Bistability in the Bacillus subtilis K-state (competence) system requires a positive feedback loop

Abstract

High expression of the transcriptional activator ComK occurs in 10-20% of the cells in stationary phase cultures of Bacillus subtilis strain 168. ComK drives the expression of more than 100 genes constituting the semidormant K-state, distinct from sporulation and vegetative growth. Among the genes so activated are those that permit competence for genetic transformation. We have addressed the origin of bistability in expression of ComK. We show that bistability requires positive autoregulation at the promoter of comK, but not a potential toggle switch, in which ComK represses the promoter of rok and Rok represses the promoter of comK. We further address the source of the noise that results in the stochastic selection of cells that will express comK. A revised model for the regulation of comK expression is proposed that partially explains bistability.

Fig. 1

Regulation of competence. A. Summary of competence regulation. Arrows and perpendiculars represent positive and negative regulation respectively. ComK binds to its promoter as a dimer of dimers. B. Illustration of the putative toggle switch, in which ComK represses the transcription of rok and Rok represses the transcription of comK.

Fig. 2

Bistability during the development of competence. A. The construction (BD2711) used to demonstrate bistability. The arrows represent positive regulation. B. The frequency distributions of fluorescence intensities in cells carrying comK-gfp (strain BD2711) are plotted at various times during the development of competence. T₀ refers to the time of departure from exponential growth. At each time point, the fluorescent intensities of 500 cells were evaluated.

Fig. 3

Constructs used to test the requirement for the autoregulatory loop for bistability. The arrowheads indicate positive regulation. A. Construct containing the loop (strain BD4010). B. Construct lacking the loop (strain BD4011).

Fig. 4

Distributions of ComK-GFP in cells without (A–E, strain BD4011) and with (F–J, strain BD4010) the ComK autoregulatory loop. The representative micrographs consist of fluorescent images overlaid with bright field images. All of the micrographs were recorded just before T₀. In panel F, the arrow points to a rare cell expressing comK-gfp. The concentrations of IPTG used for each panel are indicated. For each IPTG concentration, the fluorescent intensities of 250 cells were determined.

Fig. 5

Effect of competence regulatory mutations on the expression of comK-lacZ (A) and of comK-gfp (B–K). In A, the following strains are shown: wild-type (○; strain BD1991), multicopy degU (■; strain BD4017), codY (▲; strain BD2607), rok (□; strain BD4016) and multicopy comS (△; strain BD4018). B–F exhibit cells taken at T₀ and G–K exhibit cells taken at T₂. The strains used were wild-type for regulatory genes (B, G; strain BD2711), carried a multicopy plasmid with degU (C, H; strain BD4013), were codY (D, I; strain BD4014), rok (E, J; strain BD4012) or carried a multicopy plasmid with comS (F, K; strain BD4015).

Fig. 6

Effect of rok knockout mutation on bistability in the strain BD4010 background (Fig. 3A). The distributions for the rok (BD4019, shaded) and rok⁺ (BD4010, black) strains are shown. The fluorescent intensities of 250 cells were evaluated for each IPTG concentration.