Compounds and chemical synthesis

Boro-PBPi compounds were synthesized at Venatorx Pharmaceuticals, Inc. and BioDuro-Sundia as described in the ‘Chemistry experimental details’ section in the Supplementary Information. Ceftriaxone (USP, 1098184), azithromycin (USP, 1046056), ciprofloxacin (Sigma-Aldrich, 17850), cefoperazone (Alfa Aesar, J65185), zoliflodacin (MedChemExpress, HY-17647) and gepotidacin (MedChemExpress, HY-16742) were used. The Molecular Operating Environment (Chemical Computing Group) software was used for computational chemistry, inhibitor design and the calculation of physicochemical properties, including pK_a.

Bacterial strains and growth conditions

The strains used in this study are listed in Supplementary Table 14. FA1090, H041 (WHO X), MS11 and F89 were obtained from A.E.J. (Uniformed Services University of the Health Sciences). The N. gonorrhoeae WHO reference strains (WHO F, WHO G, WHO L, WHO M, WHO K, WHO L, WHO Y and WHO Z)^53,35 were obtained from the CDC. WHO Q (G7944)³⁶ was obtained from the National Collection of Type Cultures (NCTC 14208). FA19 was obtained from R. A. Nicholas (University of North Carolina). A set of 50 CDC strains (AR bank, 165–214)⁵⁴ was obtained from the CDC, of which six strains were confirmed not to be N. gonorrhoeae and were excluded from the study. All N. gonorrhoeae strains were cultured at 37 °C in a 5% CO₂ humidified environment.

Whole-genome sequencing and analysis

Whole-genome sequencing and genomic analyses were conducted on the parent and isolated mutant strains. DNA extraction, Illumina library preparation and Illumina sequencing (paired read with 2 × 150 bp) were performed by GENEWIZ from Azenta Life Sciences. Whole-genome-sequencing analysis of the FASTQ files provided by GENEWIZ was performed using Geneious Prime versions 2022.1.1 and 2025.1.3 (Biomatters Inc.). The reads were trimmed using BBDuK Adapter/Quality Trimming version 38.84 (Brian Bushnell) to yield approximately ten million reads. The trimmed reads were mapped to the corresponding reference genomes (GenBank accession numbers provided in Supplementary Table 15) and genetic alterations were identified, especially in the penA gene encoding PBP2 (the major target of CRO), genes encoding the major porin (PorB) and the efflux pump gene cluster (mtrR–mtrCDE). For CDC strains⁵⁴ whose assembled genomes were not available, sequence reads were assembled to generate contigs that were used as the reference genome. To identify genetic alterations in the genome sequences of mutants compared with those of the parent strain, trimmed reads were mapped to the corresponding reference genome sequences (Supplementary Table 15) and differences in genetic alterations between the parent and mutant strains were identified. A BLAST search was performed against the NCBI database (https://blast.ncbi.nlm.nih.gov/) using short amino acid sequences containing the identified PBP2 Y543C substitution to determine the presence of any PBP2 proteins with the Y543C substitution. The raw reads are deposited in GenBank under BioProject accession number PRJNA1353147.

Minimum inhibitory concentration assays

To determine the ability of test compounds to inhibit the growth of bacterial strains, MIC assays were performed using both liquid- and agar-based microdilution methods, as described previously^34,55,56, with modifications. Briefly, cryo-preserved bacterial cultures of clinical strains were streaked for isolation and cultured for 24 h on Chocolate Agar. The Chocolate Agar was made by first separately autoclaving 72 g l⁻¹ (2×) GC agar base (Remel, R453502) and 2% (2×) haemoglobin (Remel, R451402) at 121 °C for 20 min for sterilization. After cooling to approximately 50 °C, the 2×GC agar base and 2×haemoglobin solutions were combined in a 1:1 vol/vol ratio and 1% IsoVitaleX Enrichment (BD, 211876) was added to the solution. Freshly cultured colonies were used for inoculum preparation. In the liquid broth-based MIC assay, twofold serial dilutions of the test compounds were made in a 96-well plate with a final volume of 75 μl per well at twofold the final desired concentration in Fastidious Broth (FB; prepared according to Cartwright et al. with a final pH 7.2 ± 0.2)⁵⁵. To make the assay inoculum, a direct suspension was prepared by aseptically swabbing 10–15 colonies from the agar plates into culture tubes containing 2 ml fresh sterile saline. After the dilution plates were set up, direct suspensions were diluted in a cuvette containing sterile saline and the optical density at 600 nm (OD₆₀₀) was measured. Direct suspensions were diluted in FB to make assay inocula (approximately 1 × 10⁶ CFU ml⁻¹) and 75 μl assay inocula were added to the 96-well plates prepared with compound dilutions, yielding a starting bacterial concentration of 5 × 10⁵CFU ml⁻¹. The plates were incubated for approximately 24 h. The broth MIC was determined visually as the lowest concentration that completely inhibited bacterial growth.

