Development of a Tag/Catcher-mediated capsid virus-like particle vaccine presenting the conserved Neisseria gonorrhoeae SliC antigen that blocks human lysozyme

ABSTRACT Virus-like particles (VLPs) are promising nanotools for the development of subunit vaccines due to high immunogenicity and safety. Herein, we explored the versatile and effective Tag/Catcher-AP205 capsid VLP (cVLP) vaccine platform to address the urgent need for the development of an effective and safe vaccine against gonorrhea. The benefits of this clinically validated cVLP platform include its ability to facilitate unidirectional, high-density display of complex/full-length antigens through an effective split-protein Tag/Catcher conjugation system. To assess this modular approach for making cVLP vaccines, we used a conserved surface lipoprotein, SliC, that contributes to the Neisseria gonorrhoeae defense against human lysozyme, as a model antigen. This protein was genetically fused at the N- or C-terminus to the small peptide Tag enabling their conjugation to AP205 cVLP, displaying the complementary Catcher. We determined that SliC with the N-terminal SpyTag, N-SliC, retained lysozyme-blocking activity and could be displayed at high density on cVLPs without causing aggregation. In mice, the N-SliC-VLP vaccines, adjuvanted with AddaVax or CpG, induced significantly higher antibody titers compared to controls. In contrast, similar vaccine formulations containing monomeric SliC were non-immunogenic. Accordingly, sera from N-SliC-VLP-immunized mice also had significantly higher human complement-dependent serum bactericidal activity. Furthermore, the N-SliC-VLP vaccines administered subcutaneously with an intranasal boost elicited systemic and vaginal IgG and IgA, whereas subcutaneous delivery alone failed to induce vaginal IgA. The N-SliC-VLP with CpG (10 µg/dose) induced the most significant increase in total serum IgG and IgG3 titers, vaginal IgG and IgA, and bactericidal antibodies.

T he World Health Organization (WHO) Global Health Sector Strategy on sexually transmitted infections (STIs) notes vaccines as key innovations needed for sustaina ble STI control (1).
Among STI, gonorrhea is the second most reported notifiable disease in the USA after chlamydia with a total of 710,151 reported cases in 2021 (2,3).Globally, ~82.4 million new gonorrhea infections occurred in 2020, but these statistics are underestimated due to frequent asymptomatic infections (4)(5)(6).Neisseria gonorrhoeae (Ng), the Gram-nega tive bacterium and etiological agent of gonorrhea, is categorized as a high-priority pathogen for research and development efforts (7)(8)(9).The Centers for Disease Control and Prevention recommends ceftriaxone for the treatment of uncomplicated gonorrhea, but failures with this therapy have occurred, and multidrug-resistant Ng strains are rising globally (10)(11)(12)(13)(14)(15)(16)(17)(18).In addition to high prevalence and antibiotic resistance, the need for developing an effective gonorrhea vaccine is exacerbated by the brunt of gonorrhea, including infertility and its ability to augment the transmission and acquisition of HIV (19).In women, gonorrhea may lead to pelvic inflammatory disease, miscarriage, preterm birth, and ectopic pregnancies.In males, this STI presents as uncomplicated urethritis but can ascend to the epididymis or testes (20).Gonorrhea primarily affects the genitouri nary tract, but other mucosal surfaces can be involved, and disseminated disease may also occur (21)(22)(23)(24)(25). Neonatal conjunctivitis can be acquired from the infected birth canal, which if left untreated, can result in corneal scarring and blindness (25)(26)(27).
Two gonorrhea vaccines, composed of killed Ng and purified pilin, failed in clinical trials decades ago (28)(29)(30), illustrating the difficulty Ng poses to traditional vaccine design.The long-standing barriers to developing an effective Ng vaccine include remarkable antigenic variability, highly sophisticated strategies for modulating and evading host innate and adaptive immune responses, and the lack of established correlates of protection (31)(32)(33)(34)(35)(36).To address the first challenge, we carried out proteomics and bioinformatics to identify conserved vaccine antigens (37)(38)(39)(40).We selected the 34 gonorrhea protein antigens that were discovered through proteome-based reverse vaccinology studies and traditional approaches and carried comprehensive analyses of their sequence variation among over 5,000 clinical Ng isolates deposited in the Neisseria PubMLST database (5,(37)(38)(39)(40).Among the most conserved antigens we identified was a surface-exposed lysozyme inhibitor of c-type lysozyme, lipoprotein SliC (41).Our analysis showed that SliC is exceptionally well conserved, and over 96% of isolates have an identical SliC allele.The gene sliC (locus NEIP0196) has a total of 12 alleles and 22 single-nucleotide polymorphisms.There are only eight different amino acid sequences with 11 single amino acid polymorphisms distributed in <4% Ng isolates globally (40).In addition, utilizing ΔsliC, ΔsliC/p::sliC* (S83A K103A; a SliC unable to bind lysozyme), and lysozyme knockout (LysMcre) mice, we have shown experimentally that SliC provides a significant survival advantage for Ng during mucosal infection that is dependent on its function as a lysozyme inhibitor (41).Together, these data provide a premise for incorporating SliC in a gonorrhea vaccine.
We recognize, however, that subunit protein vaccines often fail due to low immu nogenicity caused by small antigen size, instability, or improper presentation to the immune system (42,43).Moreover, considering the mechanisms Ng uses to evade the human immune system, an effective vaccine may need to induce a stronger/different type of immune response compared to that elicited during infection (8,35,36).Subunit vaccines based on virus-like particles (VLPs) have been shown to induce potent B-cell responses in humans (44,45), which has led to the licensure of several successful vaccines, including hepatitis B, human papillomavirus (HPV), malaria, and hepatitis E vaccines.Intriguingly, a single dose of the HPV vaccine elicited highly durable (poten tially lifelong) antibody responses in humans (46).This ability is unprecedented by any other subunit vaccine and is believed to rely on the structural characteristics of the L1 antigen, which self-assembles into semi-crystalline capsid VLP (cVLP).Their antigenic similarity to virions makes them highly immunostimulatory (47).Specifically, their size (20-200 nm) and particular nature allow for passive drainage into lymph nodes, uptake by professional antigen-presenting cells, including B-cells, and innate immune system activation (48).Besides, their repetitive surface structure enables effective B-cell receptor crosslinking and B-cell activation (45,47,(49)(50)(51).Finally, they lack genetic material and are thus non-infectious and safe.Critically for vaccine development, the intrinsic immunogenicity of cVLPs extends to protein antigens, which are displayed at high density in an orderly fashion on the cVLP (52).This is especially apparent for antigens that are otherwise weak immunogens (53,54).
On that basis, we formulated SliC with cVLPs using the clinically validated Tag/ Catcher-AP205 cVLP platform (55).The Tag/Catcher-AP205 cVLP uses a highly effective split-protein-based conjugation system, which was developed by the separation of a bacterial pilin protein into a reactive peptide (Tag) and corresponding protein-binding partner (Catcher) (42,56).Upon mixing in the solution, the Tag and Catcher rapidly form a spontaneous isopeptide bond.This platform was developed by genetically fusing AP205 capsid to the split-protein Tag or Catcher, thus displaying 180 copies on the cVLP surface.The Tag/Catcher-AP205 has been utilized to display structurally and functionally diverse antigens, ranging in size from small peptides to large proteins (57).Importantly, the resultant VLP-displayed antigens induced antibody titers of higher quality, affinity, and avidity (42,(58)(59)(60).Recently, this vaccine technology using Tag/Catcher-AP205 decorated with the receptor-binding domain of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein was shown to elicit strong virus neutralization activity and is a highly promising candidate vaccine for preventing coronavirus disease 2019 (COVID-19), currently in a phase-3 clinical trial (NCT05329220) (55).The power of the Tag/Catcher-AP205 cVLP is well-documented but completely novel to the gonorrhea vaccine field (42,45,46,55,(58)(59)(60)(61)(62)(63)(64)(65)(66).

