Interplay among cyclic diguanylate, HapR, and the general stress response regulator (RpoS) in the regulation of Vibrio cholerae hemagglutinin/protease

Abstract

Vibrio cholerae secretes the Zn-dependent metalloprotease hemagglutinin (HA)/protease (mucinase), which is encoded by hapA and displays a broad range of potential pathogenic activities. Expression of HA/protease has a stringent requirement for the quorum-sensing regulator HapR and the general stress response regulator RpoS. Here we report that the second messenger cyclic diguanylic acid (c-di-GMP) regulates the production of HA/protease in a negative manner. Overexpression of a diguanylate cyclase to increase the cellular c-di-GMP pool resulted in diminished expression of HA/protease and its positive regulator, HapR. The effect of c-di-GMP on HapR was independent of LuxO but was abolished by deletion of the c-di-GMP binding protein VpsT, the LuxR-type regulator VqmA, or a single-base mutation in the hapR promoter that prevents autorepression. Though expression of HapR had a positive effect on RpoS biosynthesis, direct manipulation of the c-di-GMP pool at a high cell density did not significantly impact RpoS expression in the wild-type genetic background. In contrast, increasing the c-di-GMP pool severely inhibited RpoS expression in a ΔhapR mutant that is locked in a regulatory state mimicking low cell density. Based on the above findings, we propose a model for the interplay between HapR, RpoS, and c-di-GMP in the regulation of HA/protease expression.

Fig. 1.

Regulation of HA/protease production by c-di-GMP. (A) Production of HA/protease. Strain C7258, containing plasmids pAT1568 (VieA, EAL domain), pAT1615 [VieA(E170A), inactive EAL domain], and pAT1662 (VCA0956, GGDEF domain), was grown in TSB medium, and PDE or DGC expression was induced as described in Materials and Methods. Cultures were incubated for 16 h, and the production of HA/protease was determined. Each value represents the mean for three independent cultures. Error bars indicate standard deviations. **, P < 0.01. (B) Motility assay. Overnight cultures of C7258 transformants were stabbed into swarm agar with or without (control) l-arabinose, and the plates were incubated at 30°C. Each of the resulting swarm diameters is the mean for three independent cultures. Error bars indicate standard deviations (*, P < 0.05; **, P < 0.01 for comparison of induced versus control cultures).

Fig. 2.

Effect of manipulating the c-di-GMP pool on hapA, hapR, and rpoS mRNA expression. Strain C7258, containing plasmids expressing VieA, VieA(E170A), and VCA0956, was grown in TSB medium, and PDE or DGC expression was induced as described in Materials and Methods. Total RNA was extracted, and hapA, hapR, and rpoS mRNAs were measured by qRT-PCR. Each value represents the mean for three independent cultures. Error bars indicate standard deviations (*, P < 0.05; **, P < 0.01 for comparison of induced versus control cultures).

Fig. 3.

Effect of c-di-GMP on HapR expression in different genetic backgrounds. Strain SZS007, containing a HapR-dependent luxC-lacZ fusion (WT), and the isogenic SZS012 (ΔluxO), HX02 (ΔvpsT), AJB505 (ΔvqmA), and HX01 (hapRp*) mutants were transformed with plasmids expressing VieA and VCA0956. Transformants were grown in TSB medium, PDE or DGC expression was induced as described in Materials and Methods, and β-galactosidase activity was determined. Each value represents the mean for three independent cultures. Error bars indicate standard deviations (**, P < 0.01 for comparison of induced versus control cultures).

Fig. 4.

Induction of RpoS in quorum-sensing mutants. (A) Strain AJB502, containing a chromosomally integrated rpoS-FLAG reporter, was grown to an OD₆₀₀ of 2.0 (time zero), and samples were taken at different time points for Western blot analysis of RpoS-FLAG expression. (B) Strains AJB504 (ΔluxO) and AJB503 (ΔhapR), containing a chromosomally integrated rpoS-FLAG reporter, were grown to an OD₆₀₀ of 2.0, and cells were collected at different time points for Western blot analysis of RpoS-FLAG expression. Band intensities (BI) were determined using Kodak 1D image analysis software.

Fig. 5.

Effect of c-di-GMP on RpoS expression at a high cell density. Strain AJB502, which produces rpoS-FLAG, was transformed with plasmids expressing VieA and VCA0956. Transformants were grown in TSB medium to an OD₆₀₀ of 2.0 (time zero), and PDE or DGC expression was induced as described in Materials and Methods. Induction of RpoS-FLAG was determined by Western blotting.

Fig. 6.

Regulation of RpoS expression by c-di-GMP in the absence of HapR. Strains AJB504 (ΔluxO), AJB503 (ΔhapR), and HX03 (ΔhapR ΔvpsT), containing the rpoS-FLAG reporter and the VCA0956 expression vector, as well as strains AJB503 and N16961-FLAG, containing the QrgB expression vector, were grown to an OD₆₀₀ of 2.0 (time zero). At this stage, cultures were divided in halves; one half was induced by adding l-arabinose for VCA0956 or IPTG for QrgB, and the other half was used as a control. Samples were taken at different time points for Western blot analysis of RpoS-FLAG expression.

Fig. 7.

Model for the interplay between c-di-GMP, HapR, and RpoS in the regulation of HA/protease expression. The second messenger c-di-GMP negatively affects the expression of HapR. This regulation is independent of LuxO but requires the c-di-GMP binding protein VpsT and VqmA. At a high cell density, HapR enhances RpoS (pathway A) and both regulators diminish the c-di-GMP pool, generating a double-negative regulatory loop that further enhances the expression of HapR, HA/protease production, and motility (see the text for details). We propose that in the absence of quorum sensing, c-di-GMP also inhibits the expression of RpoS (pathway B).