Computational Identification of Potential HMG-CoA Reductase Modulators Through Drug Repurposing: Structural Insights and Therapeutic Implications.

Abstract

Introduction: HMG-CoA reductase (HMGCR) is a validated lipid-lowering target, but statin intolerance, variable response, and drug-drug interactions justify exploration of mechanistically distinct modulators. This study prioritized approved drugs that may bind predicted non-orthosteric HMGCR pockets.

Methods: A computational workflow combining homology modeling (SWISS-MODEL; GMQE = 0.95), CB-Dock pocket detection, AutoDock Vina docking, 100 ns all-atom molecular dynamics (MD) simulations (GROMACS 2023.2), MM/PBSA binding-free-energy estimation, and in silico ADMET screening was applied to Fluspirilene, Lenvatinib, Siponimod, and Lumacaftor.

Results: Two putative non-orthosteric pockets were identified with calculated volumes of 394 Å3 (Pocket 1) and 305 Å3 (Pocket 2). Fluspirilene showed the most favorable Pocket 1 docking score (-9.4 kcal/mol) and MM/PBSA binding-free-energy estimate (-65.2 ± 3.8 kcal/mol), whereas Lumacaftor showed the most favorable Pocket 2 score (-8.9 kcal/mol; MM/PBSA: -55.7 ± 4.1 kcal/mol). MD trajectories supported retention of the selected poses throughout the 100 ns simulation window.

Discussion: The predicted interactions involved hydrophobic contacts, π-associated interactions, hydrogen bonding, and electrostatic contacts at sites spatially distinct from the canonical statin/HMG-CoA catalytic region. These findings support a non-orthosteric binding hypothesis but do not establish functional allosteric inhibition without enzyme kinetics, direct binding assays, and structural validation.

Conclusion: The study identifies computationally prioritized candidate scaffolds for putative non-orthosteric HMGCR modulation and defines the biochemical, kinetic, and structural validation required before any therapeutic interpretation.