Transition of Dephospho-DctD to the Transcriptionally Active State via Interaction with Dephospho-IIAGlc

Abstract

Exopolysaccharides (EPSs), biofilm-maturing components of Vibrio vulnificus, are abundantly produced when the expression of two major EPS gene clusters is activated by an enhancer-binding transcription factor, DctD₂, whose expression and phosphorylation are induced by dicarboxylic acids. Surprisingly, when glucose was supplied to V. vulnificus, similar levels of expression of these clusters occurred, even in the absence of dicarboxylic acids. This glucose-dependent activation was also mediated by DctD₂, whose expression was sequentially activated by the transcription regulator NtrC. Most DctD₂ in cells grown without dicarboxylic acids was present in a dephosphorylated state, known as the transcriptionally inactive form. However, in the presence of glucose, a dephosphorylated component of the glucose-specific phosphotransferase system, d-IIA^Glc, interacted with dephosphorylated DctD₂ (d-DctD₂). While d-DctD₂ did not show any affinity to a DNA fragment containing the DctD-binding sequences, the complex of d-DctD₂ and d-IIA^Glc exhibited specific and efficient DNA binding, similar to the phosphorylated DctD₂. The d-DctD₂-mediated activation of the EPS gene clusters' expression was not fully achieved in cells grown with mannose. Furthermore, the degrees of expression of the clusters under glycerol were less than those under mannose. This was caused by an antagonistic and competitive effect of GlpK, whose expression was increased by glycerol, in forming a complex with d-DctD₂ by d-IIA^Glc. The data demonstrate a novel regulatory pathway for V. vulnificus EPS biosynthesis and biofilm maturation in the presence of glucose, which is mediated by d-DctD₂ through its transition to the transcriptionally active state by interacting with available d-IIA^Glc. IMPORTANCE Transcription regulation by bacterial two-component systems is achieved by a response regulator upon its transition to the transcriptionally active form via kinase activity of its cognate sensor under specific conditions. A well-known response regulator, DctD, is converted to its phosphorylated form when DctB senses ambient dicarboxylic acids. Phospho-DctD induces expression of its regulon, including the gene clusters for biosynthesis of exopolysaccharides (EPSs), the essential constituents of biofilm matrix. In the absence of dicarboxylic acids, however, DctD-mediated induction of these EPS gene clusters and biofilm maturation was observed if glucose was supplied. This suggests that dephospho-DctD could play a role in activating the transcription of target genes. A component of glucose-phosphotransferase system, IIA^Glc, was present in a dephosphorylated state in the presence of glucose. Dephospho-DctD formed a complex with dephospho-IIA^Glc and was converted to a transcriptionally active state. These findings suggest the other response regulators could also have alternative pathways of activation independent of phosphorylation.

Keywords: biofilm; dephospho-DctD; dephospho-IIAGlc; exopolysaccharides; glucose.

FIG 1

Biofilm formation and EPS production by V. vulnificus in the presence of fumarate, glucose, mannose, or glycerol. (A and B) Biofilm formation. Wild-type V. vulnificus cells were statically incubated for 48 h in AB medium supplemented with fumarate, glucose, mannose, or glycerol. The biofilms that developed on the borosilicate tubes were estimated by staining with crystal violet (A). The associated dyes were dissolved and measured by spectrophotometry at 550 nm (B). The P values for comparison with the AB-fumarate incubation are indicated (**, P < 0.001; *, 0.001 ≤ P < 0.01; ns, not significant). A comparison between AB-mannose and AB-glycerol incubations was also included with a P value above a corresponding horizontal line. (C and D) EPS production. Wild-type V. vulnificus cells grown on the AB agar plates supplemented with fumarate, glucose, mannose, or glycerol, were subjected to the process for EPS extraction, as previously described (28). The resultant EPS extracts run in a 5% stacking polyacrylamide gel were visualized by staining with Stains-All (C). The carbohydrate contents in each extract were measured by the phenol-sulfuric acid method (64). The estimated carbohydrate contents were expressed as mM glucose equivalents per cell masses equivalent to an OD₅₉₅ of 1.0 (D). P values are presented as described above.

FIG 2

Expression of three EPS gene clusters of V. vulnificus in the presence of fumarate, glucose, mannose, or glycerol. Wild-type V. vulnificus cells carrying a luxAB transcriptional reporter fused with each EPS gene cluster (EPS-I [A], EPS-II [B], or EPS-III [C]) (28) were grown in AB-fumarate, -glucose, -mannose, or -glycerol medium supplemented with 3 μg/mL tetracycline. At an OD₅₉₅ of 0.4, cells were harvested, and their luciferase activities were measured using a luminometer. The degree of cluster expression was expressed as a normalized value: the relative light unit (RLU) divided by the cell mass (OD₅₉₅) of each sample. The P values are presented as described in the legend to Fig. 1.

