Molecular dynamic simulations and structure-based pharmacophore development for farnesyltransferase inhibitors discovery

Abstract

Farnesyltransferase is one of the enzyme targets for the development of drugs for diseases, including cancer, malaria, progeria, etc. In the present study, the structure-based pharmacophore models have been developed from five complex structures (1LD7, 1NI1, 2IEJ, 2ZIR and 2ZIS) obtained from the protein data bank. Initially, molecular dynamic (MD) simulations were performed for the complexes for 10 ns using AMBER 12 software. The conformers of the complexes (75) generated from the equilibrated protein were undergone protein–ligand interaction fingerprint (PLIF) analysis. The results showed that some important residues, such as LeuB96, TrpB102, TrpB106, ArgB202, TyrB300, AspB359 and TyrB361, are predominantly present in most of the complexes for interactions. These residues form side chain acceptor and surface (hydrophobic or π–π) kind of interactions with the ligands present in the complexes. The structure-based pharmacophore models were generated from the fingerprint bits obtained from PLIF analysis. The pharmacophore models have 3–4 pharmacophore contours consist of acceptor and metal ligation (Acc & ML), hydrophobic (HydA) and extended acceptor (Acc2) features with the radius ranging between 1–3 Å for Acc & ML and 1–2 Å for HydA. The excluded volumes of the pharmacophore contours radius are between 1–2 Å. Further, the distance between the interacting groups, root mean square deviation (RMSD), root mean square fluctuation (RMSF) and radial distribution function (RDF) analysis were performed for the MD-simulated proteins using PTRAJ module. The generated pharmacophore models were used to screen a set of natural compounds and database compounds to select significant HITs. We conclude that the developed pharmacophore model can be a significant model for the identification of HITs as FTase inhibitors.