ABSTRACT
New antibiotics are needed to combat rising levels of resistance, with new Mycobacterium tuberculosis (Mtb) drugs having the highest priority. However, conventional whole-cell and biochemical antibiotic screens have failed. Here we develop a strategy termed PROSPECT (primary screening of strains to prioritize expanded chemistry and targets), in which we screen compounds against pools of strains depleted of essential bacterial targets. We engineered strains that target 474 essential Mtb genes and screened pools of 100-150 strains against activity-enriched and unbiased compound libraries, probing more than 8.5 million chemical-genetic interactions. Primary screens identified over tenfold more hits than screening wild-type Mtb alone, with chemical-genetic interactions providing immediate, direct target insights. We identified over 40 compounds that target DNA gyrase, the cell wall, tryptophan, folate biosynthesis and RNA polymerase, as well as inhibitors that target EfpA. Chemical optimization yielded EfpA inhibitors with potent wild-type activity, thus demonstrating the ability of PROSPECT to yield inhibitors against targets that would have eluded conventional drug discovery.
Subject(s)
Antitubercular Agents/classification , Antitubercular Agents/isolation & purification , Drug Discovery/methods , Gene Deletion , Microbial Sensitivity Tests/methods , Mycobacterium tuberculosis/drug effects , Mycobacterium tuberculosis/genetics , Small Molecule Libraries/pharmacology , Antitubercular Agents/pharmacology , DNA Gyrase/metabolism , Drug Resistance, Microbial , Folic Acid/biosynthesis , Molecular Targeted Therapy , Mycobacterium tuberculosis/cytology , Mycobacterium tuberculosis/enzymology , Mycolic Acids/metabolism , Reproducibility of Results , Small Molecule Libraries/classification , Small Molecule Libraries/isolation & purification , Substrate Specificity , Topoisomerase II Inhibitors/isolation & purification , Topoisomerase II Inhibitors/pharmacology , Tryptophan/biosynthesis , Tuberculosis/drug therapy , Tuberculosis/microbiologyABSTRACT
Two-component sensors are widely used by bacteria to sense and respond to the environment. Pseudomonas aeruginosa has one of the largest sets of two-component sensors known in bacteria, which likely contributes to its unique ability to adapt to multiple environments, including the human host. Several of these two-component sensors, such as GacS and RetS, have been shown to play roles in virulence in rodent infection models. However, the role and function of the majority of these two-component sensors remain unknown. Danio rerio is a recently characterized model host for pathogenesis-related studies that is amenable to higher-throughput analysis than mammalian models. Using zebrafish embryos as a model host, we have systematically tested the role of 60 two-component sensors and identified 6 sensors that are required for P. aeruginosa virulence. We found that KinB is required for acute infection in zebrafish embryos and regulates a number of virulence-associated phenotypes, including quorum sensing, biofilm formation, and motility. Its regulation of these phenotypes is independent of its kinase activity and its known response regulator AlgB, suggesting that it does not fit the canonical two-component sensor-response regulator model.