Growth of bacterial strains

Neisseria strains were cultured in PPM medium (15 g l⁻¹ proteose pepton, 1 g l⁻¹ soluble starch, 5 g l⁻¹ NaCl, 4 g l⁻¹ KH₂PO₄, 1 g l⁻¹ K₂HPO₄, pH 7.5) for liquid culture and on GC plates (BD Difco, GC Medium Base), both supplemented with 1% IsoVitale vitamin mix (100 g l⁻¹ glucose, 10 g l⁻¹ glutamine, 26 g l⁻¹ l-cysteine, 100 mg l⁻¹ carboxylase, 250 mg l⁻¹ NAD, 500 µl l⁻¹ Fe(NO₃)₃, 150 mg l⁻¹ arginine, 3 mg l⁻¹ thiamine-HCl, 10 mg l⁻¹ vitamin B12, 13 mg l⁻¹ p-amino benzoic acid, 1.1 g l⁻¹ l-cystine, 1 g l⁻¹ adenine, 500 mg l⁻¹ uracil, 30 mg l⁻¹ guanine). Strains were either cultivated at 37 °C and 220 rpm (liquid medium) or at 37 °C and 5% CO₂ (solid medium). Opa protein expression by gonococcal strains and Opa-mediated binding to human CEACAMs was determined essentially as previously decribed⁴³.

Escherichia coli strains, P. aeruginosa and K. pneumoniae were cultured in LB medium (10 g l⁻¹ tryptone, 5 g l⁻¹ yeast extract, 5 g l⁻¹ NaCl, 15 g l⁻¹ agar for plates; pH 7.0). P. aeruginosa PAO1 mutants were obtained from Colin Manoil (University of Washington, Genome Sciences, Seattle, WA).

Lactobacillus strains used in this study were grown on Columbia blood agar (42.5 g l⁻¹ Columbia agar base, 5% defibrinated horse blood) for 24–48 h at 37 °C with 5% CO₂ or in prereduced MRS liquid medium (10 g l⁻¹ casein peptone, 10 g l⁻¹ meat extract, 5 g l⁻¹ yeast extract, 20 g l⁻¹ glucose, 1 g l⁻¹ Tween 80, 1 mg l⁻¹ resazurin, 2 g l⁻¹ K₂HPO₄, 5 g l⁻¹ Na-acetate, 2 g l⁻¹ (NH₄)₃ citrate, 0.2 g l⁻¹ MgSO₄ × 7H₂O, 0.05 g l⁻¹ MnSO₄ × H₂O, pH 6.2–6.5; 0.5 g l⁻¹ cysteine-HCl was added after the medium cooled under a CO₂ atmosphere) at 37 °C without shaking. Limosilactobacillus vaginalis was cultured in the same liquid medium with shaking at 150 rpm.

The origins of all strains used in this study are listed in Supplementary Tables 3 and 4.

Cross-streak assay

Bacteria were grown on agar plates and incubated overnight at 37 °C and 5% CO₂. Bacteria were collected and diluted in PBS to an OD₅₅₀ of 0.5 (Neisseria) or OD₆₀₀ of 0.5 (Pseudomonas, Klebsiella, Escherichia). Filter strips soaked with the indicated bacterial suspensions were stamped on GC or LB plates in a cross pattern. Plates were incubated overnight at 37 °C and 5% CO₂. Images of plates were taken using the ChemiDoc Touch Imaging System from BioRad, and inhibition zone length was determined with Inkscape 0.91.

Co-culturing of N. gonorrhoeae with P. aeruginosa culture supernatants

Of an overnight culture of P. aeruginosa PAO1, 120 μl was used to inoculate 8 ml LB medium. The culture was incubated for 24 h at 37 °C at 150 rpm to allow maximum production of AQNOs¹⁴. Then, bacteria were pelleted and the supernatant containing the Pseudomonas-conditioned medium was sterile filtered and saved. N. gonorrhoeae MS11 was grown on GC plates overnight at 37 °C and 5% CO₂. Bacteria (4 × 10⁶) from a 4-h preculture (in PPM, 37 °C, 220 rpm) were centrifuged on poly-l-lysine-coated coverslips (3,300g × 15 min) and incubated with a 1:1 mixture of PPM and Pseudomonas-conditioned medium for 3 h. Samples were fixed and processed for scanning electron microcopy.

