Vibrio vulnificus RtxA Is a Major Factor Driving Inflammatory T Helper Type 17 Cell Responses in vitro and in vivo

T helper type 17 (Th17) cells are a subset of pro-inflammatory T helper cells that mediate host defense and pathological inflammation. We have previously reported that host dendritic cells (DCs) infected with Vibrio vulnificus induce Th17 responses through the production of several pro-inflammatory cytokines, including interleukin (IL)-1β and IL-6. V. vulnificus produces RTX toxin (RtxA), an important virulence factor that determines successful pathophysiology. In this study, we investigated the involvement of RtxA from V. vulnificus in Th17 cell induction through the activation and maturation of DCs. The increased expression of the DC surface marker CD40 caused by V. vulnificus wild-type infection was reduced by rtxA gene mutation in V. vulnificus. The mRNA and protein levels of Th17 polarization-related cytokines also decreased in V. vulnificus rtxA mutant-infected DCs. In addition, the co-culture of Th cells and DCs infected with rtxA mutant V. vulnificus resulted in reduction in DC-mediated Th17 responses. Th17 cell responses in the small intestinal lamina propria decreased in mice inoculated with V. vulnificus rtxA mutant as compared to those inoculated with the wild-type strain. These decreases in DC maturation, Th17-polarizing cytokine secretion, and Th17 responses attributed to rtxA mutation were restored following infection with the rtxA revertant strain. Furthermore, the mutation in the hlyU gene encoding the activator of rtxA1 gene reproduced the results observed with rtxA mutation. Taken together, V. vulnificus, by means of RtxA, induces inflammatory Th17 responses, which may be associated with adaptive responses of the host against V. vulnificus infection.


INTRODUCTION
Vibrio vulnificus is a halophilic estuarine bacterium that causes disease by infection of wounds or ingestion of contaminated seafood (1). V. vulnificus is the causative agent of several diseases, including necrotizing fasciitis, gastroenteritis, and primary septicemia.
Recent studies have elucidated the adaptive immune response against V. vulnificus infection (8,9). Previously we demonstrated that V. vulnificus infection induces Th17 responses via maturation and activation of dendritic cells (DCs). In addition, V. vulnificus infection following oral ingestion results in the induction of Th17 cell response in the small intestinal lamina propria (8). Furthermore, V. vulnificus infection induces Th1 and T follicular helper (Tfh) cells and VvhA is involved in these responses. (9). However, the specific virulence factor of V. vulnificus necessary for the induction of Th17 cells is unclear.
RtxA is a member of the multifunctional-autoprocessing repeats in toxin, a subgroup of RTX toxin family with tandem nonapeptide repeats near the C-terminal region (10). RtxA is exported via the modified type I secretion system (11). Several studies have evaluated the cytotoxic and cytopathic effects of RtxA, and reported that RtxA was related to the in vivo growth of bacteria (12), host cell necrosis (12), apoptosis (13), inflammasome activation (14), actin aggregation (15), phagocytosis inhibition (16), and the production of reactive oxygen species (17). The role of RTX toxin in pathogenesis has been investigated in V. vulnificus and other bacterial species. However, its effect on host adaptive immune responses against V. vulnificus infection remains unclear. Therefore, we investigated whether RtxA influences Th17 cell responses following V. vulnificus infection.
We found that the rtxA mutant of V. vulnificus induced lower levels of DC maturation and activation than wild-type (WT) V. vulnificus. Furthermore, the ability of V. vulnificus to induce Th17 cell responses became diminished following mutation of the rtxA gene, consistent with the observation of reduced expression and secretion of Th17 cell-polarizing cytokines. Involvement of RtxA in Th17 cell induction was confirmed through the recovery of the decreased Th17-related responses following infection with an rtxA revertant. Furthermore, the mutation of the hlyU gene, an anti-repressor of rtxA gene, resulted in defective Th17 cell responses. Taken together, our results suggest that V. vulnificus induces Th17 cell responses in vitro and in vivo through RtxA.

