Glioblastoma is an aggressive and treatment-resistant brain tumor with poor prognosis, frequently driven by aberrant activation of receptor tyrosine kinases such as fibroblast growth factor receptor 3 (FGFR3). Genetic alterations including FGFR3–TACC3 fusions and post-transcriptional deregulation mediated by tumor-suppressive microRNAs (miR-99a and miR-100) contribute to sustained FGFR3 signaling, glioma progression, and therapeutic resistance. In this study, a structure-based computational pipeline was employed to identify natural inhibitors targeting FGFR3. A phytochemical library comprising 25 plant-derived compounds was screened using ADME criteria, resulting in seven candidates with favorable gastrointestinal absorption and drug-likeness. Among them, Guggulsterone emerged as the top-ranked compound, exhibiting the highest docking affinity (− 10.1 kcal/mol) toward the FGFR3 kinase domain and forming stable hydrogen bonding and hydrophobic interactions with key active-site residues. The FGFR3 structure used showed high stereochemical quality (ERRAT score: 96.5; 93% residues in favored Ramachandran regions). Molecular dynamics simulations (500 ns) demonstrated stable complex formation, with RMSD convergence after 40 ns, low RMSF values, and consistent radius of gyration (1.96–2.04 nm). Persistent intermolecular hydrogen bonds (two to four) and stable solvent accessibility were observed throughout the simulation. MMGBSA binding free energy calculations predicted a favorable interaction (ΔGtotal = − 30.44 kcal/mol), primarily driven by van der Waals and electrostatic contributions. Pharmacophore analysis revealed two hydrogen bond acceptors and seven hydrophobic features, while density functional theory calculations indicated moderate chemical reactivity (HOMO–LUMO gap: 0.1797 a.u.). Toxicity assessment using ProTox 3.0 classified Guggulsterone as low-toxic (LD50 = 2300 mg/kg, Class 5). Collectively, these in-silico findings suggest Guggulsterone as a promising natural FGFR3-binding scaffold warranting further experimental validation in FGFR3-driven glioma models.
Javaid et al. (Sun,) studied this question.