Integrating multi-omics data with genome scale metabolic modeling for the analysis of Pseudomonas aeruginosa persister cells

Pseudomonas aeruginosa is a Gram-negative multi-drug resistant bacterial pathogen, capable of forming environmentally-resistant subpopulations known as persister cells. These cells are transient phenotypic variants that are able to tolerate antimicrobial treatment and have been implicated in the recalcitrant nature of chronic infections and in the resistance to disinfection. While persister cells are classically associated with reduced metabolic activity, the characteristics of their metabolism are not well understood. In this work, we performed an experimental and computational systems-level analysis to characterize the metabolic state of persister cells. To accomplish this, we deeply profiled both wildtype and persister samples of P. aeruginosa with transcriptomic sequencing and metabolomic analyses. This revealed a distinct metabolic repertoire in persister cells, marked by an increase in central metabolism activity. To aid in the analysis, integration of both the transcriptomic and metabolomic datasets with a P. aeruginosa genome-scale metabolic network reconstruction (GENRE) provided condition-specific models of the persister and untreated states. We then used the model of persister cell metabolism to hypothesize single-gene targets of persister cell viability. Experimental testing of model predictions suggested that persister cells are robust to single-gene deletions and that combinatorial targeting strategies may be necessary to completely inhibit the persister phenotype. Using this approach, we gained insight into P. aeruginosa persister cell metabolism and highlighted possible combinatorial gene targeting strategies to inhibit the development of the persister cell phenotype. Overall design: RNA was extracted from untreated and persister Pseudomonas aeruginosa (PA14) cells immediately before exposure, 5 and 24 hours after exposure to 10X MIC conditions.