Mice
All animals used in this study were purchased from Orient Bio Inc. (Seoul, Korea). Seven-to 11-week-old female C57BL/6 mice were used for experiments. All mice were capable of accessing a standard laboratory chow diet (cat. no. 1314; Altromin Spezialfutter GmbH & Co. KG, Lage, Nordrhein-Westfalen, Germany) and water. The animals were housed in an SPF facility under a strict light cycle (lights on at 07:00 a.m. and off at 07:00 p.m.) at 22 ± 1 • C and 52.5 ± 2.5% relative humidity, and all animal experiments were ethically performed in accordance with the guidelines of the Korea University Institutional Animal Care and Use Committee (Seoul, Korea; approval no. KUIACUC-2016-170, 2017-113).

Bacterial Strains, Plasmids, and Culture Conditions
All V. vulnificus strains and plasmids used in this study are listed in Table 1. Unless otherwise noted, V. vulnificus strains were cultured in Luria-Bertani (LB) medium supplemented with 2.0% (wt/vol) sodium chloride (NaCl; LBS) at 30 • C. For the infection experiments, the bacteria were cultured overnight, followed by their incubation in fresh LBS medium at 30 • C. The culture was diluted to ∼1.8 × 10 8 CFU/mL in LBS and centrifuged at 2,420 × g for 3 min at room temperature. The cells were resuspended in antibiotic-free growth medium before infection of DCs or in phosphate-buffered saline (PBS) before oral administration into mice.

Construction of the rtxA Mutant and Revertant Strain
To inactivate rtxA gene in vitro, the region from 15-to 1,068bp downstream of the translational initiation codon of rtxA was deleted using the polymerase chain reaction (PCR)-mediated linker-scanning mutation method, as previously described (7,25). In brief, Briefly, pairs of primers RTXA01-F and -R (for amplification of the 5 ′ amplicon) or RTXA02-F and -R (for amplification of the 3 ′ amplicon) were designed and used ( Table 2). The resulting rtxA mutant was amplified by PCR using the mixture of both amplicons as the template and RTXA01-F and RTXA02-R as primers. The 1.2-kb DNA fragment carrying nptI encoding for aminoglycoside 3 ′ -phosphotransferae and conferring resistance to kanamycin (26) was inserted into a unique BamHI site present within the rtxA in pKK1621. The rtxA::nptI cartridge from the resulting construction (pKK1625) was liberated and ligated with SpeI-SphI-digested pDM4 (24) to generate pKK1628 ( Table 1). Escherichia coli S17-1 λpir, tra strain (22) containing pKK1628 was used as a conjugal donor to V. vulnificus MO6-24/O to generate the rtxA mutant KK1604 ( Table 1). The conjugation and isolation of the transconjugants were conducted using a previously described method (27).
To complement the rtxA mutation, rtxA revertant was constructed as previously described (16). In brief, the region containing the intact sequences of rtxA was amplified by PCR using RTXA01-F and RTXA02-R as primers. The amplified intact rtxA was ligated into SpeI-SphI-digested pDM4 to generate pKK1629 ( Table 1) and pKK1629 was transferred into the rtxA mutant KK1604 to generate the rtxA revertant KK1606 by conjugation.

Preparation of Murine Bone-Marrow-Derived Dendritic Cells (BMDCs)
We generated BMDCs using the method originally described by Inaba et al. (28) with some modifications (8). In brief, the isolated bone-marrow cells were flushed out from the femurs and tibiae of mice, and red blood cells were depleted with RBC lysis buffer containing 0.15 M NH 4 Cl, 1 mM KHCO 3 ,