Scanning electron microscopy

Bacteria were grown on GC plates and incubated overnight at 37 °C and 5% CO₂. Bacteria (4 × 10⁶) from a 4-h preculture (in PPM, 37 °C, 220 rpm) were centrifuged on poly-l-lysine-coated coverslips (3,300g × 15 min). NQNO was added at the respective concentration, and bacteria were incubated for 3 h at 37 °C and 5% CO₂. Bacteria were fixed with 2 × 300 µl fixans (3% formaldehyde, 2% glutardialdehyde, 0.09 M sucrose, 0.01 M CaCl₂, 0.01 M MgCl₂ in 0.1 M HEPES, pH 7.2) for 5 and 25 min, respectively. Samples were washed twice with 1 ml 0.1 M HEPES. Dehydration was performed in a graded series of ethanol (30%, 50%, 70%, 80%, 90%, 96%, 3 × 100%) for at least 10 min each. Dehydrated samples were critical-point dried with liquid CO₂ using a Baltec CPD 030 (Baltec). Finally, samples were mounted on stubs with silver-coating polish (drying overnight at room temperature), sputtered with 6 nm platinum using Quorum Q 150R ES (Quorum Technologies) in a low-pressure argon atmosphere and imaged using a Zeiss Auriga 40 Crossbeam FIB-FESEM.

Synthesis of AQNOs

The saturated AQs were synthesized via Conrad–Limpach cyclization and the unsaturated trans-Δ¹-NQ by Camps cyclization as described previously¹⁹. The N-oxides were obtained by locking the 4-hydroxyquinolins as ethyl carbonates followed by N-oxidation with mCPBA (m-chloroperoxybenzoic acid) and deprotection. A list of names, abbreviations and structures of all synthetic compounds is given in Supplementary Table 5. The details of compound synthesis together with the characterization and confirmation by NMR spectroscopy and high-resolution mass spectrometry are described in Supplementary Data File 1.

Bacterial growth assays in the presence of AQs, AQNOs or antibiotics

Bacteria were grown on agar plates as described above. Before exposure, bacteria were precultured in 5 ml liquid medium for at least 2 h, then collected and resuspended in PBS. Optical density at 550 nm (OD₅₅₀; N. gonorrhoeae and commensal Neisseriae) or 600 nm (OD₆₀₀; all other microorganisms) was determined and 4 × 10⁷ bacteria were inoculated in 5 ml of the respective growth medium. AQ or AQNOs dissolved at 5 mM in DMSO were added to the indicated final concentrations. DMSO was adjusted to 1% (v/v) final concentration in all samples including controls. Samples were incubated at 37 °C for 8–10 h until control cultures reached stationary phase or until they reached an OD of 2.5. Optical density was determined every 0.5 h. Growth inhibition was quantified by the calculation of the area under the curve (AUC) using Prism5 (GraphPad) with the start OD at T₀ subtracted. The calculated AUC of the DMSO control was set at 1. To visualize multiple samples, a colour gradient was used to indicate strong inhibition/no growth (white) to weak inhibition/full growth (dark green).

Stability of NQNO in solvent and in cell culture medium

Two samples of NQNO were dissolved in deuterated methanol (CD₃OD) to reach a final concentration of 25 mM. Samples were kept at 25 °C or 37 °C for 28 days, proton NMR spectra were measured with the same acquisition parameters every week and the spectra compared with the ones from day 0.

Stability of 100 µM NQNO in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum at 24 °C and 37 °C was evaluated over a period of 4 days. Samples were taken at the beginning (day 0) and every second day. Samples were analysed by LC–MS and the integrals of the detected NQNO (by using extracted ion chromatograms) compared to the integral of the control (100 µM NQNO in H₂O), which was set to 100%.

ATP measurement

Neisseria species were grown in the presence of NQNO or DMSO for 10, 20, 30, 45 or 120 min. Bacteria were collected by centrifugation for 10 min at 4,000 rpm and 4 °C, resuspended in 1 ml PBS and snap frozen. Samples were lysed by sonication for 1 min at maximum intensity on ice. Supernatant was collected by centrifugation for 10 min at 13,500 rpm at 4 °C. ATP assay was performed using 10 µl of supernatant, 100 µl ATP assay buffer (25 mM Tris pH 7.8, 4 mM EGTA, 20 mM MgSO₄, 1 mM dithiothreitol, pH 7.8), 20 µM luciferin and 1 µg firefly luciferase (Sigma, L9420; 10 × 10¹⁰ units mg⁻¹ protein) per well. Luminescence was measured using a Thermo Fisher Varioskan Flash spectrophotometer.

Determination of NADH levels

Neisseria species were grown in the presence of the indicated concentrations of NQNO, 3Me-NQNO, Antimycin A or DMSO for 1 h. DMSO concentration was adjusted to 1% in all samples. Resazurin assay was performed using 1 × 10⁶ bacteria and 0.02 g l⁻¹ resazurin sodium (Sigma-Aldrich) in PBS per well. Fluorescence was measured after 2 h of incubation in flat-bottom transparent 96-well plate using excitation at 459 nm and emission at 590 nm in a Varioskan Flash spectrophotometer (Thermo Fisher).

