Monoclonal Antibodies against Vibrio vulnificus RtxA1 Elicit Protective Immunity through Distinct Mechanisms

Abstract

Vibrio vulnificus causes rapidly progressing septicemia with an extremely high mortality rate (≥50%), even with aggressive antibiotic treatment. The bacteria secrete multifunctional autoprocessing repeats-in-toxin (MARTX) toxins, which are involved in the pathogenesis of Gram-negative Vibrio species. Recently, we reported that immunization with the C-terminal region of V. vulnificus RtxA1/MARTX_Vv, RtxA1-C, elicits a protective immune response against V. vulnificus through a poorly defined mechanism. In this study, we generated a panel of new monoclonal antibodies (MAbs) against V. vulnificus RtxA1-C and investigated their protective efficacies and mechanisms in a mouse model of infection. Prophylactic administration of seven MAbs strongly protected mice against lethal V. vulnificus infection (more than 90% survival). Moreover, three of these MAbs (21RA, 24RA, and 47RA) demonstrated marked efficacy as postexposure therapy. Notably, 21RA was therapeutically effective against lethal V. vulnificus infection by a variety of routes. Using Fab fragments and a neutropenic mouse model, we showed that 21RA and 24RA mediate protection from V. vulnificus infection through an Fc-independent and/or neutrophil-independent pathway. In contrast, 47RA-mediated protection was dependent on its Fc region and was reduced to 50% in neutropenic mice compared with 21RA-mediated and 24RA-mediated protection. Bacteriological study indicated that 21RA appears to enhance the clearance of V. vulnificus from the blood. Overall, these studies suggest that humoral immunity controls V. vulnificus infection through at least two different mechanisms. Furthermore, our panel of MAbs could provide attractive candidates for the further development of immunoprophylaxis/therapeutics and other therapies against V. vulnificus that target the MARTX toxin.

INTRODUCTION

Vibrio vulnificus, a member of the normal marine microbiota, is an important human pathogen associated with the consumption of seafood and wound infection following exposure to water containing this pathogen (1–4). The bacterial infection frequently progresses to severe skin lesions and septicemia in people with hepatic disorders or immunocompromised conditions (5). Despite support care and aggressive antibiotic therapy, V. vulnificus septicemia has a greater than 50% mortality rate; this rate increases to more than 90% for patients with septic shock (5–8). During the past decade, the incidence of V. vulnificus infection has increased worldwide, probably due to the warming of coastal waters (9, 10).

V. vulnificus produces a wide range of potential virulence factors required for survival and growth, including capsular polysaccharide (VvPS), iron assimilation systems, flagella, pili, VvhA, VvpE, and the multifunctional autoprocessing repeats-in-toxin (MARTX) toxin (MARTX_Vv, or RtxA1) (8, 11). Among them, V. vulnificus RtxA1/MARTX_Vv, a large secreted protein, belongs to the repeats-in-toxin (RTX) toxin family, which has been identified in a number of Gram-negative bacterial pathogens (12, 13). V. vulnificus RtxA1/MARTX_Vv has N- and C-terminal repeat regions and multiple activity domains related to its specific toxin activity (12–15). Recent studies have shown that the bacterial RtxA1/MARTX_Vv toxin is involved in apoptosis, necrosis, actin aggregation, the production of reactive oxygen species, and pore formation in the host cell membrane (16–22). In addition, the rtxA1 mutants have been shown to be defective in translocation from the intestine to the bloodstream, more sensitive to phagocytosis, and to have reduced colonization ability in a mouse model of infection (17, 18, 21). These results suggest that the V. vulnificus RtxA1/MARTX_Vv toxin is associated with bacterial cytotoxicity and survival during V. vulnificus infection, which also indicate the potential of the multifunctional bacterial toxin as a preventive and therapeutic target for V. vulnificus infection.

A recent study suggested that a humoral immune response raised against the C-terminal region (amino acids [aa] 3491 to 4701) of V. vulnificus RtxA1/MARTX_Vv, RtxA1-C, is sufficient for survival against V. vulnificus (23). Recombinant RtxA1-C, which was shown to effectively stimulate an immune response, conferred strong protection to actively immunized mice. Furthermore, both preexposure prophylaxis and postexposure therapy with immune serum against RtxA1-C protected naive mice from V. vulnificus challenge (23). However, the prophylactic and/or therapeutic potential of monoclonal antibodies (MAbs) against RtxA1-C has not yet been investigated. Moreover, the mechanism of anti-RtxA1-C MAb-mediated antibacterial immunity has yet to be defined.

To gain insight into the potential protective effect and the mechanism(s) of anti-RtxA1-C MAbs, we generated and characterized a panel of new MAbs against RtxA1-C. Using three recombinantly expressed fragments of RtxA1-C and different mouse models of infection, we mapped a panel of new MAbs to one of three regions in RtxA1-C and also examined the contributions of these MAbs to protection against V. vulnificus infection. Several distinct MAbs against RtxA1-C provided more than 90% prophylactic protection, and three of these MAbs (21RA, 24RA, and 47RA) exhibited significant efficacy (greater than 90% survival rate) as postexposure therapy in a mouse model. In subsequent mechanistic studies using Fab fragments and a neutropenic mouse model, we found that the therapeutic efficacy of 47RA required the IgG Fc region and some neutrophil functions, whereas the therapeutic benefits of 21RA and 24RA did not. Furthermore, postinfection treatment with 21RA significantly decreased the bacterial load in the blood. Taken together, these studies support the validation of V. vulnificus RtxA1/MARTX_Vv as a target for MAb-based therapies against V. vulnificus. Furthermore, this study demonstrates that protective MAbs can restrict V. vulnificus infection through distinct mechanisms.

MATERIALS AND METHODS

Ethics statement for animal use.

All mouse experiments were performed according to the guidelines and protocols approved by the Institutional Animal Care and Use Committee at Chonbuk National University (approval no. CBU 2013-0008 and CBU 2014-00022). All experiments were designed to minimize the numbers of animals used, and every effort was made to minimize pain and distress to the animals.

MAb generation.

Previously, we described the generation of 10 MAbs (1RA, 2RA, 3RA, 4RA, 5RA, 7RA, 8RA, 9RA, 10RA, and 11RA) against RtxA1-C from the V. vulnificus M06-24/O strain (24). To develop additional novel anti-RtxA1-C MAbs for the present study, RtxA1-C was expressed recombinantly in Escherichia coli, purified, and complexed with a monophosphoryl lipid A (MPL) adjuvant (Sigma Adjuvant System; Sigma-Aldrich, MO). BALB/c mice were then intraperitoneally (i.p.) primed and boosted four times (20 μg/mouse) over 3-week intervals. Approximately 1 month after the final boost, splenocytes were harvested from mice 3 days after an additional intravenous boost with purified RtxA1-C. The harvested splenocytes were fused to P3X63Ag8.653 myeloma cells with polyethylene glycol to generate hybridomas. After selection in hypoxanthine-aminopterin-thymidine (HAT) medium, hybridomas producing MAbs against RtxA1-C were selected and cloned by limiting dilution.

MAbs were harvested or purified from either mouse ascitic fluid or hybridoma culture supernatants by ammonium sulfate precipitation and protein A chromatography according to the manufacturer's instructions (Amicogen, Jinju, South Korea).