FIG 3

Expression of ntrC and dctD₂ genes of V. vulnificus in the presence of fumarate, glucose, mannose, or glycerol. Wild-type V. vulnificus cells carrying a luxAB transcriptional reporter fused with the regulatory region of the glnA-ntrB-ntrC operon (32) (A) or the regulatory region of the dctB₂D₂ operon (16) (B) were grown for 2 h in AB medium supplemented with glycerol (22 mM) and tetracycline (3 μg/mL) and then subcultured to AB-fumarate, -glucose, -mannose, and -glycerol media supplemented with 3 μg/mL tetracycline. Bacterial cells were harvested at every 30 min, and light production was measured and presented as described in the legend to Fig. 2. The arrows indicate the point at which V. vulnificus cells pregrown in AB-glycerol were challenged with different carbon sources.

FIG 4

Expression of the EPS-II and EPS-III gene clusters in various mutant V. vulnificus strains defective in DctD₂ and/or IIA^Glc. V. vulnificus strains carrying mutant DctD₂ (i.e., ΔdctD₂, a phospho-form of DctD [dctD_D57E], and a dephospho-form of DctD [dctD_D57Q]), mutant IIA^Glc (i.e., Δcrr and a dephospho-form of IIA^Glc [crr_H75Q]), or deletion of two ORFs (Δcrr ΔdctD₂) were used to compare the expression of two EPS gene clusters. Each strain carrying either EPS-II (A) or EPS-III (B) was inoculated in AB-fumarate and AB-glucose media. After incubation for 4 h, the cell densities and light production were measured, and the extents of the gene clusters’ expression were presented as described in the legend to Fig. 2. The P values for comparison with the samples in fumarate and glucose are indicated: **, P < 0.001; ns, not significant.

FIG 5

Specific interaction between DctD₂ and IIA^Glc proteins. (A to D) Bacterial two-hybrid system. Two plasmids, pUT18c-dctD₂ and pKT25-crr (Table S1), were constructed and cotransformed to E. coli BTH101, as described in Materials and Methods. This transformant was grown in M9 medium supplemented with glucose (Glc) (A and B) or phosphoenolpyruvate (PEP) (C and D), and then the resultant β-galactosidase activities were examined. For comparison, the negative and positive controls (E. coli BTH101 harboring pUT18c/pKT25 and pUT18c-zip/pKT25-zip, respectively), were included in these assays: shown are blue colonies on the agar plates supplemented with 40 μg/mL X-Gal (A and C) and the specific β-galactosidase activities in Miller units (B and D), as described in Materials and Methods. (E) In vitro interaction between d-IIA^Glc and d-DctD₂. To examine the role of the phosphorylated states of DctD₂ in specific interaction with d-IIA^Glc, both phosphorylated (p-) and dephosphorylated (d-) forms of recombinant DctD₂ were prepared: the original DctD₂ (DctD_WT), p-DctD₂ (DctD_D57E), and d-DctD₂ (DctD_D57Q) (Table S1). Various combinations of DctD₂ and d-IIA^Glc proteins were mixed, and the reaction mixtures were run in a native gel. Lane 1, DctD_WT (5 μM); lane 2, DctD_D57E (5 μM); lane 3, DctD_D57Q (5 μM); lane 4, d-IIA^Glc (5 μM); lanes 5 to 8, d-IIA^Glc (5 μM) with DctD_D57Q at a concentration of 0.04, 0.2, 1, or 5 μM; lanes 9 to 12, DctD_D57Q (5 μM) with d-IIA^Glc at a concentration of 0.04, 0.2, 1, or 5 μM; and lane 13, DctD_D57E (5 μM) with d-IIA^Glc (5 μM). Each protein band and the newly emerged band are indicated with arrows.

FIG 6

Binding affinities of p-DctD₂ and d-DctD₂/d-IIA^Glc complex to the regulatory region of the EPS-II cluster. The labeled DNA probe (50 nM), covering −391 to +61 relative to the TIS of the EPS-II gene cluster (16), was incubated with DctD_WT (A), DctD_D57E (B), and DctD_D57Q (C) at concentrations ranging from 100 to 700 nM. To examine the role of d-IIA^Glc in DNA binding of DctD, the same mixtures were also added with 500 nM d-IIA^Glc (D, E, and F). The reaction mixtures were resolved in 6% native polyacrylamide gels. Lane 1, probe only; and lanes 2 to 9, probe with DctD proteins at a concentration of 100, 150, 200, 300, 400, 500, 600, or 700 nM DctD_WT. DNA probes bound by p-DctD (A, B, D, and E) or a complex of d-DctD with d-IIA^Glc (D and F) are indicated with arrows.

