Multidrug-resistant Neisseria gonorrhoeae infection in heterosexual men with reduced susceptibility to ceftriaxone, first report in Thailand

The global rapid emergence of azithromycin/ceftriaxone resistant Neisseria gonorrhoeae threatens current recommend azithromycin/ceftriaxone dual therapy for gonorrhea to ensure effective treatment. Here, we identified the first two N. gonorrhoeae isolates with decreased ceftriaxone susceptibility in Thailand. Among 134 N. gonorrhoeae isolates collected from Thai Red Cross Anonymous Clinic, Bangkok, two isolates (NG-083 and NG-091) from urethral swab in male heterosexual patients had reduced susceptibility to ceftriaxone (MICs of 0.125 mg/L). Both were multidrug resistant and strong biofilm producers with ceftriaxone tolerance (MBEC > 128 mg/L). NG-083 and NG-091 remained susceptible to azithromycin (MIC of 1 mg/L and 0.5 mg/L, respectively). Reduced susceptibility to ceftriaxone was associated with alterations in PBP2, PBP1, PorB, MtrR, and mtrR promoter region. NG-083 belonged to sequence type (ST) 7235 and NG-091 has new allele number of tbpB with new ST. Molecular docking revealed ceftriaxone weakly occupied the active site of mosaic XXXIV penicillin-binding protein 2 variant in both isolates. Molecular epidemiology results revealed that both isolates display similarities with isolates from UK, USA, and The Netherlands. These first two genetically related gonococcal isolates with decreased ceftriaxone susceptibility heralds the threat of treatment failure in Thailand, and importance of careful surveillance.


Results
N. gonorrhoeae clinical isolates from urethral swabs in male patients had reduced susceptibility to ceftriaxone. Among the 134 N. gonorrhoeae clinical isolates collected during 2016-2019, two isolates (NG-083 and NG-091) which were isolated in 2017 from urethral swabs in male heterosexual patients at Thai Red Cross AIDS Research Centre, Anonymous Clinic, had reduced susceptibility to ceftriaxone (agar dilution MICs of 0.125 mg/L). Both isolates were resistant to penicillin G, tetracycline, ciprofloxacin, and gentamicin, while remaining susceptible to azithromycin ( Table 1). The patients were a 30 and 32-year-old long term foreign resident men (South African and Australian) in Thailand, symptomatic with dysuria, or penile discharge at the time of specimen collection, and negative for HIV. Both were isolated cases and no relation to each other epidemiologically. One patient develops symptoms one week after unprotected oral sex with a female, while the second patients origin of infection was not disclosed. Both were treated for gonorrhea with Ceftriaxone 250 mg, and Azithromycin 250 mg as per current CDC guidelines 5 . Patients di not return to the clinical follow-up and therefore unable to be verified as having been cured or not.
N. gonorrhoeae isolates with reduced susceptibility to ceftriaxone are associated with alterations in PBP2, PBP1, PorB, MtrR, and mtrR promoter region. The isolates NG-083 and NG-091 with reduced susceptibility to ceftriaxone harbor specific ceftriaxone resistance patterns (Table 2). NG-083 and NG-091 had an L421P substitution in PBP1, the mosaic PBP2 patterns XXXIV ( Fig. 1) (Supplementary Figs. S2 and S3), an adenine deletion in the 13-bp inverted repeat sequence of the mtrR promoter region, an H105Y substitution in the MtrR repressor, and G120K and A121N substitutions in PorB porin. Mosaic XXXIV PBP2 variants contain up to 52 amino acid alterations compared with the wild-type PBP2 (Table 3). Figure 2 summarises the mosaic XXXIV PBP2 variant showing the location of all amino acid alterations drawn in Pymol using the crystal structure of a soluble form of N. gonorrhoeae wild-type PBP2 (extracted from rcsb.org/PDB: 3EQU). Table 1. Patients summery and antimicrobial susceptibility profile against the N. gonorrhoeae isolates with reduced susceptibility to ceftriaxone (n = 2), Thailand, 2016Thailand, -2018. HIV human immunodeficiency viruses, MIC minimum inhibitory concentration, IM intramuscular, MBEC minimal biofilm eradication concentrations were categorized as responsive reaching about 90% of the total non-viable cells within a given antibiotic concentration range.      www.nature.com/scientificreports/ antagonistic effects were not found in all antibiotic combinations against both isolates (Table 5). Additive effects were found with ceftriaxone plus azithromycin, fosfomycin, or gentamicin. Indifferent was found in ceftriaxone plus ertapenem. At 0.5X MIC of ceftriaxone, bacterial growth was rapidly inhibited within 8 to 12 h in both isolates, but the concentration at 0.125X and 0.25X MIC was unable to inhibit bacterial growth within 24 h (Fig. 5). However, 0.125X MIC of ceftriaxone plus 0.5X or 1X MIC of azithromycin, and 0.25X MIC of ceftriaxone plus 0.5X or 1X MIC of azithromycin showed effective killing against NG-083 and NG-091.
N. gonorrhoeae isolates with reduced susceptibility to ceftriaxone produce strong biofilms with ceftriaxone tolerance. The isolates NG-083 biofilm had significantly more biomass than the NG-091 biofilm (P = 0.004) (Fig. 6). Both isolates were strong biofilm producers with an atypical ellipse-shaped structure or clumps, and well tolerated to a ceftriaxone concentration up to 128 mg/L (Fig. 7). The analysis showed that the live/dead bacterial ratio of NG-083 and NG-091 biofilms treated with ceftriaxone MIC (0.125 mg/L) concentration was not different in comparison with non-treated group (P > 0.05) (Fig. 7). Both isolates displayed minimal biofilm eradication concentration for ceftriaxone around 128 mg/L concentration (Table 1).

