Generation of recombinant gonococcal strains and GMMA thereof

The parent isolate for preparation of the vaccine strain was FA1090 derived from a patient with disseminated invasive gonorrhea⁷⁶. The lpxL1, rmp and lgtF genes of FA1090 were respectively deleted through homologous recombination with pBS-ΔlpxL1 kanR⁷⁷, pBS-Δrmp eryR and pBS-lgtF::cmR constructs containing upstream and downstream regions of the genes flanking an antibiotic resistance cassette. All polymerase chain reactions (PCR) were performed with the KAPA Hifi 2X master mix (Roche) and primers reported in Supplementary Table 3. For the generation of the pBS-Δrmp eryR construct, amplification of the up and downstream regions of the rmp gene was performed with the primer couples UpIII-FOR/REV and DpIII-FOR/REV using 50 ng of genomic DNA of the FA1090 strain as template. Amplification of the erythromycin resistance gene (eryR) was done with the primer couple EryR_gono_SmaI-Fw/Rev using as template 10 ng of a plasmid carrying the gene. PCR products were purified using the kit Wizard SV Gel and PCR Clean-Up System (Promega), according to the manufacturer’s instructions, and cloned as XbaI-SmaI (UP rmp), SmaI (eryR) and SmaI-XhoI (DOWN rmp) into a pBluescript vector digested with XbaI-XhoI (NEB). Amplification of upstream and downstream regions of the lgtF gene was performed with the primer couples lgtF-UP-fw/rv and lgtF-DO-fw/rv, respectively, using 50 ng of genomic DNA as template; the amplification of the chloramphenicol resistance gene (cmR) was done with the primer couple cloKOF/R using 10 ng of a synthetic DNA template (Geneart). PCR products were transformed into E. coli MACH-1 competent cells (Thermo Scientific) immediately following amplification. Screening of transformants was done by colony PCR using T7prom/pETseqRv primers. The PCR product amplified using the primers Lpx UP Fwd and LpxL1 DO Rev and the plasmid pBS-ΔlpxL1 kanR as a template, the pBS-Δrmp eryR linearized plasmid or the PCR product of the positive lgtF clone were used for the transformation of the FA1090, the FA1090 ΔlpxL1 or the FA1090 ΔlpxL1Δrmp strain, respectively. Transformations were carried out by spotting a mixture of 30 µl of bacterial resuspension in phosphate-buffered saline (PBS) and 30 µl of DNA onto a GC agar +1% isovitalex plate for 5–6 h and transformants were selected on GC agar +1% isovitalex plates with either kanamycin 40 µg/ml, erythromycin 2 µg/ml or chloramphenicol 2 μg/ml. All transformants were tested by PCR analysis using Accuprime Taq Polymerase (Thermo Scientific) and with external primers (primer couples LpxL1 est FW/REV, UP_CHECK_NGO1577-Fw/DW_CHECK_NGO1577-Rev, and lgtF-ext-F/R for ΔlpxL1, Δrmp and ΔlgtF, respectively) to check the correct event of double recombination. The genome of the final FA1090 ΔlpxL1Δrmp vaccine producer strain was obtained using both Illumina and PacBio sequencing and checked with respect to regions relative to the gene deletions.