In vitro Infection Protocol
Dendritic cells grown for 7 days were seeded onto a 24-well plate (1.5-2 × 10 6 cells/mL) in an antibiotic-free growth medium. Before infection, the bacteria were centrifuged at 2,420 × g for 3 min at room temperature, resuspended, and adjusted to 1.5-2 × 10 8 CFU/mL in antibiotic-free RPMI 1640 media. The DCs were infected with V. vulnificus WT and mutants at various multiplicities of infection (MOI; the ratio of bacteria number to BMDC number) for various durations. In particular, DCs were infected at an MOI of 1 for 30 or 60 min to evaluate the expression of surface markers and the secretion of cytokines, respectively or for 60 min at an MOI of 10 for co-culture with naïve CD4 + T cells. After infection, the cells were washed twice with PBS and incubated for 20 h in an antibiotic-containing growth medium at 37 • C under 5% CO 2 .

In vivo Infection Protocol and Lamina Propria Cell Isolation
To ablate normal flora, mice were given drinking water containing rifampicin (50 µg/mL) for 24 h. Before infection with bacteria, food and water were eliminated from the cages of mice to empty their stomachs for 12 h. In the subsequent step, 100 µL of bacterial suspension containing 1 × 10 7 CFU of V. vulnificus was orally inoculated immediately following the administration of 50 µLof 8.5% (w/v) sodium bicarbonate (NaHCO 3) . Mice were sacrificed at day 2 post-infection for the isolation of cells from the lamina propria of small intestines, as previously described (29). In brief, the small intestines from the mice were washed in cold PBS to clear feces. Fat tissues and Peyer's patches were removed and the intestines were longitudinally cut, followed by washing with cold PBS. The intestines were then cut into pieces (2-3 cm) and incubated in RPMI medium containing 1 mM ethylenediaminetetraacetic acid (EDTA) with gentle stirring at 37 • C for 15 min, followed by washing with warm PBS. The incubation in EDTA-containing medium was performed twice. The tissues were subsequently finely cut and incubated in RPMI containing 0.1 mg/mL of collagenase D (Roche Diagnostics, Basel, Switzerland) at 37 • C for 30 min with gentle stirring. The incubated supernatants were collected using a 70-µm cell strainer, and the unfractionated cells were centrifuged at 400 × g for 3 min at 4 • C. The lamina propria lymphocytes were isolated with 40 and 85% Percoll gradient media (GE Healthcare Life Sciences, Little Chalfont, UK) by gradient centrifugation.

Flow Cytometric Analysis
Antibodies used for both surface and intracellular staining were diluted at 1:250.

Cytokine Assays
Dendritic cells were infected with V. vulnificus at the indicated MOIs and times, washed and seeded into a 96-well plate (2 × 10 4 cells/well), followed by further incubation for 20 h. The cell supernatant was obtained to measure the levels of secreted cytokines. The quantities of IL-1β, IL-6, and IL-23 in the culture supernatants were determined using Mouse ELISA Ready-Set-Go! kits (IL-1β; eBioscience, Inc., San Diego, CA, USA) and a Mouse IL-6 ELISA kit (BD Biosciences, San Diego, CA, USA) or sandwich ELISA with anti-mouse IL-23p19 monoclonal antibody (clone 5B2) for plate coating and biotinylated anti-mouse IL-12/23 p40 monoclonal antibody (clone C17.8). A standard curve was generated using recombinant IL-23 (eBioscience, Inc., San Diego, CA, USA). The levels of secreted IL-17A in the culture supernatants were determined using a Mouse ELISA Ready-Set-Go! kit (IL-17A; eBioscience, Inc).

Statistical Analysis
Statistical analysis was performed with unpaired Student's t-test for pairwise comparisons or one-way analysis of variance (ANOVA) with a Bonferroni t-test for multiple comparisons in Sigmaplot version 12.5 (Systat Software Inc., Washington, USA). P < 0.05 was considered statistically significant.