Characterization of MAbs.

Enzyme-linked immunosorbent assays (ELISAs) were performed to determine the immunoglobulin (IgG) subclasses of the MAbs and to map their binding regions to one of three RtxA1-C fragments (24). For MAb subclass determination, microtiter plates (Maxi-Sorp; Nalge Nunc International, NY) were incubated overnight at 4°C with polyclonal rabbit antibodies against each mouse IgG isotype (Invitrogen, CA). The plates were then washed with washing buffer (phosphate-buffered saline [PBS], 0.05% Tween 20, 0.05% bovine serum albumin [BSA], 0.025% NaN₃) and incubated with blocking buffer (PBS, 0.05% Tween 20, 3% BSA, 0.025% NaN₃) for 1 h at 37°C. After being washed with washing buffer, the plates were incubated with anti-RtxA1-C MAbs, washed, and incubated with biotin-conjugated goat anti-mouse IgG1, IgG2a, IgG2b, IgG3, or IgM antibodies (Invitrogen). Next, horseradish peroxidase (HRP)-conjugated streptavidin (Sigma-Aldrich) was added, and the plates were incubated for 1 h at 4°C. After washing, HRP was detected by the addition of its substrate, 3,3′,5,5′-tetramethylbenzidine (TMB) (Sigma-Aldrich). Reactions were stopped by the addition of 1 M H₂SO₄, and the optical densities at 450 nm were measured with a microplate reader.

To map the binding regions of the MAbs, microtiter plates were coated with equal amounts of different fragments of RtxA1-C (23, 24), produced recombinantly in E. coli, and incubated in blocking buffer. After washing, the plates were incubated with the different MAbs against RtxA1-C. Bound MAbs were then detected with biotin-conjugated goat anti-mouse IgG antibodies and HRP-conjugated streptavidin, as described above.

Preparation of Fab fragments of MAbs.

Purified MAbs were cleaved by papain after determination of the optimal digestion conditions in pilot assays. After digestion, the reactions were stopped by the addition of 55 mM iodoacetamide (Sigma-Aldrich). The Fab fragments in PBS from the crude digestion were recovered as uncleaved IgG and Fc fragments were removed using a protein G affinity column and by size exclusion chromatography on a Hi-load 16/60 Superdex 75 column (Amersham Biosciences, NJ). The purified Fab fragments were concentrated and confirmed by SDS-PAGE analysis, Western blotting, and ELISA.

Bacteria.

All V. vulnificus strains were grown in heart infusion (HI) medium with 2.5% NaCl at 37°C. For antibody-mediated protection studies, the M06-24/O strain was grown overnight in a rotary shaker at 37°C, subcultured in HI with 2.5% NaCl, harvested upon reaching the mid-log phase of growth, and then washed three times with cold PBS. Age-matched mice were infected with V. vulnificus, diluted in PBS, through the intraperitoneal (i.p.), subcutaneous (s.c.), or intragastric (i.g.) route. To evaluate cross-protection efficiency of an anti-RtxA1-C MAb against different V. vulnificus strains, V. vulnificus strains ATCC 29307 and CMCP6 were cultured, washed, diluted, and infected i.p. as described above.

Three clinical strains (ATCC 29307, CMCP6, and YJ016) and one environmental strain (Evn1) were examined for cross-reactivity with the MAbs raised against RtxA1-C from the M06-24/O strain. CMM744, an rtxA1 mutant of the M06-24/O strain with an out-of-frame gene deletion from bp 286 to 4755 of RtxA1/MARTX_Vv, was used as a negative control (18, 24). V. vulnificus strains Evn1 and YJ016 were generous gifts from J. D. Oliver and L. I. Hor, respectively. The ATCC 29307 strain was purchased from the American Type Culture Collection (ATCC, VA).

Western blot analysis.

To examine the cross-reactivity of our anti-RtxA1-C MAbs against V. vulnificus RtxA1/MARTX_Vv toxins from diverse clinical and environmental strains, we prepared secreted V. vulnificus RtxA1/MARTX_Vv toxins from bacterial culture supernatants as described previously (24). Briefly, V. vulnificus was subcultured until the mid-log growth phase, and the culture supernatants were harvested. Then, the secreted V. vulnificus RtxA1/MARTX_Vv toxins were obtained by acetone precipitation of the supernatants, after adjusting each to the same optical density at 600 nm. The acetone precipitates were dissolved in nonreducing sample buffer without β-mercaptoethanol and incubated for 3 min at 37°C. Solubilized proteins were resolved by SDS-PAGE and transferred to a nitrocellulose membrane (Invitrogen). The membrane was then blocked in 5% nonfat dry milk in TBST (20 mM Tris-HCl [pH 8.0], 150 mM NaCl, 0.5% Tween 20), washed in TBST, and incubated with anti-RtxA1-C serum or each of the MAbs (10 μg/ml). After being washed with TBST, the membrane was incubated with a 1/3,000 dilution of secondary antibody conjugated with HRP (Invitrogen), washed, developed with enhanced luminol-based chemiluminescent (ECL) Western blotting reagent (Amersham Bioscience), and detected with a LAS-1000 luminescent image analyzer (Fujifilm, Tokyo, Japan).

Mouse experiments.

All wild-type female CD1 mice were obtained from a commercial source (Orient Bio, Inc., a branch of Charles River Laboratory, Seongnam, South Korea). Age-matched female mice were infected with the appropriate doses of V. vulnificus i.p., s.c., or i.g. at a given time point before or after the administration of a single dose of purified MAbs, ascitic fluid, Fab fragments, or PBS by i.p. injection. For s.c. infection, 8-week-old CD1 mice were anesthetized with an i.p. injection of xylazine and ketamine and then were inoculated with 1 × 10⁴ CFU of the V. vulnificus M06-24/O strain by footpad injection. For i.g. infection, 8-week-old CD1 mice were i.g. fed 300 μl (1 × 10⁸ CFU) of a bacterial suspension in PBS using a 20-gauge animal feeding needle after anesthetization.

Iron overload and reduction of neutrophils in mice.

For iron overload experiments in mice, 8-week-old female CD1 mice were pretreated i.p. with 450 μg of ferric ammonium citrate (approximately 80 μg of elemental iron) (Sigma-Aldrich) 50 min before the bacterial infection (25).

To reduce neutrophils in a mouse model, neutropenia was induced using cyclophosphamide (CYC) (Sigma-Aldrich), a well-known cytostatic and immunosuppressant drug, as previously described (21, 26). Briefly, 8-week old female CD1 mice were administered CYC i.p. (150 mg/kg). The neutropenic mice were infected i.p. with 1 × 10⁵ CFU of the V. vulnificus M06-24/O strain 3 days after CYC treatment.

Quantitation of blood bacterial load.

For kinetic analysis of the bacterial load in the blood, 8-week-old female CD1 mice were inoculated i.p. with 1 × 10⁶ CFU of the V. vulnificus M06-24/O strain. Then, a single dose of purified MAbs (500 μg) was administered at 1 h postinfection. Whole-blood samples were obtained by phlebotomy of the axillary vein at 1, 3.5, and 6 h after i.p. infection. The blood samples were then serially diluted and plated to count the numbers of viable bacteria.

