In silico Screening for Identification of Novel Aurora Kinase Inhibitors by Molecular Docking, Dynamics Simulations and Ligand-Based Hypothesis Approaches

Abstract

Aurora kinase family members are involved in a wide variety of cell cycle events including centrosome separation, cytokinesis, kinetochore formation, spindle assembly, chromosomal segregation and microtubule dynamics. Typically, dysfunction of Aurora proteins is associated with aneuploidy, cell death and mitotic arrest, leading to tumorigenesis. This spurred a vast interest to identify pharmacologically active small-molecule inhibitors of Aurora proteins. In this study, we isolated four novel inhibitors through virtual screening and docking analyses. These hits were subsequently characterized by molecular dynamic simulations to monitor their binding stabilities at ATP binding site. To aid characterizing novel and more potent inhibitors for Aurora kinases, we explored selective features of our ligand dataset by ligand-based pharmacophore modeling approach. The best pharmacophore model was then employed for performing virtual screenings of libraries isolated from Princeton and Uorsy databases. On the basis of common pharmacophore features, Lipinski’s rule of five, absorption, distribution, metabolism and elimination properties, hits were short listed and refined by molecular dockings. Finally, the selected compounds were validated on the basis of binding capabilities, consensus scoring and activity values. We propose that the novel inhibitors described in this study may warrant characterization in designing active lead for clinical studies that may serve as anticancer drugs in future.