A systems analysis of drug tolerance in Mycobacterium tuberculosis

PROPOSAL SUMMARY This project will address the critical need for accelerated development of multidrug regimen to achieve fast and complete clearance of Mycobacterium tuberculosis (Mtb), thereby lowering the likelihood for the emergence of antimicrobial resistance. Mtb dynamically adapts to extra- and intracellular host environments by adopting heterogeneous physiologic states, with varied susceptibility profiles to frontline antitubercular drugs. In the first four years of the R01, we have made progress towards dissecting this capability of Mtb by developing technologies to (i) uncover regulatory mechanisms that drive the pathogen into dormant states in host-simulated environments (controlled bioreactors) and directly within host cells (Path-seq), (ii) sort and characterize at single cell resolution translationally-dormant persister-like subpopulations within isogenic cultures (PerSort), (iii) uncover and characterize context-specific vulnerabilities within regulatory and metabolic networks (EGRIN2 and PRIME), and (iv) rationally formulate novel synergistic drug combinations (DRonA and MLSynergy). Using these capabilities and their applications reported across sixteen publications, we discovered that heterogeneous drug tolerant subpopulations co-exist within an isogenic culture of Mtb, even in the absence of drug treatment. Furthermore, we discovered that stressful environments and treatments activate additional drug tolerance networks, which may potentiate the emergence of resistance. Based on these findings, we hypothesize that we can achieve fast and complete clearance of Mtb infection with a combination of drugs that target vulnerabilities across heterogeneous drug tolerant subpopulations that co-exist in varied combinations and proportions depending on host- and treatment-contexts. To test this hypothesis, we will mechanistically characterize how the heterogeneous population structure of Mtb changes dynamically in response to host-relevant environmental cues and drug treatments. We will then uncover and characterize vulnerabilities within regulatory and metabolic networks that support and drive transitions to drug tolerant states. Using machine-learning techniques, we will predict and validate synergistic drug combinations targeting multiple vulnerabilities to cripple heterogeneous environment- and drug-induced states of Mtb. By performing time kill curves, we will investigate whether validated combinatorial interventions accomplish complete and faster clearance of heterogeneous Mtb subpopulations in diverse contexts. Altogether, the proposed activities will identify novel drug targets, and novel drug combinations for fast and complete clearance of a heterogeneous Mtb population. Given that phenotypic heterogeneity as a means for tolerating and resisting drugs is a universal phenomenon, the systems biology framework developed in this project will be generalizable to the discovery of effective multidrug regimen for diverse infectious diseases and even cancers.