Cell-Based Discovery Methods Furnish New Compounds Effective Against Gram-Negative Bacteria

Bacterial resistance to antibiotics inevitably results from their clinical use, and we have to continuously develop new antibiotics to stay ahead in this biological arms race. It is particularly important and challenging to develop new antibiotics for Gram-negative bacteria, which possess an outer membrane (OM) with a continuous outward layer of lipopolysaccharide (LPS). This structure forms a permeability barrier that protects bacterial targets from antibiotic engagement, and its assembly requires tens of conserved proteins to cooperate in the biosynthesis and transportation of LPS, many of which are essential for bacterial viability. The work reported here aims to develop approaches to discover new antibiotics specifically against Gram-negative bacteria. In order to find cell-penetrating compounds and to capture inhibitors of a variety of targets, I focused on cell-based assays instead of in vitro methods. Approximately 700,000 compounds were screened for antibacterial activity against E. coli, and close to 1,500 compounds were active. Multiple cell-based methods were developed to screen that subset for compounds with activity against promising targets/pathways, including LPS biogenesis. This thesis discusses in detail one of the cell based assays developed. This particular cell-based filter was based on the assumption that compounds with greater activity against Gram-negative bacteria, which are generally less permeable than Gram-positive bacteria, are likely to act on targets specific to the former such as those involved in maintaining the outer membrane permeability barrier. We identified a compound with good activity against wt E. coli and weak activity against Acinetobacter baumannii but no activity against S. aureus or B. subtilis. Surprisingly, all resistance-conferring mutations in E. coli and in A. baumannii mapped to the active site of phenylalanine tRNA synthetase. The compound was confirmed to inhibit purified WT enzyme but did not inhibit a purified mutant enzyme. The compound was not active against mammalian cells, but was active against M. tuberculosis. In combination, it was shown to prevent resistance to a different tRNA synthetase inhibitor that was demonstrated to be safe in clinical trials, but was withdrawn due to development of resistance. Structure-activity relationship (SAR) studies of the hit compound rapidly led to a five-fold improvement in antibacterial activity and broadened spectrum. Further improvements are expected.Chemistry and Chemical Biolog