Detection of ROS

Log-phase bacteria were collected by centrifugation for 10 min at 4,000 rpm at room temperature and washed in oxidative burst buffer (8 g l⁻¹ NaCl, 0.2 g l⁻¹ KCl, 0.62 g l⁻¹ KH₂PO₄, 1.14 g l⁻¹ Na₂HPO₄, 1 g l⁻¹ glucose, 50 mg BSA in double-distilled H₂O; pH 7.2). Bacteria at 1 × 10⁷ per well were seeded in a white 96-well plate in 50 µl oxidative burst buffer. NQNO, Antimycin A or DMSO at 2× final concentration were premixed in a separate plate in 50 µl oxidative burst buffer containing 100 µM lucigenin (Santa Cruz, CAS 2315-97-1). The lucigenin solution was added to the bacteria and luminescence was recorded using a Thermo Fisher Varioskan Flash spectrophotometer.

Selection for NQNO and 3Me-NQNO resistance

To select for NQNO or 3Me-NQNO resistance, N. gonorrhoeae MS11 was grown continuously in the presence of sublethal concentrations of NQNO or 3Me-NQNO⁶⁴. To this end, 5 ml PPM medium containing 2.5 µM NQNO or 0.1 µM 3Me-NQNO was inoculated with N. gonorrhoeae strain MS11 at an OD₅₅₀ of 0.2, and incubated for 8 h at 37 °C and 220 rpm (conditioned bacteria). After this growth phase, conditioned bacteria were either streaked on GC agar plates without antibiotic or used to inoculate 5 ml PPM containing 0, 5, 10, 25 or 50 µM NQNO or 0.2, 0.5, 1, 2.5 or 5 µM 3Me-NQNO. Agar plates and liquid cultures were incubated at 37 °C and 5% CO₂. Growth of the conditioned bacteria in PPM was monitored by reading the OD₅₅₀ after 14 h. The conditioned gonococci grown overnight on GC agar were used for the next round of conditioning with 2.5 µM NQNO or 0.1 µM 3Me-NQNO for 8 h, and then again evaluated for sensitivity to higher concentrations of NQNO or 3Me-NQNO. This procedure was repeated for a total of 10 days (NQNO) or 14 days (3Me-NQNO). Single clones MS11-R1 and MS11-R2 were derived after 8 days of conditioning in 2.5 µM NQNO.

Genome determination and comparison

Paired-end libraries (250 bp) were prepared from isolated genomic DNA of parent strain MS11-N309 and its NQNO-resistant mutants MS11-R1 and MS11-R2 using the Nextera XT DNA Library Preparation kit (Illumina). The libraries were sequenced on a MiSeq sequencing system (Illumina) using v.2 sequencing chemistry. FastQC (v.0.11.5)⁶⁵ was used to analyse the quality of the raw sequencing data. The raw reads were trimmed using Sickle v.1.33 (https://github.com/najoshi/sickle). SPAdes (v.3.10.1)⁶⁶ was used for the assembly of the raw reads. The raw FASTq files and draft genome sequences of the three strains MS11_N309 (accession no. SAMN19108268), MS11-R1 (N568) (accession no. SAMN19108269) and MS11-R2 (N569) (accession no. SAMN19108270) were combined in Bioproject number PRJNA728975 and are publicly available from the NCBI GenBank. The accessible links are: N. gonorrhoeae MS11-N309: https://www.ncbi.nlm.nih.gov/nuccore/JAHBBP000000000; N. gonorrhoeae MS11-R1: https://www.ncbi.nlm.nih.gov/nuccore/JAHBBO000000000; N. gonorrhoeae MS11-R2: https://www.ncbi.nlm.nih.gov/nuccore/JAHBBN000000000. For gene prediction and automatic annotation, we employed Prokka (v.1.12)⁶⁷ and the NCBI Prokaryotic Genome Annotation Pipeline (PGAP release 5.2)⁶⁸. For subsequent comparative genomic analysis, we used the PGAP annotation.

Genome comparison

The quality trimmed reads of the parent strain MS11-N309 and the two NQNO-resistant mutants MS11-R1 and MS11-R12 were aligned to the reference genome of N. gonorrhoeae MS11 (accession number NC_022240; https://www.ncbi.nlm.nih.gov/nuccore/NC_022240.1/) with the BWA-MEM algorithm from the Burrows–Wheeler Aligner (BWA) software package (v.0.7.17)⁶⁹. The produced alignment was sorted with SortSam, and PCR duplicates were marked with MarkDuplicates using Picard (v.2.17.3) (http://broadinstitute.github.io/picard/) for downstream analysis.