Statistical analysis.

All data were analyzed with Prism software (GraphPad Software, Inc., CA). For survival analysis, Kaplan-Meier survival curves were analyzed by the log rank and Mantel-Haenszel tests. For the experiment involving blood bacterial load, statistical significance was determined using the nonparametric Mann-Whitney test. P values of <0.05 were considered statistically significant.

RESULTS

In prior studies, we generated 10 MAbs (1RA to 11RA) against the C-terminal region (amino acids [aa] 3491 to 4701) of V. vulnificus RtxA1/MARTX_Vv, RtxA1-C, and demonstrated that the passive transfer of anti-RtxA1-C serum protected mice against lethal V. vulnificus infection (23, 24). However, the protective potential of these previously generated MAbs against RtxA1-C was not determined. To comprehensively investigate the efficacies and mechanisms of anti-RtxA1-C MAbs against V. vulnificus infection, we also developed a panel of new MAbs against V. vulnificus RtxA1/MARTX_Vv in this study.

Generation of anti-RtxA1-C MAbs.

To generate new anti-RtxA1-C MAbs, we immunized BALB/c mice with purified RtxA1-C, fused splenocytes to a nonsecreting myeloma cell line, and produced more than 1,000 hybridomas. After screening of the hybridomas, 27 additional new MAbs (12RA to 52RA) that recognized RtxA1-C were generated (Table 1). To identify the locations at which the MAbs bound RtxA1-C, all MAb binding activities were mapped by ELISA using three recombinantly produced RtxA1-C fragments, as previously described (24): FR-I (aa 3491 to 3980), FR-I-II (aa 3491 to 4380), and FR-I-II-III (aa 3491 to 4701 [RtxA1-C]). Out of the 37 MAbs that recognized V. vulnificus RtxA1-C, 10 MAbs recognized only FR-I-II-III, 13 MAbs bound FR-I-II and FR-I-II-III but not FR-I, and 14 MAbs detected all three RtxA1-C fragments (FR-I, FR-I-II, and FR-I-II-III) (Table 1). The isotype of each MAb was also identified by heavy-chain-specific ELISA (Table 1). Most of the MAbs belonged to the IgG1 (16/37) or IgG2a (18/37) subclass, whereas three of the MAbs were determined to be IgG2b. Overall, we generated a panel of new MAbs against RtxA1-C and mapped the binding region of each MAb to one of three fragments of V. vulnificus RtxA1-C (Table 1).

TABLE 1.

Characterization of the MAbs against RtxA1-C^a

Control or antibody	Isotypeb	Binding activityc			Protective activityd
		FR-I	FR-I-II	FR-I-II-III	% of surviving mice	Antibody source
PBS control		−	−	−	30
Antibodies
14NS1	IgG2a	−	−	−	33	PUR
1RA	IgG1	−	+	+	20	ASC
2RA	IgG2a	+	+	+	60	ASC
3RA	IgG1	+	+	+	70*	ASC
4RA	IgG1	−	−	+	50	PUR
5RA	IgG2a	−	+	+	60	ASC
7RA	IgG1	−	+	+	40	ASC
8RA	IgG2b	+	+	+	30	ASC
9RA	IgG1	−	+	+	20	ASC
10RA	IgG1	−	−	+	20	PUR
11RA	IgG2a	+	+	+	80*	ASC
12RA	IgG2a	−	+	+	60	ASC
13RA	IgG2a	−	+	+	100*	ASC
14RA	IgG1	−	+	+	50	ASC
16RA	IgG2a	−	−	+	40	ASC
19RA	IgG2a	+	+	+	70*	ASC
20RA	IgG1	−	+	+	30	ASC
21RA	IgG2a	+	+	+	100*	ASC/PUR
22RA	IgG2b	+	+	+	30	ASC
23RA	IgG2a	+	+	+	20	ASC
24RA	IgG2a	+	+	+	100*	ASC/PUR
26RA	IgG1	−	+	+	30	ASC
27RA	IgG2a	−	−	+	30	ASC
28RA	IgG2a	−	+	+	30	ASC
29RA	IgG1	−	+	+	50	ASC
30RA	IgG1	−	−	+	50	ASC
32RA	IgG2b	+	+	+	30	ASC
38RA	IgG2a	−	+	+	40	ASC
40RA	IgG1	+	+	+	30	ASC
41RA	IgG1	−	−	+	50	ASC
42RA	IgG2a	−	−	+	70*	ASC
44RA	IgG1	−	+	+	60	ASC
45RA	IgG2a	−	−	+	90*	ASC
46RA	IgG2a	+	+	+	90*	ASC
47RA	IgG2a	+	+	+	90*	PUR
48RA	IgG1	−	−	+	40	ASC
50RA	IgG2a	+	+	+	100*	ASC
52RA	IgG1	−	−	+	60	ASC

^aThe isotype and binding activity data from 1RA, 2RA, 3RA, 4RA, 5RA, 7RA, 8RA, 9RA, 10RA, and 11RA have been adapted from our previous publication (24).

^bMAb isotypes were determined by heavy-chain-specific ELISAs.

^cThe binding activities of the anti-RtxA1-C MAbs were determined by ELISAs with recombinantly produced RtxA1-C fragments, as previously described (24). The fragments used were FR-I (RtxA1, amino acids [aa] 3491 to 3980), FR-I-II (RtxA1, aa 3491 to 4380), and FR-I-II-III (RtxA1, aa 3491 to 4701). Binding activities are designated as follows: +, immunoreactivity; −, no immunoreactivity.

^dFor prophylactic studies with anti-RtxA1-C MAbs, 8-week-old female CD1 mice were intraperitoneally (i.p.) administered a single dose of purified MAbs (1 mg) or ascitic fluid (200 μl) 4.5 h before infection with 1 × 10⁶ CFU of V. vulnificus strain M06-24/O i.p. The infected mice were then followed for survival over 96 h. As a control MAb, mice were treated with 14NS1, an anti-West Nile virus NS1 MAb (61). Asterisks indicate significant differences compared with the PBS-treated control group (P < 0.05). ASC, ascitic fluid; PUR, purified antibody. “ASC/PUR” indicates that the same result was obtained when the antibody was passively transferred either as ascitic fluid or in a purified form in independent experiments.

In vivo protection of anti-RtxA1-C MAbs against V. vulnificus infection.

