The two-component system CpxAR controls biofilm formation by directly regulating the T3SS needle tip protein EseB in Edwardsiella piscicida

Abstract

The type III secretion system (T3SS) translocon protein EseB (needle tip protein) forms filamentous appendages on the surface of Edwardsiella piscicida to facilitate autoaggregation and biofilm formation. By contrast, another T3SS translocon protein EseC inhibits biofilm formation by sequestering EseC's chaperone EseE, which also functions as a positive regulator of the escC-eseE operon, in which EseB is encoded. The two-component system (TCS) EsrAB and the regulator EsrC tightly and positively regulate the T3SS in E. piscicida. The TCS CpxAR provides an adaptive response to external environmental changes. In this study, we have shown that disruption of the histidine kinase CpxA (sensor) instead of CpxR (response regulator) significantly reduces biofilm formation in E. piscicida. CpxR is negatively regulated by CpxA, and significant amounts of CpxR accumulate in E. piscicida in the absence of CpxA. CpxR, together with EsrB and EsrC, directly binds the promoter of the cpxR-cpxA operon to promote CpxR transcription and expression. The elevated phosphorylated CpxR (CpxR-P) binds to the promoter of the escC-eseE operon to repress eseB transcription and expression, while EseE, EsrB, and EsrC bind directly to the same promoter to promote EseB transcription and expression. E. piscicida is an enteric pathogen that senses microbiota-derived indole in the gut lumen. EseB filament-mediated biofilm formation in E. piscicida is inversely proportional to exogenous indole. Together, CpxR inhibits while EsrB, EsrC, and EseE stimulate transcription and expression of the escC-eseE operon, thereby coordinately controlling EseB filament-mediated biofilm formation in E. piscicida in response to environmental stimuli.IMPORTANCEEdwardsiella piscicida is primarily an enteric pathogen of fish and can form a biofilm to resist the lethal effects of host or antimicrobial agents. The assembly of filamentous appendages on the bacterial surface, mediated by the type III secretion system (T3SS) needle tip protein EseB, promotes bacterial-bacterial interactions and biofilm formation when E. piscicida is cultured in Dulbecco's modified Eagle's medium (DMEM). In this study, we have shown that the histidine kinase CpxA regulates biofilm formation in E. piscicida by negatively regulating its response regulator CpxR. Binding to the promoter of the escC-eseE operon, CpxR negatively regulates, whereas EsrB, EsrC, and EseE positively regulate the escC-eseE operon, of which EseB is encoded, coordinately regulating biofilm formation in E. piscicida.

Keywords: CpxAR; Edwardsiella piscicida; T3SS needle tip; biofilm.

Fig 1

CpxA induces autoaggregation and biofilm formation in E. piscicida. (A) Autoaggregation of E. piscicida strains in DMEM at 25.0°C under a 5.0% CO₂ atmosphere in a glass test tube at 24 hours post-subculture (hps). (B) Biofilm formed by E. piscicida strains in DMEM. E. piscicida strains were subcultured in DMEM in a 24-well plate horizontally embedded with coverslips, biofilms developed on the coverslips were fixed and stained with 0.2% crystal violet (left panel), and biofilm formation was evaluated by examining the OD₆₃₀ of the dissolved crystal violet (right panel). ***P < 0.001; *P < 0.05; NS, not significant. (C) Immunoblotting of the steady-state protein levels of EseB in E. piscicida strains. Total bacterial proteins (TBPs) from equal amounts of E. piscicida strains were probed with rabbit anti-EseB and rabbit anti-DnaK antibodies. DnaK was used as a loading control (left panel). Protein levels of EseB from bacterial pellets were quantified by densitometry and normalized to DnaK. The graphs show the relative ratios of intracellular EseB, which are averages of the results of at least three independent experiments (right panel). ***P < 0.001; NS, not significant.

