We explored possible genomic rearrangements because previous studies have described distinct genome configurations among emm3 consensus reference genomes MGAS315 and SSI-1, where SSI-1 contained large-scale genomic rearrangements compared with MGAS315 (23). Hybrid genome assembly for 14 S. pyogenes emm3.93 isolates (5 from the Netherlands and 9 from England) revealed that 12 isolates shared a genomic configuration that differed from MGAS315 by a ≈200-kb inversion around the terminus (ter) and from SSI-1 by a ≈521-kb inversion around the origin of replication (ori) (Figure 3, panel B; Appendix Figure 2). Of the 12 isolates, 11 belonged to emm3.93 surge isolates in each of the 3 clades and 1 was obtained from a patient with invasive disease from 2017 in England (Figure 3, panel A). The 200-kb genomic inversion around the ter is prophage-mediated and results in the movement of phage-encoded superantigen gene ssa and holin genes between prophages MGAS315.1 and MGAS315.2 (24) (Figure 3, panel B). The ≈521-kb inversion around the ori is spanned by 2 repeat regions of 6,200 bp (Figure 3, panel B). The inversions around the ori and the ter can occur independently, as indicated by strain ABC020055675, which lacks prophages MGAS315.1 and MGAS315.2 (Figure 3, panel B; Appendix Figure 2). Of interest, the genomic configuration of the remaining 2 isolates was identical to the consensus MGAS315 genome (Appendix Figure 2). One isolate was obtained from a patient with invasive disease in 2017 in the Netherlands with a common ancestor to clade 3 and the other from a patient with noninvasive infection from clade 1 (Figure 3, panel A). Although we analyzed a limited number of strains, our observations suggest the described genome configuration in the 11 recent emm3.93 isolates confers a survival advantage to S. pyogenes.
We report the rapid synchronous upsurge of S. pyogenes emm type 3.93 in the Netherlands and England during November 2023-March 2024. Epidemiologic analysis showed that emm3.93 was isolated more often from children 6-17 years of age with iGAS in both countries. The data from the Netherlands revealed a significant association between emm3.93 and pneumonia or pleural empyema, which was corroborated by an increase of emm3.93 among pleural isolates in England. Although the numbers were low, emm3.93 showed an increased risk for causing meningitis, which was significant in the data from England. Furthermore, emm3.93 was associated with streptococcal toxic shock syndrome in the Netherlands. Despite those concerning associations, the data from England showed no significant elevation in 7-day mortality for emm3.93 compared with mortality associated with other contemporaneous emm types.
Genomic analysis revealed 3 distinct clades among recent emm3.93 surge isolates. Clades 1 and 2 consisted only of isolates from England, whereas emm3.93 clade 3 contained a mix of isolates from the Netherlands and England. This analysis contrasts with previous analysis of the post-COVID-19 emm1 iGAS isolates, which showed predominantly country-specific surges of emm1 and little evidence of transmission of UK isolates to the Netherlands (4,5). This finding suggests that emm3.93 has had cross-border transmission events.
Analysis of complete genomes revealed a shared genomic configuration in all but 1 recent isolate from the 3 observed clades. When compared with other publicly available emm3 consensus reference genomes, the genomic configuration differed by either a ≈200-kb inversion around the ter or a ≈521-kb inversion around the ori. An identical genomic configuration was also identified in an invasive emm3.93 isolate from England from 2017, suggesting that this genome configuration was already circulating in the population before the surge. Of interest, emm3.93 caused a small surge of iGAS in England in 2018 (22) but did not expand to the levels observed in 2023-2024, nor did other countries report an increase in iGAS caused by emm3.93. We speculate that this 2017 isolate from England could have been a precursor to the 2023-2024 surge isolates but did not have a major advantage in the pre-COVID-19 climate because of sufficient population immunity.
Inverted repeats between prophages MGAS315.1 and MGAS315.2 have likely permitted the observed prophage-mediated homologous recombination. This phenomenon has been observed before in emm3.1 strains, which contained an alternative large-scale genomic rearrangement of 1 Mb because of chromosomal inverted repeats and was associated with invasive disease (23). However, because the inversion around the ori is also detected in an isolate without prophages MGAS315.1 and MGAS315.2 and without the inversion around the ter, we propose an updated view of the relationship between genomic configurations within emm3 S. pyogenes.
Because only single complete genomes are available of each of the reference strains, only a snapshot of the genomic configuration is being represented and then determined as consensus. Although our analysis of the individual isolates did not indicate a mixed population of inverted and noninverted genomes, care should be taken when interpreting the stability of an inversion that may occur in recombination hot spots with inverted homologous regions (25), because the switch back to its previous configuration may occur stochastically. For example, a recent emergent emm82 strain was detected to display an array of genome configurations among closely related isolates (26). Regardless, further information on the frequencies and selective pressures that cause inversions needs to be determined to make strong associations with epidemiologic data. Our dataset provided 2 observations that may suggest the observed genome configuration in the surge isolates conferred a survival benefit or affected virulence (27). First, an isolate from 2017 sharing a close common ancestor with the strains in clade 3 lacked the ≈200-kb inversion around the terminus and did not expand among iGAS patients. Second, we identified a single recent isolate in clade 1 without the surge-associated genomic configuration. Because the isolate was collected from a noninvasive wound swab, we speculate that specific genomic inversions are preferentially selected for during an infection pathway. The orientation in the recent surge isolates may affect virulence potential by shifting the location of the phage-encoded superantigen gene ssa, which was previously shown to be pyrogenic and toxic in rabbits (28). Additional research is needed to demonstrate if the observed genomic inversion affects ssa expression. Regardless, this clade 1 isolate demonstrates the dynamic nature and stochasticity of genome rearrangements within the emm3 population.
The first limitation of our study is the difference between the clinical data collected in the Netherlands and England with the lack of formal registration for death after infection in the Netherlands. The second limitation is the low number of genomes sequenced from isolates from the Netherlands collected during the recent outbreak. Although all 4 recent emm3.93 isolates from the Netherlands are in a single clade, other nonsequenced isolates might fall into the other outbreak-associated clades.
In conclusion, combined epidemiologic and molecular surveillance is key to detect emergence of new or more virulent S. pyogenes variants and to rapidly assess their clinical and epidemiologic relevance. emm typing revealed a replacement of emm1.0 by emm3.93 as the dominant type from November 2023 onwards, continuing over a few months. It is possible that population immunity to emm1.0 developed during 2023, leaving increased susceptibility to potentially new virulence traits of the new clades of emm3.93. Alternatively, the genomic inversion may have resulted in a competitive advantage of emm3.93 over emm1.0. In the context of the current upsurge of S. pyogenes emm3.93, international collaboration proved invaluable to assess the spread of the new clades and provide a more comprehensive picture of the associations of emm3.93 with both clinical manifestations and case fatality. In addition, a combination of long- and short-read sequencing was necessary to reveal the extent of genetic diversity for the emerged emm3.93 clades.