To determine whether MAbs against RtxA1-C provide protection against lethal V. vulnificus infection in vivo, we evaluated their inhibitory activity in a mouse model of V. vulnificus infection. For this study, we used 8-week-old CD1 mice, which have a baseline survival rate of ∼30% after V. vulnificus infection. Mice were infected intraperitoneally (i.p.) with 1 × 10⁶ CFU of V. vulnificus 4.5 h after administration of a single dose of purified MAb (1 mg) or ascitic fluid (200 μl) i.p. Interestingly, a single dose of 13RA, 21RA, 24RA, or 50RA conferred 100% protection against the lethal V. vulnificus challenge (P ≤ 0.0001); in contrast, control MAb 14NS1-treated mice exhibited a survival rate of only 33% (P = 0.678) (Table 1 and Fig. 1A). The inhibitory effects of 21RA and 24RA were not dependent on antibody preparation; administration of 21RA and 24RA as either ascitic fluid or purified MAb resulted in equivalent protection (data not shown). In addition, mice that were treated with 13RA, 21RA, or 24RA did not exhibit any visible signs or symptoms at 16 h postinfection. Similarly, passive transfer of 45RA, 46RA, 47RA, or 50RA MAbs increased the survival rate to 90 to 100%; however, mice treated with these MAbs showed fur ruffling and slightly reduced activity by 36 h postinfection (Table 1 and Fig. 1A). Surprisingly, when administered i.p. as an immunoprophylactic regimen, a single dose of an anti-RtxA1-C MAb, 21RA (1 mg), provided 90% protection (P < 0.0001) to the mice that had been infected with a higher dose of V. vulnificus (8 × 10⁶ CFU). In contrast, all PBS-treated control mice died within 8 h of the infection (Fig. 1B). As expected, the protective efficacy of 21RA was dose dependent, as smaller amounts elicited less protection or more severe symptoms. Administration of a single dose of more than 4 μg (0.15 mg/kg) of 21RA resulted in significant protection (Fig. 1C) (80 to 100% survival, P ≤ 0.0389). Although administrations of 4 μg and 20 μg provided equal protection (80% survival), mice treated with 4 μg required more time for complete recovery than mice that had received 20 μg (64 h versus 32 h, respectively).

FIG 1.

Prophylactic efficacies of anti-RtxA1-C MAbs. (A and B) Survival of mice treated with anti-RtxA1-C MAbs. A single dose of purified MAbs (1 mg) or ascitic fluid (200 μl) was passively transferred to 8-week-old CD1 mice via the intraperitoneal (i.p.) route, 4.5 h prior to i.p. infection with 1 × 10⁶ (A) or 8 × 10⁶ (B) CFU of V. vulnificus. Survival curves were constructed using data from more than two independent experiments, each with five mice. As shown in Table 1, administration of 13RA, 21RA, 24RA, 45RA, 46RA, 47RA, or 50RA significantly improved the survival rate to 90 to 100% against infection with 1 × 10⁶ CFU of V. vulnificus (P ≤ 0.0009). Passive transfer of 21RA conferred 90% protection against challenge with 8 × 10⁶ CFU of V. vulnificus (P < 0.0001). (C) Dose-dependent protection conferred by the 21RA MAb. Eight-week-old CD1 mice were infected i.p. with 1 × 10⁶ CFU of V. vulnificus 4.5 h after the administration of a single dose of purified 21RA MAb (0.8, 4, 20, 100, or 500 μg). Survival curves were constructed using data from two independent experiments, each with five animals. Statistical differences compared with PBS-treated mice were as follows: 0.8 μg, P = 0.6226; 4 μg, P = 0.0389; 20 μg, P = 0.0148; 100 μg, P = 0.0055; and 500 μg, P < 0.0001. (D) Long-lasting protection by the 21RA MAb. Six-week-old CD1 mice were passively transferred a single dose of purified 21RA (500 μg) i.p. and then infected i.p. with 1 × 10⁶ CFU of V. vulnificus 14 days later. Mice pretreated with 21RA were significantly protected compared with the control PBS-treated mice (P = 0.001). Survival curves were constructed using data from two independent experiments, each with five to six animals.

To confirm that passive transfer of anti-RtxA1-C MAbs provided long-lasting protection against V. vulnificus infection, we assessed the duration of the prophylactic effect with MAb 21RA. Six-week-old CD1 mice were i.p. administered a single dose of 21RA (500 μg) or PBS and then challenged i.p. with 1 × 10⁶ CFU of V. vulnificus 14 days later. Interestingly, passive transfer of 21RA MAb 14 days before the bacterial infection conferred 90% protection (n = 10, P = 0.001), compared with ∼18% survival in the PBS-treated control group (Fig. 1D). Taken together, these results strongly suggest that MAbs against RtxA1-C are beneficial in prophylaxis against V. vulnificus infection.

Therapeutic efficiency of anti-RtxA1-C MAbs.

Although antiserum against RtxA1-C was previously shown to confer therapeutic immunity against V. vulnificus infection (23), the homogeneity and consistency of anti-RtxA1-C MAbs could provide more practical therapeutic applications compared with polyclonal antibodies. Therefore, we evaluated a number of our MAbs for their abilities to control an established V. vulnificus infection. Eight-week-old CD1 mice were inoculated i.p. with 1 × 10⁶ CFU of V. vulnificus and then administered i.p. a single dose of purified MAb (0.5 or 1 mg [18.5 or 37 mg/kg]) or ascitic fluid (200 μl) 1 h after infection. As observed in our prophylactic study (Fig. 1A), administration of PBS to mice resulted in a baseline survival rate of 30%. Notably, a single dose of 21RA resulted in a 100% survival rate (P = 0.0001) in a preestablished V. vulnificus infection (Table 2 and Fig. 2A). Similarly, a single dose of 24RA or 47RA significantly improved the survival rate to 90% (P = 0.0008). However, no significant survival advantage was observed when 13RA, 45RA, or 50RA was administered 1 h postinfection (20 to 40% survival, P ≥ 0.5675) (Table 2 and Fig. 2A), even though these MAbs provided high prophylactic efficacies (Table 1 and Fig. 1A). We also investigated the therapeutic efficacies of anti-RtxA1-C MAbs at a later and perhaps more clinically relevant time point, 2 h after i.p. infection, at which time the infected mice already manifested bacteremia, reduced activity, and ruffled fur (see Fig. 4) (data not shown). Notably, a single dose of 21RA or 24RA at 2 h postinfection provided significant protection (21RA, 80% survival, P = 0.0041; 24RA, 90% survival, P = 0.0073) (see Fig. S1 in the supplemental material).

TABLE 2.

Therapeutic activities of MAbs against RtxA1-C^a

Control or antibody	% of surviving mice (no. of survivors/total)b	P value	Antibody sourcec
PBS control	30 (16/53)
Antibodies
13RA	20 (2/10)	0.5962	ASC
21RA	100 (10/10)	0.0001	PUR (0.5 mg)
24RA	90 (9/10)	0.0008	PUR (1 mg)
45RA	30 (3/10)	0.6078	PUR (1 mg)
47RA	90 (9/10)	0.0008	PUR (1 mg)
50RA	40 (4/10)	0.5675	ASC

^aTo determine the therapeutic efficacies of the MAbs, 8-week-old CD1 mice were challenged intraperitoneally (i.p.) with 1 × 10⁶ CFU of V. vulnificus strain M06-24/O. Then, a single dose of purified MAbs (0.5 mg or 1 mg) or ascitic fluid (200 μl) was passively transferred i.p. 1 h postinfection. Mice were followed for 96 h. P values were determined by comparison to the PBS-treated control group.

^bNumbers in parentheses represent the number of survivors and the total number of infected mice.

^cValues in parentheses indicate the amount of MAb administered to each mouse. ASC, ascitic fluid; PUR, purified antibody.

FIG 2.