Fig 2

CpxA represses the expression and secretion of the T3SS translocon protein EseB. (A) Immunoblotting of EseB expression and secretion in E. piscicida strains. Total bacterial proteins (TBPs) and extracellular proteins (ECPs) from equal amounts of E. piscicida WT, ∆cpxA, and ∆cpxA[cpxA] strains were probed with rabbit anti-EseB and rabbit anti-DnaK antibodies. DnaK, a cytosolic chaperone, was used as the loading control. The immunoblotting data shown are representative of three independent experiments. (B) Immunofluorescence staining of E. piscicida strains with mouse anti-EseB antibody. E. piscicida WT, ∆cpxA, ∆cpxA[cpxA], and ∆eseB strains were subcultured in DMEM in a 24-well plate horizontally embedded with coverslips. The biofilm formed on the coverslips was immunofluorescently stained, and images were captured using a confocal laser scanning microscope. Scale bar, 5.0 µm, scale bar for magnification, 15.0 µm. (C) The immunofluorescence images of the E. piscicida WT, ∆cpxA, and ∆cpxR strains, each being introduced with pFPV-escC_{-200 to -1}-gfp. At 24 hps in DMEM, the image for each strain was captured under a confocal microscope. Scale bar, 50.0 µm. (D) Fluorescence intensity of the GFP signal in E. piscicida WT, ∆cpxA, and ∆cpxR strains, each being introduced with pFPV25-escC_{-200 to -1}-gfp. Fluorescence intensity indicates the protein level of GFP in each strain examined at 24 hps in DMEM. ***P < 0.001. (E) The mRNA levels of escC, eseB, escA, eseC, eseD, and eseE in the escC–eseE operon in E. piscicida strains were investigated by qRT-PCR. Multi-reference genes (rpoB and gyrB) were used to determine the relative transcript levels of each gene. Data are expressed as mean ± SD. One-way ANOVA in SPSS was used to calculate the P values as compared to the WT strain. ***P < 0.001; NS, not significant.

Fig 3

CpxA negatively regulates CpxR, and phosphorylated CpxR represses the transcription and expression of EseB. (A) The mRNA levels of cpxR in E. piscicida WT and ∆cpxA strains were examined by qRT-PCR. Multi-reference genes (rpoB and gyrB) were used to determine the relative transcript levels of cpxR. Data are expressed as mean ± SD. One-way ANOVA in SPSS was used to calculate the P values as compared to the WT strain. ***P < 0.001. (B) Fluorescence intensity of the GFP signal of the WT strain or of the ∆cpxA strain being introduced with pFPV25-cpxR_{-226 to -1}-gfp. The GFP signal reflects the strength of the regulated cpxR promoter in different E. piscicida strains in DMEM at 24 hps. Data shown are representative of three independent experiments. ***P < 0.001. (C) Immunoblotting of the steady-state protein levels of CpxR in E. piscicida strains. Total bacterial proteins from similar amounts of E. piscicida strains were harvested at 24 hps in DMEM and probed with rabbit anti-CpxR and rabbit anti-DnaK antibodies (left panel). DnaK was used as a loading control. CpxR protein levels were quantified by densitometry and normalized to DnaK. The graphs show the relative ratios of intracellular CpxR, which are the averages of at least three independent experiments (right panel). ***P < 0.001; NS, not significant. (D) Immunoblotting for expression and secretion of EseB from E. piscicida strains ∆cpxR/pJN-cpxR-2HA and ∆cpxR/pJN-cpxR_D51A-2HA. The TBPs and ECPs from equal amounts of E. piscicida strains were probed with rabbit anti-EseB, rabbit anti-HA (CpxR), and rabbit anti-DnaK antibodies. CpxR expression was induced by supplementation with L-arabinose at final concentrations of 1.0 mM, 10.0 mM, and 50.0 mM. To study the phosphorylated CpxR (HA), proteins were isolated by Mn²⁺ Phos-Tag SDS-PAGE gel electrophoresis before transfer to PVDF membrane. The images are representative of three independent experiments. (E) Immunoblotting of the steady-state protein levels of EseB, phosphorylated and non-phosphorylated CpxR-2HA in E. piscicida strains ∆cpxA∆cpxR/pJN-cpxR-2HA and ∆cpxA∆cpxR/pJN-cpxR_D51A-2HA induced by L-arabinose supplementation at final concentrations of 1.0 mM and 10.0 mM. The TBPs from equal amounts of E. piscicida cultures were isolated by Mn²⁺ Phos-Tag SDS-PAGE gel electrophoresis to isolate the phosphorylated and non-phosphorylated CpxR-2HA (left panel) or CpxR_D51A-2HA (right panel) before transfer to a PVDF membrane and probed with rabbit anti-EseB, rabbit anti-HA (CpxR), and rabbit anti-DnaK antibodies. The images are representative of three independent experiments.