Bacteria and culture conditions
The serum-resistant Ng FA1090 (PorB1B) and the Ng 2016 WHO reference strains were used in the studies (37,67).The ∆sliC knockout and complementation strain ∆sliC/P::sliC were constructed previously using the Ng FA1090 (37).
Ng strains were maintained on gonococcal (GC) agar [GC base (GCB); Difco] with Kellogg's supplement I and 12.5 µM ferric nitrate or on chocolate agar plates, as indicated in the text, in a 5% CO 2 atmosphere at 37°C for 18-20 h (68).After passage on GCB, transparent and non-piliated colonies were cultured in GCB liquid (GCBL) medium supplemented with Kellogg's supplement I and 12.5 µM ferric nitrate (69,70).
Escherichia coli strain MC1061 was utilized as a host during genetic manipulations, whereas E. coli BL21(DE3) was used for overproduction of rSliC, N-SliC, and C-SliC.E. coli strains were maintained on Luria-Bertani (LB) agar or cultured in LB broth.

Genetic manipulations
Cloning procedures, oligonucleotides, and gene blocks were designed using SnapGene software version 2.8 (GSL Biotech LLC).Primers and gene blocks were synthesized by Integrated DNA Technologies.Q5 High-Fidelity DNA polymerase, DNA ligase, and NEBuilder HiFi DNA Assembly Master Mix were obtained from New England Biolabs (NEB).All resulting genetic constructs were verified by Sanger Sequencing at the Center for Quantitative Life Sciences at Oregon State University.

Expression and purification of N-SliC and C-SliC
Recombinant N-SliC and C-SliC were purified from 3 L cultures of E. coli BL21(DE3) (71) carrying pET22-N-SliC and pET22-C-SliC, respectively.Bacteria were incubated at 37°C, 210 rpm until OD 600 of ~0.6 was reached.Cultures were then shifted to 18°C for 1 h, and protein expression was induced with 0.1 mM IPTG.After overnight incubation, the cells were pelleted at 5,000 × g for 15 min at 4°C.Bacteria were suspended in binding/lysis buffer (50 mM Na 2 HPO4, pH 8, 200 mM NaCl, 25 mM imidazole, and 10% glycerol) supplemented with protease inhibitor mini tablets (Pierce) and lysed by passing through French Press at 1,500 psi.Cellular debris was removed by centrifugation at 8,000 × g for 15 min at 4°C, and the obtained supernatant was passaged through 0.45 µm nylon membrane (EZ flow).The cleared cell lysate was applied to a 5-mL IMAC Nickel column (Bio-Rad), and the recombinant proteins were purified on an NGC Medium-Pressure Liquid Chromatography System (Bio-Rad) using binding/lysis buffer and elution buffer (50 mM Na 2 HPO4, pH 8, 200 mM NaCl, 250 mM imidazole, and 10% glycerol).Elution fractions containing either N-SliC or C-SliC were pooled and applied onto a Vivaspin 20 centrifugal concentrator (GE HealthCare).Samples were supplemented with DTT and EDTA to final concentrations of 1 and 0.5 mM, respectively.The His Tag was removed by overnight incubation at 4°C with TEV protease in a 1:100 ratio.Dialysis was performed using a 1:4 ratio of binding/lysis buffer by placing TEV-digested samples into snakeskin dialysis tubing (3.5 MWCO) with low stirring overnight at 4°C.Samples were applied to a 5-mL IMAC Nickel column (Bio-Rad) to remove TEV.The removal of His Tag was confirmed by immunoblotting.Protein samples were concentrated as described above and subjected to size exclusion chromatography using a HiLoad 16/600 Superdex 75-pg column (GE HealthCare) with phosphate-buffered saline (PBS) as running buffer to isolate N-SliC and C-SliC.Protein purity was confirmed by sodium dodecyl sulfate-polyacryla mide gel electrophoresis (SDS-PAGE), and the protein concentration was measured using the Bio-Rad DC Protein Assay.The rSliC was purified as described previously (41).

