Quorum sensing employs a dual regulatory mechanism to repress T3SS gene expression

Abstract

The type III secretion system (T3SS) is a needle-like complex used by numerous bacterial pathogens in host infection to inject exotoxins into the host cell cytoplasm. The T3SS is a known virulence factor in the shrimp pathogen Vibrio campbellii. The ~40 genes comprising the V. campbellii T3SS are regulated by a network of transcription factors in response to changes in the cell's environment: cell density (quorum sensing; QS), temperature, calcium, and host cell contact. Under positive environmental stimuli, the master T3SS transcription factor ExsA activates the expression of the four structural T3SS operons required for needle formation. Previous studies identified a key role of the master QS transcription factor LuxR: repression of exsA transcription via DNA binding at the exsBA promoter. Here, we uncovered a new regulatory role of LuxR: post-translational repression of ExsA activity via transcriptional repression of the gene encoding the anti-anti-activator ExsC. In V. campbellii, ExsC is a positive regulator of T3SS transcription; deletion of exsC decreases ExsA-dependent transcription activation of the T3SS structural promoters. Through genetic epistasis and in vitro biochemical assays, we show that LuxR directly binds the exsC promoter upstream of ExsA and represses transcription of exsC. Our findings collectively show that V. campbellii responds to high cell density signals to shut down ExsA-dependent expression of the T3SS via two mechanisms. We postulate that this dual regulatory mechanism by LuxR enables both the rapid inactivation of existing ExsA protein and blocks its further synthesis, leading to a rapid shutdown of T3SS activity at high cell density.

Importance: Vibrio campbellii utilizes the type III secretion system (T3SS) as a mechanism of pathogenesis, which is a highly studied "injectisome" complex that delivers exotoxins into host cells during infection. The T3SS pathogenicity island in V. campbellii comprises ~40 genes that are organized into four structural operons. In this study, we determined that quorum sensing-a method of bacterial communication-regulates T3SS genes both at the transcriptional and post-translational levels to shut down T3SS gene expression at high population densities.

Keywords: T3SS; Vibrio; Vibrio campbellii; quorum sensing; type III secretion.

Fig 1

Schematic models of the Exs regulatory cascade and the organization of the exs regulatory genes in V. campbellii. (A) Schematic model of the regulatory networks that control T3SS gene expression in other organisms. The diagram shows the current model in the T3SS field in which secretion is induced upon T3SS contact with the host cell membrane during infection or upon chelation of calcium ions from media along with the addition of magnesium. The schematic shows the internal Exs regulatory cascade determined in P. aeruginosa and V. parahaemolyticus, which comprises ExsA, ExsD, ExsC, and ExsE. In the absence of induction, transcriptional regulator ExsA is bound to anti-activator ExsD, and anti-anti-activator ExsC is bound to ExsE. Upon induction of the system, ExsE is exported, freeing up ExsC to bind to ExsD, and ExsA protein is free to bind to T3SS promoters I–IV to activate transcription. (B) The schematic shows gene organization of T3SS regulatory genes exsA, exsB, exsC, exsE, and exsD in V. campbellii BB120 and their corresponding VIBHAR_XXXXX locus tags.

