA and comEC are activated. ComEA as well as ComEC then contribute to the DNA uptake process, probably due to their ability to shuffle the DNA through the periplasmic space and the inner membrane, respectively.