Machine Learning-Driven Drug Repurposing Investigation to Identify New Tubulin Inhibitors Against Cancer

Tubulin is one of the major structural proteins that exists either as free forms or as microtubules. These microtubules have unique characteristics, termed as ‘exquisite dynamic behavior’, that is associated with the polymerization and depolymerization of the microtubules or the tubulin units. In cancer cells, the tubulin-mediated dynamicity is lost owing to the elevated expression and functioning of microtubules. These lead to uncontrolled cancer cell proliferation, evading controlled cell cycle checks, survival, and resistance to existing cancer chemotherapy. Considering this, the tubulins/microtubules are validated drug targets in cancer. The agents targeting the microtubules are called microtubule-targeting agents (MTAs). These agents are further subclassified into three groups and act at 7 druggable sites identified so far. However, the majority of inhibitors approved are phytochemicals and are semisynthetic-derived, which are associated with poor pharmacokinetics and toxicities that hamper their clinical use. The rapidly growing cancer cases globally demands rapid drug discovery. One of the rapid and recommended ways for drug discovery is drug repurposing or repositioning. To achieve this, we employed machine learning (ML)-driven, integrated with in silico-based high-throughput virtual screening of 4500 FDA-approved compounds to identify new tubulin inhibitors interacting with the Colchicine Binding Site (CBS) with anticancer potential. The virtual screening workflow was supported by structure-based drug design approaches including three stages of molecular docking, MM-GBSA binding free energy calculations and molecular dynamics (MD) simulations, leading to a refined library of potential tubulin inhibitors. The best hit compounds were tested biochemically to verify tubulin inhibition, as well as tested against a panel of cancer cell lines, whereby Omeprazole and Podofilox demonstrated promising results. The outcomes of this work may lay a solid foundation for further drug development projects, targeting tubulin and cancer via drug repurposing.