Discussion
We identified a pair of gonococcal isolates with decreased susceptibility to ceftriaxone, ciprofloxacin, tetracycline, penicillin G, gentamicin, and ertapenem, while remaining susceptible to azithromycin. Both isolates appeared to be unrelated, with for example two different sequence types according to the NG-MAST but same sequence type according to Ng-STAR. The genome sequencing analysis revealed alterations in PBP2, PBP1, PorB, MtrR, and mtrR promoter region in both isolates. Molecular docking studies suggested ceftriaxone weakly occupies the active site of mosaic XXXIV PBP2 variant, which may explain the decreased ceftriaxone activity. Interestingly, both isolates with reduced susceptibility to ceftriaxone demonstrated strong biofilm production that was associated with ceftriaxone tolerance at concentrations higher than the MIC. Molecular epidemiology results revealed that both isolates display similarities with isolates from the UK, USA, and the Netherlands. The most effective combination was ceftriaxone plus azithromycin which showed bactericidal activity for both isolates.  28 , and Canada 29 . Moreover, these two isolates are genetically closely related to two N. gonorrhoeae isolated from the urethra of male patients from Las Vegas in USA in 2009, which also exhibited reduced susceptibility to cefixime, with most of the isolates having mosaic penA allele XXXIV 30 . As given the 8 years between specimen collection of the Las Vegas isolates and two isolates in our study, it's difficult to relate an epidemiological connection between these two and the Las Vegas isolates. However, Thailand is a popular getaway for sex tourism 3 , there is the possibility that these two isolates (NG-083 and NG-091) may have transmitted through travelers. However, because we did not have samples from other sexual partners of these two men to provide broader context, is difficult to established clear epidemiological relations.
To our knowledge, no isolates have previously been reported in Thailand with decreased ceftriaxone susceptibility. The data from the National Antimicrobial Resistance Surveillance Center, Thailand (NARST) 31 , and other studies from Thailand 32 showed that 100% of N. gonorrhoeae isolates were susceptible to cefixime and ceftriaxone. A report from the United Kingdom documented a gonorrhea treatment failure with the isolate exhibiting a ceftriaxone MIC = 0.25 µg/mL and azithromycin MIC = 1 µg/mL 12,33 . In comparison, our isolates demonstrate similar susceptibility to azithromycin, while their ceftriaxone MICs were only a single dilution lower than the ceftriaxone MIC from the UK isolate, suggesting that if the strains identified in Thailand develop higher   34 . In addition, the improper use or misuse of antibiotics, and overthe-counter drug usage in Thailand, may have contributed to the development of higher ceftriaxone MICs 35 . It is possible that the persistence of such isolates within biofilm, seen with infections involving higher MICs, reflected a slower killing of N. gonorrhoeae and may severely complicate gonorrhea treatment. The threat of multidrug-resistant gonorrhea leading to treatment failure in Thailand with the current recommended dual therapy remains present. Our findings emphasized the value of the CDC recommendations for laboratories to maintain culture-based methods to detect antimicrobial-resistant N. gonorrhoeae, particularly for patients with possible treatment failure. In addition, we also emphasized the potential value of genomic monitoring to detect antimicrobial-resistant N. gonorrhoeae. It is important that clinicians be on high alert so that treatment failures can be identified and reported promptly to the local department of health. Rapid detection and effective treatment may prevent sequelae, allow partners to be identified and treated in a timely manner, and prevent or slow further transmission of resistant strains. Furthermore, given the study findings, we advised continued practice of dual therapy with ceftriaxone 250 mg plus azithromycin 1gm in Thailand. However, it is necessary to have new strategies such as resistance guided therapy for N. gonorrhoeae infections and contact tracing to identify the origin of the gonococcal infections as preventive measures. Finally, there are urgent need for development of new or combination therapies to tackle the multidrug resistant N. gonorrhoeae infections.
The strengths of this study are that all infections were diagnosed by culture, which has 100% specificity for gonorrhea. Phenotypic antimicrobial susceptibility testing. The isolates obtained from patients were analysed using both broth micro-dilution and agar plate dilution methods (Supplementary methods). MIC interpretive criteria were in accordance with the Clinical and Laboratory Standards Institute (CLSI) 36 or European Committee on Antimicrobial Susceptibility Testing (EUCAST) 37 when available. In the present study in vitro decreased susceptibility to ceftriaxone was defined as having an MIC of > 0.064-0.125 mg/L.