Therapeutic activities of anti-RtxA1-C MAbs. (A and B) Therapeutic effects after intraperitoneal (i.p.) and subcutaneous (s.c.) infections. (A) At 1 h postinfection i.p. with 1 × 10⁶ CFU of V. vulnificus, 8-week-old CD1 mice were passively transferred a single dose of purified MAbs (0.5 or 1 mg), ascitic fluid (200 μl), or saline (PBS) i.p. As shown in Table 2, 21RA, 24RA, and 47RA exhibited significant therapeutic efficacies (P ≤ 0.0008), but 13RA, 45RA, and 50RA did not. Survival curves were constructed using data from more than two independent experiments, each with five to six animals. (B) Eight-week-old CD1 mice were s.c. inoculated via the footpad with 1 × 10⁴ CFU of V. vulnificus and then i.p. administered a single dose of 21RA (500 μg) or PBS 1 h postinfection. The 21RA MAb provided significant therapeutic benefit compared with the PBS control after s.c. infection (n = 11, P = 0.0001). Survival curves were constructed using data from two independent experiments, each with five to six mice. (C and D) Therapeutic activity of the 21RA MAb in iron-overloaded mice after i.p. (C) and intragastric (i.g.) (D) infection. Eight-week-old mice were injected i.p. with ferric ammonium citrate 50 min before i.p. infection with 20 CFU (C) or i.g. inoculation with 1 × 10⁸ CFU (D) of V. vulnificus. At 1 h postinfection, mice were treated with a single dose of 21RA (1 mg) or PBS i.p. The survival curves for iron-overloaded mice treated with 21RA and PBS were significantly different after i.p. (n = 10, P = 0.0002) and i.g. (n = 10, P = 0.0212) infection. Survival curves were constructed using data from two independent experiments, each with five mice.

FIG 4.

Effect of the 21RA MAb on blood bacterial load over time. Eight-week-old CD1 mice were treated with a single dose of purified 21RA (500 μg) or PBS at 1 h postintraperitoneal (post-i.p.) infection with 1 × 10⁶ CFU of V. vulnificus. Whole blood was harvested at the indicated times postinfection, and the bacterial loads in the blood were determined by the plating method. Data are expressed as the average CFU per milliliter of whole blood from five to seven mice per time point. Asterisks indicate values significantly different from the control value (P < 0.05). The dashed line horizontal represents the limit of sensitivity of the assay.

Since V. vulnificus infection often occurs through contaminated wounds, and the susceptibility to and virulence of V. vulnificus infection have been shown to differ depending on the route of infection (4, 6, 27, 28), we studied the therapeutic efficacy of an anti-RtxA1-C MAb after subcutaneous (s.c.) inoculation of V. vulnificus. Eight-week-old CD1 mice were infected s.c. with a lethal dose of V. vulnificus (1 × 10⁴ CFU) by footpad injection and then administered a single dose of 21RA (500 μg) i.p. 1 h after infection. As expected, s.c. infected mice exhibited longer survival times than i.p. infected mice, as previously described (27, 28). The mean survival times were 32 h and 16 h after s.c. and i.p. infection, respectively (Fig. 2A and B). Interestingly, the administration of 21RA 1 h post-s.c. infection provided significant protection (91% [10/11] survival; P = 0.0001) (Fig. 2B). In contrast, 9% (1/11) of the s.c. infected mice in the PBS control group survived. However, extensive edema surrounding the site of injection was observed in all 21RA-treated mice (11/11) at 5 h postinfection. Of the 21RA-treated mice, 55% (6/11) exhibited some symptoms of sickness, including slight fur ruffling and reduced activity, by 16 h postinfection. However, all surviving mice had completely recovered by 48 h postinfection.

Given that the presence of preexisting liver disease and hemochromatosis with impaired iron metabolism increases the incidence of V. vulnificus septicemia with a high degree of mortality (29–32), we investigated the therapeutic efficiency of an anti-RtxA1-C MAb in a mouse model of iron overload. For the treatment of iron overload, 8-week-old CD1 mice were inoculated i.p. with ferric ammonium citrate 50 min before V. vulnificus inoculation. The iron-treated mice were injected i.p. with 20 CFU of V. vulnificus and then a single dose of 21RA (1 mg) was passively transferred i.p. 1 h after infection. Similar to what has been observed in other iron overload studies (25, 27, 33), the iron-overloaded mice were highly susceptible to V. vulnificus infection. The administration of a single dose of 21RA 1 h postinfection provided 100% protection (n = 10; P = 0.0002) to the iron-overloaded mice, compared with a 20% survival rate in the PBS-treated mice (Fig. 2C). Furthermore, we evaluated the therapeutic activity of an anti-RtxA1-C MAb against i.g. infection in the iron overload mouse model, since patients with hepatic diseases often develop fulminating septicemia following the ingestion of uncooked seafood (7, 34, 35). Eight-week-old CD1 mice pretreated with ferric ammonium citrate were infected i.g. with 1 × 10⁸ CFU of V. vulnificus and then a single dose of 21RA MAbs (1 mg) was passively transferred by i.p. injection 1 h postinfection. Under these conditions, a single dose of 21RA provided significant therapeutic efficacy (70% [7/10] survival, P = 0.0212) (Fig. 2D). In contrast, PBS-treated mice exhibited only a 20% survival rate. Overall, by using different infection routes of V. vulnificus and by including an iron overload mouse model of infection, our data unequivocally demonstrate that MAbs against RtxA1-C can protect mice against V. vulnificus-induced mortality in therapeutic postexposure trials.

Mechanism of protection by anti-RtxA1-C MAbs.

To determine whether the therapeutic effects of our MAbs were linked to antibody effector functions, we assessed the role of the Fc portion in antibody-mediated protection against V. vulnificus. Survival studies were performed with Fab fragments generated from therapeutic MAbs by papain digestion. Eight-week-old CD1 mice were inoculated i.p. with 1 × 10⁶ CFU of V. vulnificus. At 1 h postinfection, each Fab fragment was passively transferred at an amount equimolar to that of the binding sites of the MAbs administered in the MAb therapeutic studies (21RA Fab, 330 μg; 24RA Fab, 660 μg; 47RA Fab, 660 μg). The Fab fragments derived from 21RA and 24RA maintained the therapeutic effects of the full-length antibodies (21RA Fab, 100% [16/16] survival, P < 0.0001; 24RA Fab, 87% [13/15] survival, P = 0.0028) (Fig. 3A). In contrast, Fab fragments generated from 47RA did not provide any therapeutic efficacy (40% [6/15] survival, P = 0.929) compared with 39% (13 of 33) survival in the PBS-treated control group (Fig. 3A); in contrast, the corresponding full-length 47RA MAb (containing the Fc portion) was highly therapeutic (Table 2 and Fig. 2A). These results suggest that the Fc region might determine the therapeutic potential of 47RA through its interaction with the Fc receptor(s) on innate immune cells, whereas 21RA and 24RA are likely to inhibit bacterial infection in an Fc receptor-independent manner.

FIG 3.

