Identification of Effective Cyclin-Dependent Kinase 3/Cyclin E Inhibitors Using Multi-Level Computational Screening and Simulation

Cyclin-dependent kinase 3 (CDK3) plays a crucial role in regulating cell cycle progression through the G0/G1 and G1/S phases. Although extensively lowered in normal tissue, CDK3 is overexpressed in multiple cancers, making it a promising target for anticancer therapy. Despite its significance, no established inhibitors of CDK3 are currently available. In this study, a library of 204 derivatives of five potent CDK3 inhibitors—Purvalanol A, NU6102, R547, Dinaciclib, and RO-3306—was screened for potential CDK3/cyclin E1 inhibition using a multi-tiered computational pipeline. Molecular docking with AutoDock 4.2 and AutoDock Vina identified five top-scoring compounds, which were further evaluated through absorption, distribution, metabolism and excretion (ADME) profiling, toxicity prediction, and quantum chemical analysis. Frontier molecular orbital (FMO) and global reactivity descriptor studies suggested favourable electronic properties for receptor binding. Molecular dynamics (MD) simulations (100 ns in duplicates) and molecular mechanics Poisson-Boltzmann surface area (MM-PBSA) calculations revealed stable ligand-receptor interactions and favourable binding energetics. Two leading candidates, CID_11212010 and CID_25211747, were subjected to extended 1200 ns MD simulations, with CID_25211747 showing exceptional stability and minimal conformational fluctuation. ONIOM (Our own N-layered Integrated molecular Orbital and molecular Mechanics) calculations further validated the strong binding affinity of CID_25211747 with CDK3. Collectively, these results nominate CID_25211747 as a promising lead compound towards the development of effective CDK3 antagonists, offering valuable insight for future experimental drug design.