A genome wide approach to identify factors asociated with vancomycin non-susceptibility in Staphylococus aureus

Vancomycin (VAN) represents the standard of care in the treatment of Methicillin-Resistant S. aureus (MRSA) infections and widespread use has driven antimicrobial resistance. While fully resistant S. aureus strains have been identified, vancomycin-intermediate S. aureus (VISA) and heterogeneous vancomycin-intermediate susceptible S.aureus (hVISA) strains which contain resistant subpopulations despite appearing susceptible, are more common. These strains are associated with poorer outcomes including persistent infections or prolonged bacteremia. Because hVISA can be challenging to identify, prevalence is unclear and varies significantly as currently reported. Identifying the prevalence of hVISA is vital to understanding the clinical impact of these infections. Further, whole genome sequencing (WGS) offers a tool to identify factors such as multi locus sequence types (MLSTs) or single nucleotide polymorphisms (SNPs) that may aid in identifying or predicting the presence of hVISA. This study therefore utilized a stepwise phenotypic approach to elucidate the prevalence of hVISA among a collection of clinical isolates derived from hospitals in the states of Texas and California and employed WGS to identify MLSTs and SNPs associated with hVISA. VAN E-tests, VAN impregnated BHI agar and population analysis profile-area under the curve tests were used to identify the prevalence of hVISA in a collection of 320 clinical isolates. Additionally, a subset of these isolates underwent WGS to identify MLSTs while their genomes were compared to the reference (N315) to identify SNPs. The overall prevalence of hVISA in this study was 3% and was most common among isolates with an MIC of 2µg/mL. The predominant MLST among hVISA isolates was ST5, contributing to 70% of the isolates. Further, by comparing hVISA to VSSA isolates matched by year, location, MIC and MLST, we identified SNPs in four candidate genes that were exclusive to hVISA. Finally, a phylogenetic analysis demonstrated the heterogeneity of isolates displaying the hVISA phenotype. While this study highlights the complex and multifactorial nature of hVISA, it does provide insight into its prevalence and factors that may provide utility while identifying these infections. These tools may aid in identifying patients suffering from hVISA infections and allow clinicians to make more rapid and informed decisions during treatment.