Protective mechanisms of anti-RtxA1-C MAbs. (A) Therapeutic efficacies of Fab fragments in CD1 mice. Eight-week-old CD1 mice were administered a single dose of Fab fragments (21RA Fab, 330 μg; 24RA Fab and 47RA Fab, 660 μg) or PBS as a control 1 h after an intraperitoneal (i.p.) challenge with 1 × 10⁶ CFU of V. vulnificus. The survival curves for mice treated with Fab fragments and PBS were significantly different for 21RA Fab fragments (n = 16, P < 0.0001) and 24RA Fab fragments (n = 15, P = 0.0028) but not for 47RA Fab fragments (n = 15, P = 0.929). Survival curves were constructed using data from more than three independent experiments, each with five to six mice. (B) Therapeutic efficacy of MAbs in neutropenic mice. Neutropenic mice were induced by treatment of 8-week-old CD1 mice with cyclophosphamide (150 mg/kg) 3 days before infection. The neutropenic mice were infected i.p. with 1 × 10⁵ CFU of V. vulnificus, and a single dose of purified MAbs (21RA, 0.5 mg; 24RA and 47RA, 1 mg) or PBS was given i.p. 1 h postinfection. The statistical differences for the survival curves of the MAb-treated mice compared with the PBS-treated mice were as follows: 21RA, P < 0.0001; 24RA, P < 0.0001; and 47RA, P = 0.0019. In addition, the survival rates of 47RA-treated and 21RA-treated or 24RA-treated mice were significantly different (n = 10, P = 0.0121). Survival curves were constructed using data from two independent experiments, each with five to six mice.

Since a variety of immune cells, such as macrophages and neutrophils, play an important role in defense against bacterial infection (21, 36–39), we investigated whether neutrophils are associated with the therapeutic efficacies of three MAbs (21RA, 24RA, and 47RA) using a neutropenic mouse model of infection. Eight-week-old CD1 mice were pretreated i.p. with cyclophosphamide (CYC) to reduce their neutrophil populations. Three days after CYC treatment, the neutropenic mice were challenged i.p. with 1 × 10⁵ CFU of V. vulnificus and then i.p. administered a single dose of purified MAbs (0.5 or 1 mg) 1 h after infection. At baseline, the neutropenic PBS-treated control mice showed increased susceptibility to V. vulnificus infection (8.3% [1/12] survival) (Fig. 3B), consistent with a previous study (21). Interestingly, the administration of 21RA (0.5 mg), 24RA (1 mg), or 47RA (1 mg) showed significant therapeutic efficacy in neutropenic mice (21RA, 100% [10/10] survival, P < 0.0001; 24RA, 100% [10/10] survival, P < 0.0001; 47RA, 50% [5/10] survival, P = 0.0019) (Fig. 3B). However, the therapeutic activity of 47RA in neutropenic mice was significantly reduced compared with the activities of 21RA and 24RA (P = 0.0121), which suggests that neutrophils might partially contribute to 47RA-mediated protection against V. vulnificus. Overall, these results suggest that different MAbs against V. vulnificus RtxA1/MARTX_Vv mediate in vivo protection by distinct pathways.

Effect of an anti-RtxA1-C MAb on blood bacterial load.

To gain further insight into the mechanism(s) by which MAbs against RtxA1-C protect against an established V. vulnificus infection in vivo, we evaluated the effect of one therapeutic MAb, 21RA, on the bacterial load in the blood by using a mouse model of acute sepsis. Eight-week-old CD1 mice were infected i.p. with 1 × 10⁶ CFU of V. vulnificus and then administered a single dose of 21RA (500 μg) or PBS 1 h postinfection. The bacterial loads in the blood were analyzed 1, 3.5, and 6 h after the i.p. V. vulnificus infection. In the mouse model of acute sepsis, bacteremia was observed in both PBS-treated and 21RA-treated mice 1 h postinfection (Fig. 4); all (10/10) of the mice under each condition exhibited detectable levels of bacteria in the blood. In addition, the bacterial loads between the two groups were not significantly different (P = 0.421) (Fig. 4). At 3.5 h postinfection, only 50% (3/6) of the 21RA-treated mice had detectable bacteria in the blood, whereas 100% (6/6) of the PBS-treated mice had a measurable V. vulnificus titer. Mice that had received the 21RA MAb had an approximately 7.8-fold-lower level (P = 0.026) of V. vulnificus in the blood 3.5 h postinfection compared with the PBS-treated mice (Fig. 4). Moreover, only 14% (1/7) of the 21RA-treated mice still had detectable infectious bacteria in the blood 6 h postinfection, whereas bacteria were still detectable in 100% (6/6) of the PBS-treated mice. Six hours postinfection, a 100-fold-lower level of V. vulnificus was detected in the blood of the 21RA-treated mice (P = 0.0012) compared with the PBS-treated mice (Fig. 4). Taken together, these bacteriological analyses suggest that treatment of mice with 21RA increases the clearance of V. vulnificus from the blood, which might contribute to the increased survival from an established V. vulnificus infection in 21RA-treated mice.

Immunoreactivity and cross-protection of anti-RtxA1-C MAbs against various V. vulnificus strains.

Our therapeutic MAbs were raised against a recombinant RtxA1-C protein from the M06-24/O strain. However, since efficient preclinical applications of our therapeutic MAbs could be facilitated by the cross-reactivity and cross-protection of these MAbs against other V. vulnificus strains, we first evaluated the immunoreactivities of our anti-RtxA1-C MAbs against native secreted V. vulnificus RtxA1/MARTX_Vv toxins from diverse clinical strains (ATCC 29307, CMCP6, and YJ016) and an environmental strain (Env1). For these experiments, the RtxA1/MARTX_Vv toxins were obtained from bacterial culture supernatants of the different strains. To evaluate the extents of cross-reactivity of our MAbs against other V. vulnificus strains, secreted proteins were precipitated with cold acetone from the supernatants of different strains and then analyzed by Western blotting with each MAb. The three MAbs (21RA, 24RA, and 47RA) with strong therapeutic efficacy against the M06-24/O strain recognized the intact RtxA1/MARTX_Vv protein (approximately 500 kDa) and also immunoreacted with a smaller band of ∼130 kDa in all V. vulnificus strains tested (four clinical strains and one environmental strain) (Fig. 5A; see Fig. S2 in the supplemental material), as observed in other studies (18, 20, 24, 40). In the culture supernatants of CMCP6 and YJ016, an additional band with an apparent molecular mass of 170 kDa was also recognized by the three MAbs, presumably due to toxin variants with different arrangements of effector domains generated by V. vulnificus rtxA1 gene recombination (Fig. 5A) (41). Consistent with the data obtained with the three cross-reactive MAbs, RtxA1-C-specific antiserum detected the intact RtxA1/MARTX_Vv protein (∼500 kDa) and the smaller bands (∼130 and/or ∼170 kDa) from all strains tested, indicating that these smaller bands could be cleavage products of the intact RtxA1/MARTX_Vv protein (Fig. 5A; see Fig. S2). Importantly, neither these three MAbs nor the RtxA1-C-specific antiserum recognized any specific band in the culture supernatant from an rtxA1-deficient V. vulnificus strain (CMM744) (Fig. 5A; see Fig. S2). Overall, these results demonstrate that the three MAbs against the V. vulnificus M06-24/O strain appear to exhibit broad cross-reactivity against various V. vulnificus strains.

FIG 5.