To generate small scale GMMA or nOMV preparations, cell-free supernatants were recovered by centrifugation at 8000 × g for 15 min at 4 °C from overnight cultures of the gonococcal WT or mutant strains in Meningitis Chemically-defined Medium I (MCDMI-mod, composed as follows: 15 g/L Soy peptone (BBL Phytone), 5.8 g/L NaCl, 2.56 g/L MgSO₄–7H₂0, 5.24 g/L K₂HPO₄–3H₂0, 3 g/L L-Glutamic Acid, 0.20 g/L L-Arginine, 0.5 g/L L-Serine, 0.3 g/L L-Cysteine, 0.25 g/L L-Glycine, 0.01 g/L Fe(III) Citrate, 0.02 g/L CaCl₂, 5 g/L Na-(DL)Lactate, 0.34 g/L Betaine-H20, 0.5 g/L Vitamin Mix(5X) consisting of 0.5 g/L Thiamine, 0.5 g/L Riboflavin, 0.5 g/L Pyridoxine, 0.5 g/L Niacinamide) at 37 °C and filtered through 0.2-μm Sartobran P H9 filter to ensure bacteria removal. After filtration, MgCl₂.6H₂O was added to a final molarity of 1 mM and 50 U/L of benzonase was added for DNA removal and incubated at 4–8 °C with stirring overnight. A tangential flow filtration step using 300 K Sartocon slice cassettes was used for retention of the GMMA and buffer exchange in PBS, followed by an ultracentrifugation step (30,000 × g for 15 h) to pellet the GMMA which were subsequently resuspended in PBS and sterile-filtered using a 0.2 µm filter, and the product was stored at –20 °C.

To generate the three batches of Ng GMMA from FA1090 ΔlpxL1Δrmp strain Ng GMMA 1, Ng GMMA 2 and Ng GMMA 3, medium scale preparations were performed. One Erlenmeyer flask containing MCDMI-mod, pH 6.9 was inoculated with a vial of strain FA1090 ΔlpxL1Δrmp seed preparation and incubated at 37 °C, under agitation. When the culture reached exponential growth, a fermenter containing 1.5 L of MCDMI-mod medium was inoculated with a starting optical density at 590 nm (OD₅₉₀) equal to 0.3–0.4. The fermentation occurred under the following conditions: temperature 37 °C, air flow rate 1 vvm, pH 6.9 controlled through concentrate acid as titrant, dissolved oxygen tension (DOT) set point 30% controlled through impeller stirring speed. Upon depletion of carbon sources in the medium, the supernatant was recovered by centrifugation (8000 × g for 15 min at 4 °C) and filtered through 0.2 μm, to ensure bacterium removal. One mM MgCl₂ and 50 kU/L benzonase were added to hydrolyze DNA potentially released in the supernatant. The material was then purified through tangential flow filtration (300 KDa PES cassettes with around 50 kg supernatant/m² of membrane surface) and the diaretentate was eluted from a chromatography step, using a multimodal resin (CaptoCore 700) in buffer PBS pH 7.4. Total protein content was estimated by the Lowry method, and the product was stored at –70 °C after a final 0.2 μm filtration.

Characterization of GMMA

Total protein content of GMMA was estimated by the Lowry method. The purity of Ng GMMA and Ng GMMA ΔltgF were evaluated by sodium-dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and size exclusion chromatography module (SE-HPLC). In particular, SE-HPLC checked the purity of GMMA preparations, allowing determination of generally occurring contaminants such as soluble proteins (reported as fluorescence (FLR)) and DNA (reported as µg DNA/µg protein).

Multiangle light scattering (MALS) coupled with high-performance liquid chromatography-size exclusion chromatography (SEC) was used to assess GMMA particle radius. The LOS content in Ng GMMA was measured by SE-HPLC analysis by quantifying the reactive carbonyl groups of the saccharide moiety (OS) generated after acid hydrolysis to remove the lipid A and derivatized with semicarbazide, as previously reported⁷⁸. LOS was characterized by immunoblotting using murine mAb 2C7 as the antibody⁷⁹. Lipid A was extracted from Ng GMMA using mild-acid hydrolysis and analyzed with an Ultraflex matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometer (Bruker Daltonics) in negative-ion reflectron mode as previously reported⁷⁸.

