Tracking carbapenem resistance plasmids in hospital Enterobacteriaceae using long-read sequencing
Robert A. Moran 1,2, Mahboobeh Behruznia 1, Mark I. Garvey 1,3, Alan McNally 1
1 University of Birmingham, United Kingdom
2 Returning to Australia in 2025
3 University Hospitals Birmingham NHS Foundation Trust, United Kingdom
Birmingham’s Queen Elizabeth Hospital (QE) is a 1200-bed specialist tertiary centre that has one of the highest rates of carbapenem use in the UK. Bloodstream infections (BSIs) caused by carbapenemase-producing Enterobacteriaceae (CPE) at the QE have a 12-month all-cause mortality rate of 70%. To better understand plasmid-associated carbapenem resistance here, we have been conducting genomic surveillance using only Oxford Nanopore long-read sequence data. This approach is allowing us to rapidly assemble complete or near-complete whole-genome sequences. We defined and tracked plasmid lineages over an 18-month period, by comparing complete plasmid sequences with a particular focus on the smaller MGEs that shape their topologies. Tracking structural variation within plasmid lineages has providedinsights into their spatiotemporal and inter-host dispersal in the hospital setting, as well as to our understanding of how plasmids evolve as they transmit horizontally. Across 130 QE CPEs (6 genera; 10 species; 65 STs), we found several different plasmids carrying carbapenemase genes. These included internationally-reported plasmids such as the L-type pOXA-48 lineage carrying blaOXA-48, and multiple F-type lineages carrying blaNDM variants. Of particular interest was the blaKPC-2-bearing N-type lineage pQEB1, defined here using the unique structure of its antibiotic resistance region, which has not yet been seen outside of England’s West Midlands. pQEB1 was found in 15 different bacterial hosts from 17 patients in multiple QE wards. Twelve structural variants of pQEB1 were observed, with variation resulting from IS26-mediated inversion events or acquisitions of insertion sequences and/or transposons, including two carrying arsenic or mercury resistance genes. In most cases, acquired elements were also found in the chromosome or co-resident plasmids of the host carrying a particular pQEB1 variant. We hypothesise that circulating plasmids regularly exchange smaller MGEs with their bacterial hosts, and conclude that evolving plasmid structures can serve as practical markers for high-resolution genomic surveillance.