Differential Metabolic Functionality in Antibiotic-Resistant Pseudomonas aeruginosa Revealed with an Integrated Computational and Experimental Approach

Changes in bacterial metabolism accompanying the development of antibiotic resistance remain poorly understood. In this study, we performed a single-carbon source utilization screen on lab-evolved antibiotic-resistant Pseudomonas aeruginosa to investigate these changes. The metabolic capabilities of piperacillin-resistant, tobramycin-resistant, and ciprofloxacin-resistant P. aeruginosa as well as paired ancestral and media-evolved control lineages were evaluated by measuring growth curves on 190 unique carbon sources. Our resulting 950 growth curves revealed that resistant lineages exhibit mainly decreased catabolic function with occasional gains of function compared to antibiotic-sensitive P. aeruginosa. In resistant lineages we also observed changes in growth dynamics, including growth rate and time to reach mid-exponential phase. A genome-scale metabolic network reconstruction of P. aeruginosa strain UCBPP-PA14, iPau1129, was used to contextualize whole-genome sequencing data of the resistant lineages. The model was used to predict the impact of resistance mutations on loss of catabolic function. For example, five genes deleted in the piperacillin-resistant lineage were predicted to drive loss of the ability to utilize L-leucine. Model predictions were experimentally validated with a transposon mutant library. Our results show that metabolism is altered through the evolution of antibiotic resistance. Instances where our model failed to correctly predict genotype-phenotype relationships highlight gaps in our current understanding of P. aeruginosa metabolism. Our combined computational and experimental framework can be applied to identify metabolic limitations in other antibiotic-resistant pathogens. Drug-driven metabolic limitations have the potential to be targeted to select against antibiotic-resistant populations.