Genotyping of Mycoplasma genitalium Suggests De Novo Acquisition of Antimicrobial Resistance in Queensland, Australia

LETTER

Mycoplasma genitalium is a sexually transmitted organism that causes nongonococcal urethritis in men and pelvic inflammatory disease in women. In Australia, there is concern over the emergence and spread of antimicrobial resistance (AMR) in M. genitalium. While high levels of AMR have been reported in men who have sex with men (1, 2), it is not clear if AMR is specific to a particular strain and there is no information on the diversity of M. genitalium strains in Australia.

We successfully genotyped 89/111 M. genitalium-positive samples collected from 2013 to 2019 (majority of samples from 2017 to 2019) obtained from different patients as part of a previous study (1). We also genotyped eight additional samples from four patients to investigate changes in genotypes over time. Genotyping consisted of PCR and Sanger sequencing of mgpB and MG_309, as described elsewhere (3–5). Sequence typing was compared to epidemiological data and the presence of macrolide and parC quinolone AMR markers as previously described (1). This study was approved by the Children’s Health Queensland human research ethics committee (HREC/12/QRCH/139).

Of the 89 individual patient samples, we identified 39 mgpB sequence types, of which 21 were previously identified and 18 were novel, i.e., not previously published or available within GenBank. Of the novel mgpB types, 15 were identified only once in our patients, two occurred twice, and one was identified across five different samples. Similarly, for MG_309, we identified 23 sequence types, of which eight were novel repeat number variants and one was a new strain type. For the purposes of this study, the mgpB and MG_309 data were combined to assign a genotype for each sample, as previously described (6).

Results of M. genitalium genotyping for individual patient samples can be seen in Fig. 1. In brief, we saw 68 different genotypes across 60 male and 29 female patients, and they were observed in 1 to 5 samples each (average 1.3 samples per genotype). While the presence or absence of resistance markers to macrolides and/or quinolones was stable for some genotypes, in most cases, the specific AMR mutations showed a more random distribution, both within and across different genotypes. For example, the two samples belonging to genotype 61 had identical macrolide and quinolone AMR profiles, whereas genotypes 8 (n = 5) and 30 (n = 5) both comprised samples with wild-type and mutant parC sequences (Fig. 1). Moreover, the most common parC resistance mutation, S83I, was observed across 32 different genotypes, followed by D87N (7 genotypes) and D87Y (3 genotypes). When comparing gender, specimen site, and location, we did not see any correlation with M. genitalium genotype. When looking at consecutive samples from the same patient, we observed no change in genotype and AMR profiles in three of the four patients, consistent with chronic M. genitalium infection or reinfection with the same strain (days between samples ranged from 8 to 99 days). In one patient, we observed different genotypes in samples collected 104 days apart, suggesting reinfection with a different M. genitalium strain.

FIG 1.

Neighbor-joining tree of Mycoplasma genitalium genotyping from 89 individual patient samples. Genotypes were determined based on the presence of single nucleotide polymorphisms (SNPs) in the mgpB gene fragment and tandem repeat unit length of the MG_309 gene. Concatenated sequences were used to construct the phylogenetic tree, including results of the ResistancePlus MG macrolide resistance screening and Sanger sequencing of parC for detection of quinolone resistance.

It should be noted that the genotyping loci used here include an adhesin (mgpB) and an immunogenic surface-exposed lipoprotein (MG_309) (7, 8), which are likely to be under greater selective pressure to change as they are in constant contact with the host immune system during infection. The highly diverse genotypes/strains reported here are consistent with previous reports from geographically diverse regions of the world (4, 9, 10). Within our study, the lack of association between genotypes and AMR suggests that there are potential inherent limitations in utilizing this genotyping scheme for ongoing use in our local population. Additional whole-genome sequencing studies may be warranted to fully understand the emergence and spread of M. genitalium in Queensland, Australia, both at the population level and within specific patient groups.

In summary, we confirm the presence of a wide range of M. genitalium genotypes in Queensland, Australia. Geographically and temporally diverse samples contained the same genotype, suggesting that there is ongoing transmission of some strains; however, the variation in AMR mutations across genotypes is highly suggestive of de novo acquisition of resistance rather than being underpinned by clonal spread of specific resistant strains.

Data availability.

Novel mgpB and MG_309 sequences were deposited into GenBank (accession numbers MT476043 to MT476061).