The Genome Analysis Toolkit’s (GATK) (v.3.8.0) RealignerTargetCreator and IndelRealigner tools were used to perform realignment around the Indels⁷⁰. The GATK BaseRecalibrator was used to perform base quality score recalibration. Variant calling was then performed using the GATK HaplotypeCaller with default parameters, except for the ploidy that was set to 1 according to GATK best practices. The sorted single-nucleotide polymorphisms and InDels were then filtered using GATK VariantFiltration with the filtering criteria recommended by GATK. To further remove sequencing bias, only variants with a minimum read depth of 10 were considered. The effects of the variants were then annotated using SnpEff (v.4.3s)⁷¹.

Cultivation of eukaryotic cells

Immortalized human vaginal epithelial cells (hVECs, line MS74) were obtained from A. J. Schaeffer (Feinberg School of Medicine, Northwestern University, Chicago, IL) and were derived from vaginal tissue of a post-menopausal woman. The cell line was created through immortalization of the cells with human papilloma virus 16, E6 and E7 genes according to ref. ⁷². hVECs were cultured on gelatin-coated cell culture dishes in DMEM supplemented with 10% fetal calf serum, 1% non-essential amino acids and 1% pyruvate at 37 °C in 5% CO₂, and subcultured when 60% confluency was reached. HeLa S3 (DSMZ ACC 161) and ME-180 (ATCC HTB-33) cervical carcinoma cells were cultured in DMEM supplemented with 10% fetal calf serum and passaged every 2–3 days.

Measurement of metabolic activity of human cells

Plates (96-well) were coated overnight at 4 °C with 0.1% gelatin in PBS before use. hVECs, HeLa cells or ME-180 cells (2 × 10⁴) were seeded in 100 µl of their respective growth medium supplemented with the indicated concentrations of NQNO, 3Me-NQNO or DMSO. Cultures were grown for 1 or 2 days at 37 °C and 5% CO₂. Next, 10 µl of MTT solution (12 mM in PBS) were added to each well and cells were incubated for an additional 2 h at 37 °C. After removing the MTT-containing growth medium, 100 µl isopropanol was added to each well. The formazan produced by cellular metabolism was allowed to dissolve overnight and OD₅₅₀ was measured using a Varioskan Flash spectrophotometer (Thermo Fisher).

Detection of epsilon1 zeta1 genes

Genomic DNA was extracted from Neisseria strains. Presence of the toxin–antitoxin system epsilon1 zeta1 was tested via PCR by using primers epsilon1 forward (5′-TATCATATGAATAAAGTTGAGCCC-3′) and zeta1 reverse (5′-TATAAGCTTTTATCTGCTGATGGATTTTTTGGC-3′) at 58 °C annealing temperature and 90 s elongation time. Opa loci were amplified with primers Opa_for (5′-GCAGATTATGCCCGTTACAG-3′) and Opa_rev (5′-GTTTTGAAGCGGGTGTTTTCC-3′) at 55 °C annealing temperature and 45 s elongation time. PCR fragments were separated in 1.2% agarose gel using 1 kb DNA ladder for size comparison

Membrane integrity assay

Bacteria were grown overnight at 37 °C and 5% CO₂ and used to inoculate PPM medium. After 2 h culture, bacteria at an OD₅₅₀ of 0.04 were collected and washed in 1 ml PBS. Bacteria were incubated with 50 µM NQNO or 1% DMSO as solvent control at 37 °C and 330 rpm for 30 min. Staining was done using 1.67 µM SYTO9 green fluorescent nucleic acid stain (Invitrogen) and 1.4 µM propidium iodide (Sigma-Aldrich). Bacteria were fixed on a coverslip by centrifugation (2,000 g, 2 min, room temperature). Samples were analysed using a Leica SP5 confocal microscope by excitation at 600–630 nm for propidium iodide and at 454–510 nm for SYTO9 and acquired in xyz mode with a 1,024 × 1,024 pixel format and 100 Hz scanning speed. All images were analysed in ImageJ/Fiji software. The percentage of PI-stained bacterial cells in relation to SYTO9-stained cells was enumerated from independent microscope pictures (n = 5 or 6) for each condition.

