Progressive evolution of Streptococcus equi from Streptococcus equi subsp. zooepidemicus and adaption to equine hosts

<i>Streptococcus equi</i> subsp. <i>equi</i> causes the equine respiratory disease 'strangles', which is highly contagious, debilitating and costly to the equine industry. <i>S. equi</i> emerged from the ancestral <i>Streptococcus equi</i> subsp. <i>zooepidemicus</i> and continues to evolve and disseminate globally. Previous work has shown that there was a global population replacement around the beginning of the twentieth century, obscuring the early genetic events in this emergence. Here, we have used large-scale genomic analysis of <i>S. equi</i> and its ancestor <i>S. zooepidemicus</i> to identify evolutionary events, leading to the successful expansion of <i>S. equi</i>. One thousand two hundred one whole-genome sequences of <i>S. equi</i> were recovered from clinical samples or from data available in public databases. Seventy-four whole-genome sequences representative of the diversity of <i>S. zooepidemicus</i> were used to compare the gene content and examine the evolutionary emergence of <i>S. equi</i>. A dated Bayesian phylogeny was constructed, and ancestral state reconstruction was used to determine the order and timing of gene gain and loss events between the different species and between different <i>S. equi</i> lineages. Additionally, a newly developed framework was used to investigate the fitness of different <i>S. equi</i> lineages. We identified a novel <i>S. equi</i> lineage, comprising isolates from donkeys in Chinese farms, which diverged nearly 300 years ago, after the emergence of <i>S. equi</i> from <i>S. zooepidemicus</i>, but before the global sweep. Ancestral state reconstruction demonstrated that phage-encoded virulence factors <i>slaA</i>, <i>seeL</i> and <i>seeM</i> were acquired by the global <i>S. equi</i> after the divergence of the basal donkey lineage. We identified the equibactin locus in both <i>S. equi</i> populations, but not <i>S. zooepidemicus</i>, reinforcing its role as a key <i>S. equi</i> virulence mechanism involved in its initial emergence. Evidence of a further population sweep beginning in the early 2000s was detected in the UK. This clade now accounts for more than 80% of identified UK cases since 2016. Several sub-lineages demonstrated increased fitness, within which we identified the acquisition of a new, fifth prophage containing additional toxin genes. We definitively show that acquisition of the equibactin locus was a major determinant in <i>S. equi</i> becoming an equid-exclusive pathogen, but that other virulence factors were fixed by the population sweep at the beginning of the twentieth century. Evidence of a secondary population sweep in the UK and acquisition of further advantageous genes implies that <i>S. equi</i> is continuing to adapt, and therefore, continued investigations are required to determine further risks to the equine industry.