A genome-wide association study to identify the genetic loci underlying carbapenem resistance in Acinetobacter...

Carbapenem-resistant Acinetobacter baumannii (CRAB) presents a significant public health concern, and is recognized as a key contributor to the global antimicrobial resistance (AMR) crisis. This study applied a bacterial genome-wide association study (bGWAS) framework to identify genetic factors beyond the canonical carbapenemase OXA-23, that may contribute to carbapenem resistance.

A. baumannii isolates with publicly available whole-genome sequencing (WGS) data and carbapenem-specific antimicrobial susceptibility testing (AST) results were compiled (n = 1,601). Unitigs were extracted from de novo assemblies, and single-nucleotide polymorphisms (SNPs) were called from raw sequence reads mapped to seven selected reference genomes. The unitig- and SNP-based genome-wide association analyses were carried out, including conditional analyses based on the known resistance determinants. Structural analysis was performed for the select candidate mutations.

The population was highly clonal and comprised ten lineages, including three predominantly resistant lineages (ST1, ST2, and ST32). Known CRAB-associated loci, including OXA-23, the RND efflux pump regulator AdeL, and OXA-66 variants (Ile129Leu and Leu167Val), were readily identified. Structural analysis of mutations identified outer-membrane proteins as potential candidate loci, however these loci were not supported in the conditional analyses across all reference genomes. The candidate loci supported in the conditional analysis across reference genomes included an amino acid substitution in a LysR-type regulator adjacent to the biofilm-associated operon pgaABCD, and an intergenic substitution located between enzymes involved in LPS remodeling.

OXA-23 is the major determinant of carbapenem resistance in A. baumannii. Additional resistance mechanisms may become significant in the absence of OXA-23, including efflux pump regulation (AdeL), membrane permeability (OprD, DcaP), lipopolysaccharide (LPS) transport (PqiB), LPS remodeling (acyltransferase, glycosyltransferase), and biofilm formation (LysR). Integration of the conditional analysis based on known resistance determinants and a multi-reference mapping framework may improve the detection and prioritization of candidate associations.