For the agar-based method, GC agar was used to determine the MIC. After the initial strain incubation for about 24 h, bacterial colonies were re-streaked for isolation on fresh Chocolate Agar and cultured for 18–24 h. To prepare GC agar, the GC agar base (Remel, R453502) dissolved in water was autoclaved at 121 °C for sterilization and once cooled to approximately 50 °C, it was supplemented with IsoVitaleX Enrichment (BD, 211876) and 1 M filter-sterilized ferric nitrate (Fisher Scientific, S25320) to achieve final concentrations of 1% and 12 µM, respectively. The determined amount of test compound stock solution was pipetted directly into a 100 mm × 15 mm Petri dish for each desired test concentration and 20 ml of the prepared GC agar (before solidification) was added to the Petri dish containing the test compound. The agar plates were swirled to dissolve the compound and vented for solidification in a biological safety cabinet until dry. For the preparation of assay inoculum, a direct suspension was first prepared by aseptically swabbing several colonies from the agar plates and suspending them in 2 ml sterile cation-adjusted Mueller–Hinton broth (CAMHB); these were then diluted with sterile CAMHB and adjusted to an OD₆₀₀ of 0.1 (0.5 McFarland standard equivalent). A Steer’s replicator was used to plate up to 32 spots of 2 µl inocula on a single agar plate from a separate 32-well plate containing inocula (500 µl each). The agar plates were vented in the biological safety cabinet until they were dry, inverted and incubated for 24 h. The agar MIC was determined as the lowest concentration that completely inhibited the bacterial growth.

Time-kill assay

The time-kill assay for N. gonorrhoeae was performed as described⁵⁷, with the following modifications. The bacterial inoculum for the assays was prepared by suspending colonies of N. gonorrhoeae cultured on Chocolate Agar for approximately 24 h in sterile saline. The cell suspension was adjusted to an OD₆₀₀ of 0.1 and diluted 1:500 in FB broth (final cell density of approximately 1 × 10⁵ CFU ml⁻¹; 90 µl each) in a 96-well plate, which was then incubated with shaking at 200 rpm for 4 h to allow the bacteria to reach the logarithmic phase of growth. After the initial growth phase, 10 µl of the appropriate 10× drug dilution was added to each well (total volume of 100 µl) to achieve final drug concentrations of 4× and 16× MIC. For the untreated control, 10 µl FB was added. Samples were taken at −4 h (time of inoculation), 0 h (time of compound addition) as well as 2, 4, 6 and 24 h following the addition of the compound and the CFU ml⁻¹ was determined for each sample. Growth curves were analysed by plotting the log₁₀(CFU ml⁻¹) versus time using GraphPad Prism version 10.3.2.

Frequency of resistance

N. gonorrhoeae strains (ATCC 49226, H041 and WHO Q) were cultured on Chocolate Agar for about 24 h and colonies were suspended (approximately 5 × 10⁸ CFU ml⁻¹) in PBS. A 0.1-ml aliquot of the cell suspension was spread on ten GC agar plates (100 mm diameter) supplemented with 1% IsoVitaleX, 12 µM ferric nitrate and test compounds (4× and 16× MIC) for ATCC 49226 and H041. Chocolate Agar was used for WHO Q because the strain grew poorly on GC agar. The viable cell count in each suspension was determined by plating serial tenfold dilutions onto the corresponding agar. The plates were incubated for 24 h, the visible colonies were counted and spontaneous mutation frequency was calculated.

