Toward ROS-potentiating anti-infectives: integrated computational and experimental analyses of bacterial ROS metabolism

Abstract

Antibiotic resistance is a growing public health concern, and existing approaches to antibiotic discovery are unable to keep pace with the decline of effective therapies. Anti-virulence therapies represent an alternative to traditional antibiotics, and target the ability of pathogens to infect the host rather than essential cellular functions, which is projected to slow resistance development. One anti-virulence strategy involves sensitizing bacteria to immune-derived reactive oxygen species (ROS), the importance of which is highlighted by attenuated virulence in many pathogens when one or more detoxification system is perturbed. Compounds have been identified that sensitize Staphylococcus aureus to killing by immune cells through the inhibition of an antioxidant synthesis pathway, suggesting that sensitization to ROS is a promising therapeutic avenue. Due to the complexity of ROS production and detoxification, computational approaches are useful in predicting and interpreting ROS network perturbations.

Steady-state modeling can be used to identify mechanisms that alter endogenous ROS production, which can impact survival to exogenous oxidants. Based on a previous observation that inefficiencies in ATP production or usage led to increased endogenous ROS production and oxidant sensitivity, we explored the impact of futile cycles on those processes. Futile cycling increased endogenous production of ROS and sensitivity to oxidative stress, which demonstrated an important connection between energy metabolism and oxidative stress. Focusing on the membrane-permeable, immune oxidant H2O2, we recognized that the outcome of H2O2 exposure is dictated by a kinetic competition between detoxification and damage pathways. Therefore, we constructed and experimentally validated a systems-level, dynamic model of H2O2 detoxification in E. coli. This model was able to capture the dynamics of wild-type and mutant H2O2 clearance, and predict detoxification dynamics following a physiologically relevant environmental perturbation (carbon source starvation). More recently, we combined this dynamic model with an analogous model of the NO• response network in E. coli, and used it to explore the mechanism underlying delayed NO• clearance in the presence of transient H2O2 stress. Collectively, the work in this thesis emphasizes the utility of integrating computational and experimental approaches to improve understanding of phagosomal stresses on bacteria, which will aid in the realization of ROS-potentiating anti-infectives.