N. gonorrhoeae standard strains FA1090 (ATCC700825) and FA19 (BAA-1838) and two N. gonorrhoeae strains from clinical isolates (Department of Medical Laboratory, Affiliated Hospital of Zunyi Medical University) were plated overnight on gonococcal base (GCB) agar (Oxoid, Basingstoke, UK) at 37°C with 5% CO₂. To obtain sialylation of N. gonorrhoeae, a final concentration of 50 μg/mL CMP-N-acetylneuraminic acid (CMP-NANA) (Sigma) was added to gonococcal base liquid (GCBL) medium and grown at 37°C for 18 h in an atmosphere of 5% (v/v) CO₂ (30, 31). In some experiments, bacteria grown for 18–24 h were resuspended in phosphate-buffered saline (PBS) using a Sensititre nephelometer to prepare the corresponding bacterial suspensions.

Sequences of the App passenger domain (ngo2105), NHBA protein (ngo1958), and MetQ protein (ngo2139) were obtained from the NCBI database (GenBank: NC_002946.2). Genes encoding the passenger domain (43–1190 aa), NHBA protein (1–429 aa), and MetQ protein (23–288 aa) fragments were amplified from N. gonorrhoeae FA1090 through polymerase chain reaction using the primers listed in Supplementary Table S1 . The above gene fragments were amplified and cloned into the pCold-TF vector (plasmid map is shown in Supplementary Figure S1 ) to obtain template plasmids pCold-TF-passenger, pCold-TF-nhba, and pCold-TF-metQ.

The pCold-TF-passenger, pCold-TF-nhba, and pCold-TF-metQ, and pCold-TF plasmids were transformed into E. coli BL21 (DE3) cells using heat shock. Bacterial strains transformed with recombinant plasmids were cultured on Luria–Bertani agar plates supplemented with ampicillin (100 µg/mL) and then incubated at 37°C for 12–15 h. A single colony was inoculated into Luria–Bertani broth supplemented with ampicillin in a conical flask and then incubated at 37°C for 12–15 h with shaking at 250 rpm. When the bacterial culture reached an optical density of 0.60 at 600 nm (OD₆₀₀), 0.05 mM isopropyl beta-D-thiogalactoside was added to induce recombinant protein expression, which continued for 10 h at 15°C. Recombinant proteins were purified using a Ni²⁺-nitrilotriacetic acid column, and imidazole was removed. Lipopolysaccharide was removed from the recombinant protein preparations using an EtEraser™ Endotoxin Removal Kit (Xiamen Bioendo Technology Co. Ltd, Xiamen, China). After the recombinant protein was treated with endotoxin removal, LPS content in the recombinant protein was detected using a gram-negative lipopolysaccharide detection kit (Xinuo Biopharmaceutical Co, LTD, Tianjin, China), which was lower than 0.5 EU/mL. Protein purity was analyzed via 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie Blue Fast staining (Epizyme, China). The protein concentration was determined using a BCA kit (Solarbio, Beijing, China), followed by storage at −80°C until use.

Female BALB/c mice (6–8 weeks old) of specific-pathogen-free (SPF) grade were purchased from the Experimental Animal Center of Zunyi Medical University (Zunyi, China). The mice (n = 6) were randomly divided into six groups: App + CTB, NHBA + CTB, MetQ + CTB, App + NHBA + MetQ (trivalent + CTB), pCold-TF + CTB, and PBS + CTB. Mice were immunized intranasally on days 0, 14, 28, and 42 (32). Then, 10 µg of App, NHBA, MetQ, or pCold-TF protein or 10 µg each of App, NHBA, MetQ, or their mixture (App + NHBA + MetQ) were suspended in 20 μL of PBS and mixed with 10 μg of CTB adjuvant (Absin, Shanghai, China) to a total volume of 30 μL for mouse immunization. Tail vein blood was collected from the mice 1 week after each vaccination, and vaginal secretions were collected on days 35 and 49. For vaginal secretions, 50 μL of sterile PBS was aspirated into the posterior vaginal fornix three to five times, repeated twice, and the fluids were combined. Finally, after centrifugation, the supernatants of the serum and vaginal secretions were stored at −80°C until further analysis.