Isolation of zoliflodacin-resistant mutants and mutation identification

Zoliflodacin-resistant mutants were isolated by selection of WHO K (MIC = 0.12 µg ml⁻¹) on GC agar supplemented with 4× MIC of zoliflodacin. Colonies arising on agar were streaked on Chocolate Agar and a single colony was tested for broth MIC to confirm the elevated MIC of zoliflodacin. Short-read genome sequencing was performed on the parent and isolated mutant strains (GENEWIZ). Genome analysis from FASTQ files provided by GENEWIZ was performed using Geneious Prime version 2022.1.1 (Biomatters Inc.). The reads were trimmed using BBDuK Adapter/Quality Trimming version 38.84, yielding approximately ten million reads. Read mapping was performed with the Geneious Mapper using the WHO K reference genome (GenBank ID: GCF_900087865.2) to provide >100 mean coverage of the entire chromosome, followed by Geneious single-nucleotide-polymorphism analysis, resulting in the identification of the GyrB(D429N) mutation present in each mutant.

Serial passage experiments with ceftriaxone and boro-PBPi 21

Glycerol stocks of the strains tested (ATCC 49226, WHO L, CDC-0197 and WHO Q) were cultured on fresh Chocolate Agar for approximately 24 h at 37 °C with 5% CO₂. The arising bacterial colonies were re-streaked on fresh Chocolate Agar for second isolation and cultured for 18–24 h for use in the assay. To prepare the assay inoculum, a direct suspension was made by aseptically swabbing several colonies from the agar plates and suspending them in 2 ml sterile CAMHB; the direct suspensions were then diluted with sterile CAMHB and adjusted to an OD₆₀₀ of 0.1 (0.5 McFarland standard equivalent). Each inoculum (10 µl) was spotted on GC agar (0.6 ml; 3–4 mm depth) containing the desired concentrations (0.25, 0.50, 1, 2, 4 and 8× agar MIC) of CRO and 21 in 24-well flat-bottomed plates. After the plates were incubated for 20–24 h at 37 °C with 5% CO₂, susceptibility was recorded as the lowest concentration that completely inhibited bacterial growth. For passaging with the compound, cells that grew robustly on agar containing the highest concentration of each compound were swabbed, suspended in 0.5–1 ml fresh CAMHB and the suspension was adjusted to an OD₆₀₀ of 0.1. The cell suspension (10 µl) was spotted on GC agar containing the compound in 24-well plates, followed by incubation for 20–24 h at 37 °C with 5% CO_2. Passaging experiments were serially performed ten times (one passage per day) and susceptibility was monitored daily. After days 1 and 10 as well as any day when a change in susceptibility was observed, the agar MIC was determined as described earlier.

Expression and purification of N. gonorrhoeae PBP2

In the N. gonorrhoeae PBP2 inhibition assay, constructs comprising the TPase domain of PBP2 from the FA19 and H041 strains (tPBP2^FA19 and tPBP2^H041) were used. The cloning, expression and purification of these PBP2 constructs were previously described^29,58. Importantly, the TPase domains of PBP2 were acylated by fluorescently labelled penicillin V, Bocillin-FL (Thermo Fisher Scientific) at the same rate as that of full-length PBP2 (ref. ⁵⁸).

Measurement of in vitro inhibition of N. gonorrhoeae PBP2

The inhibitory potency towards wild-type PBP2^FA19 was assessed by determining the inhibitor concentration required to reduce PBP2^FA19 binding to Bocillin-FL by 50% (IC₅₀) using a fluorescence polarization competitive equilibrium binding assay⁵⁹. Enzyme titration/saturation binding experiments were initially performed to establish the assay conditions for competitive binding. A solution of 0.2 µM Bocillin-FL in assay buffer (50 mM HEPES–NaOH pH 8.0), 300 mM NaCl and 10% (vol/vol) glycerol) was prepared and saturation binding was performed by mixing 40 µl PBP2 at different concentrations (0–24 µM) with 40 µl of the 0.2 µM Bocillin-FL solution in individual wells of black 384-well microplates. The fluorescence was monitored immediately following mixing using a Cytation 3 plate reader (BioTek) with a fluorescence polarization cube containing polarizing filters with 485 nm excitation and 520 nm emission. The instrument gain was set to achieve minimum fluorescence values of 50–60 and each measurement was the average of three flashes. Fluorescence was measured continuously for up to 120 min and the response was stabilized within 10 min, with a dose-dependence on the PBP2^FA19 concentration. For the competition binding assays, twofold serial dilutions of compounds were prepared in assay buffer and mixed with Bocillin-FL (final concentration of 0.1 µM) in black 384-well microplates. PBP2^FA19 was added to a final concentration of 0.25 µM and fluorescence was immediately measured at 1-min intervals for up to 10 min. Fluorescence values at the reaction end point (typically within 8 min) were normalized to the maximal response plotted against inhibitor concentration and fitted to a four-parameter inhibitor-response curve to derive the IC₅₀.

