Spo0A~P imposes a temporal gate for the bimodal expression of competence in Bacillus subtilis

Abstract

ComK transcriptionally controls competence for the uptake of transforming DNA in Bacillus subtilis. Only 10%-20% of the cells in a clonal population are randomly selected for competence. Because ComK activates its own promoter, cells exceeding a threshold amount of ComK trigger a positive feedback loop, transitioning to the competence ON state. The transition rate increases to a maximum during the approach to stationary phase and then decreases, with most cells remaining OFF. The average basal rate of comK transcription increases transiently, defining a window of opportunity for transitions and accounting for the heterogeneity of competent populations. We show that as the concentration of the response regulator Spo0A∼P increases during the entry to stationary phase it first induces comK promoter activity and then represses it by direct binding. Spo0A∼P activates by antagonizing the repressor, Rok. This amplifies an inherent increase in basal level comK promoter activity that takes place during the approach to stationary phase and is a general feature of core promoters, serving to couple the probability of competence transitions to growth rate. Competence transitions are thus regulated by growth rate and temporally controlled by the complex mechanisms that govern the formation of Spo0A∼P. On the level of individual cells, the fate-determining noise for competence is intrinsic to the comK promoter. This overall mechanism has been stochastically simulated and shown to be plausible. Thus, a deterministic mechanism modulates an inherently stochastic process.

Figure 1. Transcription rates from the comK promoter.

(A) The relative luminescence readings corrected for OD for the comK promoter and the OD readings for the growth curves are presented for the wild type (purple dotted curves) and ΔcomK (red curves) strains. Each pair of curves is connected to its Y-axis by a black arrow. (B) Expression from the comK promoter in a ΔcomK background compared with expression from the same promoter in a ΔcomK Δspo0A background. (C) Expression from PcomK in ΔcomK, ΔcomK Δspo0A, ΔcomK Δrok and the ΔcomK Δrok Δspo0A backgrounds.

Figure 2. comK expression in the absence of Spo0A and expression of a synthetic promoter (PsynthA) containing the −35 and −10 motifs of SigA-dependent promoters.

(A) Expression from PcomK in a ΔcomK background compared with the expression in a ΔcomK Δspo0A strain (see Figure 1B). The scale on the left of the graph is used to plot the data from the ΔcomK Δspo0A strain. (B) The expression from PcomK in a ΔcomK Δrok Δspo0A background is plotted in parallel with the expression from a ‘core’ PsynthA promoter. The scale on the left applies to PcomK. The PsynthA-luc fusion was placed at the ectopic amyE locus.

Figure 3. Organization of the comK promoter region.

(A) Putative Spo0A binding sites for activation (A1, 2 and 3) or inhibition (R1 and 2) are shown, as are the two ComK binding sites , the −35 and −10 promoter motifs, the start site (+1) of transcription and the initiation codon for translation. The Spo0A binding sites are predicted based on the published consensus . (B) Schematic representation of the comK promoter showing the putative Spo0A binding sites in relation to the ComK boxes and the region in which Rok binds . (C) Mutagenesis of the Spo0A binding sites. For each box the mutagenized sequences are shown below the wild type sequences. In panels A and C residues that differ from the Spo0A binding consensus sequence are shown in lower case.

Figure 4. Spo0A binds directly to sequences in the comK regulatory region.

(A) The PcomK sequence is displayed with the extents of the wild type probes encompassing A1, 2 and 3 (green) and R1 and 2 (purple) indicated. (B) A radiolabeled DNA fragment containing A1, 2 and 3 was incubated with the indicated concentrations of 0A∼P and the mixture was resolved by polyacrylamide gel electrophoresis. In the lower section of the panel, an otherwise identical probe with the A2, 2 and 3 mutations (Figure 3) was used. (C) In this panel the probe fragment in the upper gel contained R1 and R2. The probe in the lower section carried R1 and R2 mutations (Figure 3). (D) The wild type probe used in panel B was incubated with Rok at the indicated concentrations and the results of gel electrophoresis are shown in the upper panel. In the lower panel an identical probe was used in which the A2 mutation diagrammed in Figure 3C was introduced.

Figure 5. Effect of mutations in the putative Spo0A binding sites on comK expression.

The effects of mutations in R1 (panel A), R2 (panel B), A3 (panel C) or A1 (panel D) in the ΔcomK background are compared with the expression from the wild type comK promoter.

Figure 6. Rok and Spo0A bind at A2.

Panel A shows the effect of an A2 mutation on comK expression, alone and in combination with a null-mutation in rok. The effect of a spo0A knockout mutation is shown for comparison and panel B compares the effect of a rok mutation alone.

Figure 7. Effect of combined mutation of A1, 2, and 3 on PcomK expression.

Panel A shows the effect of this triple mutation. Panel B shows its effect in Δrok and Δspo0A backgrounds.

Figure 8. Co-expression of PcomK-cfp.

With PsdpA-yfp (panel A) or with Pspo0A-yfp (panel B). Cells were segmented microscopically and the average pixel intensities in the CFP and YFP channels were recorded for each cell. The green boxes surround competence-expressing cells, with the lower limit of the boxes set from the threshold value (36) derived from Figure 8. These cells comprise 12.6% of the 11,238 cells measured for panel A and 14.2% of the 14,369 cells measured for panel B. The numbers of cells with CFP values in excess of the threshold were deposited into bins and displayed as a histogram, normalized to the total number of cells within each bin of YFP values. These data have been plotted on a histogram as a percent of total competent cells, with similar results (Figure S7).

Figure 9. Stochastic simulation data match the experimentally observed comK expression.

(A) The average comK output of 5000 Gillespie simulations reproduces the relative amplitudes of the uptick in ΔcomK, ΔcomKΔrok, ΔcomKΔrokΔspo0A and ΔcomKΔspo0A mutants (compare to Figure 1C). (B) The curves are normalized to a peak amplitude of 1 and plotted on the same axes to emphasize the sharp downturn in expression in the simulated ΔcomK strain (compare to Figure S5). The details of the simulation are presented in Text S1.

Figure 10. Cartoon of the uptick mechanism.

The top portion of the figure shows a graphical representation of the Rok and OA∼P concentrations, as well as the availability of RNAPol and the rate of comK basal transcription during the transition to stationary phase. (RNAPol availability is used as a plausible stand-in for the cause of the global increase in transcription we have observed). The peak rate of transcription coincides with T₀, the time of departure from exponential growth. When the concentrations of available RNAPol and of OA∼P are low (1) Rok is dominant and the rate of comK transcription is also low. As the concentration of OA∼P increases further, Rok is antagonized at sites A1, A2 and A3 and at the same time RNAPol becomes more available. As a result the rate of comK transcription increases (2). Finally, the OA∼P concentration reaches a level that is able to repress at R1 and R2 and comK transcription slows (3). In reality, of course, three demarcated periods of time do not exist. Note that the concentration of Rok remains constant throughout and both RNAPol and OA∼P work to counteract its effects. Rok works at an unidentified site in addition to A1–3, shown here between A3 and R1. For simplicity, the availability of RNAPol is shown as constant after T₀, although our data would suggest that it varies somewhat.