Project Summary. Discovering new antibiotics for Gram-negative bacteria is uniquely challenging, due to their double-membrane structure that acts as a permeability barrier to drugs and as an anchor for efflux pumps. Efforts that target one membrane protein or one efflux pump at a time are ineffective, due to rapid rise of resistance mutations. We will target the TrmD-catalyzed m1G37 methylation of tRNA to inhibit biosynthesis of multiple classes of membrane proteins, with the potential to accelerate bactericidal action. TrmD is a bacteria-specific S- adenosyl-methionine (AdoMet)-dependent methyl transferase that controls accuracy of the protein-synthesis reading frame. Loss of TrmD increases +1 frameshifting and causes cell death. We have shown that genes for multiple membrane proteins and efflux pumps in E. coli and in other Gram-negative bacteria contain TrmD- dependent codons near the start of the reading frame. We hypothesize that targeting TrmD will reduce protein synthesis of all of these genes, thus offering a novel solution to an unmet need. While AstraZeneca (AZ), GSK, and academic labs have attempted to target TrmD by screening small molecular compound libraries, isolated hits lack the cell-permeability needed to exhibit an antibacterial effect. Here, we propose to screen a large collection of microbial extracts and fractions for cell-permeable and TrmD-targeting natural products (NPs) that are potent and selective over the human counterpart Trm5. We will use a cell-based assay, consisting of a 1:1 mix of an E. coli (Ec) TrmDmCh strain (dependent on trmD for survival and expressing mCh (mCherry) as a fluorescence marker) and an Ec Trm5YFP strain (dependent on trm5 for survival and expressing YFP), in a high- throughput screening (HTS) campaign to isolate NPs that selectively inhibit the TrmDmCh strain. We perform this assay in Ec tolC+ cells, which maintain the entire Gram-negative efflux machinery including the major efflux pump encoded by tolC, to screen for NPs that are cell-permeable and resistant to efflux. A pilot screen with this tolC+ cell-based assay has identified an attractive hit, demonstrating the HTS-readiness of the assay. In Aim 1, we will use this tolC+ cell-based assay to screen 74,770 actinobacterial extracts and fractions available at The Scripps Research Institute (TSRI). We will assess hits in secondary assays, remove false positives, evaluate their activity at the whole-cell level, and test them for permeability and efflux in a panel of Gram-negative bacteria. In Aim 2, we will de-replicate the top 20 hits to isolate the active NPs, determine their structures, and use a combination of genome sequencing and mining to identify their biosynthetic gene clusters (BGCs) for developing biotechnology platforms to scale up their production. In Aim 3, we will test active NPs for conferring TrmD- deficient phenotypes in whole-cell assays, determine their potency, selectivity, mechanism of action, and assess their risk of resistance. These NPs represent novel leads in a new paradigm of antibiotic discovery that addresses the multi-drug resistance problem of Gram-negative bacteria.

Project Narrative We have shown that targeting TrmD, a conserved and essential tRNA methyl transferase in bacteria, reduces protein synthesis in general and biosynthesis of the Gram-negative double-membrane structure in particular, suggesting a dual-strategy to accelerate bactericidal action. While many small molecular compound libraries have been screened to target TrmD, isolated hits lack the cell-permeability necessary to exhibit an antibacterial effect. We propose to screen a large collection of microbial extracts and fractions to target TrmD, using a novel and HTS-ready Ec tolC+ cell-based assay, to identify NPs that are cell-permeable, resistant to efflux, and selective of targeting, which will serve as leads for the next generation of antibiotics with a novel mechanism of action distinct from those currently in clinical use.