As the interaction between PBP2^H041 and Bocillin-FL did not produce an adequate response in the fluorescence polarization assay, the inhibitory potency towards tPBP2^H041 was assessed by determining the concentration of the compound required to inhibit the binding of Bocillin-FL to PBP2^H041 by 50% (IC₅₀) using an SDS–PAGE-based competition binding assay. Twofold serial dilutions of each compound were prepared in the assay buffer, mixed with 1 µM tPBP2^H041 and incubated at ambient temperature for 60 min. Bocillin-FL (1 µM) was added and the reaction mixtures were further incubated for 60 min before resolution by SDS–PAGE using 10% NuPAGE Bis-Tris mini protein gels (Invitrogen). The amount of Bocillin-FL incorporated into tPBP2^H041 at each inhibitor concentration was detected by fluorescence scanning of the SDS–PAGE gels with an Azure 600 imager (Azure Biosystems, Inc.) and quantified using the ImageJ software⁶⁰. Data were normalized to the maximum fluorescence, plotted against the inhibitor concentration and fitted to a four-parameter inhibitor-response curve to derive the IC₅₀.

Membrane preparation and PBP-binding assay using the membrane

Membrane preparation and PBP-binding assays were performed as previously described^61,62, with modifications. Glycerol stocks of the ATCC 49226 and H041 strains were cultured on fresh Chocolate Agar for approximately 24 h at 37 °C with 5% CO₂. The arising bacterial colonies were picked and suspended in FB to an OD₆₀₀ of approximately 0.7, which was inoculated with 1% into 325 ml FB supplemented with 1/100 volume 4.2% sodium bicarbonate. The culture was incubated at 37 °C with shaking at 160 rpm until an OD₆₀₀ of 0.6–0.8 was reached and cooled on ice. The cells were harvested by centrifugation at 5,000g and 4 °C for 10 min, washed twice with ice-cold PBS and maintained frozen at −80 °C. The cells were resuspended in PBS, sonicated and centrifuged at 15,000g and 4 °C for 30 min. The supernatant was ultracentrifuged at 136,000g and 4 °C for 1 h, washed and the pellet was resuspended in 0.5 ml PBS. The protein concentration of the membrane was approximately 1 mg ml⁻¹.

In the PBP-binding assay, the solutions containing 8 µg of the membrane and 1 µl of threefold serially diluted inhibitors (CRO and boro-PBPi 21 at final concentrations of 0, 0.4, 1.2, 3.6, 11, 33, 100 and 300 µM) were incubated at room temperature for 30 min, followed by the addition of 1 µl of 100 µM Bocillin-FL (final concentration of 10 µM) and incubation at room temperature for 30 min. Finally, the samples were mixed with 3.3 µl of 4×SDS sample buffer and incubated at 45 °C for 30 min. The PBPs were separated on SDS–PAGE gels and visualized by fluorescence (excitation, 472 nm; emission, 513 nm) using an Azure 600 imager (Azure Biosystems). The IC₅₀ value of each inhibitor was determined using Fiji ImageJ⁶³.

