Vibrio alginolyticus strains, including wild-type strain ZJO and associated deletion mutants (Table 1), were routinely grown in trypticase soy broth (TSB) or on 1.5% TSB agar plates (TSA) at 30°C. E. coli S17λpir was used in gene deletion experiments and was cultured in LB medium. Expression vector pMMB207 (Morales et al., 1991) was used in complementation experiments and suicide plasmid pDM4 (Milton et al., 1996) was used to generate gene knockouts. Vector pACYCDuet-1 (Novagen) was used to examine protein-protein interactions. When appropriate, ampicillin (100 μg mL⁻¹) or chloramphenicol (34 μg mL⁻¹) was added.

All deletions were made by allelic exchange following a method described previously (Milton et al., 1996). Briefly, primers ExsA_A1_F, ExsA_A1_R, ExsA_A2_F, and ExsA_A2_R (Table 2) were used to amplify two fragments flanking the coding sequence of exsA. Both amplified fragments incorporated 10-bp overlapping sequences in addition to BglII and SalI restriction site sequences, respectively. These products were used as templates for “splicing by overlap extension” (SOE) PCR to produce a hybrid product using primers ExsA_A1_F and ExsA_A2_R. The full-length fragment was digested with BglII and SalI and then ligated into the suicide vector pDM4 (digested with the same enzymes), generating pDM4_exsA_A1F+A2R. The resultant plasmid was electroporated into E. coli S17-1λpir and transferred into V. alginolyticus strain ZJO by conjugation. Successful transconjugants were selected on TSA plates with ampicillin and chloramphenicol. Secondary cross-over was detected by subsequent plating on agar containing 10% sucrose. For this latter selection, strains that successfully excised the plasmid sequence through secondary cross-over no longer propagate the plasmid-encoded sacB (Gay et al., 1985) and thus can grow in the presence of sucrose. The exsA deletion strain was confirmed by PCR with primers ExsA_int_F and ExsA_int_R and designated as ΔexsA. Construction of exsC, exsD and exsE deletion mutants were performed in the same manner using corresponding primers (Table 2).

To complement the exsA gene, primers ExsA-comp-F and ExsA-comp-His-R were used to amplify the complete exsA sequence with a 6X histidine sequence added at the C-terminus. The resulting amplicon was digested with EcoRI and BamHI and ligated into the expression vector pMMB207, which had been digested with the same enzymes to generate the plasmid pMMB207_exsA_His. This plasmid was then electroporated into E. coli S17-1λpir, resulting in the strain S17_pMMB207_exsA_His, and conjugated into the ΔexsA mutant strain and wild-type strain ZJO resulting in ΔexsA:pexsA and ZJO:pexsA, respectively. Construction of expression vectors for exsC, exsD and exsE was performed in the same manner with corresponding primers (Table 2).

Fathead minnow (FHM) epithelial cells were cultured in M199 medium (HyClone) supplemented with 10% (v/v) fetal bovine serum (HyClone) at 28°C. For LDH assays growth media (10% FBS) was replaced by fresh M199 supplemented with 1% FBS prior to infection to reduce background LDH activity. For complementation strains of V. alginolyticus, overnight culture was diluted 1:100 into fresh TSB with chloramphenicol and incubated at 30°C with shaking until an OD₆₀₀ of ~0.6 was obtained. Isopropyl β-D-1-thiogalactopyranoside (IPTG; 1 mM) was then added to induce protein expression. After 3 h the induced bacteria were pelleted by centrifugation (3220 × g; 30 min) and resuspended in an equivalent volume of M199 with 1% FBS. Wild-type and deletion mutants were also treated in the same manner, but without antibiotics. Bacterial suspensions were added to cell monolayers in a 12-well plate at a multiplicity of infection (m.o.i.) of ~100. Plates were centrifuged at ~600 × g for 2 min to synchronize contact with host cells. Supernatants were collected at 1.5 h post infection and LDH activity was measured with a CytoTox 96 Non-Radioactive Cytotoxicity Assay (Promega). Maximum LDH release was achieved by lysis of cells using 10X Lysis Solution and spontaneous LDH release was measured from uninfected cells. Percent cytotoxicity was calculated as follows: %Cytotoxicity=Test LDH release – Spontaneous releaseMaximum release – Spontaneous release×100

