Escherichia coli phylogeny drives co-amoxiclav resistance through variable expression of TEM-1 beta-lactamase.
William Matlock, Gillian Rodger, Emma Pritchard, Matthew Colpus, Natalia Kapel, Lucinda Barrett, Marcus Morgan, Sarah Oakley, Katie L Hopkins, Aysha Roohi, Drosos Karageorgopoulos, Matthew B Avison, A Sarah Walker, Samuel Lipworth, Nicole Stoesser
Abstract
Open AccessCo-amoxiclav (amoxicillin and clavulanate) is a commonly used combination antibiotic, with resistance in Escherichia coli associated with increased mortality. The class A beta-lactamase blaTEM-1 is often carried by resistant E. coli but exhibits high phenotypic heterogeneity, complicating genotype-phenotype predictions. We curated a dataset of n = 377 diverse E. coli isolates where the only acquired beta-lactamase was blaTEM-1. We generated hybrid assemblies and co-amoxiclav minimum inhibitory concentrations (MICs), and blaTEM-1 qPCR expression data for a subset (n = 67/377). We first tested whether intrinsic expression of blaTEM-1 varied between E. coli lineages, for example, from regulatory system differences, which are challenging to genomically quantify. Using genotypic features, we built a hierarchical Bayesian model for blaTEM-1 expression, controlling for phylogeny. Expression varied across the phylogeny, with some lineages (phylogroups B1 and C, ST12) expressing blaTEM-1 more than others (phylogroups E and F, ST372). Next, we built a second model to predict isolate MIC from genotypic features, again controlling for phylogeny. Phylogeny alone shifted MIC past the clinical breakpoint in 19% (55/292) of isolates with greater-than-chance probability, mostly representing ST12, ST69 and ST127. A third causal model confirmed that phylogenetic influence on blaTEM-1 expression drove variation in MIC. We speculate that intergenic variation underlies this effect.