Pharmacophore modeling and molecular dynamics simulations to study the conformational stability of natural HER2 inhibitors in breast cancer therapy.

Abstract

HER2-positive breast cancer remains a significant clinical challenge, often exhibiting resistance to standard therapies. This study applies a comprehensive in silico approach to identify the natural compounds with potential inhibitory effects on HER2, focusing on pharmacophore modeling, virtual screening, molecular dynamics (MD) simulations, and binding affinity estimation. Initially, 24 known HER2 inhibitors from the BindingDB database were analyzed using Schrödinger's Phase module to generate a pharmacophore model, highlighting one hydrophobic (H) and three aromatic rings (RRR) features essential for HER2 binding. Screening against the Coconut Database, comprising 406,076 natural compounds, yielded 60,581 hits that matched the HRRR pharmacophore. These hits underwent a rigorous docking workflow with Glide (HTVS, SP, and XP modes), narrowing the candidates to 757 compounds with high binding affinity. Further refinement using Lipinski's rule of five produced a final set of 12 compounds exhibiting drug-like properties. 500-ns MD simulations evaluated these complexes' stability and dynamic behavior, while MM-GBSA calculations confirmed strong binding affinities dominated by van der Waals and electrostatic interactions. Compounds CNP0116178, CNP0356942, and CNP0136985 demonstrated superior binding profiles compared to the reference, marking them as lead candidates for HER2 inhibition. This study underscores the efficacy of computational methods in early-stage drug discovery and highlights promising candidates for further experimental validation and optimization. These findings offer a basis for developing targeted HER2 therapies and demonstrate the potential of natural compounds in advancing breast cancer treatment.