X-ray crystallography

The crystal structure of the transpeptidase domains of PBP2 derived from the CRO-reduced-susceptibility strain 35/02 of N. gonorrhoeae (tPBP2^35/02) and resistant strain H041 has been reported³⁰. Proteins were purified and concentrated to 13 mg ml⁻¹ in Tris–HCl (pH 7.8) with 10% glycerol and 500 mM NaCl, and then crystallized in the same way, as reported³⁰. These crystals occupy the P2₁2₁2₁ space group with one molecule in the asymmetric unit. Complexes of tPBP2^35/02 or tPBP2^H041 with 12, 15 and 21 were generated by soaking crystals with 1 µl of 10 mM boro-PBPi dissolved in PBS for 4 h (12 and 15) or 5–10 min (21), followed by flash-freezing in liquid nitrogen. For 15, a mixture of the S,R- and R,R-diastereomers was used for soaking; only the R,R form was observed in the electron density. For complexes with tPBP2^35/02, diffraction data were collected at a wavelength of 1.00 Å on an Eiger-16M detector at the SER-CAT 22-ID beamline at the Advanced Photon Source (Argonne, IL, USA); 200° of data were collected in 0.25° oscillations with an exposure time of 0.2 s per frame and a crystal-to-detector distance of 200 mm. For the tPBP2^H041–21 complex, data were collected at a wavelength of 0.978 Å on an Eiger2 9M detector at the NYX 19-ID beamline (National Synchrotron Light Source II). A total of 180° of data were collected in 0.25° oscillations with 0.1 s exposure per frame and a crystal-to-detector distance of 200 mm. The datasets were processed using HKL2000 (ref. ⁶⁴) and structures were solved by refinement against the apo structure of tPBP2^35/02 or tPBP2H^H041. Bound inhibitors were modelled into the |F_o| − |F_c| difference electron density map, followed by iterative cycles of model building using COOT⁶⁵ and refinement with PHENIX⁶⁶ for the tPBP2^35/02 structures and REFMAC5 (ref. ⁶⁷) for tPBP2^H041. The crystallographic data collection and model refinement statistics are listed in Supplementary Table 16.

Approval of facilities and procedures for animal studies

The facilities and procedures for animal studies were accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International and assured from the Office of Laboratory Animal Welfare. All animal rooms used for the PK and efficacy studies were maintained at a temperature range of 20–24 °C and humidity of 30–70%, with 12 h light–dark cycles, unless otherwise specified.

Pharmacokinetic analysis of boro-PBPi 18 in mice

Boro-PBPi 18 was formulated at a maximum concentration of 100 mg ml⁻¹ in PBS adjusted to pH 6–7 using NaOH. The PK study was conducted at the BioDuro-Sundia DMPK group (Jiangsu, China) using female BALB/c mice (7–9 weeks old; Vital River Laboratories). The study protocol (BioDuro BD-202102114) was approved by the Institutional Animal Care and Use Committee (IACUC). Boro-PBPi 18 was administered once via IV injection (3 mg kg⁻¹) or SC injection at 10, 30, 100, 300 or 1,000 mg kg⁻¹ (three mice per group, allocated randomly). Blood microsamples (30 µl each) were collected at following the time points following administration: 0.083, 0.25, 0.5, 1, 2, 4, 8 and 24 h. Plasma from each K₂EDTA-treated whole-blood sample was prepared using centrifugation at 2,000g and 4 °C for 5 min. The concentration of boro-PBPi in the plasma was measured using an ultra-performance liquid chromatography–tandem mass spectrometry (UPLC–MS/MS) platform (Waters Acquity UPLC and Sciex 6500). Each plasma sample was diluted in 5% trichloroacetic acid in water, followed by centrifugation at 2,000g and room temperature for 15 min. The supernatant was diluted twofold with 50% acetonitrile and loaded onto an Avantor ACE 5 C4 column (50 mm × 2.1 mm). Mobile phase (MP) A was 5 mM ammonium acetate and 0.05% formic acid in water and MP B was 0.1% formic acid in acetonitrile. Compound 18 was eluted in a linear gradient from 5% to 95% MP B over 1.2 min at a flow rate of 0.6 ml min⁻¹ and the elution was subjected to MS/MS analysis using multiple reaction monitoring in positive-ion mode with a transition of 611.10 to 429.00. Buspirone (5 ng ml⁻¹) and tolbutamide (50 ng ml⁻¹) in 5% trichloroacetic acid were used as the internal standard. The lower limit of quantification for 18 was 2 ng ml⁻¹.

Pharmacokinetic analysis of boro-PBPi 21 in mice

Boro-PBPi 21 was formulated at a maximum concentration of 30 mg ml⁻¹ in PBS adjusted to pH 6–7 using NaOH. The PK study was conducted at the BioDuro-Sundia DMPK group using male CD-1 mice (7–9 weeks old; Vital River Laboratories). The IACUC-approved protocol number was BioDuro BDW-2201-0007. The PK study of 21 was conducted in a manner similar to that of 18. The bioanalytical method used to quantify 21 was performed using a UPLC–MS/MS platform (ExionLC AD system and a SCIEX 7500). Each sample prepared from the plasma was loaded onto an ACQUITY UPLC peptide BEH C18 column (1.7 μm, 300 Å, 50 mm × 2.1 mm). MP A was 1% formic acid in water and MP B was 1% formic acid in acetonitrile. Compound 21 was eluted at a flow rate of 0.5 ml min⁻¹ in a linear gradient from 5% to 55% of MP B over 0.1 min and an isocratic hold at 55% MP B for 0.8 min, followed by a linear gradient to 95% of MP B over 0.1 min; the elution was subjected to MS/MS analysis using multiple reaction monitoring in positive-ion mode with a transition of 647.4 to 465.0. Tolbutamide (0.5 ng ml⁻¹ in 5% trichloroacetic acid) was used as an internal standard. The lower limit of quantification for 21 was 2 ng ml⁻¹.

