The diverse trajectories of bacterial heterogeneity and adapted subpopulations in invasive Staphylococcus aureus infections

Abstract Background Within-host Staphylococcus aureus adaptation can lead to antibiotic treatment failure through the emergence of mutations that cause antibiotic resistance or immune evasion. Yet, given the inherent heterogeneity of bacterial populations, it is not known whether these adaptive mutations emerge at low frequency and subsequently sweep to dominance as infection progresses. Methods To assess bacterial within-host evolution during invasive S. aureus infections and assess the mutational trajectory of infecting bacterial sub-populations, we collected 669 clinical samples from 103 patients with S. aureus bacteraemia or other sterile-site infections and sequenced up to five colonies per sample. Within-patient S. aureus lineages were identified based on a threshold of fewer than 100 mutations. We identified adaptive mutations by tracking convergent evolution among lineages. Results Nasal swabs displayed significantly higher genetic diversity than clinical samples (median nucleotide diversity 3.6 x 10 − 7 vs 0). Pathoadaptive mutations involved 16 genes and were found in 26/122 (21%) within-host lineages. Tracking adapted subpopulations revealed divergent trajectories, with a minority of selective sweeps and frequent non-sweeping events or adaptations at low frequency, particularly within the known adaptive locus agr . Similar evolutionary trajectories were found when we investigated de novo transposon insertions, which were sporadic and contributed marginally to within-host adaptation. Conclusions Our findings show that the expansion of the genomic analysis to subpopulations improves the power of within-host evolution studies and reveals unexpected adaptive trajectories that might reflect the complex selective environment of invasive bacterial infections. Trial registration ACTRN12623001222639.