Computational studies on structurally diverse dipeptidyl peptidase IV inhibitors: an approach for new antidiabetic drug development

Abstract

Dipeptidyl peptidase IV (DPP-IV) deactivates the natural hypoglycemic incretin hormone GLP-1. Inhibition of this enzyme restores glucose homeostasis in diabetic patients making it an attractive target for the development of new antidiabetic drugs. With this in mind, we suggested an in silico work flow for the identification of novel DPP-IV inhibitors. Ligand-based and structure-based pharmacophores were designed using HipHop program provided in catalyst and ligandScout 3.0 software, respectively. Generated models were validated by receiver operating characteristic curve analysis, Guner–Henry scoring method and by pharmacophore-based screening of marketed DPP-IV inhibitors. Ligand-based pharmacophore model A scored 0.8 AUC value, 0.865 Guner–Henry score and gave all marketed DPP-IV inhibitors as hits through screening while structure-based pharmacophore B scored 0.77 AUC value, 0.66 Guner–Henry score and gave four marketed DPP-IV inhibitors as hits (except alogliptin) out of five. These validated pharmacophores have effectively been used in search of three databases, Maybridge hitfinder collection, Chemdiv, and Asinex. Resulting hits were subjected to molecular docking using ligandfit program. Five hit compounds namely Asinex ASN 09417841, AW 00785, ChemDiv 0173-0023, ChemDiv 0276-0112, and ChemDiv 8010-1357 scored high Ligscore1 and −PLP1 score comparable to standard drug sitagliptin. Good interactions were found with important residues like Glu205, Glu206, Tyr662, Phe357, Arg358, Tyr666 etc. They were reported as novel virtual leads to design potent DPP-IV inhibitors.