Work

Drug repurposing and African natural product screening, targeting the Parkinson‘s-associated LRRK2 G2019S mutant kinase using molecular docking and molecular dynamics approaches

Year: 2026

Type: article

Abstract: The G2019S mutation in leucine-rich repeat kinase 2 (LRRK2) is the most prevalent genetic contributor to Parkinson’s disease and is associated with increased kinase activity. Targeting this mutation remains a key therapeutic strategy; however, structure-guided identification of repurposable inhibitors remains limited. In this study, we implemented an integrated computational pipeline combining structure-based virtual screening, pharmacophore modeling, molecular dynamics (MD) simulations, and MM-GBSA free-energy calculations to prioritize candidate inhibitors of the LRRK2 G2019S mutant. A total of 9,407 compounds from DrugBank and African natural product libraries were screened against the ATP-binding pocket of the recently resolved cryo-electron microscopy structure of the mutant kinase (PDB ID: 8TZC). A structure-based pharmacophore model derived from the co-complex inhibitor MLi-2 was used for complementary screening. Selected candidates were further evaluated using 300-ns MD simulations, binding stability metrics, and per-residue energy decomposition. Virtual screening of 9,407 compounds (2,614 FDA-approved drugs, 4,923 East African and 1,870 North African natural products) against the ATP-binding pocket of LRRK2 G2019S, validated by MLi-2 redocking (RMSD = 0.095 Å); identified 14 candidates with docking scores more favorable than the reference inhibitor MLi-2 (− 10.1 kcal/mol) and hinge-region hydrogen bonding with GLU1948 and/or ALA1950 (four DrugBank, six East African, and four North African compounds). A complementary structure-based pharmacophore model derived from the LRRK2 G2019S–MLi-2 co-complex was screened against the DrugBank subset, yielding 81 pharmacophore-matched hits, of which four compounds, Rimonabant, Amlexanox, Bromocriptine, and Cilostazol, showed strong convergence across pharmacophoric fit, docking affinity, and hydrogen bonding at GLU1948/ALA1950. Representative top-ranked candidates from both approaches included, among FDA-approved drugs, Acalabrutinib (-10.6 kcal/mol), Atovaquone (-10.6 kcal/mol), Belumosudil (-10.4 kcal/mol), Rimonabant (-9.8 kcal/mol), and Bromocriptine (-9.2 kcal/mol); and, among natural products, Gallagic acid (-11.0 kcal/mol), 7,7′-dihydroxy-6,8′-bicoumarin (-10.7 kcal/mol), 10-hydroxy-10,7′-(chrysophanol anthrone)-chrysophanol (-10.7 kcal/mol), Anastatin A (-10.7 kcal/mol), and 5,7,2′,4′-tetrahydroxy-6-(3-methylbut-2-enyl)isoflavanone (-10.3 kcal/mol). Across the 300-ns MD simulations, Acalabrutinib and the chrysophanol anthrone dimer maintained the most stable backbone conformations, with mean RMSD values of approximately 1.066 and 1.022 nm, respectively, closely approaching the stability profile of the reference inhibitor. MM-GBSA calculations confirmed MLi-2 as the strongest overall binder (ΔG = − 46.91 ± 3.44 kcal/mol), while Rimonabant showed the most favorable binding free energy among repurposed drugs (− 36.61 ± 5.53 kcal/mol) and Gallagic acid was the top-performing natural product (− 33.18 ± 8.70 kcal/mol), sustaining a persistent hydrogen-bonding network throughout the simulation. These convergent docking, pharmacophore, dynamic stability, and binding free-energy analyses identify Rimonabant, Gallagic acid, Acalabrutinib, and Atovaquone as the most promising repurposable candidates against LRRK2 G2019S. Collectively, this study utilizes the recently resolved cryo-EM structure of the LRRK2 G2019S mutant to implement an integrated structure-based screening strategy and prioritize multiple repurposable drug candidates and natural scaffolds with favorable structural and energetic profiles. The findings provide mechanistic insight into mutation-specific ligand binding and establish a computational foundation for future experimental validation and LRRK2-targeted drug discovery in Parkinson’s disease.

Source: Scientific Reports

DOI: 10.1038/s41598-026-71061-9

Authors: Omnia Saeed, Martina B. William, Mohamed Salama, Tamer M. Ibrahim, Raghda Ramadan

Institutions: Nile University, American University in Cairo, Assiut University, Trinity College Dublin, Kafrelsheikh University

Language: English

Topic: Computational Drug Discovery Methods

Subfield: Computational Theory and Mathematics

Field: Computer Science

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $2,690