Discovery of Novel Tubulin Inhibitors via Structure-Based Hierarchical Virtual Screening

To discover novel tubulin inhibitors, we performed structure-based virtual screening against the colchicine binding pocket. In combination with a hierarchical docking and scoring procedure, the structural information of an additional subpocket in colchicine site was applied to filter out the undesired docking hits. This strategy automatically resulted in 63 candidates meeting the structural and energetic criteria from a screening library containing approximately 100,000 diverse druglike compounds. Among them, nine molecules were chosen for experimental validation, which all share the similar binding pose and contain an enriched scaffold bearing thiophene core. Encouragingly, five compounds are active in tubulin polymerization assay. The most potent inhibitor, 2-(2-fluorobenzamido)-3-carboxamide-4,5-dimethylthiophene, is structurally distinct to any known colchicine site binders and has higher ligand efficiency than colchicine. On the basis of its predicted binding pose, we systematically probed its binding characteristics by testing series of structural modifications. The obtained structure-activity relationship results are consistent with our binding model, and the inhibition activities of two analogues are improved by 2-fold. We expect that the novel structure discovered in the present study may serve as a starting point for developing tubulin inhibitors with improved efficacy and fewer side effects. We also expect that our hierarchical strategy may be generally applicable in structure-based virtual screening campaigns.