Fig 2

T3SS gene expression can be measured using a PvopN-gfp reporter. (A) GFP reporter assay results are shown for isogenic V. campbellii BB120 strains wild-type (WT), ΔexsA, ΔluxR, and ΔexsA ΔluxR containing the P_vopN-gfp reporter (pPP26) and either no additional plasmid (–), the P_tactheo-exsA expression plasmid (pexsA, pPP25) induced with 10 µM IPTG and 1 mM theophylline, the P_tactheo-luxR expression plasmid (pluxR, pPP60) induced with 1 mM IPTG and 1 mM theophylline, or an empty vector control plasmid (EV). (B) RT-qPCR measurements of vopN transcripts from cells collected at HCD (OD₆₀₀ = 1.0) compared to the internal control hfq gene. The strains assayed were wild-type (WT), ΔexsA, ΔluxR, and ΔexsA ΔluxR. (C) GFP reporter assay results are shown for isogenic V. campbellii BB120 strains wild-type (WT), ΔluxR, ΔexsA ΔluxR, ΔluxO, and luxO D61E containing the P_vopN-gfp reporter. (D) GFP reporter assay results are shown for isogenic V. campbellii BB120 strains wild-type (WT) or ΔluxRΔexsA containing the P_vopN-gfp reporter. CRISPRi was used to knockdown gene expression in wild-type strains containing plasmids with guide RNAs targeting exsD (exsDi), exsE (exsEi), and exsC (exsCi) via induction with 100 µM IPTG. For panels A–C, the error bars represent the standard deviation of the mean. A one-way analysis of variance (ANOVA) test was performed on normally distributed data (Shapiro–Wilk test), followed by Tukey’s multiple comparisons test. Different letters indicate significant differences between strains in pairwise comparisons (P < 0.05; n = 3). For panel D, the error bars represent the standard deviation of the mean. A one-way ANOVA test was performed on normally distributed data (Shapiro–Wilk test), followed by Tukey’s multiple comparisons test. Different letters indicate significant differences between strains in pairwise comparisons (P < 0.05; n = 6).

Fig 3

LuxR and ExsA bind and regulate exsC. (A) GFP reporter assay results are shown for strains ΔexsAΔexsB and ΔexsAΔexsB ΔluxR containing P_vopN-gfp reporter (pPP26) and either the P_tactheo-exsA expression plasmid (pexsA, pPP25) induced with 10 µM IPTG and 1 mM theophylline or an empty vector control plasmid (EV). (B) GFP reporter assay results are shown for strain ΔexsAΔexsB containing the P_exsC-gfp reporter (pPP51) and either no additional plasmid (–), P_tactheo-exsA expression plasmid (pexsA, pPP25) induced with 10 µM IPTG and 1 mM theophylline, the P_tactheo-luxR expression plasmid (pluxR, pPP60) induced with 1 mM IPTG and 1 mM theophylline, or an empty vector control plasmid (EV). For both panels, the error bars represent the standard deviation of the mean. A one-way ANOVA test was performed on normally distributed data (Shapiro–Wilk test), followed by Tukey’s multiple comparisons test. Different letters indicate significant differences between strains in pairwise comparisons (P < 0.05; n = 3). (C, D) DNaseI footprinting assays were performed with a 425 bp DNA probe corresponding to the sequence spanning the exsC and exsB promoters. The final concentration of the DNA probe used in the reaction mix was 20 nM. (C) A DNaseI-only treated control reaction (top; -LuxR) was compared with a reaction containing 0.25 µM LuxR (bottom, +LuxR). (D) A DNaseI-only treated control reaction (top, -ExsA) was compared with a reaction containing 1 µM ExsA (bottom, +ExsA). The red boxes indicate the ExsA-binding site (BS), and the blue boxes indicate LuxR BSs a, b, 1, and 2. The gray boxes indicate the −10 and −35 sites.

Fig 4

Deletion of exsC and exsD is epistatic to luxR. (A, B) GFP reporter assay results are shown for isogenic V. campbellii strains all with a ΔexsAΔexsB background and containing either the P_tactheo-exsA expression plasmid (pexsA, pPP25), or the P_tactheo-exsA P_tactheo-exsC double-expression plasmid (pexsA, pexsC, and pPP62), and the P_vopN-gfp chromosomal reporter (pPP26). Other isogenic strain mutations include ΔexsC, ΔluxR, and complementation with the P_tactheo-exsC expression plasmid (pexsC, pPP59), P_tactheo-luxR expression plasmid (pluxR, pPP60), or the empty vector control plasmid (EV). ExsA alone/ExsA-ExsC were induced with 10 µM IPTG and 1 mM theophylline, ExsC alone was induced with 100 µM IPTG and 1 mM theophylline, and LuxR was induced with 1 mM IPTG and 1 mM theophylline. (A, B) Error bars represent the standard deviation of the mean. A one-way ANOVA test was performed on normally distributed data (Shapiro–Wilk test), followed by Tukey’s multiple comparisons test. Different letters indicate significant differences between strains in pairwise comparisons (P < 0.05; n = 3 (A) or n = 4 (B)). (C, D) GFP reporter assay results are shown for isogenic V. campbellii strains all with a ΔexsAΔexsB background and containing the P_tactheo-exsA expression plasmid (pexsA, pPP25) and the P_vopN-gfp chromosomal reporter. Other isogenic strain mutations include ΔexsD, ΔluxR, and complementation with either the P_tactheo-exsD expression plasmid (pexsD, pPP80), P_tactheo-luxR expression plasmid (pluxR, pPP60), or the empty vector control plasmid (EV). ExsA was induced with 10 µM IPTG and 1 mM theophylline, ExsD was induced with 100 µM IPTG and 1 mM theophylline, and LuxR was induced with 1 mM IPTG and 1 mM theophylline. Error bars represent the standard deviation of the mean. A one-way ANOVA test was performed on normally distributed data (Shapiro–Wilk test), followed by Tukey’s multiple comparisons test. Different letters indicate significant differences between strains in pairwise comparisons (P < 0.05; n = 3).