The parameters derived from non-compartmental PK analysis were as follows: plasma concentration at time 0 after IV administration (C₀), plasma t_½, plasma C_max, AUC of the concentration versus time curve from t = 0 to infinity (AUC_inf), volume of distribution at steady-state (V_SS), clearance (CL), time to maximum plasma concentration (T_max) and bioavailability from the SC dose (F_SC). In each of these instances, variability was related to concentrations below the lower limit of quantification, which restricted the characterization of the terminal phase in at least one mouse per group (Supplementary Table 2).

Efficacy assessment of boro-PBPi 18 in the FA1090 vaginal infection model

Efficacy analysis was performed by Eurofins Pharmacology Discovery Services, through support by CARB-X and the Pre-Clinical Services group at the National Institute of Allergy and Infectious Diseases, using ovariectomized and 17β oestradiol-treated female BALB/c mice⁶⁸. The study protocol (PDS IM005-07302018) was approved by the IACUC. Groups of five immunocompetent female ovariectomized BALB/c mice (5–6 weeks old) were used for the efficacy study. The mice were obtained from BioLASCO Taiwan and acclimated for three days. After acclimation, an ovariectomy was performed on 4-week-old mice. The period of surgical recovery and acclimation was at least seven days. The mice were treated with meloxicam (SC injection; 20 mg kg⁻¹) if signs of pain or distress were observed during this period. Before infection, the mice were administered SC injections of oestradiol solution (0.23 mg per mouse) two days before infection (day −2) as well as on the day of infection (day 0). To minimize the indigenous vaginal bacteria, the animals were treated (q12h) with streptomycin (1.2 mg per mouse) and vancomycin (0.6 mg per mouse) by intraperitoneal injection along with 0.4 mg ml⁻¹ trimethoprim sulfate supplied in the drinking water. These antibiotic treatments started two days before infection and continued daily until the end of the study. On day 0, the mice were inoculated intravaginally with N. gonorrhoeae FA1090 (ATCC 700825; 0.02 ml (2.56 × 10⁶ CFU) per mouse) under anaesthesia induced by intraperitoneal injection of 80 mg kg⁻¹ pentobarbital, followed by rinsing the vagina with 50 mM HEPES (pH 7.4; 30 μl). The mice were administered SC injections of 18 twice a day with a 12 h interval (q12h) starting at 2 h post infection for one day (five mice per group, allocated randomly). At 2 or 26 h after infection, the mice were killed by CO₂ asphyxiation to harvest their vaginal lavage fluid. For vaginally infected mice, vaginal lavage was performed twice with 200 μl GC broth containing 0.05% saponin to recover vaginal bacteria. Lavage samples from each mouse were pooled in a total volume of 500 μl. Bacterial burden in the lavage fluids was determined by performing tenfold serial dilutions and plating 0.1 ml of each dilution onto Chocolate Agar plates. The bacterial burden (CFU ml⁻¹) of the lavage fluid was calculated.