Immunoreactivity and cross-protection efficiency of anti-RtxA1-C MAbs against different V. vulnificus strains. (A) Immunoreactivity of anti-RtxA1-C MAbs against V. vulnificus strains. Native secreted V. vulnificus RtxA1/MARTX_Vv toxins from four clinical strains (ATCC 29307, CMCP6, YJ016, and M06-24/O) and one environmental strain (Env1) were obtained from bacterial culture supernatants by precipitation with cold acetone. Precipitated proteins were resolved by SDS-PAGE (10% Tris-glycine gels) and analyzed by Western blotting with an RtxA1-C-specific antiserum or with 21RA, 24RA, or 47RA MAb. CMM744 indicates the rtxA1-deficient mutant of the V. vulnificus M06-24/O strain. The positions of the protein markers (in kDa) and the antibodies used for Western blotting are indicated on the left and the right sides of the blots, respectively. “α-RtxA1-C” indicates the RtxA1-C-specific antiserum. (B and C) Cross-protection of an anti-RtxA1-C MAb, 21RA, against various V. vulnificus strains. Mice were inoculated intraperitoneally (i.p.) with 8 × 10⁶ CFU of the ATCC 29307 strain (B) or 1 × 10⁶ CFU of the CMCP6 strain (C) and then administered i.p. a single dose of 21RA MAb (0.5 mg) at 1 h postinfection. In the saline-treated groups, infection with the V. vulnificus ATCC 29307 or CMCP6 strain resulted in survival rates of 50% and 27%, respectively. Compared with the saline-treated control groups, the statistical differences for the survival curves of the 21RA-treated mice were as follows: ATCC 29307, 100% survival rate, P = 0. 0104; CMCP6, 90% survival rate, P = 0.0044. Survival curves were constructed using data from two independent experiments, each with five to six mice.

As the cross-reactive MAbs (21RA, 24RA, and 47RA) exhibited strong therapeutic activity against the V. vulnificus M06-24/O strain (Fig. 2A and Table 2), we further assessed the therapeutic effect of 21RA against various strains of V. vulnificus. Studies were performed with 8-week-old CD1 mice, which exhibit survival rates ranging from 27 to 50% upon lethal infection with different V. vulnificus strains. Mice were challenged i.p. with V. vulnificus ATCC 29307 (8 × 10⁶ CFU) or CMCP6 (1 × 10⁶ CFU) and then administered i.p. a single dose of 21RA MAb (0.5 mg) at 1 h postinfection. Notably, one dose of 21RA MAb provided significant therapeutic efficiencies against V. vulnificus strains ATCC 29307 and CMCP6, compared with saline-treated control groups (ATCC 29307, 100% survival rate, P = 0. 0104 [Fig. 5B]; CMCP6, 90% survival rate, P = 0.0044 [Fig. 5C]). Overall, these results suggest that cross-reactive MAbs against RtxA1-C could mediate cross-protection against different V. vulnificus isolates.

DISCUSSION

V. vulnificus is a well-known food-borne pathogen that causes gastroenteritis, septicemia, and necrotizing fasciitis (5, 42). V. vulnificus-induced septicemia is associated with very high rates of mortality, even with appropriate antibiotic treatments, since V. vulnificus grows particularly rapidly in vivo and is highly invasive (5, 6, 11). Therefore, the development of alternative treatment modalities that are prophylactically and therapeutically effective against V. vulnificus infection is urgently required. In this study, we generated a panel of new MAbs against RtxA1-C, which corresponds to the C-terminal region (amino acids [aa] 3491 to 4701) of V. vulnificus RtxA1/MARTX_Vv. Using immunoassays and recombinantly produced truncations of RtxA1-C, these MAbs were characterized with respect to their isotype, binding region on RtxA1-C, and extent of cross-reactivity against RtxA1/MARTX_Vv toxins from various V. vulnificus strains (Table 1 and Fig. 5A). Prophylactic studies in vivo demonstrated that 30% (11/37) of these anti-RtxA1-C MAbs conferred significant protection against V. vulnificus infection in a mouse model (Table 1). In particular, a single therapeutic dose of 21RA, 24RA, or 47RA significantly protected mice against V. vulnificus-induced mortality (Table 2, Fig. 2, and Fig. 5B and C). Studies with Fab fragments and a neutropenic mouse model demonstrated that these therapeutic MAbs elicited protective immunity through distinct pathways (Fig. 3). Furthermore, bacteriological study indicated that the therapeutic activity of 21RA is associated with enhanced clearance of V. vulnificus from the blood (Fig. 4).

Previous studies have shown that antisera against surface antigen and virulence factors, including VvPS and RtxA1/MARTX_Vv, are critical mediators of passive protection against V. vulnificus (23, 43–46). However, less information is available regarding the epitopes or determinants that elicit protective antibodies. In this study, we used E. coli-expressed fragments of RtxA1-C to identify the locations of the MAb epitopes. Based on previous secondary structure prediction and putative domain mapping (24), we generated three distinct fragments of RtxA1-C: FR-I (aa 3491 to 3980), FR-II (aa 3981 to 4380), and FR-III (aa 4381 to 4701). Of our MAbs, 14, 13, and 10 bound to FR-I, FR-II, and FR-III, respectively (Table 1). Interestingly, three MAbs (21RA, 24RA, and 47RA) with high therapeutic efficacy against intraperitoneal (i.p.) infection recognized the FR-I region (Tables 1 and 2 and Fig. 2A). However, no clear correlation was observed between the prophylactic efficacy of a MAb and its binding localization (Table 1 and Fig. 1A): FR-I, 3RA, 11RA, 19RA, 21RA, 24RA, 46RA, 47RA, and 50RA; FR-II, 13RA; and FR-III, 42RA and 45RA. Although more studies are needed to precisely determine the relationship between MAb binding and function, our data suggest that binding differences might contribute somewhat to the observed differences in MAb function.

Infections caused by V. vulnificus are characterized by two hallmark clinical manifestations: fulminating septicemia and rapidly progressive cellulitis (5, 6, 11, 42). Fulminating septicemia occurs primarily in patients with preexisting hepatic disease after the ingestion of contaminated seafood, and rapidly progressive cellulitis can result from infection of seawater-associated wounds. Healthy people are also at risk for developing cellulitis after wound infections by V. vulnificus, which could cause edema, necrosis, and septicemia. Given these findings, we further investigated the therapeutic efficacies of our anti-RtxA1-C MAbs in different animal models. Surprisingly, the 21RA MAb was therapeutically efficacious in both a subcutaneous (s.c.) infection model (Fig. 2B) and also in an iron overload mouse model using i.p. infection (Fig. 2C). Moreover, therapeutic administration of 21RA significantly increased survival when V. vulnificus was injected intragastrically (i.g.) in an iron overload mouse model (Fig. 2D). These results suggest that our anti-RxtA1-C MAbs show promising clinical potential against V. vulnificus infection.

