Novel Multilocus Sequence Typing and Global Sequence Clustering Schemes for Characterizing the Population Diversity of Streptococcus mitis

Streptococcus mitis is a common oral commensal and an opportunistic pathogen that causes bacteremia and infective endocarditis; however, the species has received little attention compared to other pathogenic streptococcal species. Effective and easy-to-use molecular typing tools are essential for understanding bacterial population diversity and biology, but schemes specific for S. mitis are not currently available. We therefore developed a multilocus sequence typing (MLST) scheme and defined sequence clusters or lineages of S. mitis using a comprehensive global data set of 322 genomes (148 publicly available and 174 newly sequenced). We used internal 450-bp sequence fragments of seven housekeeping genes (accA, gki, hom, oppC, patB, rlmN, and tsf) to define the MLST scheme and derived the global S. mitis sequence clusters using the PopPUNK clustering algorithm. We identified an initial set of 259 sequence types (STs) and 258 global sequence clusters. The schemes showed high concordance (100%), capturing extensive S. mitis diversity with strains assigned to multiple unique STs and global sequence clusters. The tools also identified extensive within- and between-host S. mitis genetic diversity among isolates sampled from a cohort of healthy individuals, together with potential transmission events, supported by both phylogeny and pairwise single nucleotide polymorphism (SNP) distances. Our novel molecular typing and strain clustering schemes for S. mitis allow for the integration of new strain data, are electronically portable at the PubMLST database (https://pubmlst.org/smitis), and offer a standardized approach to understanding the population structure of S. mitis. These robust tools will enable new insights into the epidemiology of S. mitis colonization, disease and transmission.