Vibrio parahaemolyticus ExsE is requisite for initial adhesion and subsequent type III secretion system 1-dependent autophagy in HeLa cells

Abstract

Vibrio parahaemolyticus pandemic serotype O3 : K6 causes acute gastroenteritis, wound infections and septicaemia in humans. This organism encodes two type III secretion systems (T3SS1 and T3SS2); host-cell cytotoxicity has been attributed to T3SS1. Synthesis and secretion of T3SS1 proteins is positively regulated by ExsA, which is presumptively regulated by the ExsCDE pathway, similar to Pseudomonas aeruginosa. Herein we deleted the putative exsE from V. parahaemolyticus and found constitutive expression of the T3SS1 in broth culture as expected. More importantly, however, in a cell culture model, the ΔexsE strain was unable to induce cytotoxicity, as measured by release of lactate dehydrogenase (LDH), or autophagy, as measured by LC3 conversion. This is markedly different from P. aeruginosa, where deletion of exsE has no effect on host-cell cytolysis. Swarming and cytoadhesion were reduced for the deletion mutant and could be recovered along with T3SS1-induced HeLa cell cytotoxicity by in cis expression of exsE in the ΔexsE strain. Loss of adhesion and swarming motility was associated with the loss of flagella biogenesis in the exsE-deficient strain. Mouse mortality was unaffected by the deletion of exsE compared with a wild-type control, suggesting that additional adhesins are important for intoxication in vivo. Based on these data, we conclude that ExsE contributes to the negative regulation of T3SS1 and, in addition, contributes to regulation of an adherence phenotype that is requisite for translocation of effector proteins into HeLa cells.

Fig. 1.

Deletion of exsE does not negatively affect synthesis or secretion of Vp1656. (a) The indicated strains were used to infect HeLa cells cultured in DMEM supplemented with 10 % (v/v) FBS at an m.o.i. of 100 for 4 h followed by centrifugation to separate cell pellet and supernatant fractions. Cell-associated and secreted proteins obtained from whole-cell lysate and trichloroacetic acid-precipitated cell-free supernatant, respectively, were separated using 4–20 % SDS-PAGE, transferred to PVDF and probed using anti-Vp1656 and anti-DnaK antibodies. DnaK was used as a loading control. This experiment was replicated three times and a representative blot is shown here. (b) RT-PCR was used to detect mRNA transcripts for a subset of T3SS1 genes after infection. RT+ reactions included reverse transcriptase; RT− reactions contained no reverse transcriptase. The latter indicated no evidence for contaminating DNA in the reaction.

Fig. 2.

Deletion of exsE results in loss of HeLa cell cytotoxicity. The indicated strains were used to infect HeLa cells cultured in DMEM supplemented with 1 % FBS (v/v) at an m.o.i. of 100 for 4 h and assayed for T3SS1-dependent cytotoxicity. Percentage cytotoxicity was calculated based on the release of LDH relative to the uninfected control (0 %) and wild-type NY-4 strain (100 %). The reported values represent the mean±sem for three independent replicates. Asterisks, statistically significant difference in mean value compared with wild-type NY-4 by one-way ANOVA (P<0.05).

Fig. 3.

Deletion of exsE results in loss of LC3 conversion. The indicated strains were used to infect HeLa cell monolayers at an m.o.i. of 100. Whole-cell lysates were separated using 4–20 % SDS-PAGE, transferred to a membrane and probed using anti-LC3 and anti-actin antibody. Densitometry was used to determine relative LC3-II accumulation and to calculate the LC3-II/I ratio. The reported values represent the mean±sem for three independent experiments. Asterisks, statistically significant difference in mean value compared with wild-type NY-4 by one-way ANOVA (P<0.05).

Fig. 4.

Deletion of exsE results in loss of adhesion. The indicated strains were used to infect HeLa cell monolayers for 30 min. Slides were washed to remove non-adherent bacteria and stained using Diff-Quik. (a) Representative fields for each strain are shown here (63×). Bar, 20 µm. (b) Attached bacteria were enumerated by direct count and the reported values represent the mean±sem for three biological replicates. Asterisks, statistically significant difference in mean value compared with wild-type NY-4 by one-way ANOVA (P<0.05).

Fig. 5.

Deletion of exsE results in a non-swarming phenotype. Swarm agar was inoculated with the indicated strains and incubated at 37 °C for 8 h. In cis complementation of exsE recovered the swarming phenotype. The assay was repeated three times with similar results and a representative photograph is shown here.

Fig. 6.

Deletion of exsE does not affect in vivo lethality. Indicated strains were used to infect C57-B6 mice as described by Piñeyro et al. (2010). Mice were observed over the course of 24 h and mortality was recorded. Kaplan–Meier survival plots for all strains are shown here; asterisks represent values with a statistically significant difference in mortality compared with wild-type NY-4 by log-rank analysis.

Fig. 7.

Deletion of exsE does not affect relative transcription levels of exsA or lafA. HeLa cell monolayers were infected with either NY-4 or ΔexsE at an m.o.i. of 100. Total RNA was extracted and cDNA was synthesized for qPCR analysis as described in the main text. Reported values represent the mean±sem for relative expression level of four biological replicates at each time point. Asterisks, statistically significant difference in mean value compared with t = 0 h by one-way ANOVA.