N. gonorrhoeae FA1090, FA19, and the two clinical strains were inoculated on GCB plates and incubated at 37°C for 24 h. Colonies were picked and placed in PBS and adjusted to 0.5 McFarland units. The bacterial precipitates were washed with PBS and added to 1× loading buffer, then boiling at 100°C for 15 min. Bacterial lysates were separated via SDS-PAGE and transferred to polyvinylidene fluoride membranes. The expression of App passenger, NHBA, and MetQ in different N. gonorrhoeae strains were analyzed through immunoblotting using App, NHBA, or MetQ antisera as the primary antibodies (1:2,000) and horseradish peroxidase-labelled goat anti-mouse IgG (ZSGB-Bio, China) as the secondary antibody (1:5,000). Finally, a high-sensitivity chemiluminescent substrate detection kit (Epizyme, Cambridge, MA, USA) was used to analyze App, NHBA, and MetQ expression in different N. gonorrhoeae strains.

Antibody titers in serum and vaginal secretions were assessed using enzyme-linked immunosorbent assays (ELISAs). Nunc-Immuno plates were coated with purified antigen (96 wells of purified antigen for App, NHBA, MetQ, or pCold-TF at 10 μg/mL), incubated overnight at 4°C, and washed with PBS-T (PBS containing 0.1% Tween 20). Antisera from the immunized groups were diluted with blocking buffer (100 μL/well) and incubated at 37°C for 1 h in the wells. The plates were then washed three times with PBS-T. Horseradish-peroxidase-labelled goat anti-mouse IgG, IgG1, IgG2a, IgG2b, IgG3, and IgA were diluted at 1:5,000 and added to the plates, which were left to stand for 1 h at 37°C. The plates were washed three times, and a 3,3’,5,5’-tetramethylbenzidine solution (Solarbio) was added for 30 min for color development. The reaction was terminated by adding a termination solution (Solarbio). Absorbance was recorded at 450 nm, and the antibody titer was defined as the highest dilution of the sample with an absorbance value 2.1 times that of the negative control.

For the serum bactericidal assay (SBA) (33), N. gonorrhoeae FA1090 and FA19 were inoculated into GCB plates and incubated for 18–24 h at 37°C with 5% CO₂. To obtain sialylation of N. gonorrhoeae strain FA1090 was cultured in GCBL medium (containing 50 μg/mL CMP-NANA) in 5% (v/v) CO₂ at 37°C for 18 h. The number of selected colonies was adjusted to 4 × 10⁴ colony-forming units (CFU)/45 μL. Then, 45 μL of heat-inactivated antisera at different dilution titers (mixture of three mouse sera samples) was added, and the plates were incubated at 37°C with 5% CO₂ for 15 min. Thereafter, 25% human serum was added as a complement source, followed by incubation at 37°C, 5% CO₂ for 30 min. The entire reaction mixture was diluted and spread onto GCB plates, with the CFU counted on the following day. The SBA bactericidal titers had the highest antibody dilutions, which resulted in more than 50% N. gonorrhoeae killing. The untreated group (0, without serum and complement), the lowest dilution of PBS-immunized group antiserum (PBS, with complement) and pCold-TF-immunized group antiserum (pCold-TF, with complement) were set as controls. The bacterial survival rate of the untreated group (0, without serum and complement) was set to 100%, and the bacterial survival rate of each group was calculated by the number of colonies in each serum group/number of colonies in the untreated group ×100%.

As described previously (26), an antibody-mediated adhesion inhibition assay was performed using human cervical cancer epithelial cells (ME-180, ATCC HTB33, Fenghui Biotechnology Co, Ltd, Hunan, China). ME-180 cells were cultured in RPMI 1640 with 10% fetal bovine serum for 24 h to form a monolayer of fused cells in 24-well tissue culture plates. Bacterial suspensions were prepared by inoculating N. gonorrhoeae FA1090 onto GCB plates and incubating overnight. Colonies with pili were then selected. The bacteria were pre-incubated with heat-inactivated antiserum in RPMI 1640 medium (HyClone, Logan, UT, USA) for 30 min at 37°C. ME-180 cells were washed three times with PBS, and the antiserum-pre-treated bacterial suspension was added at a multiplicity of infection of 10:1 to the ME-180 cells. This was followed by co-incubation at 37°C and 5% CO₂ for 3 h. To remove the non-adherent bacteria, we washed the solution three times with PBS. The ME-180 cells were lysed with 1% saponin. The lysates were serially diluted and plated on GCB plates to determine the bacterial number of CFU. Adhesion rates were calculated as the ratio of the number of CFU that adhered to the cells to the initial number of CFU.