Luciferase assay with hTLR4-transfected HEK293 NF-κB-reporter cell lines

HEK293-hTLR4-transfected adherent cells were maintained in Dulbecco’s modified Eagle medium (DMEM) supplemented with 4.5 g/l glucose and HEPES (GIBCO), 10% fetal bovine serum (FBS, Hyclone), 1% penicillin/streptomycin/glutamine, and specific antibiotics for the different cell lines: puromycin (5 µg/ml Invivogen), blasticidin (10 µg/ml; Invivogen), hygromycin (250 µg/ml, Roche). For the NF-κB luciferase assay, 25,000 cells/well were seeded in 90 µl of complete DMEM without antibiotics in 96-well μClear luciferase plates (Greiner Bio) and incubated for 24 h at 37 °C, 5% CO₂ before 10 µl of serial 3-fold dilutions of the different samples were added. After incubation for 5 h at 37 °C, supernatants were removed by aspiration, and cells were lysed for 20 min at room temperature (RT) using 20 µl/well of 1:5 diluted Passive Lysis Buffer (Promega). Eleven three-fold dilutions of tested samples were prepared in PBS from a serial dilution starting at 9 µg/ml for Ng GMMA, Shigella sonnei GMMA (kindly provided by the GSK Vaccines Institute for Global Health, Siena, Italy) and MenB OMV (detergent-extracted OMV from meningococcus NZ98/254 strain from commercial 4CMenB vaccine – GSK batch product), 1 µg/ml for FA1090 WT nOMV (native OMV). Ten microliters of serially diluted stimuli were added to the cells in a final volume of 100 µl (final concentration 0.9 and 0.1 µg/ml). A negative control (PBS only) and a positive control (USP Reference Standard Endotoxin, RSE) were included in each plate. Luciferase activity was detected using 100 µl/well of Luciferase Assay Reagent (Promega), and emitted light was immediately quantified using a Tecan Spark luminometer.

IL-6 release assay

Purified PBMC from healthy donors were plated in a 96-well plate at the concentration of 1.5 × 10⁵ cells/well in RPMI supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, 10 mM HEPES, minimum essential medium with non-essential amino acids, 2 mM L-glutamine and 1% (v/v) FBS. Three-fold dilutions of the same OMV and GMMA samples as in the previous assay were prepared in PBS from a starting concentration equal to 9 μg/ml. Amount of 20 μl of the stimuli were added to the cells in a final volume of 200 μl (final concentration 0.9 μg/ml). PBMC of the selected donors were incubated overnight with the stimuli at 37 °C in a humidified atmosphere containing 5% CO₂. Quantitative determination of IL-6 in PBMC supernatants stimulated with different samples was performed using an electrochemiluminescence immunoassay system from Meso Scale Discovery (MSD) platform according to the manufacturer’s instructions.

Immunogenicity in animal models

Animal husbandry and experiments were ethically reviewed and carried out in accordance with European Directive 2010/63/EU, in compliance with relevant guidelines (Italian Legislative Decree N. 26/2014) and the GSK’s policy and guidelines on the care, welfare and treatment of animals, in GSK animal facilities located in Siena, Italy (AAALAC accredited). The ethical protocol P004/26/01 was reviewed by the local GSK ethical committee. The study refers to the research project AWB 2020_05 approved by the Italian Ministry of Health.

Female CD1 mice, 7-week-old, were purchased from Charles River Laboratories and kept in a controlled environment (individually ventilated cages; 22 ± 3 °C; 12 h/12 h light/dark cycle). Our mice were commonly group-housed, so we only used females as they exhibit generally a reduced level of aggressiveness in medium to long housing conditions, which can affect scientific outcome⁸⁰. To guarantee a more heterogenous and representative immune response, CD1 mice, which are outbred mice, were used in this study, instead of the more classical Balb/c or C57BL/6 inbred mice. Animals and their housing and husbandry were checked daily, and their well-being and health status were recorded in a dedicated logbook according to the local standard operating procedures. Final bleeding was performed under general anesthesia and animals were euthanized by cervical dislocation before recovery from anesthesia.

Mice were immunized intraperitoneally on Days 1, 29 and 57 with 10 µg/mouse of either of the three different batches of Ng GMMA (NgG1, NgG2 and NgG3) adsorbed to aluminum hydroxide (alum); 3 mg/ml), with 4CMenB (10 µg/mouse in terms of OMV component), or alum alone (10 animals per group; 200 µl/mouse). Sera were collected on Days 0, 56 (4wp2) and 72 (2wp3) and vaginal washes were collected on Day 74. In a different study using the same protocol, to evaluate the potential effects of LOS, animals were immunized with NgG or FA1090Δlpxl1ΔrmpΔlgtF GMMA (10 µg/mouse) adsorbed to alum (NgGΔlgtF; 3 mg/ml), or alum alone (3 mg/ml).

