Metabolic constraints on the evolution of antibiotic resistance

Abstract

Despite our continuous improvement in understanding antibiotic resistance, the interplay between natural selection of resistance mutations and the environment remains unclear. To investigate the role of bacterial metabolism in constraining the evolution of antibiotic resistance, we evolved Escherichia coli growing on glycolytic or gluconeogenic carbon sources to the selective pressure of three different antibiotics. Profiling more than 500 intracellular and extracellular putative metabolites in 190 evolved populations revealed that carbon and energy metabolism strongly constrained the evolutionary trajectories, both in terms of speed and mode of resistance acquisition. To interpret and explore the space of metabolome changes, we developed a novel constraint‐based modeling approach using the concept of shadow prices. This analysis, together with genome resequencing of resistant populations, identified condition‐dependent compensatory mechanisms of antibiotic resistance, such as the shift from respiratory to fermentative metabolism of glucose upon overexpression of efflux pumps. Moreover, metabolome‐based predictions revealed emerging weaknesses in resistant strains, such as the hypersensitivity to fosfomycin of ampicillin‐resistant strains. Overall, resolving metabolic adaptation throughout antibiotic‐driven evolutionary trajectories opens new perspectives in the fight against emerging antibiotic resistance. image Bacterial metabolism constrains the evolution of antibiotic resistance. A modeling approach is developed to interpret the functionality of metabolic rewiring in resistance‐evolving E. coli growing on glycolytic or gluconeogenic carbon sources from metabolomics data. Large‐scale untargeted metabolome profiling reveals metabolic adaptations in 190 evolved antibiotic‐resistant E. coli populations, in part as compensation for consequences of the primary resistance mechanisms. Carbon and energy metabolism strongly constrain the evolutionary trajectories, both in terms of speed and mode of resistance acquisition. A novel constraint‐based modeling approach, together with genome re‐sequencing of resistant populations, identifies condition‐dependent compensatory mechanisms.