Infection With rtxA Mutant Strain of V. vulnificus Affects DC Maturation and Activation
We previously demonstrated that infection with WT V. vulnificus results in the induction of DC maturation and activation, leading to Th17 cell stimulation (8).
Although several studies have identified virulence factors of V. vulnificus, the specific virulence factors involved in Th17 responses are unknown. We evaluated the abilities of several virulence factors from V. vulnificus to induce Th17 cell responses by infecting DCs at a multiplicity of infection (MOI) of 1 for 30 min with the WT strain and mutant strains with defects in ahpCl, rtxA, and vvhA genes encoding peroxiredoxin, RtxA, and hemolysin, respectively. The expression of maturation/activation-related cell surface markers, including CD40, CD80, and major histocompatibility complex (MHC) II, was analyzed with flow cytometry. As shown in Figures 1A,B, expression levels of these markers were reduced in rtxA mutant-infected DCs compared to those in DCs infected with the WT strain. In particular, CD40 expression level was significantly reduced in rtxA mutant-infected DCs. On the contrary, no differences in CD40 expression level were observed between DCs infected with ahpCl and vvhA mutant strains and WT-infected DCs. These data suggest that RtxA may act as a virulence factor of V. vulnificus that induces maturation and activation of DCs.

DCs Infected With rtxA Mutant Strain Show Defective Secretion of Th17 Polarization-Related Cytokines
In our previous study, the expression and secretion of IL-1β and IL-6, the major Th17-polarizing cytokines, increased in DCs upon infection with the WT strain (8). To investigate the role of some virulence factors in the expression of cytokines related to Th17 cell induction, we infected DCs with WT and mutants of V. vulnificus at an MOI of 1 for 30 or 60 min and evaluated their expressions at the mRNA (Figure 2A) and protein ( Figure 2B)   Figure 2A, the mRNA expressions of IL-6 and IL-23p19 were reduced in the rtxA mutant-infected group but not in cells infected with ahpCl and vvhA mutant strains, except for IL-1β expression in the vvhA mutant group. The secretion levels of IL-1β and IL-6 were significantly decreased in rtxA mutant-infected DCs compared to those in cells infected with other mutant strains ( Figure 2B). In contrast, the expression level of transforming growth factor beta (TGF-β), a Th17 cellrelated cytokine, was similar between mutant-infected and WTinfected DCs (Figure 2A). These results indicate that RtxA may be involved in Th17 cell induction caused by WT V. vulnificus.

Induction of Th17 Responses Is Reduced in vitro and in vivo After RtxA-Deficient V. vulnificus Infection
Dendritic cells are antigen-presenting cells involved in the polarization of naïve CD4 + T cells into each subset such as Th1, Th2, and Th17, depending on the cytokines they produce (30). To determine whether DCs infected with the rtxA mutant strain have a weaker ability to induce Th17 cell responses, we infected DCs with the WT or rtxA mutant strain at an MOI of 10 for 60 min, and subsequently co-cultured these cells with naïve CD4 + T cells at a ratio of 1:5 (DC: CD4 + T). We evaluated the population of IL-17-expressing CD4 + T (Th17) by flow cytometry (Figure 3A) and the levels of secreted IL-17A in the supernatant by ELISA ( Figure 3B). As shown in Figures 3A,B