Previous studies have demonstrated that antisera against virulence factors, including capsular polysaccharide (VvPS) and RtxA1/MARTX_Vv of V. vulnificus, can potently inhibit V. vulnificus infection in both passive transfer prophylactic and therapeutic models in vivo (23, 45). However, the protective mechanisms of these antisera have yet to be fully elucidated. In this study, we used Fab fragments and a neutropenic mouse model to gain insight into the mechanisms of our therapeutic MAbs against V. vulnificus RtxA1/MARTX_Vv. Although more studies are needed to definitively establish the mechanisms of these therapeutic MAbs, the therapeutic MAbs might block V. vulnificus invasion, facilitate immune system effector function, or inhibit a specific immunomodulatory or bacteriologic function of V. vulnificus RtxA1/MARTX_Vv. For 47RA, Fc-mediated effector functions appear to play an important role in its therapeutic effect against V. vulnificus, since treatment with 47RA Fab fragments, which are unable to activate Fc receptor-dependent pathway(s), did not provide any therapeutic advantage (Fig. 3A). This suggests that Fc receptor-expressing immune cells, including macrophages, NK cells, and neutrophils, could engage 47RA to mediate its therapeutic effect. As expected, the therapeutic efficacy of 47RA in neutropenic mice, which harbor a reduced neutrophil population, was reduced to 50% compared to the efficacies of 21RA and 24RA (Fig. 3B). Consistent with our data, neutropenic mice have been shown to exhibit significantly increased sensitivity to V. vulnificus infection (21). Furthermore, the phagocytic activity of neutrophils in patients with chronic liver diseases has been shown to be lower than that in healthy individuals, and survival of V. vulnificus in these patients was shown to be inversely correlated with the extent of neutrophil-mediated phagocytosis (47). Given that V. vulnificus RtxA1/MARTX_Vv can be delivered by direct contact of the bacteria with the target cells, we hypothesize that 47RA might target the RtxA1/MARTX_Vv from V. vulnificus attached to target cells, such as neutrophils. Following opsonization with 47RA, the interaction of the MAb and Fc receptor(s) on neutrophils could trigger Fc-mediated phagocytosis and clearance of V. vulnificus by activating signaling pathways that induce the formation of phagosomal membranes around the bacteria. In support of this hypothesis, MAbs against Bordetella pertussis adenylate cyclase toxin, a member of the RTX family of bacterial exotoxins, have been shown to promote Fc-dependent phagocytosis of B. pertussis by neutrophils (48). Also, a study of protective MAbs against a Bacillus anthracis toxin suggested that both the Fc fragment and the host Fcγ receptor(s) play an important role in the protective activity of these MAbs (49). However, the therapeutic effects of 21RA and 24RA in this study were both Fc independent and neutrophil independent, since the Fc regions of these MAbs were not required for protection and these MAbs completely maintained their therapeutic benefits in neutropenic mice (Fig. 3). Although we did not precisely identify the specific mechanism(s) of 21RA and 24RA, their therapeutic effects could be mediated by the direct inhibition of a previously implicated function of V. vulnificus RtxA1/MARTX_Vv, since this toxin has been linked to various bacteriologic and immunomodulatory effects, including antiphagocytic activity, cellular necrosis and apoptosis, colonization, and invasion into the bloodstream from the gastrointestinal tract (17, 18, 21, 22, 38, 40, 50). In addition, more recent studies have supported these potential functions of V. vulnificus RtxA1/MARTX_Vv (51–54). Interestingly, many RTX toxins mainly target leukocytes expressing the β-subunit (CD18) of β2 integrins and induce necrosis, cell lysis, or signaling cascades that lead to apoptosis in toxin concentration-dependent manners (52, 55–58). For example, the B. pertussis adenylate cyclase toxin, which targets phagocytes with CD11b/CD18 integrin receptor, promotes the internalization of integrins and raft components, hampers macrophage adhesion capacity by increasing the intracellular cyclic AMP (cAMP) concentration and inducing calcium-dependent processes, and suppresses phagocyte bactericidal functions (52). Based on these potential functions of bacterial toxins and our results (Fig. 3 and 4), we speculate that binding of these MAbs to V. vulnificus RtxA1/MARTX_Vv could inhibit one of the cytotoxic functions of this toxin against target cells, such as neutrophils and macrophages. This inhibition could occur by one or more of the following mechanisms. (i) These MAbs could block the interaction of V. vulnificus RtxA1/MARTX_Vv with a toxin receptor on neutrophils and macrophages, thereby inhibiting the internalization and antibacterial signaling of the bacterial toxin, or (ii) internalized MAbs bound to V. vulnificus RtxA1/MARTX_Vv could block the function of the MAb-binding region and/or preclude its interaction with intracellular targets. Although future follow-up studies on the detailed mechanisms underlying the protective effects of 21RA and 24RA against V. vulnificus will be necessary to discriminate between these possibilities, given that these MAbs recognized the FR-I region (aa 3491 to 3980) (Table 1), which harbors a cysteine protease domain, they are likely to inhibit the function of the FR-I region in an Fc-independent manner. It is also possible that our therapeutic MAbs inhibit V. vulnificus infection through a combination of different mechanisms. Based on the data presented here, our ongoing studies are aimed at testing these hypotheses and defining the precise mechanisms through which the 21RA, 24RA, and 47RA MAbs protect against V. vulnificus infection. Interestingly, we found that 47RA did not compete with either 21RA and 24RA for binding to RtxA1-C in MAb competition studies (T. H. Lee and K. M. Chung, unpublished observations), which provides a clue that MAbs recognizing distinct regions of RtxA1-C might elicit protective immunity against V. vulnificus through different mechanisms. Taken together, the data presented here indicate a novel potential pathway of protection by anti-RtxA1-C MAbs after V. vulnificus infection.

For rapid preclinical and clinical implementation, candidate antibodies must effectively inhibit bacterial infection and exhibit a broad antibacterial spectrum. A previous study demonstrated that antiserum against VvPS provided significant protection when administered before and after lethal V. vulnificus infection (45). However, a potential limitation for the application of antiserum against VvPS as a protective agent would be the existence of a large number of serotypes and carbotypes of VvPS, which could limit the cross-reactivity of anti-VvPS serum (45, 59, 60). The use of MAbs against RtxA1-C, the C-terminal region of V. vulnificus RtxA1/MARTX_Vv, could avoid this potential concern because the C-terminal region of V. vulnificus RtxA1/MARTX_Vv is highly conserved (12, 14, 15, 41). In support of this, three therapeutic MAbs (21RA, 24RA, and 47RA) against the V. vulnificus M06-24/O strain recognized both intact and processed RtxA1/MARTX_Vv products in different clinical and environmental strains (ATCC 29307, CMCP6, YJ016, and Evn1) (Fig. 5A; see Fig. S2 in the supplemental material). Notably, these MAbs cross-reacted with an ∼130-kDa cleavage product of RtxA1/MARTX_Vv, which might be associated with RtxA1/MARTX_Vv-mediated cell death (Fig. 5A) (40). In therapeutic postexposure trials in mice, a cross-reactive MAb, 21RA, also provided cross-protection against various clinical strains (Fig. 5B and C). Given these findings, MAbs against RtxA1-C could be highly clinically relevant in the development of novel therapeutic MAbs against V. vulnificus infection.

In summary, we have demonstrated that prophylaxis or therapy with MAbs against RtxA1-C effectively inhibited lethal V. vulnificus infection through at least two independent mechanisms. Our protective MAbs against V. vulnificus may facilitate the generation of novel therapeutic MAbs capable of providing superior clinical protection against V. vulnificus. Moreover, improved clinical outcomes could also be obtained by combining therapeutically efficacious MAbs that act through independent mechanisms and/or by administering therapeutic MAbs in combination with antibiotics.