Intralineage amino acid variation is present in all surface bound proteins. PD173074 concentration Low levels of variation (proportion of variable sites < 0.05) exist in 22 surface proteins, whilst SdrD, Spa and SraP have higher levels of intralineage variation. Across all
proteins there are small levels of intralineage variation in host-interface domains (proportions of variable amino acid sites vary from 0.000 to 0.078) (Additonal file 1 Table S1). Interestingly, intralineage levels of variation differ between lineages in host-interface domains of a small subset of surface bound proteins. For example, the FN-1 binding domain of FnBPA has a proportion of variable amino acid sites of 0.032, 0.016 and 0.008 for CC5, CC8 and CC30 respectively, whilst there is an interlineage variation of 0.139. Such variation could support S. aureus lineage adaption to hosts and environments, and/or S. aureus evasion of the host immune response. An example of a highly variable surface protein is FnBPA. The distribution of protein domain variants of FnBPA across CC lineages shows evidence of recombination. (Additonal file 3 Table S3). For the purposes of this paper we define a domain variant as any domain with a Alvocidib cost sequence encoding https://www.selleckchem.com/products/rg-7112.html one amino acid difference. In addition, we define a domain that has greater than 5% of variable amino acids as a major variant
within a domain. The data shows that a range of major and/or minor sequence variations exist for the N terminus of the variable region domain, the fibrinogen (FG) and elastin (ELN) binding domain and the fibronectin (FN-1) binding domain (Additonal file 3 Table S3). Within each CC lineage only one major sequence variant exists for each FnBPA domain, and therefore the whole gene is lineage-specific. Surprisingly, the same major sequence variant of a domain Cobimetinib supplier is often found in unrelated lineages. Furthermore, whilst a lineage may share a major sequence variant of one domain with one unrelated lineage, it may share a major sequence variant at an adjacent domain with a different unrelated lineage. This shows that the fnbpA gene has a mosaic structure and indicates the fnbpA locus
is evolving through recombination, in addition to point mutation. Loughman et al.  have previously identified FnBPA sequence variants from human strains of lineages that have not had their genome sequenced (CC12, CC15, CC25, CC55, CC59, CC101, CC121 and CC509) and classified seven isotypes. They have shown that all isotypes have human fibrinogen binding activity, but that isotype I (found in CC8, CC15 and CC55) binds weakly to elastin. Inclusion of these partial gene sequences [GenBank: AM749006-15], corresponding to amino acid residues 1- 565, in our analysis suggests these gene variants are typical. Interestingly, they prove that no animal S. aureus strain has a major domain variant that is not found in a human S. aureus lineage.