The serogroup B meningococcal outer membrane vesicle-based vaccine 4CMenB induces cross-species protection against Neisseria gonorrhoeae

Abstract

There is a pressing need for a gonorrhea vaccine due to the high disease burden associated with gonococcal infections globally and the rapid evolution of antibiotic resistance in Neisseria gonorrhoeae (Ng). Current gonorrhea vaccine research is in the stages of antigen discovery and the identification of protective immune responses, and no vaccine has been tested in clinical trials in over 30 years. Recently, however, it was reported in a retrospective case-control study that vaccination of humans with a serogroup B Neisseria meningitidis (Nm) outer membrane vesicle (OMV) vaccine (MeNZB) was associated with reduced rates of gonorrhea. Here we directly tested the hypothesis that Nm OMVs induce cross-protection against gonorrhea in a well-characterized female mouse model of Ng genital tract infection. We found that immunization with the licensed Nm OMV-based vaccine 4CMenB (Bexsero) significantly accelerated clearance and reduced the Ng bacterial burden compared to administration of alum or PBS. Serum IgG and vaginal IgA and IgG that cross-reacted with Ng OMVs were induced by 4CMenB vaccination by either the subcutaneous or intraperitoneal routes. Antibodies from vaccinated mice recognized several Ng surface proteins, including PilQ, BamA, MtrE, NHBA (known to be recognized by humans), PorB, and Opa. Immune sera from both mice and humans recognized Ng PilQ and several proteins of similar apparent molecular weight, but MtrE was only recognized by mouse serum. Pooled sera from 4CMenB-immunized mice showed a 4-fold increase in serum bactericidal50 titers against the challenge strain; in contrast, no significant difference in bactericidal activity was detected when sera from 4CMenB-immunized and unimmunized subjects were compared. Our findings directly support epidemiological evidence that Nm OMVs confer cross-species protection against gonorrhea, and implicate several Ng surface antigens as potentially protective targets. Additionally, this study further defines the usefulness of murine infection model as a relevant experimental system for gonorrhea vaccine development.

Fig 1. Serum and vaginal antibody titers against F62 OMVs from 4CMenB-immunized mice.

Groups of 25 BALB/c mice were immunized three times, three weeks apart with 250 μl or 4CMenB by the IP or SC routes or with PBS or Alum (IP route). Serum and vaginal antibody titers on day 31 and day 52 (ten days after the 2^nd and 3^rd immunization, respectively) against F62 OMVs were measured by ELISA. Shown are serum (A) total Ig, (B) IgG1, (C) IgG2a, (D) IgA and (E) IgG1/IgG2a ratios for on days 31 and 52 (left column). Vaginal (F) total Ig, (G) IgG1, (H) IgG2a and (I) IgA on day 31 are shown in the right column. No difference was found in any sample or Ig tested between control animals receiving PBS and those receiving alum only. *, p<0.05; **, p<0.01; ***, p<0.0001. Results from the repeat experiment were similar.

Fig 2. 4CMenB has in vivo efficacy against Ng.

Mice were immunized three weeks apart with 250-μL doses of 4CMenB by the IP (blue) or SC (red) route or with PBS (black) or alum (purple) by the IP route and challenged with Ng strain F62 three weeks after the final immunization. Shown are the combined data from two independent trials (total n = 38–41 mice/group). (A) Percentage of culture-positive mice over time; (B) Average CFU per ml of a single vaginal swab suspension; (C) total bioburden over 7 days expressed as area under the curve; (D) Percentage of mice that cleared infection by day 7 post-challenge. *, p < 05; **, p < 0.01; ***, p < 0.0001.

Fig 3. Pooled antiserum from 4CMenB-immunized mice is bactericidal against a serum-sensitive and a serum-resistant Ng strain.

Serial dilutions of pooled serum from mice vaccinated with 250 μl of 4CMenB (blue lines) or Alum alone (black lines) by the IP route were incubated with 10⁴ CFU of the challenge strain F62 or strain FA1090 in microtiter plates as described in the Methods. After 5 min, NHS or heat-inactivated (HI)-NHS (final concentration, 10%) was added. After 55 min incubation at 37°C, the number of viable Ng in each well was determined by duplicate culture on GC agar. (A, C) Recovery of F62 (A) or FA1090 (C) from each condition expressed as the number of CFU from wells incubated with test serum divided by the number recovered from wells containing PBS instead of test serum, X times 100. Solid lines indicate NHS was used as the complement source; dotted lines represent data from wells tested in parallel with HI-NHS. The dotted line at 50% survival was drawn to identify the bactericidal₅₀ titers. (B, D) Recovery of viable F62 (A) or FA1090 (B) from a 1:480 or 1:240 dilution of serum, respectively, from 4CMenB-immunized mice or unimmunized mice for each of three independent experiments, expressed as above. A significant difference was observed between the immunized and unimmunized serum (**, p < 0.01; ***, p < 0.0001).

Fig 4. Serum and vaginal antibodies from 4CMenB-immunized mice recognize Ng outer membrane proteins by western immunoblot.

(A) Pooled antiserum from mice immunized with 250 μL of 4CMenB by the IP (250IP, upper panels) or SC (250SC, lower panels) route were tested against OMVs (app. 20 μg per lane) from 7 different Ng strains fractionated on 4–20% Tris-glycine gels by western blot (1:10,000 dilution of primary antisera) followed by secondary anti-mouse IgG-HRP. A boosting effect is observed when comparing the band intensities for serum collected after 2 and 3 immunizations. (B) Pooled vaginal washes from immunized or control mice collected after the third immunization tested against OMVs from the F62 challenge strain by western blot (1:100 dilution), followed by secondary anti-mouse IgG-HRP or anti-mouse IgA-HRP. Pooled vaginal washes were from mice given: (1) SC250; (2) IP250; (3) PBS; (4) Alum only. (5) 250IP mouse serum (1: 10,000) used for comparison. The band recognition pattern was similar for blots incubated with serum (lanes 5) or vaginal washes (lanes 1–4), and vaginal washes from IP-immunized mice were more strongly reactive than from SC-immunized mice. All lanes were equally loaded, as determined by Ponceau S staining. Shown are representative results from at least 2 separate experiments with identical results.