Luminex immunoassay for specific IgG and IgA detection

Luminex Magplex beads were activated according to the manufacturer’s instructions and incubated for 2 h with 40 µg/ml of Ng GMMA diluted in 500 µl of 50 mM MES. After washing twice with PBS-0.05% Tween 20, beads were resuspended in 500 µl of PBS-0.05% Tween 20 + 0.5% bovine serum albumin (BSA) (assay buffer) and stored at 4 °C. Test samples and a standard mouse pooled sera were prediluted in assay buffer and then consecutive threefold dilutions were performed (50 µl/well) in a 96-well microplate (Millipore Corporation). Standard and blank controls were included in each plate. An equal volume of conjugated microspheres (3000 beads/well) was added to prediluted samples and incubated for 60 min at RT in the dark on a plate shaker at 700 rpm. After washing with PBS, phycoerythrin (PE)-labeled secondary antibody was added to reveal specific IgA (50 µl/well of 5 µg/mL R-PE goat anti-mouse IgA, Southern Biotech) and IgG (50 µl/well of a 2.5 µg/ml R-PE goat anti-mouse IgG, Fcγ fragment specific, Jackson Immunoresearch) and the plate was incubated for 60 min with shaking. After washing, the beads were suspended in PBS and analyzed with Bioplex 200 system. Data were acquired with the Bio-Plex Manager software version 6.2 (Bio-Rad Laboratories). For each analyte, the median fluorescent intensity was converted to relative light units (RLU)/ml by interpolation from a five-parameter logistic standard curve. The final titer of each sample was expressed as geometric mean of concentrations with recovery 75–125% respectively to the median concentration of all the interpolated concentrations for each sample.

Selection and characteristics of the strains used in functional assays

The rationale for the selection of relevant N. gonorrhoeae strains for the assay development and test was based on the analysis of the genetic variability of the outer membrane protein (OMP) components of Ng GMMA, with respect to a list of 59 of the more abundant OMPs through mass spectrometry analyses detection (listed in Supplementary Table 2). This list was used to define a multilocus typing schema (at protein sequence level) in a collection of more than 4000 sequenced genomes of gonococcal isolates and all strains were typed to assign protein allele identifiers and to measure a genetic distance. The resulting phylogenetic tree allowed the identification of 15 separate clusters (Supplementary Fig. 3). With the aim to present a hSBA panel representative of the general gonococcal population, single strains belonging to the most densely populated clusters (including clusters 3, 15, 4, 9 and 12) and multiple strains from the most genetically diverse cluster 1 were selected, also considering the existence of two PorB protein variants, 1a and 1b (Supplementary Table 4).

GC strains can be classified as serum-sensitive or serum-resistant depending on their susceptibility to killing by normal human serum (NHS). The strains tested in hSBA were evaluated for their complement sensitivity to non-immune human serum used as complement source during the development of the assay. Four out of eleven strains were classified as serum sensitive and therefore have been tested in hSBA after growth in medium containing CMP-NANA (which allows LOS sialylation and confers serum resistance to the bacterium. Supplementary Table 4 lists features of the different strains used in our functional assays.

A standard method for GC strains LOS immunotyping is not reported in literature. In order to determine the prevalent epitopes of LOS expressed by the selected GC strains, an immunochemical characterization of LOS was performed by western-blotting analysis on bacterial lysates grown in SBA-like condition. The following anti-LOS mAbs extensively described in literature were used: mAb 17-1-L1 (henceforth referred to as mAb L1), 4C4, L3,7,9 and 2C7. The tissue culture supernatant containing anti-LOS mAbs L1 was provided by Professor Sanjay Ram (Division of Infectious Diseases and Immunology, University of Massachusetts Medical School). Mouse mAb 2C7 was produced as recombinant internally. In particular the published variable region sequences were fused to the constant region of murine IgG2a⁶⁴. Purified mAb 4C4 and supernatant of mAb L3,7,9 are commercially available. Results are summarized in Supplementary Table 5. The data showed that the strain panel was also heterogeneous in terms of LOS structures.

