Related Experiment Video
Updated: Aug 9, 2026

Method for Identifying Small Molecule Inhibitors of the Protein-protein Interaction Between HCN1 and TRIP8b
Published on: November 11, 2016
ARG128-associated recognition of nobiletin by BACE1 revealed by triplicate molecular dynamics, MM-GBSA, and
1Liaoning University of Traditional Chinese Medicine, Shenyang, 110847, Liaoning, China.
Abstract:
Alzheimer's disease (AD) remains the leading cause of dementia worldwide, and beta-secretase 1 (BACE1) remains a high-priority target for reducing amyloid-beta production. Nobiletin, a polymethoxylated flavonoid from citrus peel, has reported neuroprotective and amyloid-lowering effects in AD models, and previous enzymatic work suggested weak BACE1 inhibition; however, its replica-level binding stability and residue-level energetic determinants remain unresolved. Here, we analyzed the nobiletin-BACE1 complex using ChEMBL-based chemical-space contextualization, qualitative QSAR uncertainty assessment, molecular docking, triplicate 200-ns all-atom molecular dynamics simulations, molecular mechanics generalized Born surface area (MM-GBSA) end-point scoring, residue-level decomposition, ProLIF interaction fingerprints, electrostatic surface analysis, and a same-protocol OM99-2 structural control. The ligand remained locally accommodated within the BACE1 cleft across the sampled trajectories, with a mean internal MM-GBSA score (DeltaG_bind) of - 15.71 +/- 0.76 kcal/mol (n = 3: - 14.20, - 16.35, and - 16.57 kcal/mol). Residue-level decomposition identified ARG128 as the dominant computed hotspot (total mean = - 11.13 +/- 0.38 kcal/mol), with favorable van der Waals (VDW mean = - 6.88 +/- 0.21 kcal/mol) and Coulombic (Coulomb mean = - 4.25 +/- 0.53 kcal/mol) components. Additional hydrophobic/aromatic packing involved VAL69, TYR71, PRO70, PHE38, and PRO129, while ASP32/ASP228 were positioned outside the dominant nobiletin contact pattern. Together, these results support a putative ARG128-associated, non-dyad-dominant catalytic-cleft recognition model that should be interpreted as a hypothesis-generating computational framework for future biochemical testing and scaffold optimization.