Fig 5. PilQ, MtrE, porin and Opa are recognized by 4CMenB antisera.

OMVs (app. 20 μg) from 6 Ng strains (Table 2) were subjected to SDS-PAGE on a 4–20% Tris-glycine gel and (A) stained with sypro ruby or (B) transferred to PVDF for western blot with the 250IP antiserum. The stained gel was aligned with the Western blot, and corresponding bands were digested and analyzed by mass spectrometry. The numbers indicated with arrows on each panel correspond to the same numbers on the Western blots except for bands 5 and 6, which were excised from a different gel but are indicated on the western based on the banding patterns. Proteins identified are described in Table 1. Among the proteins identified, known surface-exposed outer membrane proteins are: (1) PilQ, (3) BamA, (4) MtrE, (5) PorB, and (6) Opa.

Fig 6. 4CMenB -induced antibodies bind PilQ and MtrE at the surface of viable Ng FA1090 and MS11 bacteria.

Immunoprecipitations were performed with wild-type strains FA1090 and MS11 and their isogenic pilQ and mtrE mutants using antisera from mice immunized with 250 μL of 4CMenB via the IP route (250IP) or given alum only (negative control). Bacterial components bound by 4CMenB-induced antisera were subjected to SDS-PAGE (non-denaturing conditions, 4–20% Tris-glycine) and Western blotting with 4CMenB 250IP antiserum. WCLs, total cellular proteins; OMVs, outer membrane vesicles; pilQ-, FA1090ΔpilQ; mtrE-, MS11ΔmtrE. Data shown are representative of at least 2 separate experiments with identical results. The wide band around 200 kDa corresponds to the antibodies within the test antisera that are present in the antigen-antibody complexes and pulled down with the protein A/G agarose.

Fig 7. Comparison of pooled serum from 4CMenB-immunized mice and humans for cross-reactivity against Ng OMV proteins.

The reactivity of pooled serum from 4CMenB-immunized mice and humans was tested for recognition of Ng strain F62 OMV proteins by western blot. Pooled serum from unvaccinated humans was tested in parallel. Mouse antiserum was collected 10 days after the third 4CMenB immunization; human subjects were immunized with 4CMenB two times and the sera were collected between 2 and 15 months after the second immunization as described in the Methods. (A) Pooled mouse serum from 4CMenB-immunized mice, 1:10,000 dilution; (B) pooled serum from 4CMenB-immunized persons, 1:500; (C) pooled serum from unimmunized persons, 1:500. Bands of similar molecular weight that were detected by both immunized murine and human sera are indicated by the solid arrow; bands that were detected only by mouse sera are indicated by the open arrow heads. The lowest band indicated by the open arrowheads is MtrE. The faint band under the 40 kD marker indicated by the solid arrow is consistent with the apparent molecular weight of PorB. The ~ 20 kD band that is recognized by both human and mouse serum is similar in apparent molecular weight to a band that was strongly recognized by serum from a 4CMenB-immunized individual in Semchenko et al. [30]. A band of this molecular weight was also intensely recognized by IgG and IgA in vaginal washes from 4CMenB-immunized mice (Fig 4B).

Fig 8. Individual sera from 4CMenB-immunized subjects recognize several Ng OMV proteins including PilQ.

Individual sera from (A) seven 4CMenB-immunized and (B) seven unimmunized individuals were tested against Ng F62 OMV by western blot. Several bands were recognized by sera from immunized people that were not detected by sera from unimmunized subjects with the exception of bands ~ 62 kD and 119 kD, which were detected by most sera regardless of immunization status. The reactivity of individual sera from (C) five 4CMenB-immunized and (D) two unimmunized individuals was tested against Ng strain F62 OMV proteins that were fractionated by SDS/PAGE for a longer period of time to allow visualization of PilQ. An intense band above the 220 kD marker, which is consistent with the migration of PilQ, was seen in blots incubated with four of the five immunized serum samples but not when incubated with two unimmunized control serum samples. Serum from the fifth immunized individual (S11) also recognized PilQ, but much less intensely.

Fig 9. 4CMenB-immunized mice have cross-reactive antibodies against Ng NHBA.

Crude OMVs (~10 μg) prepared from Ng strains FA1090 or FA1090-ΔNhba were fractionated by SDS-PAGE and transferred to nitrocellulose membranes. Western blot was performed using sera (1:2,500) from mice immunized with 4CMenB (right panel) or given alum alone (middle panel) or PBS (left panel). A band of ~ 66 kD is recognized by the immune serum in OMVs from the wild-type strain but not in OMVs from the isogenic nhba mutant FA1090-ΔNhba.

Fig 10. Serum from 4CMenB-immunized and unimmunized individuals showed no significant difference in bactericidal activity against the challenge strain.

Human serum from ten 4CMenB-immunized and eleven unimmunized individuals was tested for bactericidal activity against Ng strain F62. The bactericidal₅₀ titer is shown for each individual (immunized: mean 1,584, median 1,200; unimmunized: mean 993, median 720). The difference in the two groups was not significant (unpaired t-test, p = 0.3). The interval between the second dose of 4CMenB and the date of serum collection ranged from 2 months (2 individuals), 4 months (6 individuals) and 15–16 months (2 individuals), with no potential correlation observed between number of months post-vaccination and bactericidal₅₀ titer.