Abstract
What dictates the extent to which a bacterial strain can evolve under antibiotic pressure remains poorly defined, despite its implications for the effectiveness of antibiotics in clinical practice. In this study, we compare the evolution of
Escherichia coli
and
Klebsiella aerogenes,
two medically relevant members of
Enterobacteriaceae
, under colistin and trimethoprim selection. We discover that the mere presence of effector genes of resistance, as well as conserved gene regulatory architecture, do not necessarily translate to similar evolvabilities during antibiotic challenge. Instead genotype-phenotype maps of adaptive mutations, which can diverge at different biological levels, govern how effectively a strain evolves resistance. Both,
E. coli
and
K. aerogenes
, demonstrated strikingly parallel capacities for adaptation to trimethoprim, as a result of conserved mechanisms of resistance, including drug target modification and efflux. On the other hand,
E. coli
was driven to extinction by colistin, while
K. aerogenes
was able to evolve a high level of colistin resistance. Like the better studied
K. pneumoniae
,
K. aerogenes
acquired mutations in the BasSR (also called PmrAB) and PhoQP two-component signalling pathways under colistin selection. These mutations increased the expression of chromosomally-encoded lipopolysaccharide (LPS) modifiers, EptA and Arn. Similar mutations in
E. coli
were unable to confer protection against colistin, despite successfully up-regulating homologous effectors genes. However, expression of MCR-1, a horizontally acquired EptA-homolog found in clinical strains, conferred high level colistin resistance to
E. coli
. This difference in the intrinsic capacity to adapt to colistin was reflected in frequencies of resistance among clinical isolates from around the world. It also altered the dependency of
Escherichia
and
Klebsiella
species on horizontal gene acquisition for colistin resistance. Our study highlights the importance of understanding pathogen-specific genotype-phenotype maps of drug resistance to better predict long-term effectiveness of antibiotic therapies and define resistance-proof treatment strategies.