Computer aided drug discovery: Design, synthesis and testing of novel anti-cancer agents

Abstract

The continuous enhancements of computational tools for drug discovery, have greatly improved the pre-clinical stages of pharmaceutical research, to a higher level of efficiency and greater speed. The combination between homology modelling methods in structure-based drug design, and the chemoinformatics techniques in lead optimization, has allowed successful development of small molecule inhibitors with 30 folds higher potency of best known ASCT2 inhibitors. Three dimensional model of ASCT2 protein was constructed via fold recognition technique, the preferred method in absence of highly similar protein sequences. This model was used in virtual screening against commercial SPECs database comprised of approximately 300 000 compounds of drug-like molecules to identify potential inhibitors and characterize active site interactions through different docking procedures. Visual analysis of the docking poses with correlation to the biological results has allowed verification of the ASCT2 active site and determine the important residues for strong ligand binding. Furthermore, the ASCT2 model has been used to validate the biological results of the lead development stage through further docking analysis. AMACR homology model was built using the three dimensional X-ray crystal structure of MCR with high sequence similarity and shared protein family. The Maestro modelling software platform offered highly accurate model predictions and further docking calculations in the virtual screening stage. The biological testing was highly reliant on general cell viability results rather than specific AMACR inhibition due to the high cost and difficulties associated with AMACR activity assays. To conclude, computer-aided drug discovery in partnership with complementary in vitro techniques have made major contributions in this thesis, to develop biologically active molecules, demonstrating the high efficiency, great speed and cost-effective advantages of molecular modelling.