Genomics of experimental diversification of Pseudomonas aeruginosa in cystic fibrosis lung-like conditions

Abstract

AbstractPseudomonas aeruginosa is among the most problematic opportunistic pathogens for adults with cystic fibrosis (CF), causing repeated and resilient infections in the lung and surrounding airways. Evidence suggests that long-term infections are associated with diversification into specialized types but the underlying cause of that diversification and the effect it has on the persistence of infections remains poorly understood. Here, we use evolve and resequence experiments to investigate the genetic changes accompanying rapid, de novo phenotypic diversification in lab environments designed to mimic two aspects of human lung ecology: spatial structure and complex nutritional content. After ∼220 generations of evolution, we find extensive genetic variation present in all environments, including those that most closely resemble the CF lung, attributable to a combination of high mutation supply rates resulting from large population sizes and the complex ecological conditions imposed by resource complexity and spatial structure. We use the abundance and frequency of nonsynonymous and synonymous mutations to estimate the ratio of mutations that are selectively neutral (hitchhikers) to those that are under selection (drivers). A significantly lower proportion of driver mutations in spatially structured populations suggests that reduced dispersal generates subpopulations with reduced effective population size, decreasing the supply of beneficial mutations and causing more divergent evolutionary trajectories. The genes most commonly mutated tend to impact regulatory functions linked to a range of CF-associated phenotypes, including one gene that confers antibiotic resistance despite the absence of antibiotic selection in our experiment, but do not appear to be specific to CF-like conditions arising from antimicrobial treatment, immune system suppression, or competition from other microbial species. Our results are consistent with models of adaptation that see the first mutations fixed during adaptation to a stressful environment being those that are broadly beneficial across a range of environments.