Structure-based inhibitor design and validation : application to Plasmodium falciparum glutathione S-transferase

Abstract

The primary aim of this study was to use a computational structure-based ligand design strategy in finding novel ligands that could act as inhibitors of PfGST as basis for future antimalarial drug development. Since there is only one PfGST isoenzyme present in the parasite and the architecture of the binding site differs significantly from its human counter part, PfGST is considered a highly attractive drug target. Inhibition of PfGST is expected to interfere at more than one metabolic site in synergy: it is likely to disrupt the glutathione-dependent detoxification process, which will lead to an increase in the cytotoxic peroxide concentration and most likely lead to an increase in the levels of ferriprotoporphyrin IX and hemin as well. S-hexyl glutathione was co-crystallized with PfGST (Harwaldt et al., 2004), consequently it was seen as one of the most important lead compounds in the development of PfGST inhibitors. The first step in the rational drug design strategy was to modify GTX, concentrating on its ability to bind competitively to the G site and the hydrocarbon chain protrudes into the H site as well. Considering the 3D structure of the enzyme, modifications to GTX were made by LUDI and NEWLEAD, resulting in a library of active site binding ligands ranked by AutoDock according to their ability to optimally bind to PfGST. Additionally, the ligands were ranked according to their affinity for binding to PfGST produced by AutoDock, LUDI and XScore. Once all the compounds were ranked by these in silico methods they were screened for acquisition or synthetic accessibility and those available were experimentally screened for activity against recombinantly expressed PfGST. Based on in silico predictions NDA was the best inhibitor followed by LAP and EDP. From the biological assay and Lineweaver-Burk analysis the order of inhibition was NDA as the best inhibitor tested, followed by LAP and EDP. EDP and LAP showed competitive inhibition but the inhibition constant values were signi_cantly lower than GTX. With respect to GSH and CDNB, NDA was found to be a non-competitive inhibitor. It was suggested therefore that NDA binds to a non-substrate Summary 93 binding site that may lead to conformational change of the enzyme and hence lead to a loss in enzyme activity. This data leads to the conclusion that the H site should be better exploited in order to find more potent inhibitors or non-substrate binding sites. It was concluded that the experimental results add confidence to the discriminative power of the structure-based ligand design strategy and that these inhibitors could form scaffolds for future antimalarial drug development.