Expression and purification of SC-AP205 cVLPs
AP205 cVLP, complementary to N-SliC and C-SliC, was expressed in E. coli One Shot BL21 Star (DE3) cells and purified via an OptiPrep (Sigma) step gradient as we previously described (42,62).

Coupling of N-SliC and C-SliC to AP205 cVLPs
We first established optimal coupling by mixing N-SliC and C-SliC with complementary SpyCatcher VLPs at a molar ratio of 1:3 for 2 h at room temperature in PBS.The uncou pled SliC protein was removed by ultracentrifugation onto an OptiPrep step gradient and dialyzed against PBS.Subsequently, SDS-PAGE and centrifugation were used to determine the coupling efficiency (by densitometry using ImageQuant) and stability of SliC-VLP complexes (42).The quality of SliC-VLP complexes was further investigated by dynamic light scattering (DynaPro NanoStar) and negative stain transmission electron microscopy (TEM) to ensure that particles are intact and monodisperse (42,62).For TEM, N-SliC-VLPs were adsorbed to 200-mesh carbon-coated grids, stained with 2% uranyl acetate (pH7.0), and analyzed with a Morgagni 268 electron microscope (42,62).

Enzyme-linked immunosorbent assays
Enzyme-linked immunosorbent assays (ELISAs) were performed as described previously (72) with minor modifications.Briefly, U-shaped high-binding 96-well microtiter plates (Nunc) were coated with purified N-SliC (125 ng/well) suspended in coating buffer (14 mM Na 2 CO 3 and 35 mM NaHCO 3 ) overnight at 4°C.Coated plates were blocked using Block Ace (Bio-Rad) dissolved in PBS containing 0.05% Tween-20 (PBST).Serum or vaginal lavage samples were serially diluted in PBST at varying starting dilutions (1:27 to 1:243) and added to each well.The wells were washed with PBST and incubated with diluted goat anti-mouse secondary antisera (1:10,000): total IgG, IgG1, IgG2a, IgG3, and IgA (SouthernBiotech) conjugated to horse radish peroxidase (HRP).Wells were washed, and reactions were developed using TMB Peroxidase EIA Substrate (Bio-Rad).End-point titers were determined using the average reading of eight wells incubated with secondary but no primary antibody plus 3 and 2 standard deviations as a baseline for serum and vaginal lavages, respectively.IgG1/IgG2a ratios were calculated based on group averages.For statistical analysis, Kruskal-Wallis test with Dunn's multiple comparisons was applied to non-transformed arithmetic data.For the comparison of the two groups, the non-parametric Mann-Whitney U test was carried out.For all analyses, P values of <0.05 were considered statistically significant.

Serum bactericidal assays
Sera from mice immunized with tested vaccines and control groups, as described in the text, were pooled and heat-inactivated for 30 min at 56°C (37,73,74).Subsequently, the sera were serially diluted in Hanks balanced salt solution (HBSS) at twofold dilutions starting from 1:64 to 1:8,192.The Ng FA1090 cells [2 × 10 4 colony forming units (CFU/ mL)] were prepared from non-piliated colonies collected from chocolate agar plates and suspended in HBSS to OD 600 ~0.1 (1 × 10 8 CFU/mL).The Ng cells (1 × 10 3 in 40 µL) were added to wells containing test sera, mixed by shaking for 15 s, and incubated at 37°C and 5% CO 2 atmosphere for 15 min before adding 10 µL of IgG/IgM-depleted normal human serum (NHS) or heat-inactivated NHS (HI-NHS) as the complement source (10%, vol/vol).Samples were incubated for an 1 h at 37°C with 5% CO 2 .Final suspensions (5 µL) were spot-plated onto chocolate agar plates and incubated overnight at 37°C and 5% CO 2 for 18-20 h for CFU enumeration.Controls included Ng alone, Ng incubated with NHS (Pel-Freez), Ng with HI-NHS (Pel-Freez), and bacteria incubated with test sera and HI-NHS.The average percentage of killing was determined from three independent experiments by calculating the differences in the number of CFUs recovered from Ng incubated with test sera and NHS and the number of CFUs recovered from Ng incubated with test sera and HI-NHS.