Human serum bactericidal assay (hSBA)

Functional antibodies were measured by hSBA against the homologous FA1090 strain and the panel of heterologous strains described in the former section. Bacterial colonies from an overnight culture were resuspended in GC + 1% Isovitalex (CMP-NANA was added to the broth medium for serum-sensitive strains: 0.5 µg/ml for F62 and GC14 strains, 0.2 µg/ml for WHO-M and MS11 strains) and incubated at 37 °C with gentle shaking until the culture reached OD₆₀₀ = 0.5. The broth culture was then diluted 1:10,000 in SBA buffer (Dulbecco’s phosphate-buffered saline [dPBS] + 1% BSA + 0.1% glucose) with the exception of the BG27 strain that was diluted 1:2500. Mouse sera, previously heat-inactivated at 56 °C for 30 min, were serially diluted in SBA buffer. The assay was assembled in a sterile 96-flat bottom well microplate in a final volume of 32 µl/well. The serial dilutions of each test sample were let to react with pre-diluted bacteria and with human serum from healthy donors, screened for lack of activity against GC strains, as complement source (16% for FA1090, BG27, BG8, SK92-679, WHO-F, WHO-G, WHO-N and 10% for F62, MS11, WHO-M, GC14). Human serum was obtained from clinical study MENB REC 2ND GEN-074 (V72_92) after subjects written consent, according to Good Clinical Practice and to the declaration of Helsinki.

The reaction mixture was incubated at 37 °C for 60 min at 160 rpm, then agar overlay medium was added to each well and the plate was incubated overnight at 37 °C with 5% CO₂ in humid atmosphere. One day later, colony-forming units (CFUs) were automatically acquired with a high-throughput image analysis system (Discovery v12 Axiolab) and were automatically counted for each well by an image analysis system (Reading AxioVision). Bactericidal titer was defined as the reciprocal of the serum dilution giving a killing >50% respect to the average number of CFU calculated in replicates of ‘without serum’ control.

Bacterial adhesion inhibition assay (BAI)

A cell-based fluorescent BAI assay was used to assess the capacity of alum-, 4CMenB-, NgGΔlgtF– and NgG-immunized murine pooled sera (heat-inactivated) to inhibit the adhesion of homologous FA1090 and heterologous SK92-679 strains to Ect1 human ectocervical cells and SV-HUC1 human urethral epithelial cells.

Cultured cells were detached from 175 cm²-flask, after resuspension to avoid cell clumping. Cell number and viability were determined by an automated counter. They were then seeded into 96-well plates (3 × 10⁴ cells/well) and cultured in F-12K Nut Mix medium (21127-022, Thermo Fisher Scientific) for SV-HUC1 cell line or keratinocytes epidermal growth factor medium (10724-011, Thermo Fisher Scientific) + 20% FBS for Ect-1 cell line till confluence. GC strains were harvested from a fresh overnight plate culture into 10 ml GC + 1% Isovitalex medium. Bacteria were grown at 37 °C under shaking till OD₆₀₀ = 0.5, then resuspended in dPBS and labeled with Oregon Green dye for 15 min at 37 °C. Afterwards, bacteria were washed to remove excess of dye, resuspended in dPBS-1% BSA, and combined, at final OD₆₀₀ = 0.1, with an equal volume of serially cell medium-diluted sera for 15 min at RT. Bacteria-sera complexes were added to cell plates and incubated for 1 h at 37 °C to allow bacterium-cell adhesion. After three washes with dPBS, samples were fixed for 20 min with 4% formaldehyde at RT and, after one washing step with dPBS, finally one volume of distilled water was added to each well. Plates were analyzed by Opera Phenix instrument (Revvity).