Molecular epidemiological typing.
The presence of carA and orf1 genes were detected by specific PCR primers (Supplementary methods Table S1) (BioDesign, Thailand) to confirm N. gonorrhoeae identification as described previously 38,39 . Detection and characterization of ceftriaxone resistance mechanisms. The ceftriaxone resistance mechanisms in N. gonorrhoeae isolates were investigated by penA, mtrR, ponA, and porB genes using PCR and automated DNA sequencing (1st BASE Inc, Malaysia) (Supplementary methods Table S3). The PCR products of genes were purified using HiYieldTMGel/PCR fragments extraction kit as described by the manufacturer (RBC Bioscience, Taiwan). Sequencing was conducted using BigDye terminator cycling conditions.  40 . The internal fragments of 2 highly polymorphic antigen-encoding loci including outer membrane proteins (por) and transferrin binding protein unit B (tbpB) 40 were amplified, purified (HiYieldTMGel/PCR) and sequenced (Supplementary methods Table S4) as previously described 40 . However, because NG-MAST no longer assigns new STs, the two N. gonorrhoeae with reduced susceptibility to ceftriaxone were further analyzed with NG-STAR. Sequence of 7 genes (penA, mtrR, porB, ponA, gyrA, parC and 23S rRNA) were extract from genome sequence of NG-083 and NG-091 isolates, then the sequence were analyzed using NG-STAR database (https:// ngstar. canada. ca/ allel es/ query? lang= en). Preparation of sequencing reactions were performed by the 1st BASE Inc, Malaysia, as described above.
Whole-genome sequencing and genome assembly. The genomes of two N. gonorrhoeae with reduced susceptibility to ceftriaxone were sequenced using Illumina sequencing platform (2 × 150 paired-end, Illumina, Inc., USA). The genomic DNA was extracted with the PureLink PCR Purification kit and quantified in   Phylogenetic analyses. The core genome from draft genomes of N. gonorrhoeae NG_83 and NG_91 clinical isolated in Thailand and from other N. gonorrhoeae WGS investigations conducted elsewhere (available on NCBI database) were determined for the number of single nucleotide polymorphisms (SNPs) by Core-Genome SNP Analysis. A reference was randomly selected among the genome sequences to generate a core genome alignment and a phylogenetic tree was constructed using a core SNP alignment. Antibiotic synergy analysis using microdilution checkerboard assay. The synergistic activities of antibiotics were screened against 2 strains of N. gonorrhoeae isolates with reduced susceptibility to ceftriaxone using checkerboard method (Supplementary methods).
Time kill assay. Two strains of N. gonorrhoeae with reduced susceptibility to ceftriaxone that showed the best synergistic activity by checkerboard method was confirmed using time-kill assay. All experiments were performed at least three times (Supplementary methods).
Biofilm formation, quantification and classification. Biofilm formation in a 96-well-microtitre-plate format was performed as described previously [49][50][51] (Supplementary methods). Two methods 52,53 were used to quantify and classify the biofilm by Crystal Violet staining and and confocal laser scanning microscopy, performed in triplicate, and repeated three times.

Minimal biofilm eradication concentrations (MBEC). Minimal biofilm eradication concentrations
(MBEC) were established as described previously [49][50][51] by adding the serially diluted antibiotics to mature biofilms and incubating at 37 °C for 24 h and then staining with PrestoBlue (Thermo Fisher Scientific). All experiments were performed in triplicate and repeated three times.

Statistical analyses.
Fisher's exact test (two-tailed) was used to verify the association between mutations in resistance determinants and penicillin resistance in N. gonorrhoeae isolates by using the R statistical pack-  www.nature.com/scientificreports/ age. Biofilm experiments were analyzed by confocal laser scanning microscope (biomass, Live/Dead ratio and biovolume inhibition). MATLAB-based tool PHLIP (without connected volume filtration) were used to calculate descriptive parameters of biofilms (including biovolume, substratum coverage, area-to-volume ratio, spatial spreading and 3D colocalization) from the integrated total of each individual slice of a thresholded z-stack as described previously 49,50,54 . The calculation of the different proportions of green (live bacteria) as well as red and yellow/colocalized (dead bacteria) biovolumes from the analyzed stacks were using the ' colocalization in 3D' value and the parameters 'red' , 'green' , and 'total biovolume' (in μm 3 ) generated by the PHLIP software as described previously 48,49,53 . A biofilm was considered affected by an antibiotic within the given concentration range when there is a constant increase in the red + yellow (RY) biovolume fraction within the given antibiotic concentration range and this fraction is at least 80% of the total biovolume. The data were compared by either unpaired two-tailed Student's t-test or unpaired two-tailed Mann-Whitney's U test. Statistical significance was accepted at p < 0.05, p < 0.01, p < 0.001, and p < 0.0001.