V. vulnificus rtxA Revertant Infection Recovers the Reduced Induction of Th17 Responses After rtxA Mutant Infection
To confirm the involvement of RtxA in the induction of Th17 responses in vitro and in vivo, we used the revertant strain of rtxA mutant. DCs were infected with V. vulnificus WT, rtxA mutant, and rtxA revertant at an MOI of 1 for 30 or 60 min or at an MOI of 10 for 60 min (Figures 4A-D). As shown in Figure 4A, the reduced expressions of maturation/activationrelated surface markers in rtxA mutant-infected DCs were increased in rtxA revertant-infected DCs. Likewise, the reduced mRNA ( Figure 4B) and protein ( Figure 4C) expressions of IL-1β and IL-6 in rtxA mutant-infected DCs were restored in rtxA revertant-infected DCs to the levels observed in WTinfected DCs. Both in vitro ( Figure 4D) and in vivo (Figure 4E), To test if the role of rtxA is regulated by HlyU, we infected DCs with V. vulnificus WT or mutant strains lacking the genes associated with rtxA gene expression, hlyU and smcR at an MOI of 1 for 30 or 60 min or at an MOI of 10 for 60 min (Figures 5A-D). As shown in Figure 5, infection with the hlyU mutant also resulted in decreased secretion of IL-1β and IL-6, the two Th17-polarizing cytokines (Figure 5C), although the decreases in the expression of DC surface markers ( Figure 5A) and cytokine mRNAs ( Figure 5B) were not significant. In addition, the differentiation of naïve CD4 + T cells into IL-17-secreting cells was reduced following hlyU mutant infection ( Figure 5D).
No significant difference was observed between smcR mutantinfected DCs and WT-infected DCs. Taken together, these results imply that RtxA confers V. vulnificus with an ability to promote Th17 responses in vitro and in vivo.