Characterization of T-cell responses

Mice were immunized intraperitoneally on Days 1 and 29 with either 4CMenB (10 µg/mouse in terms of OMV protein content), NgG (10 µg/mouse), or alum alone (10 animals per group). Five mice/group were sacrificed 12 days after the second immunization and their spleens were collected and processed individually. Splenocytes were isolated and seeded in round-bottomed 96-well plates at a density of 1.5 × 10⁶ cells/well. They were then stimulated overnight at 37 °C in RPMI containing 10% FBS, 1% penicillin, streptomycin, glutamine, 100 mM β-mercaptoethanol and anti-CD28 (BD, 2 µg/ml) with medium alone, heterologous Salmonella GMMA (kindly provided by the GSK Vaccines Institute for Global Health, Siena, Italy), Ng GMMA, SK92-679 GMMA, heat-killed (HK) FA1090 strain, HK SK92-679 strain or OMV from 4CMenB vaccine (OMV). Brefeldin A (5 µg/ml) was added for the last 4 h of incubation. Cells were then washed with PBS, stained with Live/Dead near infra-red (Invitrogen) for 20 min at RT, fixed and permeabilized with Cytofix/Cytoperm (BD) for 20 min at RT, washed with PBS and stored at 4 °C in PBS-1% BSA until analysis. For intracellular staining, the cells were washed with Perm/Wash Buffer 1X (BD) and incubated for 20 min in the presence of anti-CD16/CD32 Fc block (BD) and finally stained for 20 min at RT with anti-CD3-APC (Biolegend), anti-CD4-BV510 (Biolegend), anti-CD8-PE-CF594 (Beckton-Dickinson), anti-CD44-V21 (Beckton-Dickinson), anti-IFN-γ-BV785 (Biolegend), anti-TNF-α-PE (Miltenyi-Biotec), anti-IL-2-PECy5 (Biolegend), anti-IL-17-PECy7 (e-Bioscience), anti-IL-13-PerCPeFluor710 (Invitrogen), and anti-IL-4-PerCPeFluor710 (Invitrogen). After staining, the cells were washed with Perm/Wash buffer 1X, once with PBS, and finally resuspended in 150 µl of PBS. Samples were acquired on a LSRII flow cytometer (BD Biosciences) and data were analyzed using FlowJo software (Treestar). T-cell subsets were identified as follows: Th17 were all CD4⁺ T cells expressing IL-17; Th1 were all CD4⁺ T cells expressing IFN-γ, but not IL-17 and IL-4/IL-13; Th2 were all CD4⁺ T cells expressing IL-4/IL-13 but not IFN-γ and IL-17; Th0 were all CD4⁺ T cells expressing IL-2 and TNF-α and combinations of IL-4/IL-13 and IFN-γ.

Statistical methods

The hSBA and Luminex data were evaluated using a one-way ANOVA model, considering group as a fixed factor and allowing for heterogeneous variances. Log2 and Log10 titers were analyzed respectively, with arithmetic means back-transformed into geometric mean titers (GMT) and differences in arithmetic mean, estimated from the same variance analysis model, into geometric mean ratio (GMR). Graphs were created showing individual values, geometric means, and 95% confidence intervals. For BAI assay data, it was checked if the response was at least 30% inhibition for each lot and test. Due to high variability, each lot was tested at least in two independent tests and additional tests were performed to strengthen the results when they appeared borderline. Data from all tests were analyzed by combining all results together and individually to evaluate consistency. Further, to combine all the tests, the mean of the response among the two replicates (within plate) at each dilution was computed for all the tests and data were analyzed all together. The dose-response curve linear part was identified, and the p-value of the regression was computed. The dose-response was considered significant if the p-value was equal or lower than 0.05. The overall mean of the responses at each dilution was compared to the 30% success criterion. For intracellular staining, immunization groups were compared with one-way ANOVA separately for each in vitro stimulation, and multiple comparison was performed with the Kruskal Wallis test.