IN SILICO CONSTRUCTION AND MOLECULAR-DYNAMICS VALIDATION OF A TRIPLE-MUTANT (K103N/Y181C/Y188L) HIV-1 REVERSE TRANSCRIPTASE AND STRUCTURE-BASED SCREENING OF THYMOQUINONE-DERIVED NON-NUCLEOSIDE INHIBITORS

Resistance mutations in the non-nucleoside inhibitor binding pocket (NNIBP) of HIV-1 reverse transcriptase (RT) — most prominently K103N, Y181C and Y188L — continue to erode the durability of the entire non-nucleoside reverse transcriptase inhibitor (NNRTI) class. We report an integrated computational workflow that (i) builds and validates a single, clinically representative RT model carrying all three mutations simultaneously, and (ii) uses that model to prioritise novel chemotypes seeded from the natural-product quinone thymoquinone (TQ). The p66 catalytic subunit of wild-type RT (PDB 2BAN) was edited to introduce K103N, Y181C and Y188L, and the triple mutant (designated 2BAP) was generated by comparative modelling in MODELLER using the single-mutant crystal structures 3MED, 1JKH and 2YNF as templates. The model was energy-minimised and relaxed by a 100 ns explicit-solvent molecular-dynamics (MD) simulation in NAMD. Stereochemical quality improved markedly relative to the parent structure (PROCHECK core-region occupancy 75.2% → 96.0%; ERRAT overall quality factor 65.25 → 78.86, rising to 87.36 after MD), and the backbone RMSD plateaued near 100 ns, indicating a stable, well-packed model. A predefined 24–26 residue NNIBP grid was used for AutoDock-based docking of TQ and of three leads — a virtual-screening-derived pentacenedione (Ligand 1), a rationally assembled pyridin-2(1H)-one (Ligand 2, TQ-PZP-03) and a 3D-QSAR-benchmark pyridinone (Ligand 3, VC1) — against wild-type and the K103N, Y181C, Y188L and triple-mutant structures. All leads occupied the NNIBP and, critically, retained their binding scores across the mutant panel rather than collapsing at Y181C/Y188L. In-silico ADMET (SwissADME) and toxicity (ProTox) profiling passed the leads through standard drug-likeness filters, while an efavirenz benchmark exposed the limited applicability domain of the toxicity model for this chemical space. The validated triple-mutant model and the retained cross-mutant docking profiles together provide a structure-based rationale for advancing the pyridinone leads to experimental evaluation.