For bacteria grown in TSB, overnight culture was directly inoculated (1:100) into fresh TSB media with appropriate antibiotics and incubated for 3 h. An overnight bacterial culture was also used to infect FHM cell monolayers with an m.o.i. of 100 for 3 h. A total of 3 ml of each co-culture was collected and total RNA was isolated using a RiboPure-Bacteria kit (Ambion) followed by a secondary treatment of DNase using TURBO DNA-free kit (Ambion). Isolated RNA was reverse-transcribed into cDNA using an iScript Reverse Transcription Supermix (Bio-Rad). qPCR was performed with primer pairs (Table 2) to amplify internal fragments using SsoAdvanced SYBR Green Supermix (Bio-Rad) according to manufacturer's instructions. Cycling parameters were identical for all primer sets: 95°C for 30 s, 39 cycles of 95°C for 5 s, 55°C for 15 s, and 72°C for 30 s. Reactions were performed using the CFX 96 real-time PCR system (Bio-Rad) and relative expression was calculated using theΔΔCt method with dnaK serving as a house keeping gene (Livak and Schmittgen, 2001; Erwin et al., 2012; Nydam et al., 2014).

Adhesion assays were performed as described previously (Erwin et al., 2012). Briefly, FHM cell monolayers were grown on glass coverslips (Nunc; washed and sterilized) in six-well plates in the presence of M199 supplemented with 10% (v/v) FBS. Monolayers were then inoculated with indicated strains at an m.o.i. ~100. Importantly, V. alginolyticus strain ZJO does not adhere to glass coverslips unless host cells are present. We quantified the adherent bacteria by calculating the ratio of bacteria attached to coverslips to total bacteria as previously reported (Letourneau et al., 2011). Results were presented as average percentages from three independent replicates.

We prepared semi-solid swarming motility agar (Niu et al., 2005) using LBS [2.5% (w/v) NaCl in LB] broth with the addition of 0.5% (w/v) agar, 0.04% (v/v) sterile Tween 80 and supplemented with ampicillin. A single isolate was selected for each strain and inoculated onto a freshly prepared swarm agar plate. Plates were incubated at 30°C for 8 h and images were obtained using ChemiDoc™MP System (Bio-Rad).

To examine the potential interactions between ExsA-ExsD, ExsA-ExsE, ExsC-ExsD, and ExsC-ExsE, recombinant proteins (ExsA-His, ExsC-His, ExsD-HA, and ExsE-HA) were induced and expressed in BL21 (DE3) with vector pACYCDuet-1. Overnight bacterial culture was diluted (1:100) into fresh LB media and IPTG (1 mM) was added when the OD₆₀₀ reached ~0.6. After a 5 h incubation at 37°C, samples (~5 mL) were collected and centrifuged for 20 min at ~12,000 × g. Supernatant was decanted and the cell pellets were re-suspended in BugBuster Master Mix (1 g pellet in 5 mL reagent; Novagen) to prepare crude protein extracts. Whole-cell lysate was collected and loaded onto Ni²⁺ resin columns (Invitrogen) to allow protein binding overnight at 4°C. Columns were then washed six times with washing buffer (30 mM immidazole) and proteins were eluted with 500 μL elution buffer (300 mM immidazole). Presence of proteins was determined by western blot analysis with monoclonal anti-His antibody (1:2000; Invitrogen) and polyclonal anti-HA antibody (1:2000; Invitrogen). To control for the possibility that HA-tagged protein bound non-specifically to the resin column, ExsD-HA, and ExsE-HA proteins were expressed alone and passed through the Ni²⁺ column to serve as controls for further western blot analysis.

Fresh FHM cell culture was prepared 1 day prior to bacterial infection in a 75 cm² cell culture flask (Nunc). Overnight bacterial culture (ΔexsE:pexsE or ΔvscC:pexsE) was diluted into fresh TSB media and IPTG (1 mM) was used to induce expression of exsE for 5 h. Induced bacteria were used to infect monolayers of FHM cells for 4 h and the bacterial cell pellet was analyzed for the presence of ExsE. Cell culture media (M199) was collected and filtered (0.8 μm, Millipore). TCA (100%) was added to the filtrates (1:4) and incubated for 1 h at 4°C. Supernatant was centrifuged (22,000 × g, 30 min) and the protein pellet was washed with cold acetone to remove residual TCA. Air-dried pellets were re-suspended with PBS and samples were analyzed by western blot.