Fig 4

The co-transcribed cpxR and cpxA are positively and directly regulated by phosphorylated CpxR, and to a lesser extent by unphosphorylated CpxR. (A) The PCR products were electrophoresed and photographed. PCR was performed using the specific primer set covering cpxR and cpxA, with the total RNA, cDNA, or genomic DNA as the template. (B) Immunoblotting of the steady-state protein level of CpxA-2HA in E. piscicida strains. TBPs from equal amounts of WT/pWSK29-cpxR_{-762 to -1}-cpxA-2HA strain (with the promoter upstream of cpxR), WT/pWSK29-cpxA_{-528 to -1}-cpxA-2HA strain (with the promoter upstream of cpxA), and WT/pWSK29-cpxA-2HA strain (without the promoter) were probed with antibodies against HA (CpxA-2HA) and DnaK. The image shown is the representative of three independent experiments. (C) EMSA on phosphorylated CpxR and the DNA fragment nt −226 to +30 of cpxR. The Cy3-labeled DNA fragment (1.0 µg) was incubated with 3.0 µM CpxR, which was phosphorylated with 0.2 M lithium potassium acetyl phosphate (AcP) as a phosphate donor in the kinase buffer at 30.0°C for 1 h. The Cy3-labeled DNA fragment nt +2 to +160 of cpxR was used as a negative control probe. The protein-DNA complex was resolved on a 5% non-denaturing polyacrylamide gel. (D) EMSA on CpxR_D51A and the DNA fragment nt −226 to +30 upstream of cpxR. The Cy3-labeled DNA fragment (1.0 µg) was incubated with 5.0 µM CpxR_D51A or 5.0 µM CpxR (as the positive control). The Cy3-labeled DNA fragment nt +2 to +160 of cpxR was used as a negative control probe. The protein-DNA complex was resolved on a 5% non-denaturing polyacrylamide gel. (E) The fluorescence intensity of the GFP signal per 1 × 10⁸ CFU (colony-forming units) of the E. piscicida strain ∆cpxR/pFPV-cpxR_{-226 to -1}-gfp/pJN-cpxR-2HA strain or ∆cpxR/pFPV-cpxR_{-226 to -1}-gfp/pJN-cpxR_D51A-2HA (top panel). Fluorescence intensity of GFP in E. piscicida was measured 24 h after subculture in DMEM supplemented with L-arabinose at a final concentration of 1.0 mM, 10.0 mM, or 50.0 mM. Mn²⁺ Phos-Tag SDS-PAGE gel electrophoresis was used to isolate the phosphorylated and non-phosphorylated CpxR-2HA (left panel) or CpxR_D51A-2HA (right panel) before transfer to a PVDF membrane and probed with anti-HA and DnaK antibodies. DnaK was used to indicate similar amounts of protein loading per lane. Images shown are representative of three independent experiments.

Fig 5

EsrB and EsrC bind to upstream of cpxR and positively regulate the cpxR–cpxA operon. (A) The transcript levels of cpxR and cpxA in E. piscicida strains were analyzed by qRT-PCR. Transcription of cpxR or cpxA was normalized against the multi-reference genes rpoB and gyrB. Means ± SD of one representative experiment are shown. One-way ANOVA in SPSS was used to calculate the P values as compared to the WT strain. ***P < 0.001; NS, not significant. (B) Immunoblotting of the steady-state protein levels of CpxR or CpxA-2HA in E. piscicida strains. TBPs of WT, ∆esrB, and ∆esrC strains were resolved and probed with rabbit anti-CpxR and rabbit anti-HA (CpxA-2HA), respectively. RpoA was included to show the similar amount of protein loading per lane (left panel). Protein levels of CpxA-2HA and CpxR were quantified by densitometry and normalized to RpoA. The graphs show the relative ratios of intracellular CpxR or CpxA-2HA, which are the average of at least three independent experiments (right panel); ***P < 0.001. (C) EMSA on EsrB or EsrC and the DNA fragment nt −226 to +30 of cpxR. The Cy3-labeled DNA fragment −226 to +30 upstream of cpxR (1.0 µg) or +2 to +160 of cpxR (negative control) was incubated with the indicated concentrations of EsrB or EsrC protein for EMSA before the protein-DNA complex was resolved on a 5.0% non-denaturing polyacrylamide gel. (D) Genetic organization of the cpxR–cpxA operon in E. piscicida. The cpxR and cpxA share the same promoter (−226 to −1 upstream of cpxR), but cpxA could also be transcribed independently. (E) The cpxP gene is located in a separate operon next to the cpxA–cpxR operon, which is directly and positively regulated by CpxR, EsrB, and EsrC.

