Comparative Systems Analysis of Opportunistic Gram-Negative Pathogens

Persistent bacterial infections are a rapidly growing concern worldwide; combatting these drug-resistant infections is hampered by a limited understanding of multifactorial pathogen evolution. Pathogens must adapt to novel nutrient restrictions, stresses induced by the host environment, competition with other microbes, and therapeutic interventions such as antibiotic treatment. The lung infections of cystic fibrosis (CF) patients are an excellent model of long-term pathogen evolution. Pseudomonas aeruginosa and species of the Burkholderia cepacia complex (Bcc) are considered the most problematic CF pathogens, known for their dominance over other pathogens, ability to induce infections lasting decades, and resistance to antibiotics (factors which also contribute to their role in serious hospital-acquired infections).

With the goal of providing novel insights into potential new therapeutic targets to combat rising drug resistance, I investigate the metabolic flexibility, virulence capability, and adaptive metabolic rewiring of these pathogens using comparative systems analyses via the framework of genome-scale metabolic models. I have built two new genome-scale metabolic reconstructions of Bcc species, updated and reconciled an existing model of P. aeruginosa PAO1 and also created a new model of P. aeruginosa PA14. I couple these models with constraint-based analysis techniques and experimental phenotype screening to validate predictions and pursue model-generated hypotheses.

My work has resulted in the prediction of specific mechanistic causes for differential antibiotic resistance and capacity for virulence between B. cenocepacia and B. multivorans via quantitative examinations of genetic redundancy and predicted activity of secondary metabolite production pathways. I extended my comparative analysis approaches to the study of decades-long evolution in P. aeruginosa clinical isolates from chronic CF infections, using a novel integration of single nucleotide polymorphism- and expression-based constraints to create isolate-specific metabolic models representing early and late stage adaptation. I identified network rewiring of redox metabolism as a potentially important contributor to the successful persistence of the late stage strains. The P. aeruginosa models have also been used to evaluate genes and enzymes essential to the production of known virulence factors and tradeoffs between virulence factor production and bacterial growth, providing a parallel avenue of treatment to enhance current antibiotic approaches. My analysis of these predictions identifies novel therapeutic targets for inhibiting virulence factor production alone or in addition to growth, of which a subset are experimentally evaluated using in vitro gene knockouts. In summary, I use an integrated computational and experimental framework to conduct a comparative systems analysis of important drug-resistant pathogens, contributing novel insights into their adaptive metabolic capabilities and proposing new therapeutic approaches.