GENOMIC EVOLUTION OF STAPHYLOCOCCUS AUREUS DURING ARTIFICIAL AND NATURAL COLONIZATION OF THE HUMAN NOSE

Staphylococcus aureus can colonize the human vestibulum nasi for long periods of time, possibly extending into many years. It is unknown whether and, if so, how S. aureus adapts to this ecological niche during colonization. We set out to determine the short (1 and 3 months) and mid-term (36 months) genomic evolution of S. aureus in healthy natural carriers and artificially colonized healthy volunteers. Eighty-five serial S. aureus strains were collected from 6 natural carriers followed for three years and 6 artificially colonized volunteers followed for one month. Genome-based multi-locus sequence typing (MLST) and single nucleotide polymorphism (SNP) analysis based on whole-genome sequencing (WGS) were carried out. Mutation frequencies within resident bacterial populations over time (1 to 36 months) were quantified using core genome SNP counts (comparing groups of genomes) and pairwise SNP divergence assessment (comparing two genomes from strains originating from the same host and sharing identical global MLST types). Total core genome SNP counts (within 1-3 months) in all naturally colonizing strains varied from 0 to 757 (median 4). These strains showed random and independent patterns of pairwise SNP divergence, ranging from 0 to 44 SNPs (median 7). When the different core genome SNP counts over a period of 3 years were considered, the median SNP count was 4 (ranging from 0 to 26). Host-specific pairwise SNP divergence for the same period ranged from 9 to 57 SNPs (median 20). During short term artificial colonization (1 month) the mutation frequency was even lower (0 to 7 core genome SNPs, median 2) and the pairwise SNP distances were 0 to 5 SNPs (median 2). We recommend the WGS of multiple colonies from a single culture during outbreak infection analysis. Quantifying mutation frequencies is important for the longitudinal follow-up of epidemics of infections and outbreak management. Over periods up to three years, maximum median core genome SNP counts and SNP divergence for the strains studied were 4 and 20 SNPs or lower. Therefore, we suggest an epidemiological cut off value of 20 SNPs as a marker of S. aureus strain identity during outbreaks of infection.