Lysozyme activity assay
To determine if the addition of SpyTag affects the SliC-mediated inhibition of c-type human lysozyme (HL, Sigma), we used the EnzChek Lysozyme Assay Kit (Thermo Fisher) as described previously (41).The lysozyme assay was carried out in black flat-bottom 96-well plates.Samples containing 2.5 µM HL (Sigma) were incubated with increasing concentrations of rSliC-STN (0-5 µM) in reaction buffer comprising of 0.1 M sodium phosphate pH 7.5, 0.1 M NaCl, and 2 mM sodium azide for 30 min at 37°C.The controls contained HL alone.After incubation, the reaction was initiated by the addition of DQ lysozyme substrate.The reaction was monitored for 20 min using a Synergy HT Microplate Reader (BioTek) at excitation and emission wavelengths of 485 and 530 nm, respectively.
To examine if immunization with N-SliC/ACP elicited antigen function-blocking antibodies, N-SliC/ACP was incubated with pooled sera from immunized and control groups (1:10, vol/vol) for 30 min and the lysozyme assays were carried out as described above.

SDS-PAGE and immunoblotting
Samples, as indicated in the text, were standardized by OD 600 values (whole-cell lysates) or by protein concentration and separated by SDS-PAGE on 4%-15% Bio-Rad Criterion TGX (Bio-Rad).The whole-cell lysates were prepared from Ng strains grown concurrently on solid GC medium for 20 h in 7% CO 2 at 37°C, collected, and lysed in Laemmli SDS buffer as described previously (39,72,75).Proteins were visualized by Colloidal Coomassie staining or were transferred onto Trans-Blot Turbo nitrocellulose 0.2 µM membranes (Bio-Rad) using a Trans-Blot Turbo transfer system (Bio-Rad).Membranes were incubated overnight in PBST supplemented with 5% non-fat dry milk, washed with PBST, and probed with pooled sera (1:5,000) or vaginal lavages (1:50) from test or control mice followed by probing the immunoblots with goat anti-mouse IgG (Bio-Rad) or IgA (SouthernBiotech) conjugated to HRP (1:10,000 dilution) as described previously (3).Cross-reacting proteins were detected using ECL Prime (Amersham) and ImageQuant LAS 4000 (GE Healthcare).

Statistical analysis
Statistical analyses were performed with GraphPad Prism 9 as indicated for each experimental procedure.

Design of the Tag/Catcher-AP205 platform for gonorrhea vaccine develop ment
To explore the Tag/Catcher system and the Acinetobacter phage AP205 cVLP for GC vaccine development, we first designed gene blocks carrying SpyTag on the N-or C-terminus (STN and STC, respectively), a linker, multicloning site, TEV protease cleavage site, and a 10× His Tag (Fig. 1A) and cloned them into pET22 vector (Fig. 1B).This newly engineered pET22-STN and pET22-STC system enables cloning and production of a selected antigen with the SpyTag placed on either N-terminus or C-terminus to ensure optimal antigen folding for purification and coupling to the cVLP (Fig. 1C).In each case, an E. coli PelB signal sequence is also added to promote proper antigen folding in heterologous host.We selected the SliC antigen from Ng FA1090, a strain that carries the most broadly distributed SliC antigen variant (40), and cloned sliC (NGO1063) into pET22-STN and pET22-STC (Fig. 1B).Both SliC fusion proteins with SpyTag positioned on the N or C terminus, N-SliC or C-SliC, respectively, were successfully overproduced in E. coli and migrated on the SDS-PAGE according to the predicted molecular weight of ~15 kDa (Fig. 2A).Highly pure N-SliC and C-SliC were obtained via affinity chromatog raphy steps, removal of cleavable His 10 -tag by TEV protease, followed by size exclusion chromatography (Fig. 2B, Lanes 2 and 6).

Assembly of the Tag/Catcher SliC-VLP vaccines
To generate Tag/Catcher SliC-VLP vaccines, individual N-SliC and C-SliC proteins were combined with the AP205 capsid protein carrying the complementary Catcher (Fig. 1D  and 2B).Quality assessment of the two different SliC-VLP complexes by SDS-PAGE revealed the covalently coupled C-SliC-VLP and N-SliC-VLP (migrating at ~45 kDa) along with uncoupled cVLP, SliC-STC, and SliC-STN (Fig. 2B).Excess amounts of uncoupled cVLP were observed in a reaction containing C-SliC as compared to the reaction containing N-SliC.Additionally, the amount of coupled C-SliC-VLP was lower in comparison to the N-SliC-VLP (Fig. 2B).These results revealed that N-SliC coupled onto the cVLP to a higher efficiency.Albeit there was similar intensity between C-SliC-VLP and N-SliC-VLP bands before (−) and after (+) centrifugation, demonstrating that both vaccine formulations are stable.Dynamic light scattering of the uncoupled cVLP-Catcher, C-SliC-VLP, and N-SliC-VLP showed that all samples are monodisperse with a peak around the expected particle size (30-40 nm), indicating that the core cVLP and SliC-VLP complexes are not aggrega ted.As expected, coupling of SliC antigen to the cVLP caused a slight increase in the diameter of the particles of 47.7 and 36.9 nm for C-SliC-VLP and N-SliC-VLP, respectively, compared to 20.8 nm of cVLP (Fig. 2C).The C-SliC-VLP had a higher polydispersity (28.7%) compared to N-SliC-VLP (11.7%) and cVLP alone (12.1%), indicating that the C-SliC-VLP population is more heterogeneous and that the coupling is not optimal.Due to the higher coupling efficiency (77%) with 140 subunits out of 180 per cVLP carrying the SliC antigen, we selected N-SliC-VLP for further studies.TEM confirmed that N-SliC-VLP contains intact and monodisperse particles (Fig. 2D).To evaluate whether the addition of STN affects SliC inhibitory activity of HL, we performed titration reactions with increasing concentrations of purified N-SliC in the presence of peptidoglycan (Fig. S1A).Similar to untagged, recombinant SliC (41), N-SliC inhibited the lytic activity of HL in a dosedependent manner, with complete blocking of lysozyme function at concentrations above 1.25 µM.These evaluations showed that N-SliC antigen is functional and couples more efficiently to the cVLPs, and thus, vaccine formulation containing N-SliC-VLP was selected for further immunization studies.

