*WINNER* Computational Design of Novel Inhibitors of Dihydrofolate Reductase in Three Bacterial Species

Abstract

This project aims to design high affinity small molecule inhibitors of bacterial dihydrofolate reductase (DHFR) for the purpose of obtaining broad-spectrum antibiotics against multiple bacteria, including Bacillus anthracis (anthrax), Staphylococcus aureus, and Mycobacterium tuberculosis. Inhibitors were designed using MOE 2020 (Chemical Computing, Ltd., Montreal, Quebec, Canada) based on a previous ZINC Database search to target the active site of DHFR based on computational analysis of the energetic frustration and evolutionary importance of amino acid residues present. This analysis was conducted using the Protein Frustratometer (http://frustratometer.qb.fcen.uba.ar/; EMBNet Aargentina, Buenos Aires, Argentina) and Evolutionary Trace (http://lichtargelab.org/software/ETserver; Baylor College of Medicine, Baylor University, Houston, Texas USA). Evolutionary trace and frustration define the active site, by determining binding sites, and areas of the molecule in high energetic states, respectively. Designed inhibitors were docked into each protein using the Docking module of MOE 2020, and the binding residues were then compared to the areas of evolutionary trace and frustration to help determine if the molecules had favorable binding scores. 189 small molecules were designed to interact with these amino acid functional groups based on complementary, non-covalent functional group interactions. The ligand interactions for the top compounds in each bacteria were examined and these compounds were examined according to Lipinski’s Rule of Five, which helps to determine potential druggability. One compound was found to have favorable bonding across all three bacterial DHFR, and fourteen compounds were recognized as having favorable bonding across two bacterial DHFR.