Whole-cell ELISA experiments were carried out as described (15). The ELISA was conducted using 96-well MaxiSorp plates (Nunc) coated with N. gonorrhoeae FA1090 strains (1×10⁷ CFU). After coating, the plates were washed with PBS and blocked with PBS containing 1% bovine serum albumin (BSA) for 1 hour. Following blocking, the plates were incubated with antibody (diluted 1:1,000) for 1 hour at 37°C, 5% CO₂. After washing three times with PBS, the plates were incubated with a secondary antibody (HRP-conjugated anti-mouse, diluted 1:2,000) for 1 hour. After additional washes, the plates were developed using tetramethylbenzidine (TMB) solution (Solarbio). The reaction was terminated by adding 1 volume of 1 M HCl. Absorbance was measured at 450 nm using a Victor3 plate reader. Relative binding capacity was calculated as A450_sialylation/A450_{non-sialylation} for each antiserum.

One week after the last immunization, three mice were randomly selected from each group. Mouse spleens were crushed and resuspended through filtration in RPMI 1640 to prepare splenocyte suspensions that were supplemented with 10% fetal bovine serum, 10 kU/mL penicillin, 10 mg/mL streptomycin, and 25 μg/mL amphotericin B. The suspension was incubated at room temperature for 24 h. The cells were added to 24-well cell culture plates (1 mL/well) at 5 × 10⁶ cells/mL. Finally, splenocytes from immunocompetent mice were stimulated by adding 10 μg of purified App, NHBA, MetQ, App + NHBA + MetQ (Trivalent), or pCold-TF proteins to PBS and cultured for 72 h at 37°C and 5% CO₂. The culture medium supernatant was collected, and the levels of IL-17A and IFN-γ were determined using cytokine assay kits (Proteintech, Rosemont, IL, USA).

SPF-grade female BALB/c mice aged 6 to 8 weeks (Experimental Animal Centre, Zunyi Medical University) were randomly divided into the following six groups (n = 6): App, NHBA, MetQ, App + NHBA + MetQ (Trivalent), pCold-TF, and PBS. Mice were vaccinated via intranasal inoculation on days 0, 14, 28, and 42. The vaccine contained 10 μg of antigen and 10 μg of CTB (CTB adjuvant:antigen = 1:1; total volume 30 μL). After the final immunization, pro-estrus mice were administered subcutaneous injections of 0.5 mg of sesame-oil-soluble estradiol on days −2, 0 (day of bacterial challenge), and +2 to increase their susceptibility. Mice were treated with antibiotics to prevent the overgrowth of commensal flora. FA1090 (1 × 10⁸ CFU/mL) was prepared, and mice were infected with 20 μL of this solution to achieve an infective dose of approximately 2 × 10⁶ CFU/mouse. Each group of mice was vaginally inoculated with vaginal secretions (vaginal rinse with 50 μL of normal saline, repeated twice), which were collected daily to dilute the smear for colony counting and observe the clearance of N. gonorrhoeae colonization.

Comparisons between two independent groups were performed using the Student’s t-test. Antibody titers were analyzed using Šídák’s multiple comparison test. Kaplan–Meier plots were used to analyze the time needed for infection clearance. The log-rank (Mantel–Cox) test was used to compare the Kaplan–Meier curves of the groups. The AUC comparisons should be made by One-way ANOVA and Dunn’s multiple comparisons test.

To construct the trivalent vaccine, we used the App protein passenger domain, full-length NHBA, and truncated MetQ with the signal peptide region removed as vaccine components ( Figure 1A ). These components were expressed in E. coli BL21 with the tag protein “trigger factor” in the pCold-TF vector, resulting in pure bands of the target proteins at relative molecular masses of 180, 130, 90, and 55 kDa, respectively ( Figure 1B ). Polyclonal antibodies were prepared by immunizing mice with the three tagged recombinant proteins. Western blotting showed that the polyclonal antibodies specifically recognized natively expressed App, NHBA, and MetQ in two laboratory strains, FA1090 and FA19, as well as in randomly selected clinical isolates ( Figure 1C ). The genetic sequences of 100 strains of N. gonorrhoeae showed 99.93% 96.37%, and 99.88% conservation of app, nhba and metQ genes, respectively ( Supplementary Tables S2–S4 ), suggesting that App, NHBA, and MetQ are conserved among different strains.

Mice were immunized with tagged recombinant proteins intranasally on days 0, 14, 28, and 35, with serum and vaginal secretions collected to determine antibody titers 1 week after each immunization ( Figure 2A ). Overall, mice immunized with the trivalent vaccine tended to exhibit stronger circulating IgG and IgA antibody responses than those immunized with the monovalent vaccine ( Figures 2B–G ). Both IgG and IgA antibodies were significantly induced in vaginal washes ( Figures 3A–F ). Except for anti-NHBA IgG, the trivalent vaccine group had significantly higher IgG and IgA antibody titers in the vaginal washes after the fourth immunization than those in the monovalent vaccine group.