Conjugation of pTetM in NQNO-resistant Neisseria strains

To transfer pTetM by conjugation, the pTetM-deficient, NQNO-resistant strain N. gonorrhoeae MS11-R2 was first made resistant to rifampicin by successive growth on increasing concentrations of the antibiotic. Next, the rifampicin-sensitive donor strain (N. gonorrhoeae MS11-F3 pilus, pTetM, N554) and the acceptor strain (N. gonorrhoeae MS11-R2 Rif^R) were precultured for 3 h in PPM medium containing 100 µg ml⁻¹ DNase (AppliChem) and 0.9% NaHCO₃. Conjugation between the donor strain (4 × 10⁷ bacteria) and the acceptor strain (2 × 10⁷ bacteria) was conducted using the filter mating technique based on a 0.2-µm-pore-size nitrocellulose membrane. Gonococci placed on the membrane were co-incubated for 24 h on a GC plate before bacteria were resuspended from the membrane in 500 µl PPM medium and plated on a GC plate containing tetracycline (12.5 mg l⁻¹) and rifampicin (0.5 mg l⁻¹) to select for the pTetM conjugated acceptor strain. Two independent clones (MS11-R2 pTetM A and MS11-R2 pTetM B) were propagated.

Generation of pNEISS epsilon1/zeta1 and transformation in N. gonorrhoeae

A gene strand containing multiple copies of the neisserial DNA uptake sequence (DUS), an opa promoter and a LIC cloning site flanked on each side by 200 bp sequences of the gonococcal lactoferrin-binding protein A (LbpA) gene (which is truncated in ~50% of gonococci and is not essential for viability or virulence^73,74) was synthesized by Eurofins Genomics and inserted into the HindIII/SacI sites of pBluescript SK. The erythromycin resistance gene ErmC encoded in plasmid Hermes8 (ref. ⁷⁵) was amplified with primers ErmC-BamH1-for 5′-AATGGATCCAGGAGGAAAAAATAAAGAGGGTTATAATGAACGAG-3′ and ErmC-XbaI-rev 5′-AATTCTAGACTTACTTATTAAATAATTTATAGCTATTG-3′ and inserted into the corresponding sites of the gene strand to yield pNEISS. The epsilon1/zeta1 coding sequences were amplified by PCR from pTetM isolated from N. gonorrhoeae MS11 using primers Epsilon1-LIC-for 5′-ACTCCTCCCCCGTTAGGATTTTATTATGAATAAAGTTGAGCC-3′ and Zeta1-LIC-rev 5′-CCCCACTAACCCGTTATCTGCTGATGGATTTTTTGG-3′ and inserted via ligation-independent cloning downstream of the opa promoter into vector pNEISS.

For transformation, a fragment containing DUS, opa promotor, epsilon1/zeta1, ErmC and the LbpA flanking sequences was amplified by PCR using primers pNEISS-TROPIC-for 5′-CTCGAGGTCGACGGTATCG-3′ and pNEISS-TROPIC_rev 5′-GGGAACAAAAGCTGGAGC-3′ at 58 °C annealing temperature and 3 min elongation time. The PCR fragment was used to transform N. gonorrhoeae by electroporation. To this end, 9 × 10⁸ log-phase gonococci were washed twice with 1 ml of ice-cold electroporation buffer (EP; 15% glycerin, 272 mM sucrose, 1 mM HEPES, pH 7.2) before being taken up in 40 µl ice-cold EP. Upon addition of 100 ng of purified PCR fragment, electroporation with a single pulse of 2,300 V, 400 ohm and 25 µF at an electrode gap of 2 mm was run on an ECM630 Exponential Decay Wave electroporation system (BTX). Neisseria were allowed to recover in 1 ml prewarmed PPM medium for 150 min at 37 °C and 220 rpm. Bacteria were selected for 48 h on GC agar with 28 mg l⁻¹ erythromycin. Erm-resistant clones were checked for the presence of epsilon1/zeta1 by PCR. Chromosomal integration of epsilon1/zeta1 into the lbpA locus was verified by PCR of genomic DNA using LbpA-for 5′-CGAGGCTGAAGTTCCTGCCCG-3′ and Zeta1-rev 5′-CAATGGCAGTTACGAAATTACACC-3′ primers (Extended Data Fig. 6b). As a control, opa loci were amplified using primers Opa-for 5′-GCAGATTATGCCCGTTACAG-3′ and Opa-rev 5′-GTTTTGAAGCGGGTGTTTTCC-3′. Expression of Epsilon1 and Zeta1 was also verified by western blotting.