An equal volume of 2X Laemmli sample buffer (Bio-Rad) was added into each sample and all samples were boiled at 100°C for 5 min and loaded onto a 12% SDS-PAGE. After electrophoresis proteins were transferred to a PVDF membrane (Bio-Rad) and the membrane was blocked with 5% skimmed milk in PBS containing 0.05% (v/v) Tween 20 (PBS-T). After 1 h blocking, the membrane was probed with monoclonal anti-His or anti-HA antibody (1:2000; Invitrogen) for 1 h at room temperature. Secondary antibody (anti-mouse-DyLight 488; Thermo Scientific) was diluted 1:5000 in PBS-T with 1% milk for 1 h. Blots were washed 3X with PBS-T and images were obtained using a ChemiDoc™MP System (Bio-Rad).

An HA-sequence was inserted at the C-terminus of the chromosomally encoded exsA as described previously (Zhou et al., 2010). Briefly, the coding sequence of exsA with a HA tag was amplified from wild-type V. alginolyticus by using primers ExsA-insertion-F and ExsA-insertion-HA-R (Table 2). The resulting amplicon was digested with XbaI and SacI and ligated into suicide vector pDM4 (digested with the same enzymes), resulting in plasmid pDM4_exsA_HA_Insertion. The plasmid was electroporated into E. coli S17-1λpir, generating the strain S17_pDM4_exsA_HA_Insertion. Plasmid pDM4_exsA_HA_Insertion was then conjugated into strain ZJO and ΔexsC resulting in strains ZJO_ pDM4_exsA_HA_Insertion and ΔexsC_pDM4_exsA_HA_Insertion, respectively. Synthesis of ExsA was determined by western blot analysis using anti-HA antibody as described above.

A paired t-test was used for paired comparisons across transcriptional profiles because we were interested in the overall pattern of difference rather than specific differences of individual genes (for deletion mutants, transcriptional results for the knockout gene were excluded for these comparisons). For phenotype comparisons, a Kruskal-Wallis one-way ANOVA was used to assess treatment effects in conjunction with a Tukey-Kramer test for multiple-comparison (α = 0.05). qPCR data were log-transformed for statistical analysis. Because log₁₀(0) is undefined, 0.05 was added to zero values before transformation. Western blots were quantified using ImageJ (https://imagej.nih.gov/ij/index.html) for statistical comparison and Dnak was used to normalize band intensity data. Histograms were prepared by using SigmaPlot (ver. 12.5, Systat Software, Inc., San Jose, CA) and heat maps for qPCR data were prepared by using the gplots package in R (version 3.3.1).

To assess the applicability of the ExsACDE regulatory model in T3SS of V. alginolyticus, deletion mutants (ΔexsA, ΔexsC, ΔexsD, and ΔexsE) were co-cultured with FHM cells. Results from cytotoxicity assays were consistent with ExsA and ExsC functioning as positive regulators, while ExsD functions as a negative regulator for V. alginolyticus T3SS (Figure S1). LDH assays revealed that deleting exsE from V. alginolyticus enhanced cytotoxicity compared with the wild-type strain (Figure 1; P < 0.05), consistent with ExsE exhibiting a negative regulatory effect on V. alginolyticus T3SS-induced cell death. This conclusion is further supported when overexpression of exsE in a wild-type strain resulted in significantly reduced cytotoxicity (Figure 1; P < 0.05).

The type III secretion system (T3SS) genes in strain ZJO are transcribed when cultured with FHM cells, but are not transcribed when cultured in TSB (Figure 2, 20-fold mean increase, t = −3.12, 9 df, P = 0.012). Hereafter, we refer to FHM cells and TSB as “inducing” and “non-inducing” conditions, respectively. Deletion of exsE permitted transcription of T3SS genes under non-inducing conditions (ΔexsE, TSB) at levels comparable to the wild-type V. alginolyticus (ZJO, cell) strain when cultured under inducing conditions (Figure 2; t = −1.02, 8 df, P = 0.34). Presumably this occurs because without ExsE, ExsC is free to bind ExsD allowing any ExsA present in the cytoplasm to serve as a transcription regulator even in the absence of cell contact. Notably, under these conditions exsA expression was still nearly 15-fold less than what we observed when the wild-type strain was cultured under inducing conditions, which may indicate that protein turnover is relatively slow for ExsA. In trans expression of exsE (ZJOpexsE, cell) reduced overall gene expression by 10.8-fold (t = 2.47, 8 df, P = 0.039) compared to the wild-type strain under inducing conditions. This is to be expected if ExsE binds ExsC and allows ExsD to interfere with ExsA transcriptional regulator activity. Nevertheless, negative activity was clearly not complete because gene expression for the ZJOpexsE strain was still 8.8-fold greater than the wild-type strain under non-inducing conditions (ZJO, TSB; t = −3.1, 8 df, P = 0.015). Importantly, we also confirmed that knocking out exsA completely eliminates T3SS expression as predicted if it is the sole transcriptional regulator (Figure S2, t = −0.97, df = 8, P = 0.36). Knocking out exsC only partially reduced overall expression (Figure S2, mean 6-fold reduction, t = −5.46, 8 df, P < 0.001) while in trans expression of exsD in the presence of cells reduced expression relative to the wild-type strain with cells (Figure S2, 11.4-fold reduction, t = −2.46, 8 df, P = 0.04), but still exhibited significant overall expression relative to the wild-type strain when cultured in TSB alone (9.4-fold increase, t = 3.72, 8 df, P = 0.006). Collectively, these expression results are consistent with the ExsACDE model of transcriptional control, although positive regulation of exsA, and hence up-regulation of the entire T3SS, is probably controlled by a separate transcriptional regulation system as has been hypothesized for V. parahaemolyticus (Zhou et al., 2010).

