Clinically relevant plasmid-host interactions indicate transcriptional not genomic modifications ameliorate fitness costs of KPC carrying plasmids

The rapid dissemination of antimicrobial resistance (AMR) around the globe is largely due to mobile genetic elements, such as plasmids. They confer resistance to critically important drugs including extended spectrum beta-lactams, carbapenems and colistin. These resistance genes are frequently encoded on large complex plasmids which have evolved alongside their pathogenic host bacteria. However, much of the research on plasmid-host evolution has studied small, simple laboratory plasmids in non-pathogenic laboratory domesticated bacterial hosts. In this work, we describe the impact of two variants of a large AMR plasmid on a globally successful pathogen. We demonstrate that different to small plasmid/laboratory strain combinations, in clinically relevant AMR pKpQIL plasmids in their naturally isolated pathogenic host Klebsiella, no changes in coding domain sequences occur. Rather, the bacterium adapts to the acquisition of the AMR plasmid by adjusting bacterial and plasmid gene expression. This is sufficient to ameliorate associated fitness costs of plasmid carriage and AMR plasmids persisted within the host without selection pressure. This indicates that previous dogma indicating that removal of selection pressure (e.g. antimicrobial exposure) results in plasmid loss due to bacterial fitness costs is unlikely to be useful in the fight against AMR. We also show that AMR plasmids can impact upon the pathogen's ability to form a biofilm, an important aspect of virulence. This work uses highly relevant models to study the interaction between AMR plasmids and pathogens and reveals striking differences with work done on laboratory domesticated plasmids and strains.