Antibodies used in this study

The following primary and secondary antibodies were employed at the indicated dilutions for western blotting (WB): monoclonal antibody (mAB) against 6×His (H8, Thermo Fisher, 1:1,000 WB); mAb against GAPDH (clone GA1R, Invitrogen, 1:2,000 WB); mAb against GFP (clone JL-8, Clontech, 1:5,000); mAb α-Opa (clone 4B12/C11, Developmental Studies Hybridoma Bank, University of Iowa, USA; a generous gift from M. Achtman, WB: 1:2,000). Rabbit polyclonal antibody (pRb) against the synthetic Epsilon1 peptide C-EKNRRMMTDEAFRKEVEKRLYAG was produced by PSL (Germany), affinity purified against the cognate peptide and used at 1 µg ml⁻¹ for WB; pRb against the synthetic Zeta1 peptides AKKEYSKQRVVTNSK-C and KIVGINQDRNSEFIDK-C was produced by PSL and affinity purified using both peptides (1 µg ml⁻¹ pRb was used for WB). HRP-conjugated goat anti-mouse IgG (115-035-146, 1:10,000 WB) and HRP-conjugated goat anti-rabbit IgG (111-035-003, 1:5,000 WB) were from Jackson Immunoresearch.

Cloning and expression of Epsilon1 in E. coli

The cDNA of gonococcal epsilon1 was amplified from pTetM plasmid isolated from N. gonorrhoeae MS11 with primers 4069_NcoI_Epsilon1_fw 5′-ATACCATGGATGAATAAAGTTGAGCCCCAAG-3′ and 4113_XhoI_Epsilon1_rev 5′-TATCTCGAGTTGCTCCTTATTTGC-3′. The resulting PCR fragment was cloned into vector pET28a (Novagen, MerckBiosciences) via NcoI and XhoI restriction sites allowing expression of Epsilon1 with a C-terminal 6×His-tag. For the expression of neisserial Epsilon1 without a tag, the epsilon1 cDNA was amplified as above with the primers 4520_e-antitoxin_NcoI_for 5′-ATACCATGGGCAATAAAGTTGAGCCCCAAG-3′ and 4521_e-antitoxin_XhoI_rev 5′-ATACTCGAGTTATTGCTCCTTATTTGCCGC-3′. The amplicon harboured an extra glycine after the start methionine and a STOP codon at the 3′ end. The resulting PCR fragment was cloned into vector pCDFduett (Novagen, MerckBiosciences) via NcoI and XhoI restriction sites.

Expression of recombinant Epsilon1-6×His

Epsilon1-6×His was expressed in E. coli BL21 Rosetta (DE3) upon induction with 0.5 mM IPTG at an OD₆₀₀ of 0.75 for 4 h at 37 °C and 200 rpm in LB medium with 50 g l⁻¹ kanamycin. After centrifugation at 4,700 rpm for 20 min at 4 °C, the bacterial pellet was resuspended on ice in 20 ml buffer A (50 mM Na phosphate, 1 M NaCl, pH 8, supplemented with 10 µg ml⁻¹ Pefabloc, 10 µg ml⁻¹ aprotinin, 1 µM PMSF and 5 µg ml⁻¹ leupeptin) and sonicated at 4 °C three times for 2 min. Lysate was cleared by centrifugation for 1 h at 16,000 rpm at 4 °C. Supernatant was filtered using a 0.2-µm-pore-size filter. Epsilon1-6×His was purified using His-FF column (GE Healthcare) and dialysed against 50 mM Tris, pH 7.5, 150 mM NaCl and 10% glycerol. The amount and the purity of proteins were analysed via SDS-gel electrophoresis. To investigate stability of the purified protein, 50 ng purified Epsilon1-6×His was incubated with 1% DMSO, 50 µM NQNO, or 1 mM H₂O₂ in PBS supplemented with freshly added protease inhibitors (10 μg ml⁻¹ leupeptin, 20 μg ml⁻¹ aprotinin, 20 μg ml⁻¹ Pefabloc and 20 μM benzamidine) for the indicated time at room temperature. Samples were separated by SDS–PAGE and analysed by western blotting using mouse monoclonal anti-6×His antibodies.

Epsilon degradation assay in E. coli

pCDFduett epsilon1 was transformed into E. coli Nova Blue (DE3). Bacteria were grown on LB plates with streptomycin (50 g l⁻¹) overnight at 37 °C and used to inoculate a liquid culture (OD₆₀₀ of 0.3) in LB medium containing 50 g l⁻¹ streptomycin. Bacteria were grown until an OD₆₀₀ of 0.6 and a sample was taken before induction (BI). The remaining culture was induced with 0.01 mM IPTG for 30 min and a second sample was then taken (after induction, AI). The IPTG induction was stopped by pelleting the bacteria and suspending them in LB medium containing 50 g l⁻¹ streptomycin. The culture was split into several samples, which received 0 mM, 1 mM or 5 mM H₂O₂, 25 µM or 50 µM NQNO or 1% DMSO. Bacteria were lysed after 45 and 60 min, and degradation of epsilon1 was analysed via western blotting and densitometric analysis using ImageJ/Fiji software. The amount of epsilon1 was compared to the expression levels of GAPDH.

