Mechanisms of cephalosporin resistance in Escherichia coli from cattle and humans

Abstract

ABSTRACT Antibacterials of the β-lactam class (including penicillins, cephalosporins and carbapenems) are the most commonly used antibacterials in human and veterinary medicine. Understanding the underlying mechanisms of β-lactam resistance is therefore important. The overarching aims of the current project were i) to characterise cephalosporin resistance mechanisms in Escherichia coli isolates from dairy farms and urinary tract infection (UTI) in the absence of cephaosporinases, and to assess evidence for farm to human (zoonotic) transmission, and, ii) to generate E. coli mutants with reduced susceptibility to cephalosporins or to the carbapenem ertapenem, and to determine how the mutations acquired induce their phenotypic effects. This study demonstrates that the main cause of 3 rd generation cephalosporin (cefotaxime) resistance in farm and UTI isolates lacking mobile cephalosporinases is the hyperproduction of chromosomal AmpC due to promoter mutations; different promoter mutations were shown to induce different levels of AmpC production, and some of these additionally led to cefoperazone resistance. Cefotaxime resistance was extended to 4 th generation cephalosporin (4GC), cefoperazone and ceftazidime resistance in some cases where AmpC was mutated, resulting in an expanded-spectrum variant. Cefoperazone resistance was also caused by MarR mutation mediated AcrAB-TolC efflux pump activation in one AmpC hyperproducer isolate or TEM-1 hyperproduction in another. 4GC resistance was associated with additional production of OXA-1 in several AmpC hyperproducers. Phylogenetic analysis revealed no evidence for acquisition of farm-related AmpC hyperproducer E. coli isolates by members of the local human population, though farm to farm ii transfer was common. Presence of AmpC hyperproducers on farms was associated with the use of amoxicillin/clavulanic acid, and not with the use of cephalosporins. Resistance to the first-generation cephalosporin cefalexin was found not to be β-lactamase mediated in a large proportion of farm and UTI isolates. In these cases, we identified that the phenotype was caused by OmpF porin disruption or downregulation. Importantly, multiple regulatory mutations that cause OmpF downregulation were identified. In addition to mutation of OmpR, already known to downregulate OmpF and OmpC porin production, we report a rseA mutation, which strongly activates the Sigma E regulon, greatly increasing DegP production leading to degradation of OmpF and OmpC porins. Furthermore, we showed for the first time that mutations affecting lipopolysaccharide structure, exemplified by the loss of GmhB, essential for lipopolysaccharide heptosylation, and even the essential lipopolysaccharide biosynthetic committed enzyme LpxB, also activate DegP production, resulting in OmpF degradation. Remarkably, given the critical importance attached to such systems for normal E. coli physiology, we find evidence of gmhB, lpxB and rseA mutations in E. coli isolates derived from human infections. Finally, we show that these regulatory mutations enhance the ability of group 1 CTX-M β-lactamase to confer reduced ertapenem susceptibility, particularly those mutations that cause OmpC in addition to OmpF downregulation and that OmpC loss confers ertapenem reduced susceptibility on AmpC hyperproducer E. coli, particularly if the AmpC has an expanded spectrum of activity. Overall, this work has identified important, previously unknown mechanisms of cephalosporin resistance in E. coli, and has added to our understanding of how multiple mechanisms come together to confer late generation cephalosporin and carbapenem resistance. These findings are important for the utility of WGS analysis as a tool to predict antibacterial susceptibility.