A multi-level computational analysis identifies Ocimum basilicum phytochemicals as potent inhibitors of the Marburg virus VP40 matrix protein

Abstract

Marburg virus disease (MVD) is an extremely lethal hemorrhagic fever with no approved antiviral treatment available, which emphasizes the need to develop newer and safer drug candidates. In the current study, we used an in silico approach to identify candidate inhibitors of the Marburg virus matrix protein VP40, which plays a vital role in the viral assembly and budding process using bioactive compounds derived from Ocimum basilicum. A homology model of the Marburg virus VP40 protein was built and assessed by Ramachandran, QMEAN, ProSA, Verify3D and ERRAT analysis, and 156 bioactive compounds derived from Ocimum basilicum were then screened against it by molecular docking. Rutin ranked first, with a docking score of -10.4 kcal/mol and hydrogen-bonded and van der Waals contacts to residues of the predicted pocket. A 500 ns all-atom molecular dynamics simulation indicated a stable VP40-rutin complex, with a backbone RMSD of 1.98 ± 0.19 Å, a mean Cα RMSF of 1.10 ± 0.52 Å, and a compact, well-preserved fold (radius of gyration 19.77 ± 0.15 Å). End-state binding free energies calculated over 51 uniformly sampled trajectory frames were favourable under both solvation models, at -46.08 ± 5.96 kcal/mol (MM/GBSA) and -27.92 ± 7.63 kcal/mol (MM/PBSA), driven primarily by van der Waals and electrostatic contributions. ADMET profiling, however, revealed a substantial pharmacokinetic liability: rutin violates three of Lipinski’s criteria (molecular weight 610.52 g/mol, topological polar surface area 269.43 Å2, 10 hydrogen-bond donors and 16 acceptors) and its predicted human intestinal absorption is only 23.45%, so it is best regarded as a lead scaffold requiring structural or formulation optimisation rather than a ready-to-use agent. Validation of the docking protocol against independent molecular dynamics reference poses gave a mixed result. The predicted affinity was reproducible, and the predicted pocket was clearly preferred over decoy boxes placed elsewhere on the protein surface, but pose recovery failed. No top-ranked pose fell within the conventional 2 Å success threshold of the reference pose. Docking was therefore used only as an initial screening filter to rank compounds, and neither the binding geometry nor any mechanism of inhibition is inferred from it. All structural and energetic conclusions reported here rest on the molecular dynamics ensemble and the end-state free-energy analyses. Gas-phase DFT analysis gave a narrow HOMO-LUMO gap (4.13 eV), high softness and a moderate electrophilicity index, which offer supportive electronic evidence of charge-transfer capacity rather than independent confirmation of affinity. Taken together, these findings identify rutin as a natural lead compound against VP40, subject to the pharmacokinetic and docking limitations set out above, and provide a critically validated computational framework for anti-Marburg drug discovery.