cVLP significantly enhances SliC immunogenicity and serum bactericidal activity and promotes a Th1 response
To assess SliC as a gonorrhea vaccine antigen and the impact of presenting SliC on the cVLP, we immunized mice with N-SliC-VLP, N-SliC, or cVLP using three SC injections at 3-week intervals.Balanced Th1/Th2 responses are considered optimal for some vaccines (76), and therefore, in the first assessment of SliC as an antigen, we adjuvanted all treatments with AddaVax-a squalene-based oil-in-water emulsion formulation similar to that of MF59, which is licensed in Europe for flu vaccines-which elicits both cellular Th1 and humoral Th2 responses (77).
To examine the specificity of immune responses elicited by the N-SliC and N-SliC-VLP vaccines, we performed an immunoblotting analysis with either purified N-SliC or a panel of whole-cell extracts obtained from different Ng strains including the 2016 WHO reference isolates (Fig. 4).SliC-specific IgG and IgA were detected in serum from mice immunized with N-SliC-VLP (Fig. 4A and B, respectively), while in contrast, no signal was observed after blotting the membranes with serum or vaginal washes from N-SliC or cVLP-immunized mice.Vaginal IgG but not IgA obtained from mice that received N-SliC-VLP readily recognized purified N-SliC (Fig. 4B).Furthermore, the N-SliC-VLP vaccine elicited systemic IgG that recognized native SliC in the isogenic strain FA1090, the complemented ΔsliC/P::sliC, and, importantly, in the 14 diverse strains of the 2016 panel of WHO isolates and Ng FA6146.As expected, no signal was detected in the ΔsliC knockout strain (Fig. 4C).
Finally, we sought to assess whether SliC-containing vaccines induce functional antibodies by examining serum bactericidal killing (SBA) and interference with SliC function as a lysozyme inhibitor (41).We used pooled murine sera from each examined group in combination with IgG-and IgM-depleted NHS to assess SBA titers against the serum-resistant Ng FA1090 (37,73,74).Human complement-dependent SBA titer (50% Ng killing) was greater than or equal to fourfold in mice that received the N-SliC-VLP vaccine compared to N-SliC and cVLP groups, respectively (Table 1).
To evaluate if the anti-SliC antibody interferes with SliC function, we examined HL activity in the presence of purified SliC using a fluorescein-labeled peptidoglycan (41,79).The lytic activity of HL remained blocked in the presence of SliC regardless of whether the antigen was pre-incubated with decomplemented sera obtained from mice immunized with N-SliC-VLP, N-SliC, or cVLP (Fig. S1C).In contrast, rabbit sera against another Ng lysozyme inhibitor, ACP, restored lysozyme activity, confirming that ACP elicits functional blocking antibodies (80) whereas SliC remained active against HL in the presence of rabbit anti-SliC (Fig. S1B and C).