Epsilon degradation in N. gonorrhoeae

Neisseria spp. were grown overnight at 37 °C and 5% CO₂ on GC plates. Upon preculture in PPM for ~2 h, 4 × 10⁷ bacteria were used to inoculate 5 ml PPM. Cultures were grown for 3 h until they reached the log phase. NQNO, Antimycin A or ciprofloxacin were added at the indicated concentrations and in all samples, DMSO was adjusted to 1% final concentration. After the indicated times, bacteria were collected by centrifugation (3,000g, 8 min, 4 °C) and lysed in modified RIPA buffer (1% Triton X-100, 0.1% SDS, 1% deoxycholic acid, 50 mM HEPES, pH 7.4, 150 mM NaCl, 10% glycerol, 1.5 mM MgCl₂, 1 mM EGTA) supplemented with freshly added protease inhibitors (10 µg ml⁻¹ leupeptin, 20 µg ml⁻¹ aprotinin, 20 µg ml⁻¹ Pefabloc and 20 µM benzamidine). Lysates were cleared (13,000 rpm, 4 °C, 20 min) and the total protein content was determined via the bicinchoninic acid protein assay kit (Pierce, Thermo Fisher). Equal amounts of each lysate were separated by SDS–PAGE and analysed by western blotting with rabbit polyclonal anti-epsilon1 antiserum or anti-zeta1 antiserum, and densitometric analysis using ImageJ/Fiji software. The amount of epsilon1 was quantified relative to protein levels of the zeta toxin.

Screening of AQNO and AQ derivatives in 96-well format

Neisseria were precultured in 10 ml PPM medium for 2 h at 37 °C and 220 rpm to reach a final OD₅₅₀ of 0.3. Precultures were diluted to an OD₅₅₀ of 0.2. The 96-well compound plate was prepared by adding 1 µl of a 0.5 mM compound stock solution in DMSO to 100 µl bacterial suspension at OD₅₅₀ 0.1 in PPM medium. The final compound concentration was 5 µM in 1% DMSO. Accordingly, the solvent control was 1% DMSO. Each 96-well plate had 6 wells holding the solvent control and 8 wells holding the positive control (5 µM NQNO). The plates were then incubated for 6 h at 37 °C at 550 rpm. Growth was monitored every 30 min using OD₅₅₀ readings (Tecan Sunrise). The experiment was performed in three independent replicates. Data evaluation was done by subtracting the starting OD₅₅₀ (time point 0 h) from OD₅₅₀ at 6 h (negative values were set to zero). The calculated OD₅₅₀ was normalized to the average OD₅₅₀ of the positive control (5 µM NQNO). Values >1 indicate reduced growth inhibition compared with 5 µM NQNO, while vaues <1 indicate a growth inhibition stronger than the one observed for 5 µM NQNO.

In vivo infection of CEAtg mice

C57BL/6J mice carrying the complete human CEA gene (CEAtg mice)⁷⁶ and wild-type C57BL/6J mice (obtained from Elevage Janvier) were kept under specific pathogen-free conditions under a 12-h light cycle in the animal facility of the University of Konstanz. Experiments involving animals were performed in accordance with the German Law for the Protection of Animal Welfare. The animal care and use protocol, including the protocol for experimental vaginal infection of female mice, was approved by the appropriate state ethics committee and state authorities regulating animal experiments (Regierungspräsidium Freiburg, Germany) under permit file number G-19/147. Experimental vaginal infection of female mice with N. gonorrhoeae was performed as previously described³⁹ for prohibiting bacterial infections with NQNO (prophylactic treatment) or using an adapted schedule as described below to treat established infections with 3Me-NQNO (therapeutic treatment). For prophylactic treatment, mice were subcutaneously injected with 17-β-estradiol (50 µg per mouse) in corn oil 4 days and 2 days before infection. The drinking water was supplemented with 40 mg per 100 ml drinking water trimethoprim sulphate (Infectotrimed, Infectopharm). Mice were inoculated intravaginally with 10⁸ c.f.u. of Opa_CEA-expressing gonococci (strain N. gonorrhoeae MS11 or Ngo XDR) suspended in 10 μl of PPM medium supplemented with 5% IsoVitale vitamin mix. The CEACAM-binding Opa_CEA-expressing Ngo XDR was visually selected and functionally tested for CEACAM binding as described previously⁴³. Prophylactic treatment was started at 1 h after infection by applying 10 µl of 25 µM or 50 µM NQNO (corresponding to 0.25 nmol or 0.5 nmol NQNO) in PPM medium supplemented with 5% IsoVitale vitamin mix and 1% DMSO intravaginally. Control mice received 10 µl PPM medium supplemented with 5% IsoVitale vitamin mix and 1% DMSO. Mice were randomly assigned to the control or treatment groups, but experimenters were not blinded to the conditions of the experiment.