The IgG isotype indirectly reflects changes in the Th1/Th2 balance (26). In mice, Th1-type cytokines induce lgG2a, whereas Th2-type cytokines induce lgG1 antibody production. After the final immunization, higher titers of each IgG isotype were observed in the trivalent vaccine group than in the monovalent vaccine group ( Figures 4A–C ). All groups generated high titers of IgG1 antibody responses (IgG1 > IgG2b > IgG2a > IgG3) and had IgG1/IgG2a ratios > 1. These results suggest that mice in all groups exhibited a bias toward Th2-type humoral immune responses. The IFN-γ and IL-17A levels were significantly higher in spleen leukocyte supernatants from trivalent vaccine-, App-, NHBA-, and MetQ-immunized mice compared to those in control mice ( Figures 4D, E ). IL-17A levels were significantly higher in the trivalent vaccine group than in the monovalent vaccine group. These results suggested that the trivalent vaccine elicited strong Th1, Th2, and Th17 immune responses.

To further investigate the putative protective immune responses, we characterized the functions of antibodies elicited by the trivalent vaccines. We evaluated the bactericidal efficacy of antisera from various immunization groups against three strains, including homologous N. gonorrhoeae FA1090 and heterologous FA19. SBA showed that the antiserum from the trivalent vaccine group had stronger bactericidal activity against the three strains than that of the NHBA and MetQ groups, with bactericidal activity range of 100–800 dilution titers ( Figures 5A, B and Supplementary Figure S2 ). To examine whether sialylation has an effect on bactericidal activity, we cultured strain FA1090 in the presence of CMP-NANA and tested the bactericidal activity of antiserum against both sialylated and non-sialylated gonococci. The results showed that the bactericidal efficiency of several antisera against sialylated gonococci was lower than that of non-sialylated gonococci ( Figure 5C ). We investigated whether sialylation affected the binding of antiserum to N. gonorrhoeae FA1090 by whole bacteria ELISA. The results showed no significant differences in the binding ability of the different antisera to sialylated and non-salialylated gonococci ( Figure 5D ).

The adhesion of gonococci to epithelial cells is a key mechanism in the establishment of infection (34). Therefore, adhesion experiments were performed on ME-180 cells with strain FA1090 in the presence of antiserum. Our results showed that the antisera from each immunized group inhibited the adhesion of gonococci to ME-180 cells in a concentration-dependent manner ( Figure 6 ). When the antisera of the App, NHBA, and MetQ groups were diluted to 1:80, gonococcal adhesion was reduced by 36.76% (P < 0.001), 31.37% (P < 0.0001), and 38.74% (P < 0.01), respectively, whereas the antisera of the trivalent vaccine group reduced gonococcal adhesion by 60.22% at this dilution (P < 0.0001). These results showed that the antiserum from the trivalent vaccine group inhibited the gonococcal adhesion rate to a significantly greater extent than antisera from the monovalent vaccine groups.

The efficacy of the trivalent and monovalent vaccines was tested in a mouse model of vaginal gonococcal infection by assessing the daily bacterial loads in vaginal washes and the time to clearance of the infection. There was no significant difference in the number of colonized bacteria between the PBS and pCold-TF controls, whereas the number of colonized bacteria in the monovalent vaccine group was lower than that in the control group ( Figure 7A ). Importantly, the number of colonized bacteria was significantly less in the trivalent vaccine group than in the monovalent vaccine group ( Figure 7A ). The area under the curve (AUC) was significantly lower in the trivalent vaccine group than in the pCold-TF group ( Figure 7B ). Although the AUC of the trivalent vaccine group exhibited a downward tendency when compared with the App, NHBA, and MetQ groups, no statistically significant difference was observed. In terms of the time to infection clearance, mice in the monovalent vaccine group were completely cleared of gonococcal infection on day 9, while for those in the PBS and pCold-TF control groups, clearance was achieved on day 10 (P < 0.01, Figure 7C ). Mice in the trivalent vaccine group were completely cleared of gonococcal infection on day 8, which was significantly faster than clearance in the NHBA and MetQ groups (P < 0.01), but not significantly different from that in the App group (P = 0.0548, Figure 7C ). Overall, these results suggest that the trivalent vaccine is somewhat better than the monovalent vaccine in protecting mice against gonococcal infections of the genital tract.