DISCUSSION
V. vulnificus inhabits marine environments and can invade the host through the ingestion of contaminated seafood or via an open wound. Upon entry, it can cause a wide range of diseases, which include gastroenteritis and primary sepsis (1), and activates the innate and adaptive immune responses via diverse virulence factors (8,9). Many studies have identified several virulence factors involved in the innate immune response activation (14,(34)(35)(36)(37). However, few studies have focused on the identification of the virulence factors that induce the adaptive immune response. Therefore, in the present study, we investigated the involvement of RtxA of V. vulnificus in the induction of Th17 cell responses and demonstrated that the increase in the population of Th17 cells was related to RtxA from V. vulnificus. Adenylate cyclase toxin (ACT) of Bordetella pertussis, a member of RTX toxin family belonging to a different subgroup, induces suppressive and modulatory effects on the immune response through the inhibition of the production of pro-inflammatory cytokines or induction of Th17 cells (38,39). Our previous findings demonstrated that V. vulnificus induced Th17 cell responses by up-regulating the secretion of IL-1β and IL-6 from DCs, and that the induction of IL-6 was sufficient and necessary for the increased Th17 cell responses (8). Presently, RtxA was demonstrated to be responsible for the secretion of the pro-inflammatory cytokines, IL-1β and IL-6, and consequent Th17 responses using rtxA mutant and rtxA revertant strains of V. vulnificus. V. vulnificus RtxA reportedly protects bacteria from phagocytosis (16). Therefore, it is conceivable that the rtxA mutant is phagocytosed and destroyed by DCs, while WT V. vulnificus carrying the intact rtxA gene may evade immune responses and induce DC activation and the production of pro-inflammatory cytokines, such as IL-1β and IL-6. Thus, V. vulnificus RtxA may be involved in the induction of Th17 responses. How RtxA of V. vulnificus promotes IL-1β secretion may involve the activation of the NLRP3 inflammasome by RtxA (14). Additionally, in the previous study, B. pertussis ACT enhanced the production of IL-6 in DCs in the presence of low concentrations of LPS (40). The RtxA of V. vulnificus likely acts synergistically with other virulence factors on IL-6 production.
V. vulnificus HlyU is a homolog for the hemolysin gene regulator of V. cholerae and regulates the expression of the vvhA, vvpE, and rtxA genes (31)(32)(33). To investigate the molecular mechanisms by which rtxA is regulated in V. vulnificus, DCs were infected with hlyU mutant. The lack of HlyU failed to increase the production of IL-1β and IL-6 from DCs. Eventually, Th17 cell responses were also down-regulated. These results clearly demonstrate that the Th17-inducing capacity of RtxA is regulated by HlyU. However, the decrease in the expression (D) DCs were infected with WT and hlyU, rtxA, and smcR mutant strain at an MOI of 10 for 60 min. After 20 h, DCs were co-cultured with naïve CD4 + T cells isolated from lymph nodes for 3 days in the presence of anti-CD3ε and anti-CD28 mAbs, followed by the flow cytometric analysis of CD4 and IL-17 expressions. The data shown in (A,D) are representative of three independent experiments, and bar graphs represent the means ± SD of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.005 as determined by Student's t-test for pairwise comparisons. WT, wild-type; hlyU, hlyU mutant; rtxA, rtxA mutant; smcR, smcR mutant.
of DC surface markers and the transcriptional expression of IL-1β and IL-6 were not statistically significant. These observed discrepancies are possibly due to the other pathways that regulate the expression of rtxA and the function of hlyU gene, which is related to other virulence factors like vvhA, vvpE, and smcR. The deletion of SmcR, a negative regulator of HlyU, could not increase the secretion of IL-1β and IL-6 or Th17 responses. The absence of effect by smcR mutation may be due to the saturated function of HlyU in WT V. vulnificus or to other rtxA gene-activating pathways.
The expression of some virulence factors, including RtxA, increases upon the contact of V. vulnificus with host cells (41,42). This observation may be associated with the results shown in Figure 1, wherein no significant decrease in the expression of CD80 and MHC class II surface molecules was observed in rtxA mutant-infected DCs, even though the secretion of Th17-polarizing cytokines and Th17 responses were significantly reduced in the rtxA mutant-infected group (Figures 2, 3). Thus, further studies are needed to identify additional virulence factors other than RtxA that are likely to be involved in the upregulation of DC surface molecules and/or Th17 responses. Similar to a previous study that showed the effects of secreted outer membrane vesicles of V. cholerae on DC activation, the release of Th17 polarizationrelated cytokines, and the induction of inflammatory Th17 cells (43), other structural components and/or secreted factors of V. vulnificus may also be involved in Th17 responses induction.
Previous studies reported that the exposure of the immune cells to sub-lytic concentrations of Panton-Valentine leukocidin, a pore-forming toxin of Staphylococcus aureus, resulted in immune cell activation and the release of pro-inflammatory cytokines, rather than cell lysis (44,45). Furthermore, the amount of secreted RtxA upon contact with host cells increased as the degree of contact increased (46). Therefore, V. vulnificus RtxA may exert different effects on the host based on the nature of contact between V. vulnificus and the host cell. The close contact between the host cells and a large number of V. vulnificus results in the production of high (over lytic) concentrations of the RTX toxin, leading to host cell lysis and spread of infection into the bloodstream of the infected host. However, weak contact between the host cells and fewer V. vulnificus cells results in the production of low (sublytic) concentrations of V. vulnificus RTX toxin, leading to the induction of Th17 cellmediated responses that are likely to be involved in controlling V. vulnificus infection. The RTX toxin of V. vulnificus forms pores on the host cell membrane and causes cell lysis (13,15), and the expression of rtxA gene is induced after the contact with the host both in vitro (15) and in vivo (17). Th17 cells play an important role in maintaining mucosal barriers and contribute to pathogen clearance at mucosal surfaces (47), although their roles in V. vulnificus infection are unknown.
In summary, our study demonstrates that RtxA, one of the key virulence factors of V. vulnificus, induces the secretion of IL-1β and IL-6 from V. vulnificus-infected DCs and contributes to the induction of Th17 cells in vitro. In addition, RtxA increases the population of Th17 cells in the small intestinal lamina propria. Overall, these data establish the importance of RtxA in adaptive immune responses against V. vulnificus.

AUTHOR CONTRIBUTIONS
AL participated in the design of the study, performed all of the experiments, the data collection, and the analysis, and drafted the manuscript. MK conceived and designed the experiments. DC contributed reagents, and materials. KJ and SC constructed and provided all V. vulnificus strains. TK conceived the study and participated in its design and the coordination and also performed the data analysis and writing of the manuscript, and has full access to all the data in this study with the financial support.