Interactions of flavonoid and coumarin derivative compounds with transforming growth factor-beta receptor 1 (TGF-βR1): integrating virtual screening, molecular dynamics, maximum common substructure, and ADMET approaches in the treatment of idiopathic pulmonary fibrosis

Abstract

Context

Idiopathic pulmonary fibrosis (IPF) is a chronic and progressive lung disease characterized by very limited treatment options and significant side effects from existing therapies, highlighting the urgent need for more effective drug-like molecules. Transforming growth factor-beta receptor 1 (TGF-βR1) is a key player in the pathogenesis of IPF and represents a critical target for therapeutic intervention. In this study, the potential of plant-derived flavonoid and coumarin compounds as novel TGF-βR1 inhibitors was explored. A total of 1206 flavonoid and coumarin derivatives were investigated through a series of computational approaches, including drug-like filtering, virtual screening, molecular docking, 200-ns molecular dynamics (MD) simulations in triplicate, maximum common substructure (MCS) analysis, and absorption-distribution-metabolism-excretion-toxicity (ADMET) profiling. 2′,3′,4′-trihydroxyflavone and dicoumarol emerged as promising plant-based hit candidates, exhibiting comparable docking scores, MD-based structural stability, and more negative MM/PBSA binding free energy relative to the co-crystallized inhibitor, while surpassing pirfenidone in these parameters and demonstrating superior pharmacological properties. In light of the findings from this study, 2′,3′,4′-trihydroxyflavone and dicoumarol could be considered novel TGF-βR1 inhibitors for IPF treatment, and it is recommended that their structural optimization be pursued through in vitro binding assays and in vivo animal studies.

Methods

The initial dataset of 1206 flavonoid and coumarin derivatives was filtered for drug-likeness using Lipinski’s Rule of Five in the ChemMaster—Pro 1.2 program, resulting in 161 potential candidates. These compounds were then subjected to virtual screening against the TGF-βR1 kinase domain (PDB ID: 6B8Y) using AutoDock Vina 1.2.5, identifying the top three hit compounds—dicoumarol, 2′,3′,4′-trihydroxyflavone, and 2′,3′-dihydroxyflavone. These hits underwent further exhaustive molecular docking for refinement of docking poses, followed by 200-ns MD simulations in triplicate using the AMBER03 force field in GROMACS. Subsequently, the binding free energies were calculated using the Molecular Mechanics/Poisson-Boltzmann Surface Area (MM/PBSA) method. MCS analysis was conducted to determine shared structural features among the top three hits, while ADMET properties were predicted using Deep-PK, a deep learning-based platform. Finally, the ligand–protein interactions were further visualized, analyzed, and rendered using ChimeraX, Discovery Studio Visualizer, and Visual Molecular Dynamics (VMD) program.