FIG 7

Cellular levels of DctD₂, IIA^Glc, and GlpK proteins in V. vulnificus cells grown in fumarate, glucose, mannose, or glycerol. Protein levels of DctD₂ (A), IIA^Glc (B), and GlpK (C) in wild-type V. vulnificus cells, which were freshly grown in AB-fumarate, -glucose, -mannose, or -glycerol medium (up to an OD₅₉₅ of 0.4), were compared by Western blot analysis. For SDS-PAGE, 120, 20, and 50 μg of cell lysates were loaded to detect bands of DctD₂, IIA^Glc, and GlpK, respectively. As a negative control for each blot, lysates of the dctD, crr, or glpK deletion mutants were included. The intensities of the corresponding protein bands on each blot were quantified, and their relative amounts (normalized by the intensities of DctD₂ in cells grown in AB-fumarate [D], IIA^Glc in cells grown in AB-glucose [E], and GlpK in cells grown in AB-glycerol [F]) were plotted, with P values indicated as follows: **, P < 0.001; ns, not significant.

FIG 8

Effect of GlpK on the binding ability of the d-DctD₂/d-IIA^Glc complex to DNA probe. (A) Addition of GlpK to the probe mixed with d-DctD₂ or p-DctD₂. Labeled DNA probes (50 nM), used in Fig. 6, were mixed with either 300 nM DctD_D57Q (lanes 3 to 5) or 300 nM DctD_D57E (lanes 8 to 10). To examine the effect of GlpK on the interaction of DctD₂ and DNA probe, various concentrations of recombinant GlpK ranging from 300 to 500 nM were added, and then the reaction mixtures were resolved in a 6% native polyacrylamide gel. Lanes 1 and 6, DNA probe only; and lanes 2 and 7, DNA probe with 300 nM DctD₂. Unbound DNA and the probes bound by p-DctD₂ are indicated with arrows. (B) Addition of GlpK to the probe mixed with d-DctD and d-IIA^Glc. Labeled DNA probes (50 nM) mixed with 300 nM DctD_D57Q and 300 nM d-IIA^Glc were further mixed with various concentrations of recombinant GlpK ranging from 50 to 400 nM (lanes 3 to 9). The reaction mixtures were resolved in a 6% native polyacrylamide gel. Lane 1, DNA probe with DctD_D57Q (300 nM); lane 2, DNA probe with DctD_D57Q (300 nM) and d-IIA^Glc (300 nM); and lane 10, DNA probe with GlpK (400 nM). Unbound DNA and the probes bound by a complex of d-DctD₂ and d-IIA^Glc are indicated with arrows. (C) Addition of GlpK to the probe mixed with DctD_WT and d-IIA^Glc. Labeled DNA probes (50 nM), mixed with 300 nM DctD_WT (which included both d-DctD₂ and p-DctD₂) and 300 nM d-IIA^Glc, were further mixed with various concentrations of recombinant GlpK ranging from 50 to 400 nM (lanes 3 to 9). The reaction mixtures were resolved in a 6% native polyacrylamide gel. Lane 1, DNA probe with DctD_WT (300 nM); lane 2, DNA probe with DctD_WT (300 nM) and d-IIA^Glc (300 nM); and lane 10, DNA probe with GlpK (400 nM). DNA probes bound by p-DctD₂ or a complex of d-DctD₂ with d-IIA^Glc are indicated with arrows.

FIG 9

Biofilm formation by Δcrr and ΔglpK mutant strains of V. vulnificus in the presence of mannose. The wild-type and Δcrr and ΔglpK mutant strains were statically incubated for 48 h in AB medium supplemented with mannose, and the resultant biofilms on the borosilicate tubes were estimated by staining with crystal violet (A). For comparison, biofilms formed by the wild type in AB medium supplemented with glucose or glycerol were included. The associated dyes were dissolved and measured by spectrophotometry at 550 nm (B). The P values are indicated above corresponding horizontal lines: **, P < 0.001; *, 0.001 ≤ P < 0.01; ns, not significant.

FIG 10

DctD₂-directed regulatory pathways for expression of two gene clusters for EPS-II/III biosynthesis in response to various carbon sources. Production of EPS, which is essential for V. vulnificus to mature biofilms, is controlled at the level of transcription of the gene clusters for EPS biosynthesis (EPS clusters). Thus, the cellular abundance of transcriptionally active forms of DctD₂ responsible for regulating the EPS-II and EPS-III clusters (16) determines the degrees of EPS production and biofilm maturation. Transcription of dctD₂ is activated by NtrC, whose expression is induced under conditions with TCA intermediates (i.e., di- and tricarboxylic acids) (16) and glucose. In the presence of dicarboxylic acids, most DctD₂ is present in the transcriptionally active state as a phosphorylated form (p-DctD₂) by its cognate sensor kinase DctB₂ (42). The transcriptionally inactive, dephosphorylated form of DctD₂ (d-DctD₂) is capable of activating the transcription of EPS gene clusters in cells grown with glucose, in which d-DctD₂ forms a complex with d-IIA^Glc. This d-DctD₂-mediated transcription of the EPS-gene clusters is reduced in the presence of glycerol, due to a competitive inhibition of GlpK against the formation of the d-DctD₂/d-IIA^Glc complex.