Fig 6

CpxR, EsrB, EsrC, and EseE coordinately regulate biofilm formation by directly regulating the escC–eseE operon, in which EseB is encoded. (A) EMSA on phosphorylated and non-phosphorylated CpxR and the Cy3-labeled DNA fragment nt −200 to −1 upstream of escC. The CpxR protein (5.0 µM) was incubated with or without 0.4 M lithium potassium acetyl phosphate (AcP) as the phosphate donor in the kinase buffer at 30.0°C for 1 h before the mixture was added to the Cy3-labeled DNA fragment from −200 to −1 upstream of escC (1.0 µg) or +2 to +160 of cpxR (negative control) for EMSA. Protein-DNA complexes were resolved on a 5% non-denaturing polyacrylamide gel. (B) EMSA between the indicated concentrations of EseE and the Cy3-labeled DNA fragment from −200 to −1 upstream of escC (1.0 µg) or +2 to +160 of cpxR (negative control) before resolving the protein-DNA complex on a 5.0% non-denaturing polyacrylamide gel. (C) EMSA between the indicated concentrations of EsrB and the Cy3-labeled DNA fragment from −200 to −1 of escC (1.0 µg) or +2 to +160 of cpxR (negative control) before resolving the protein-DNA complex on a 5.0% non-denaturing polyacrylamide gel. (D) EMSA between the indicated concentrations of EsrC and the Cy3-labeled DNA fragment of −200 to −1 of escC (1.0 µg) or +2 to +160 of cpxR (negative control) before resolving the protein-DNA complex on a 5.0% non-denaturing polyacrylamide gel. (E) The promoter region of escC. The CpxR box and the EsrB box, where CpxR or EsrB binds to the escC promoter, are underlined. S1 and S2 are the predicted transcription start sites. The bold italic nucleotide motifs indicate the −35 box and the −10 box of the two promoters; the predicted RBS and the escC start codon are also labeled. (F) Biofilm formation in E. piscicida WT, ∆eseE, ∆cpxA, ∆esrB, and ∆esrC strains. E. piscicida strains were subcultured for 24 h in DMEM in a 24-well plate horizontally embedded with coverslips, and the biofilm developed on the coverslips was fixed and stained with 0.2% crystal violet. The images shown are representative of three independent experiments (top panel), and biofilm formation was assessed by examining the OD₆₃₀ of the dissolved crystal violet (bottom panel). ***P < 0.001.

Fig 7

Both the exogenous and endogenous indole suppressed biofilm formation in E. piscicida. (A) Biofilm formed by E. piscicida in the presence of exogenous indole. WT and ∆cpxA strains subcultured in DMEM in a 24-well plate were supplemented with indole at final concentrations of 0 mM, 0.3 mM, and 0.5 mM. After 24 hours, the biofilm formed on the bottom of the plate was fixed and stained with 0.2% crystal violet. Images are representative of at least three independent experiments. (B) Immunoblotting of the steady-state protein levels of EseB, EseD, EseG, EseJ, and EsrC-FLAG in E. piscicida strains subcultured for 24 h in DMEM in the presence of 0.1 mM, 0.3 mM, and 0.5 mM indole. TBPs from similar amounts of E. piscicida strains were probed with rabbit anti-EseB, rabbit anti-EseD, rabbit anti-EseG, rabbit anti-EseJ, rabbit anti-DnaK, and mouse anti-FLAG (EsrC-FLAG) antibodies. Images shown are representative of at least three independent experiments. (C) Immunoblotting of the steady-state protein levels of EseB in E. piscicida WT, ∆tnaA, ∆cpxA, and ∆tnaA∆cpxA strains subcultured for 24 h in DMEM in the presence of 0.3 mM indole. Similar amounts of bacterial lysates from each strain were probed with rabbit anti-EseB and rabbit anti-DnaK antibodies. Images shown are representative of at least three independent experiments.

Fig 8

Schematic representation of the regulation of biofilm formation in E. piscicida. Culturing E. piscicida in DMEM simulates in vivo nutrient limitation conditions that activate T3SS. Increased expression of the T3SS needle tip protein EseB promotes EseB-filament-mediated bacterial-bacterial interaction and biofilm formation. CpxA negatively regulates CpxR, whereas EsrB, EsrC, and CpxR directly and positively regulate CpxR. Phosphorylated CpxR binds to the promoter of the escC−eseE operon to negatively regulate eseB. EseE, EsrB, or EsrC binds to the promoter of the escC−eseE operon to positively regulate eseB. EsrB and EsrC, which play a major role in this regulation, are indicated by the thick arrow. Microbiota-derived indole, sensed and transduced by an unknown signaling pathway, downregulates EseB filament-mediated biofilm formation.