Loss of exsC or in trans expression of exsD or exsE are associated with reduced transcription of exsA (Figure 3A; P < 0.05), but the magnitude of exsA transcription was still well in excess of what was observed for non-inducing conditions (Figure 3A; P < 0.001). This could occur if exsA is at least partially autoregulated, but this also raises the possibility that other factors contribute to the regulation of T3SS in V. alginolyticus. To determine if protein levels were similarly affected, we quantified the synthesis of ExsA in both wild-type V. alginolyticus and the ΔexsC mutant strain by adding an in cis HA tag the 3′ terminus of ExsA. Western blot analysis indicated ExsA was produced at limited levels under non-inducing conditions (Figure 3B, lanes 2 and 4), but at higher concentrations by both wild-type and ΔexsC mutant strains when in contact with host cells (Figure 3B, lanes 1 and 3; P > 0.05). Thus, we surmise that a T3SS-independent regulatory pathway exists that can upregulate exsA expression independent of the ExsACDE regulon.

Co-immunoprecipitation experiments with His-tagged ExsC and HA-tagged ExsE demonstrated that these two proteins likely interact as expected (Figure 4, lane 9). Similar experiments indicated that ExsE does not directly bind to ExsA (Figure 4, lane 12). The expected interactions between ExsA and ExsD (Figure 4, lane 3) and ExsC and ExsD (Figure 4, lane 6) were both confirmed.

Western blot analysis indicated that V. alginolyticus ExsE was secreted under inducing conditions (Figure 5, lane 4). VscC is a structural protein that is required for a functional T3SS in V. alginolyticus (Zhao et al., 2010). When vscC was knocked out and exsE was expressed in trans (ΔvscC:pexsE), ExsE was not detected in the supernatant (Figure 5, lane 6). Collectively, these results are consistent with T3SS-dependent secretion of ExsE. No proteins bands were detected when the wild-type V. alginolyticus was cultured in inducing conditions, confirming the specificity of the antibody that was used for the western blot (Figure 5, lane 1 and 2).

Vibrio alginolyticus was cultured in cell culture media (M199) on a six-well plastic plate (Nunc) with or without polycarbonate membrane inserts. Small molecules can pass through the membrane, but physical contact is blocked when host cells and V. alginolyticus are grown on either side of the membrane. Transcription of T3SS genes was upregulated >15-fold when host and bacterial cells were cultured in this manner (Figure 6A) although the effect was not statistically significant relative to M199 alone (t = −1.95, 9 df, P = 0.08). The lack of statistical significance is probably related to partial upregulation from the M199 media (relative to TSB; t = 4.39, 9 df, P = 0.002) and dilution of the presumptive soluble factor that signals for upregulation of exsA. Qualitatively, it appeared that the abundance of ExsA was similar when ZJO was grow in direct contact with cells or when separated by the membrane (Figure 6B; P > 0.05). ExsA was detected when ZJO was cultured in M199 compared to TSB (Figure 6B; P < 0.05), which is consistent with the upregulation of T3SS genes in the media alone (Figure 6A).

Vibrio parahaemolyticus ExsE is required for adhesion to HeLa cells and a ΔexsE mutant strain exhibited a significant reduction in cyto-adherence compared with the wild-type strain (Erwin et al., 2012). We tested if V. alginolyticus ExsE contributes to bacterial adhesion to FHM cells following a standard adherence assay (Letourneau et al., 2011). After a 30 min incubation, 24.6% ± 0.03 of ΔexsE mutant cells adhered to FHM cells, which was comparable to 23.4% ± 0.06 of wild-type V. alginolyticus, indicating ExsE is not required for adhesion to FHM cells.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.