Molecular Characterization of Colistin-Resistant A. baumannii

Key Takeaways:

• Investigators report phenotypic colistin resistance confirmed using reference broth microdilution and gradient diffusion methods, with agreement against prior routine automated reporting.
• Across the resistant collection, the authors describe nonsynonymous variation concentrated in pmrCAB (especially pmrB and pmrC), alongside comparatively limited amino-acid variation in the sequenced lpx loci.
• In a typed subset, ST2/IC2 predominated and no plasmid-mediated mcr genes were detected.

The authors describe 40 non-duplicate XDR/PDR isolates from hospitalized patients across multiple wards and specimen sources, with phenotypic colistin resistance documented across the collection.

Phenotypic colistin susceptibility testing was reported as having been performed using reference broth microdilution and a gradient diffusion method, with results compared against routine resistance profiles previously generated by the VITEK 2 automated platform in the participating hospitals. Resistance was defined using a colistin breakpoint of MIC ≥4 µg/mL (equivalent to ≥4 mg/L), and the investigators report MIC values spanning 4–32 mg/L with an MIC50 of 8 mg/L and an MIC90 of 16 mg/L in their dataset. They also report full concordance between broth microdilution and gradient diffusion, with agreement against prior VITEK 2 colistin resistance reporting.

For molecular characterization, the authors report PCR amplification and Sanger sequencing of the pmrCAB operon (pmrA, pmrB, pmrC) and of lipid A biosynthesis genes lpxA, lpxC, and lpxD across the isolates. In this resistant set, they describe multiple nonsynonymous changes concentrated in pmrB and pmrC, while noting no amino acid–altering substitutions in PmrA within their analyzed region. PmrB A226V is highlighted as the predominant PmrB substitution, with additional PmrB substitutions observed in smaller numbers and a six-nucleotide insertion identified in a single isolate. For PmrC, the authors describe broad amino-acid variation across isolates, including substitutions present in most or all samples as well as isolate-specific changes. By contrast, they report comparatively limited protein-level variation in the lpx loci, listing LpxC N287D and LpxD E117K as the specific amino-acid changes identified. In the paper’s framing, this distribution aligns chromosomal variation in established pathways with the observed resistance phenotype.

Clonal context was assessed using multilocus sequence typing (Pasteur scheme) applied to a PFGE-selected subset of isolates (n=10), and the authors report predominance of ST2 (International Clone 2) among typed strains, with single representatives of ST115 (IC2) and ST1 (IC1). They also report that plasmid-mediated mcr genes were not detected in their screening.

In the authors’ interpretation, an ST2/IC2-dominated background alongside chromosomal pmrCAB variation is described as consistent with the resistant profile observed in this cohort. The paper also notes constraints on interpretation, including that MLST was performed on a PFGE-selected subset rather than the full collection, whole-genome sequencing was not performed, and functional assays were not used to directly confirm the impact of specific mutations, leaving the findings as sequence- and typing-based observations within this setting.