In the case of therapeutic treatment with 3Me-NQNO, the animals were subcutaneously injected with 17-β-estradiol (50 µg per mouse) in corn oil 4 days and 2 days before infection. To suppress commensal overgrowth during the extended infection, mice also received cloramphenicol (500 µg per mouse) and streptomycin (2.4 mg per mouse) 4 days and 2 days before infection. On day 0, CEAtg mice were inoculated intravaginally with 10⁸ c.f.u. of Opa_CEA-expressing gonococci (strain N. gonorrhoeae MS11) suspended in 10 μl of PPM medium supplemented with 5% IsoVitale vitamin mix. At 24 h after infection, treatment was started by applying 10 µl of 50 µM 3Me-NQNO (corresponding to 0.5 nmol 3Me-NQNO) in PPM medium supplemented with 5% IsoVitale vitamin mix and 1% DMSO intravaginally. Control mice received 10 µl PPM medium supplemented with 5% IsoVitale vitamin mix and 1% DMSO.

At 24 h after infection (prophylactic treatment) or 48 h after infection (therapeutic treatment), the mucosa-associated bacteria were re-isolated by cotton swabs and vaginal lavage using 20 µl PPM medium supplemented with 5% IsoVitale vitamin mix. Serial dilutions of re-isolated bacteria were plated on GC plates without antibiotics (to detect growth of commensal bacteria), on GC agar containing chloramphenicol (10 µg ml⁻¹) and erythromycin (7 µg ml⁻¹) for MS11, or on GC plates containing ciprofloxacin (10 µg ml⁻¹) for Ngo XDR, and the colonies were counted after 20 h of incubation at 37 °C and 5% CO₂.

Immunohistochemistry of tissue samples

The genital tract of infected mice was excised and immediately fixed with 4% paraformaldehyde for at least 24 h. The fixed tissue was sequentially transferred to 10% sucrose, 0.05% cacodylic acid for 1 h at 4 °C, to 20% sucrose for 1 h at 4 °C, and then to 30% sucrose at 4 °C overnight. Organs were mounted in the embedding medium (Cryo-M-Bed, Bright Instrument) and frozen at −20 °C. Sections (10-μm thick) were cut at −20 °C using a cryostat (Vacutom HM500, Microm). Sections were stained with a mouse monoclonal antibody against collagen type IV (clone M3F7, dilution 1:200) together with a polyclonal rabbit antibody against N. gonorrhoeae (dilution 1:100) (IG-511, Immunoglobe). Detection of the primary antibodies was done with a combination of Cy5-conjugated goat anti-rabbit antibody (1:250) and Cy3-conjugated goat anti-mouse antibody (1:250) (both from Jackson Immunoresearch). Cell nuclei were visualized by the addition of Hoechst 33342 (1:30,000, Life Technologies) in the final staining step.

Test of bactericidal effect of 3Me-NQNO

Neisseria spp. were inoculated at an OD₅₅₀ of 0.2 in 5 ml PPM. After 4 h of culturing, 0.5 µM or 2.5 µM 3Me-NQNO was added to the log-phase cultures. A volume of 1 ml of culture was taken after 15 min, 30 min, 1 h, 2 h and 3 h, bacteria were washed in PBS and dilutions (10⁰ to 10⁻⁶) were plated on GC plates. Colonies were enumerated after overnight incubation at 37 °C and 5% CO₂.

Statistics

All data are presented as mean ± s.e.m. Statistical significance was determined as indicated in the figure legends using a two-tailed Student’s t-test with Prism5 (GraphPad) or one-way and two-way analysis of variance (ANOVA) by multiple comparison to the DMSO sample using Dunnett’s multiple comparisons or Sidak’s multiple comparisons test using Prism7 (GraphPad) (*P < 0.05, **P < 0.01, ***P < 0.001). For the statistical evaluation of in vivo colonization, differences between groups were analysed using Mann–Whitney rank-sum test in SigmaStat4 (SysStat). For all applied statistical tests if not differently indicated, P values were set as *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 with a 95% confidence interval and two-tailed P values.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.