Complexity in bacterial cell-cell communication: quorum signal integration and subpopulation signaling in the Bacillus subtilis phosphorelay

Abstract

A common form of quorum sensing in gram-positive bacteria is mediated by peptides that act as phosphatase regulators (Phr) of receptor aspartyl phosphatases (Raps). In Bacillus subtilis, several Phr signals are integrated in sporulation phosphorelay signal transduction. We theoretically demonstrate that the phosphorelay can act as a computational machine performing a sensitive division operation of kinase-encoded signals by quorum-modulated Rap signals, indicative of cells computing a "food per cell" estimate to decide whether to enter sporulation. We predict expression from the rapA-phrA operon to bifurcate as relative environmental signals change in a developing population. We experimentally observe that the rapA-phrA operon is heterogeneously induced in sporulating microcolonies. Uninduced cells sporulate rather synchronously early on, whereas the RapA/PhrA subpopulation sporulates less synchronously throughout later stationary phase. Moreover, we show that cells sustain PhrA expression during periods of active growth. Together with the model, these findings suggest that the phosphorelay may normalize environmental signals by the size of the (sub)population actively competing for nutrients (as signaled by PhrA). Generalizing this concept, the various Phrs could facilitate subpopulation communication in dense isogenic communities to control the physiological strategies followed by differentiated subpopulations by interpreting (environmental) signals based on the spatiotemporal community structure.

Fig. 1.

Quorum signaling in B. subtilis. (A) Overview of rap-phr gene cassettes found in the chromosome (1). Dark arrows denote rap and light arrows phr. Bent arrows indicate start sites of transcription. (B) Schematic model of Rap-Phr signal integration in phosphorelay signal transduction explained in more detail in the introduction.

Fig. 2.

Model of quorum signal integration. (A) In contrast to autoinducer signaling, which elicits a direct cell response (Left), Phr-based signals integrate with (environmental) signals (Right). (B) Schematic model of phosphorelay signaling. (C) Global output behavior of an open-loop phosphorelay. Only in regime I (“signal integration regime”) is the output sensitive to changes in both input signals and the output responds to changes in their ratio (“ratiometric coupling”). In all other regimes (II–IV), the cell is insensitive to at least one input signal. The functions κ_s and π_s are functions of system parameters and determine the regime boundaries. (D) Black-box model of the phosphorelay signal integration operating as a machine to compute the quotient of the Kin-kinase κ and Phr-modulated Rap-phosphatase activity π to control the steady-state concentration of Spo0A∼Px.

Fig. 3.

Effect of feedback on signal integration (A–C) and RapA/PhrA expression (D). (A) Output behavior of a reference open-loop system without feedback spanning regime I and IV. (B) WT low Spo0A feedbacks expand the ratiometric regime I. (C) Adding a P_spo0B loop to the WT reduces severely the π sensitivity. (D) (Left) The regulatory motif of Spo0A∼P and RapA cross-repression maps onto an effective autoactivating circuit for RapA production with activation coefficient K and maximal production V. (Right) The protein production rate as a function of Rap concentration c is shown for 3 different (K,V)-pairs. The protein decay rate is shown in black. In steady state, production equals decay, and the circled intersections determine the stable steady-state concentration(s) in each case. As K increases, RapA expression is monostable “on” (blue), bistable (red), and a monostable “off” (green). The absolute concentration of the “on” state is controlled by V, which depends on ComA activation.

Fig. 4.

Emergence of subpopulation PhrA signaling. (A) Snapshots of an outgrowing colony of strain rapA/IIa at t ≈ T₀, T₂, T₁₁, and T₄₀. T₀ denotes entry into stationary phase. Bright-field images (gray) were overlaid with the fluorescence images derived from P_rapA-iyfp (yellow). (B) Bimodal distribution of fluorescence intensities within the cell population at T_4.5. (C) Log-scale growth characteristics of the colony estimated from the segmented colony area (red) and total number of segmented cells (black). (D) Time evolution of population frequencies with a particular YFP fluorescence intensity. A clear bifurcation into “on” and “off” cells emerges at approximately T₀.

Fig. 5.

Correlation between cell fate and P_rapA-iyfp and P_spoIIA-icfp-driven fluorescence for spores (A) and vegetative cells (B) within a colony. Sample trajectories within each class are shown on the top. Within each image, heat maps color code the time evolution of fluorescence intensity (red, high intensity, to blue, low intensity) for each tracked cell trajectory and relate it to the corresponding number of cell divisions after entry into stationary phase (t > T₀). Trajectories in A were arranged according to the time the prespore appeared t_s from bottom to top and in B clustered according to cell lineage relationship.