SC and IN administration of N-SliC-VLP vaccine formulated with AddaVax or CpG
To promote both systemic and mucosal immune responses, our next choice for the immunization route was a SC prime followed by IN boost.Administration of the MetQ-CpG vaccine in a similar manner induced a protective immune response and IgA at the vaginal mucosae (81).We also sought to examine the impact of vaccine dosage (2.5-10 µg/dose) and formulation on antibody responses.In addition to AddaVax, we adjuvanted the N-SliC-VLP vaccine with CpG ODN based on the enhanced Ng clearance from the murine lower genital tract after immunization with antigens administered with Th1-inducing adjuvants including MetQ-CpG, MtrE-CpG, the lipooligosaccharide (LOS) 2C7 epitope mimic with MPLA, and Ng OMVs-IL-12 or viral replicon particles with PorB (81)(82)(83)(84)(85)(86).
Following the immunizations, the immune responses to N-SliC, N-SliC-VLP-AddaVax, and N-SliC-VLP-CpG vaccines were examined by assessing the reactivity of individual murine antisera against purified N-SliC in ELISA (Fig. 5).There was a significant increase in systemic antibody titers at the two data points (Days 31 and 52) for total IgG, IgG1, IgG2a, IgG3, and IgA in all N-SliC-VLP vaccine experimental groups regardless of vaccine dose in comparison to the control groups that received PBS, cVLP-AddaVax, or cVLP-CpG (Fig. 5A through D).Similarly, to our previous study (Fig. 3 and 4), N-SliC failed to induce significant antibody titers although administration of N-SliC led to 1,000-to 100-fold rise in the SliC-specific IgG (1.01 × 10 4 geometric mean, final sera) in comparison to controls (0.001 × 10 3 for PBS, 0.002 × 10 3 for cVLP-AddaVax, and 0.01 × 10 3 for cVLP-CpG; Fig. 5A).Titers were insignificant and negligible for IgG1, IgG2a, IgG3, and IgA (Fig. 5B through E, respectively).Importantly, our vaccine dose studies showed that the N-SliC-VLP-CpG (10 µg/dose) elicited the highest total IgG titers in the final sera with the geometric mean of 1.15 × 10 7 , which was 4.6-, 1.7-, 22-, and 14.7-fold greater than the titers induced by the same vaccine at 2.5 and 5 µg/dose, and the N-SliC-VLP-AddaVax at 2.5 and 10 µg/dose, respectively (Fig. 5A).The serum IgG1 levels were the greatest in mice that received N-SliC-VLP at 10 µg/dose formulated with AddaVax or CpG and reached 2.3 × 10 5 and 1.98 × 10 5 , respectively (Fig. 5B).Statistical analyses revealed no significant differences in IgG2a titers in murine sera between the different vaccine groups; however, the titers slightly declined in final sera in mice administered with vaccines formulated with CpG in comparison to the respective samples obtained on Day 31 (Fig. 5C).The IgG1/IgG2a ratios were 0.22, 1.46, and 0.19, 0.19, and 0.96 in mice that received N-SliC-VLP-AddaVax at 2.5 and 5 µg/dose and N-SliC-VLP-CpG at 2.5, 5, and 10 µg/dose, respectively.These results suggest that the highest tested N-SliC-VLP vaccine doses formulated with AddaVax/CpG elicited more balanced Th1/Th2 a Pooled antisera obtained from mice groups that were immunized as outlined above and, in the text, and the corresponding sham-immunized controls were tested for their ability to induce SBA killing of Ng FA1090.The data represent the reciprocals of the highest serum dilution at which ≥50% killing was noted.The titers are expressed as the median values from biological replicate experiments (n = 3), and the values in parentheses designate the SBA titers.
As expected from previous investigations that used the mucosal immunization route (81,87), our ELISA experiments showed a significant increase in titers of IgG and IgA in vaginal lavages in mice administered with all vaccine formulations in comparison to SliC alone and all control groups (Fig. 5F).Corroborating the ELISA findings, SliC-specific IgG and IgA were detected in serum and vaginal lavages derived from mice administered with N-SliC-VLP vaccines formulated with AddaVax/CpG as shown by immunoblotting in Fig. 6A and B, respectively.Furthermore, all vaccine doses and formulations elicited systemic IgG that cross-reacted with SliC protein in whole-cell extracts of the 2016 panel of Ng WHO isolates and FA6146 (Fig. 6C).Specificity of the elicited antibodies was apparent by the absence of reactivity of any of the anti-SliC sera with the corresponding Ng FA1090 ΔsliC strain and the lack of reactivity of PBS-and cVLP-AddaVax/CpG-immu nized sera with SliC in isogenic wild-type bacteria and the complemented ΔsliC/P::sliC (Fig. 6).
On the basis of the IgG3 titers (Fig. 5D), we performed the human complementdependent SBA with pooled murine sera obtained from animals immunized with the N-SliC-VLP-CpG/AddaVax vaccines at 10 µg/dose.Reciprocal sera dilutions at which 50% Ng killing was observed were fourfold to eightfold higher for N-SliC-VLP-CpG in comparison to bactericidal titers generated by immunization with N-SliC, controls (PBS, cVLP-CpG/AddaVax), and N-SliC-VLP-AddaVax (Table 1).The lack of SBA activity in mice that received N-SliC-VLP-AddaVax could be attributed to several factors including different immunization routes and dosing schedule, as well as waning IgG3 titers.Indeed, sera from mice immunized intramuscularly with NspA vaccine had the strongest bactericidal activity against Ng followed by sera from the IN and intravaginal immuniza tion cohorts (88).Furthermore, in humans, detergent-extracted meningococcal OMVs vaccine induced significantly lower SBA levels when administered IN than via intramus cular route (89).
Finally, we assessed whether vaccination with N-SliC-VLP-AddaVax/CpG given via SC and IN route induces antibodies that block the SliC function as a lysozyme inhibitor (Fig. S1D).Similar to previous immunization studies (Fig. S1C), the lytic activity of HL against peptidoglycan was blocked in the presence of purified SliC irrespective of the presence of anti-SliC sera (Fig. S1D).
Together, these studies showed that N-SliC-VLP-AddaVax/CpG administered SC followed by IN boost, induced SliC-specific systemic and mucosal IgG and IgA.Further more, IgG1/IgG2a ratios depended on vaccine dosage with the lower doses (2.5-5 µg) corresponding to bias toward Th1 response and the higher doses (10 µg) eliciting a more balanced Th1/Th2 response.The higher vaccine doses resulted in greater IgG1 and IgG3 titers.The data suggest that the N-SliC-VLP-CpG (10 µg/dose) is a promising vaccine for follow-up studies due to its ability to induce the most significant increase in total serum IgG and IgG3 titers and functional antibodies with SBA activity.