Fig 5

LuxR and ExsA bind the exsC promoter in vitro. (A–D) Electrophoretic mobility shift assays (EMSAs) were performed with a 200 bp DNA probe corresponding to the exsC promoter. The final concentration of DNA used was 2 nM. Lane 1, DNA probe only control. (A) Lanes 2–8, fivefold dilution series of LuxR from 500 nM to 0.032 nM. (B) Lanes 2–8, fourfold dilution series of ExsA from 2 µM to 0.488 nM. Letters a, b, and c indicate different shifted band patterns. (C, D) Lane 2, LuxR or LuxR R17C alone at 700 nM and 1 µM concentrations, respectively. Lane 3, ExsA alone at 500 nM concentration. Lanes 4–10, ExsA was first added at a concentration of 500 nM, followed by addition of either LuxR in a twofold dilution series from 700 nM to 10.94 nM, or LuxR R17C in a twofold dilution series from 1 µM to 15.625 nM. The original images for each gel have been included in the Fig. S8A through D.

Fig 6

LuxR and ExsA co-bind at the exsC promoter. DNaseI footprinting assays were performed with a 425 bp DNA probe corresponding to the sequence spanning the exsC and exsB promoters. The final concentration of the DNA probe used in reaction mix was 20 nM. DNase I-only treated probe (top) was compared with the probe treated with 1 µM ExsA and 0.25 µM LuxR, followed by DNaseI treatment (bottom). Protection by ExsA was observed along the predicted ExsA-binding site at the exsC promoter proximal to the −35 site. Protection by LuxR was observed along the predicted LuxR-binding site at the exsC and exsB promoters.

Fig 7

LuxR represses exsC and exsB transcription through the promoter-proximal LuxR-binding sites. (A, C) Schematic of different lengths of exsC or exsB promoters fused to the gfp cassette. (B, D) GFP reporter assays with the gfp cassette fused to various lengths of the exsC or exsB promoters shown in (A) and (C). The P_exsC-gfp reporters (pPP51, pPP77, pPP83, pPP84, and pPP85) are fused to the chromosome, whereas the P_exsB-gfp reporters (pPP07, pPP82, pPP86, and pPP87) are on ectopic plasmids. Error bars represent the standard deviation of the mean. A one-way ANOVA test was performed on normally distributed data (Shapiro–Wilk test), followed by Tukey’s multiple comparisons test. Different letters indicate significant differences between strains in pairwise comparisons (P < 0.05; n = 4).

Fig 8

Models of the dual mechanism of T3SS regulation by LuxR. (A) Proposed model for activation of exsC transcription by class II activator ExsA and an unidentified class I activator (“?”) at mid-cell density. LuxR binds and displaces the unidentified class I activator by binding to LuxR sites “a/b” at high cell density. B) At mid-cell density, the absence of LuxR (and likely other regulators) enables transcription of the exsC promoter by ExsA, which releases existing ExsA bound by ExsD. ExsA activates transcription of the four T3SS structural operons, enabling synthesis of the T3SS needles. At high cell density, LuxR represses transcription of exsA from the exsBA promoter and represses transcription of exsC. Lower levels of ExsC enable ExsD to bind to ExsA and prevent its activity as a transcriptional activator of T3SS genes.