Identification of Aurora-A Inhibitors by Ligand and Structure-Based Virtual Screening.

Abstract

Aurora kinase A has been identified as one of the most attractive targets for cancer therapy because of its critical role in the regulation of the cell cycle. In order to identify active compounds with structural diversity we performed virtual screening. 3D-QSAR pharmacophore models were developed and the best model was used as a query for screening the databases. Ligand and structure-based virtual screening protocol was conducted sequentially by applying the common feature pharmacophore and molecular docking to discover potent Aurora-A inhibitors. A total of eighty-eight compounds were selected for the in vitro activities against various human cancer cell lines (DU145 and HT29). Considering the activity data, we have identified seven compounds to be considered for the next step, among which four compounds had high inhibition rate (above 50 %) at 10 µM with GI50 lower than 10 µM. Based on the cell line and enzyme assay (Aurora-A & B) result, these four compounds were used as template/query molecule for similarity search. The best result was obtained for similarity hit SH3. It had IC50 of 0.578 and 11.77 µM for Aurora-A and B respectively, which implies 20-fold selectivity over Aurora-B. The hits obtained from this screening scheme could be potential drug candidates after further optimization.