Computational modeling approaches to validate the druggability of the 26- and 28-kDa Schistosoma glutathione transferase enzymes using bromosulfophthalein as a benchmark ligand

Abstract

PHARMACOPHORE MODELS are 3-D representations of the chemical and spatial features required for interaction with a drug target. These. models offer advantages in early-phase drug design by expediting screening experiments and enabling the sampling of highly specific chemical. space subsets, such as those containing quality drug-like candidates. The glutathione transferase enzyme of Schistosoma spp. (SGST) has been identified as an attractive drug target for the novel treatment of human schistosomiasis. We observed selective inhibition of SGST by bromosulfophthalein. Bromosulfophthalein was found to complex with SGST at a drug binding site in the target dimer interface, providing a suitable benchmark for the design of discriminative SGST pharmacophores. The aim of this research is to construct, deploy and evaluate pharmacophore models of the SGST drug binding site. The objectives are: to characterise the SGST drug binding site, to develop the pharmacophore models and finally to evaluate the drug-resolving ability of the models. We observed significant differences in the drug-binding character of SGST, compared to human glutathione transferase (hGST) counterparts, particularly that SGST supports binding of phenol and sulfonate moieties. Five- and four feature pharmacophores were developed for the respective 26- and 28 kDa SGST variants. Finally, the models demonstrated remarkable ability to retrieve candidates displaying drug-like qualities. In conclusion, we characterised and developed pharmacophore models of the drug binding domains from two major SGST variants. Assessment of drug-resolving power validates the capability of the models to sample drug-like chemicals. Altogether, these accomplishments enable efficient and reliable screening toward novel drug treatment for human schistosomiasis