Efficacy assessment of boro-PBPi 21 in the H041 vaginal infection model

The in vivo efficacy of 21 was tested in female NCI BALB/c mice (6–7 weeks old) in A.E.J.’s laboratory at the Uniformed Services University, as described³³. The mice were obtained from Charles River Laboratories. This study protocol was approved by the Uniformed Services University IACUC under protocol MIC23-759. Using a simulation based on the mouse PPB (19%) and the PK profile in male CD-1 mice, four dosing regimens for 21 were selected to vary %fT > MIC from 20% to 100% in this efficacy study. Mice in the dioestrus or anoestrus stage of the oestrous cycle were randomized into six groups and pretreated with 17β oestradiol as well as antibiotics (streptomycin and trimethoprim) to suppress the overgrowth of the commensal microbiota and to support N. gonorrhoeae colonization. The pretreated mice were vaginally infected with H041 (1–2 × 10⁴ CFU per mouse) two days before treatment with 21 via SC injection, either once or alternate dosing regimens. Control groups were administered CRO (120 mg kg⁻¹ SC, q8h) or vehicle control (PBS). Vaginal mucus was quantitatively cultured for N. gonorrhoeae for eight consecutive days post treatment. A portion of the swab was also inoculated onto heart infusion agar to monitor the presence of facultative aerobic commensal microbiota. Vaginal polymorphonuclear leukocyte influx was assessed on each culture day by cytological examination of stained vaginal smears and reported as the percentage of polymorphonuclear leukocytes among 100 vaginal cells. Efficacy was measured by comparing differences in the clearance rate and average bacterial burden over eight days following treatment with 21, CRO or compound vehicle PBS.

In vitro safety and selectivity assays for boro-PBPi 21

The assays that examined PPB, cytotoxicity, haemolysis, mitochondrial toxicity and chromosomal aberrations (micronucleus assay) as well as several in vitro inhibition assays (CYP450s and human proteases) or binding (hERG) assays for 21 were performed using established methods and controls that behaved as expected. Eight cytochrome P450 enzymes (CYP1A2, CYP2C19, CYP2D6, CYP3A4, CYP2E1, CYP2B6, CYP2C19 and CYP2C8), recommended by the US FDA⁶⁹, were used to predict metabolism-mediated drug–drug interactions. A detailed description of the methods and raw data are provided in Supplementary Table 13.

Pharmacokinetic analysis of boro-PBPi 21 in Sprague Dawley rats

The PK study was performed in male Sprague Dawley rats (5–7 weeks old; Hilltop Labs Animals, Inc.) after a single IV or IM dose at QPS, LLC. This study protocol was approved by the IACUC under QPS IACUC protocol number 005. The designated animal room where the rat PK study was performed was maintained with 12 h light–dark cycles at a temperature range of 20–26 °C and relative humidity of 20–70%. All dosing formulations of 21 were freshly prepared as clear solutions. For IM dose preparation, 21 was dissolved in vehicle (45:55 (vol/vol) 1 N NaOH:1.6×PBS) by vortexing, sonication and stirring to achieve a concentration of 75 mg ml⁻¹ at pH 6.8. A 0.6 mg ml⁻¹ IV dose formulation was prepared by diluting the IM formulation with 1×PBS (pH 7.4; approximately final pH 7.1) and filtering through a 0.22-µm filter (Millex-GV). Male Sprague Dawley rats (weight of approximately 270 g) were administered 21 via IM and IV routes. Blood samples (0.3–0.4 ml) were collected via the tail vein at pre-dose (0 h) and 0.083, 0.25, 0.5, 2, 4, 8 and 24 h post treatment from IV-dosed rats (n = 3), and pre-dose (0 h) and 0.5, 1, 2, 4, 8 and 24 h post treatment from IM-dosed rats (n = 3 rats per group, allocated randomly). All blood samples were collected into tubes containing K₂EDTA on wet ice and centrifuged at approximately 3,800g and 4 °C for 3 min within 40 min of blood collection. Terminal blood samples were collected in 10-ml Vacutainer tubes and centrifuged at 3,300g and 4 °C for 15 min. All plasma samples were snap-frozen on dry ice and stored at approximately −70 °C until bioanalysis to determine the concentration of 21 in plasma. The PK data (Extended Data Fig. 3) were evaluated using non-compartmental analysis (Phoenix WinNonlin, version 8.3, Certara USA Inc.) to determine the IM bioavailability in rats. The average concentrations from three rats per dose group were included. The terminal t_½, R² and adjusted R² values were 0.63 h, 0.98 and 0.98, respectively, following IV administration of 3 mg kg⁻¹ 21 to rats, and 0.95 h, 0.93 and 0.89, respectively, following IM administration of 30 mg kg⁻¹ 21. The IM bioavailability was calculated by dividing the dose-normalized AUC_inf based on the last observed concentration (observed AUC_inf dose in h kg ng ml⁻¹ mg⁻¹) when administered IM by the observed AUC_inf when administered IV (i.e. 66% (1030.81 / 1558.40)).

Statistics and reproducibility

No statistical method was used to pre-determine sample size. No data were excluded from the statistical analyses.

Reporting summary

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