Discovery of novel inhibitors of the bacterial cell division protein FtsZ by computational drug screening coupled with bioassays

Abstract

The problem of antibiotic resistance is worsening worldwide because of the overuse of existing antibiotics. In order to overcome the crisis of antibiotic resistance, there is an urgent need for alternative antibacterial agents that have novel mechanisms of action. FtsZ is an unexploited and attractive target for antibacterial drug discovery because of its widespread conservation in the bacterial kingdom, its absence in the mitochondria of higher eukaryotes and its known biochemical activity and molecular structure. FtsZ plays an essential role in prokaryotic cell division machinery in which undergoes GTP-dependent polymerization at midcell and assembles into the dynamic Z-ring at the site of division. The Z-ring acts on a framework for the recruitment of other cell division proteins to initiate Z-ring contraction to form bacterial daughter cells. Although FtsZ shares structural and functional similarity with eukaryotic tubulin, most of the tubulin/microtubule targeting agents, paclitaxel and colchicines, do not affect the dynamic assembly of FtsZ, indicating that FtsZ can be a selective antibacterial target. Although several compounds that block bacterial cell division and/or inhibit the biochemical activity of FtsZ in vitro have been reported, so far none has demonstrated efficacy in models of infection or has entered clinical evaluation. In this thesis, a computer-aided drug screening method was employed to identify potential FtsZ inhibitors. After in silico screening of several natural product libraries, 16 potential FtsZ inhibitors were subjected to bacterial growth inhibitory tests and in vitro assays. The compound 4-(((2R,4S,5R)-5-(2-Methyl-6-(thiophen-2-yl)pyrimidin-4-yl)-quinuclidin-2-yl)methylcarbamoyl)butanoic acid (5) was found to exhibit modest antibacterial activity against a number of Gram-positive and -negative bacteria. Compound 5 perturbed the Z-ring assembly in living E. coli cells and caused the elongation of B. subtilis and E. coli cells in vivo. It also perturbed the assembly and bundling of FtsZ protofilaments by inhibiting the GTPase activity of FtsZ. The ligand-induced destabilization of FtsZ protofilaments would hamper the functioning and formation of Z-ring to cause an inhibition of bacterial cytokinesis without affecting DNA replication or nucleotide segregation. Compound 5 was also found to show no inhibition on tubulin assembly into microtubules in vitro. In order to obtain the optimal model for predicting the binding mode of compound 5 in the GTP binding site of FtsZ and for subsequent virtual screening of compound 5 derivatives, SCan Alanines and REfine (SCARE) and homology modeling techniques were applied. As a result, several best docking models were successfully built to predict the optimal binding mode of compound 5. These models were then used to screen a library of compound 5 analogues. Sixty-nine top-scoring compounds were selected and tested experimentally. Finally, compound 5-706 was identified as the most potent derivative against FtsZ, which also exhibited high antibacterial activity against a broad range of bacteria with enhanced inhibitory effect on the GTP hydrolysis of FtsZ. The combination of docking simulation and experimental bioassays leads to study the early structure-activity relationship (SAR) of compound 5 derivatives, which can facilitate the subsequent lead optimization process. To assess the potential of compound 5 and its derivatives for combination therapy in treating bacterial infections, the synergistic effects in combination with a conventionally used β-lactam antibiotic (ampicillin) or H7 helix inhibitor of FtsZ (3-MBA) were investigated. Combination of compound 5 or 5-316 with ampicillin demonstrated synergistic effects against drug-resistant S. auerus strain. For a drug-sensitive S. auerus strain, compound 5 with ampicillin or 3-MBA showed partial synergistic effect.