Antibiotic resistant pathogens are rapidly spreading, creating a human health crisis. The significance of the problem is underscored by a recent President's Executive Order ?Combating Antibiotic-Resistant Bacteria?, which calls for developing novel therapeutics. We currently lack an effective platform for antibiotic discovery, and the paucity of good starting compounds has been widely viewed as the main bottleneck for producing new therapeutics. Natural products have been the main source of novel compounds, but they were over mined half a century ago. Uncultured bacteria represent 99% of the total biodiversity. Methods to grow them have been developed by our collaborators on this Program, and several exciting compounds have been discovered (Projects I, III). What has been lacking however is an effective method to identify early on, the potential value of a compound in an extract of a producing isolate. We reason that the discovery bottleneck can be resolved by early biological dereplication, using transcriptome analysis of extracts to report on the mode of action (MOA) of a compound. This is immediately useful information, which automatically discards known and toxic compounds, and points to compounds hitting a new target, or acting in a new way against a known one. We validated this approach in a pilot study and will develop transcriptomics into a robust drug discovery dereplication method. We will build a database of transcription profiles produced by antibiotics with a known MOA and targets. This database will facilitate analysis of unknowns. In addition, we prioritized 500 antimicrobials for transcription profiling based on in silico analysis of their suitability for drug development. Compounds will be added to a growing culture of an Escherichia coli and/or a Staphylococcus aureus strain, and transcriptome profiles of the antimicrobial challenged strains will be obtained. The large database resulting from this study will considerably enrich our knowledge of antibiotic targets and will identify candidate compounds for development. In parallel, we will obtain transcription profiles produced by fractionated and unfractionated extracts from uncultured bacteria (Project I). Extracts that produce transcriptomes clearly indicating the presence of a new compound hitting a known or new target will be given priority. We will evaluate and implement several computational tools to assign MOA. The result of this project will be a large and growing database of transcription profiles of antibiotics and antimicrobial compounds and efficient determination of which compounds produced in the antimicrobial screening pipeline of Project 1 are novel and are strong lead candidates. Coupling this Project to the rich source of novel compounds of Project 1 has potential for solving the two major impediments for new drug discovery and potential to thus bring on a new golden age of antimicrobial drug discovery.
There is a looming human health crisis caused by the spread of drug-resistant bacterial pathogens. We will develop a new platform for discovering effective antibiotics by exploiting an untapped source ? uncultured bacteria, combined with an effective method to decide early on if a compound has potential value by identifying its mode of action.