Reduced susceptibility to ceftriaxone in Neisseria gonorrhoeae is spread internationally by genetically distinct gonococcal populations

Sir, Antimicrobial resistance in Neisseria gonorrhoeae is a major public health challenge, with gonococci developing resistance to multiple classes of antibiotics. There are now increasing numbers of reports of gonococci with reduced susceptibility to extended-spectrum cephalosporins (ESCs), and documented treatment failures of genital tract gonorrhoea using oral ESCs in Japan and Hong Kong.1,2 Reduced susceptibility to ESCs in N. gonorrhoeae has been associated with alterations in the gonococcal penicillin-binding protein 2 (PBP2), including mosaic PBP2 (considered a key alteration of interest) and PBP2 substitutions A501V and A501T, as well as an adenine deletion in the mtrR efflux pump promoter and mutated alleles of the PorBIb outer membrane porin protein.3–5 Of concern is a recent study demonstrating that mosaic PBP2 in combination with A501V could lead to ceftriaxone resistance.4 In this study, we used multilocus sequence typing (MLST) to further examine the genotypic diversity of 20 gonococci exhibiting a range of ceftriaxone MICs and including isolates with and without mosaic PBP2 or A501V/T (Table 1). The isolates were from a previous study5 and were retrospectively selected based on the diversity of ceftriaxone MIC, N. gonorrhoeae multiantigen sequence type (NG-MAST)6 and geographical spread; isolates were from Australia (n = 14), Japan (n = 2), Korea (n = 1) and Hong Kong (n = 1), together with two WHO reference strains (WHO-K and WHO-L).7 MLST was performed on all 20 isolates, as previously described.8,9 Briefly, abcZ, adk, aroE, fumC, gdh, pdhC and pgm sequences were amplified by PCR, and then sent for automated fluorescent sequencing at the Australian Genome Research Facility (http://www.agrf.org.au/). MLST types were determined using the Neisseria MLST database (http://pubmlst.org/neisseria/). MLST data were compared with ceftriaxone MICs, NG-MAST type and PBP2 type. Summary of results sorted by MLST sequence type Aust, Australia; HK, Hong Kong; ST, sequence type; CRO, ceftriaxone; ND, not described, these MLST types have not been previously described. Bold text indicates raised MIC. aPBP2 type according to Ito et al.3 and Whiley et al.5 A501V, substitution at amino acid 501 of PBP2 protein; mosaic, mosaic PBP2. Summary of results sorted by MLST sequence type Aust, Australia; HK, Hong Kong; ST, sequence type; CRO, ceftriaxone; ND, not described, these MLST types have not been previously described. Bold text indicates raised MIC. aPBP2 type according to Ito et al.3 and Whiley et al.5 A501V, substitution at amino acid 501 of PBP2 protein; mosaic, mosaic PBP2. A summary of results is provided in Table 1. The results show that an array of gonococcal genotypes exhibited reduced susceptibility to ceftriaxone: 14 isolates had raised ceftriaxone MICs (0.016–0.125 mg/L), occupying six different MLST and 13 different NG-MAST types. Four of these six MLST types (sequence types 1588, 1590, 1901 and 7363) were previously observed among Japanese gonococci exhibiting reduced susceptibility to cefixime.9 Of particular note was MLST type 7363, which accounted for the majority of isolates harbouring the mosaic PBP2 allele in the previous Japanese study and for all seven such isolates in this study, including four Australian isolates. These seven isolates were further distinguished using NG-MAST and comprised six different NG-MAST types, including at least three closely related types. For example, NG-MAST types 1424 and 835 (isolated in Australia in 2002 and 2004, respectively) had identical tbpB sequences and differed by only two nucleotides in the porB sequence. Likewise, NG-MAST types 1424 and 1677 (both isolated in Australia in 2002) differed by one nucleotide and a 12 base deletion in porB. Of further interest were isolates 15 and 16, which were isolated in Australia in 2002 and in Japan in 2003, respectively. Both harboured mosaic PBP2 and could not be differentiated using either MLST or NG-MAST. The six isolates possessing the PBP2 501 alteration comprised five different MLST and six different NG-MAST types. As previously suggested, mosaic PBP2 appears to be associated with a limited number of putative parent MLST types, with 7363 being a major clone.9 However, the various NG-MAST types associated with MLST type 7363 in this study suggest that the original strains with mosaic PBP2 are diversifying, and this would go some way to explaining the considerable volatility of NG-MAST types observed for gonococci with mosaic PBP2 in Sydney, Australia, in 2007 and 2008.10 In contrast, isolates with PBP2 501 substitutions occupied several different MLST and NG-MAST types, indicating that such isolates are genetically more distinct than mosaic-harbouring isolates. This also provides further evidence that PBP2 501 substitutions are a common spontaneous alteration. Combined, the above data provide further evidence of the international spread of gonococci exhibiting reduced susceptibility to ESCs and that such strains are genotypically diverse. From a molecular epidemiological perspective, these data also highlight that the enhanced discriminatory power of NG-MAST, as opposed to MLST, is more suited to monitoring the local spread of gonococci with reduced susceptibility to ESCs, including establishing their distribution via sexual networks. Overall, our findings are consistent with those of Ohnishi et al.9 and provide further evidence that emerging resistance to ceftriaxone is not due to the expansion of a single gonococcal lineage. Rather, it is evident in genetically distinct gonococcal populations, and presumably arises through a combination of genetic exchange and spontaneous mutation. This work was supported by core funding of the Queensland Paediatric Infectious Diseases Laboratory provided by the Royal Children's Hospital Foundation. None to declare. This study was conducted as part of the reference work of the Australian National Neisseria Network.