DISCUSSION
VLPs have gained mounting recognition as promising nanotools for vaccine develop ment due to their safety and ability to provide strong and long-lasting immune responses even after one administration (51,90,91).Several VLP vaccines have been developed and commercialized: Recombivax (Merck) and Engerix (GlaxoSmithKline) for hepatitis B (92,93), Hecolin (Innovax) for hepatitis E (94), and more relatedly to STDs, Gardasil (Merck), and Cervarix (GlaxoSmithKline) for HPV (45,95,96).Clinically tested VLP vaccines have been extended to influenza A, chikungunya virus, human cytomega lovirus, and human norovirus and in SARS-CoV-2 using the AP205 platform (52,97,98).Protein and peptide antigens are frequently displayed on VLPs either through the genetic fusion of epitopes to the self-assembling coat protein or through chemical conjugation to the surface of pre-assembled VLPs.These strategies have their draw backs, for example, limitation on antigen size, low-density coupling, interference with VLP assembly, and narrow, epitope-specific antibody responses (45,53,58).The Tag/ Catcher-AP205 cVLP platform composed of peptide counterparts SpyTag and SpyCatcher that form irreversible spontaneous isopeptide bond circumvent these challenges (56).The Acinetobacter phage AP205 is an attractive VLP vaccine backbone due to its unique structure in that both the N-and C-termini are surface exposed and evenly distributed on the assembled VLP, allowing for genetic fusions at both termini while maintaining stable assembly.Additionally, AP205 is an attractive VLP vaccine backbone due to intrinsic immunogenicity, lack of pre-existing immunity in humans, manufacturability, and cost-effective production in E. coli (42,45,(99)(100)(101).
Currently, no licensed gonorrhea vaccines exist, and the foci of attention in the field are on outer membrane vesicles, protein subunit vaccines, and a peptide mimic of a glycan epitope of Ng LOS 2C7 as vaccine candidates (102).As discussed above, protein subunit vaccines can greatly benefit from VLP display, but, to date, this platform has not been explored in the gonorrhea field.Herein, we showed the premise of this approach by using the Tag/Catcher-AP205 cVLP for delivery of a novel gonorrhea antigen SliC, testing different vaccine formulations, doses, and immunization routes.SliC is a remarkably conserved surface lipoprotein that contributes to bacterial defense against the host innate immune system effector, lysozyme (41).We designed and engineered two plasmids that enable expression and purification of different Ng antigens fused with SpyTag on the N-or C-terminus (Fig. 1).Furthermore, we determined that both N-and C-SliC could be efficiently overproduced and purified; however, the C-terminal SpyTag hampered the assembling efficiency of SliC to VLPs (Fig. 2A and B, respectively).In contrast, the N-SliC antigen variant readily loaded onto AP-205 cVLP forming stable multivalent complexes and retained its function as an HL inhibitor (Fig. 2B through D; Fig. S1B, respectively).
Our investigations showed that vaccines containing monomeric N-SliC failed to induce SliC-specific antibodies when administered alone/AddaVax via different immunization routes in mice (Fig. 3 to 6).The recombinant SliC protein emulsified in Freund's complete adjuvant and administered SC elicited immune responses in rabbits that recognized the purified SliC variant and native SliC in diverse panel of Ng isolates (41).The limited immunogenicity displayed by SliC in mice could be caused by loss of its conformational stability in PBS/AddaVax formulation that affects antigen process ing in antigen-processing cells from its uptake to the final presentation of antigenic peptides.Destabilization of a protein's three-dimensional structure can lead to the loss of individual epitopes or redirection of antibody responses to irrelevant epitopes that are not favored in the wild-type conformation.The presence of adjuvants may affect protein folding and even lead to protein denaturation (103).In contrast to monomeric N-SliC vaccines, the multivalent repetitive and particulate display of N-SliC via the Tag/ Catcher-AP205 cVLP (Fig. 2) significantly potentiated its immunogenicity as shown by increased kinetic of antibody responses markedly induced antibody titers in ELISA, serum and vaginal SliC-specific IgG and/or IgA, and functional antibodies with SBA activity (Fig. 3 to 6; Table 1).
The ability of Ng to evade adaptive immune responses and the lack of established correlates/mechanisms of protection in mice or humans are long-standing barriers in the gonorrhea vaccine field (35).Work by Liu et al. has suggested that Th1 responses are important for the efficacy of an intravaginal vaccine formulation composed of Ng OMVs in combination with encapsulated poly-lactic acid microspheres containing IL-12, a Th1-inducing cytokine (104).Our data suggest that the dose and delivery route of the N-SliC-VLP vaccine affect Th1 and Th2 responses as assessed by IgG1/IgG2a ratios (Fig. 3 and 5).The N-SliC-VLP-AddaVax administered via 3× SC elicited Th1-biased immune response whereas the same vaccine formulation and dose delivered by SC, IN routes induced Th1/Th2-balanced responses.Similarly, PorB DNA-based vaccine induced both Th1 and Th2 responses, depending on the method of delivery (105).Although the function of antibodies in protection against Ng is not fully elucidated, several lines of evidence suggest that antibodies may play a crucial role.In addition to Th1 respon ses, the Ng OMVs-IL-12 vaccine depends on B-cell production of antibodies (85).The efficacy of the LOS mimitope-based vaccine (TMCP2) requires IgG and complement (106), as does the efficacy of monoclonal antibodies against this LOS 2C7 epitope vaccine requires complement (73).IgG is crucial for many effector functions of bacterial vaccines, including the blocking of function (e.g., tetanus-diphtheria vaccine) and through direct SBA (e.g., Neisseria meningitidis) (107,108).Unlike the gastrointestinal tract, where IgA predominates, the IgG is the most abundant Ig in the female genital tract (109).Passively administered antibodies protect against HPV (108).Serum IgG, as generated by systemic immunization, is readily transported into the genital tract through FcRn (110).GC infections primarily localize to mucosal surfaces; thus, boost of mucosal IgA may be important for protection against Ng infection.Indeed, MetQ-CpG and Ng OMVs-IL-12 administered via IN route elicited vaginal IgA and accelerated Ng clearance from the murine lower genital tract (72,87).
We conclude that the N-SliC-VLP-CpG (10 µg/dose) administered via SC and IN is a promising vaccine for follow-up studies due to its ability to elicit systemic and mucosal IgG and IgA, boost the serum IgG and IgG3 responses, and induce functional antibodies with SBA activity.animal facilities meet the housing service and surgical standards set forth in the "Guide for the Care and Use of Laboratory Animals, " NIH Publication No. 85-23.Animals are maintained under the supervision of a full-time veterinarian.For all experiments, mice were humanely euthanized upon reaching the study endpoint using a compressed CO 2 gas cylinder, as per the institution guidelines, which follow those established by the 2020 American Veterinary Medical Association Panel on Euthanasia (https://www.avma.org/resources/pet-owners/petcare/euthanasia).

