Phenotypic Profiling of Bacterial Stress Response Networks: A Transformative Framework For Characterizing and Predicting Antibiotic Targets and Interactions

Project Abstract/SummaryThe Wellcome Trust estimates the death toll due to microbial pathogenesis to be 700,000/year. This number isexpected to rapidly increase in the next decade if the rise of antimicrobial resistance remains unaddressed. Asa first step to understanding the mechanisms of antibiotic resistance emergence, recent studies have exploredthe biological processes affected by antibiotics from a holistic cellular perspective. Results from these studieshave challenged the traditional notion of each antibiotic eliciting a specific stress, revealing communicationbetween bacterial responses that highlight the importance of probing systems-level cellular physiology andexploiting multi-dimensional phenotypes. Although many attempts have been made to characterize cellular response to antibiotics on acomprehensive scale, most of these studies suffer from the significant disadvantage of measuring bulkpopulation-level responses. As most resistant mutants are a sub-population that dominates after selectiveantibiotic bottlenecks have been applied, bulk measurements that fail to account for single-cell behavior do notcapture the entire spectrum of responses to antibiotic stress. I will leverage two key technological developments: 1) a high-throughput imaging and image analysispipeline, and 2) a CRISPR interference library of essential gene knockdowns in the model organismEscherichia coli to answer fundamental questions about the bacterial response to antibiotics. I propose to usea combination of high-throughput microscopy and plate reader-based bulk measurements of fluorescentstress-response reporters to map response dynamics in E. coli under both oxygen-rich and anoxic conditions. Iwill combine morphological parameters and stress response information to build a rich landscape forphenotypic profiling that can be utilized to identify targets of novel antibiotics, predict antagonism incombinatorial therapies, and probe the fundamental wiring between pathways. To investigate the molecularmechanisms underlying the network architecture, I will employ CRISPRi genetic tools to alter drug-targetexpression and drug efflux. My overarching goal is to eliminate a key bottleneck in drug discovery and drugadministration approaches?the identification of cellular targets for antibiotics with unknown mechanisms ofaction and prediction of combinatorial therapeutics with improved efficacy from the vantage point of stress-response activation. This study should accelerate the antibiotic discovery pipeline through rapid targetidentification while also contributing deep understanding of bacterial physiology to guide future research acrossa wide range of organisms.