ADDITIONAL FILES
The following material is available online.

Full 6 FIG 1
FIG 1 Design of the SpyTag plasmid system to explore the Tag/Catcher-AP205 cVLP platform for gonorrhea vaccine development.(A) An outline of in silico designed genetic engineering process to develop SpyTag plasmid system using pET-22b(+), sliC from Ng FA1090, and gene blocks carrying SpyTag on the N-or C-terminus (STN and STC, respectively), a linker, multicloning site, TEV protease cleavage site, and a 10× His Tag. (B) Gibson assembly was used to clone STN and STC gene blocks to pET-22b(+) yielding pET22-STN and pET22-STC that enable to fuse antigens with the SpyTag on either N-or C-terminus, respectively.E. coli PelB signal sequence is also added to promote proper antigen folding in heterologous host.The sliC gene was cloned into pET22-STN and pET22-STC to create pET22-N-SliC and pET22-C-SliC.(C) The core AP205 cVLP displaying a complementary Catcher (SpyCatcher-VLP) is expressed in E. coli and purified.The AP205 cVLP Catcher is mixed in solution with purified N-SliC or C-SliC.The Tag and Catcher rapidly react to form a spontaneous isopeptide bond leading to formation of N-SliC-VLP and C-SliC-VLP complexes.The figure was created with BioRender.com.

FIG 2
FIG 2 Assembly and quality assessment of the Tag/Catcher SliC-VLP vaccines.(A) E. coli BL21(DE3) carrying pET22-N-SliC and pET22-C-SliC (N and C, respectively) were cultured without (−) and with (+) IPTG.The whole-cell extracts were normalized by OD 600 , separated by SDS-PAGE, and stained with Colloidal Coomassie.Both SliC fusion proteins (N-SliC and C-SliC) were overproduced (pink arrow) and migrated on the SDS-PAGE according to the predicted molecular weight of ~15 kDa.(B) Purified AP205 cVLP alone (cVLP), C-SliC, N-SliC, and reaction mixtures after incubation of the cVLP and either variant of SliC (C-SliC-VLP and N-SliC-VLP) were separated by SDS-PAGE and stained with Coomassie.The covalently coupled C-SliC-VLP and N-SliC-VLP were observed together with an excess of uncoupled cVLP and SliC.Comparison of the intensity of the conjugated SliC-VLP bands before (−) and after (+) centrifugation was used to assess the aggregation state/stability of the vaccine formulation.(C) Dynamic light scattering of the uncoupled cVLP (in pink), the C-SliC-VLP (in lavender), and N-SliC-VLP (in blue).(D)N-SliC-VLPs were adsorbed to 200-mesh carbon-coated grids, stained with 2% uranyl acetate (pH7.0), and analyzed with an accelerating voltage of 80 kV, using a CM 100 BioTWIN electron microscope.

FIG 3
FIG 3 The N-SliC-VLP vaccine adjuvanted with AddaVax markedly induced antibody titers compared to corresponding vaccine containing monomeric N-SliC.(A) Systemic total IgG titers were examined in pre-immune (P) and 10 days after first (1), second (2), and third (3) subcutaneous administration of cVLP, N-SliC, or N-SliC-VLP (n = 5 mice/group).All treatments were adjuvanted with AddaVax.(B) Total IgG, IgG1, IgG2a, IgG3, and IgA antibody titers in final sera from mice immunized with cVLP, N-SliC, or N-SliC-VLP.(C) Vaginal total IgG titers in pooled vaginal washes were examined in pre-immune (P) and 10 days after first (1), second (2), and third (3) subcutaneous administration of cVLP, N-SliC, or N-SliC-VLP.(D) Total IgG, IgG1, IgG2a, and IgA titers in final pooled vaginal lavages obtained from mice immunized with cVLP, N-SliC, or N-SliC-VLP.Bar graphs represent geometric mean ELISA titers against N-SliC with error bars showing 95% confidence limits.Statistical significance was determined using Kruskal-Wallis with Dunn's multiple comparison test.For the comparison of two groups, the non-parametric Mann-Whitney U test was carried out.For all analyses, *P < 0.05.

FIG 4
FIG 4 SliC-specific systemic IgG and IgA and vaginal IgG were elicited by subcutaneous immunization with N-SliC-VLP vaccine adjuvanted with AddaVax.Female mice (n = 5) were subcutaneously immunized with cVLP, N-SliC, or N-SliC-VLP adjuvanted with AddaVax.Purified N-SliC (A and B) and whole-cell extracts obtained from the isogenic Ng strain FA1090, the ΔsliC knockout, the complemented ΔsliC/P::sliC, and a panel of geographically, genetically, and temporally diverse Ng isolates were fractionated by SDS-PAGE.Immunoblotting was performed with pooled serum (A and C) and vaginal washes (B) collected after the third immunization, followed by secondary antibodies against mouse IgG (A and C) or IgA (A and B).

TABLE 1
Human serum bactericidal activity of pooled murine antisera to N-SliC and N-SliC-VLP vaccines delivered with different adjuvants a