IN SILICO SCREENING OF NATURAL COMPOUNDS AGAINST EGFR IN GLIOBLASTOMA: DOCKING, INTERACTION PROFILING, AND PHARMACOKINETIC ASSESSMENT

Abstract

Aim: The GBM is an aggressive primary brain tumour with a dismal prognosis, resistance to treatment, and high prevalence of deregulated EGFR signalling. Clinically approved EGFR TKIs like erlotinib have therapeutic potential, but their effectiveness in GBM is restricted by mechanisms of resistance and poor penetration of the central nervous system. For the present study, an integrated computational workflow was used which integrated molecular docking, residue-level interaction analysis and in silico pharmacokinetic analysis in order to assess selected natural compounds as potential EGFR targeting candidates. The crystal structure of the EGFR tyrosine kinase domain was downloaded from the RCSB Protein Data Bank (PDB ID: 1M17) and further prepared using AutoDock Tools. The molecular docking was carried out with AutoDock Vina with erlotinib as a reference inhibitor. The docked complexes were analysed for interaction using Discovery Studio Visualizer and PlexView, and the pharmacokinetic and drug-likeness properties were analysed using SwissADME. The most highly predicted binding affinity for EGFR of the screened compounds was diosquinone (−10.1 kcal/mol), which had a higher predicted binding affinity than erlotinib (−7.2 kcal/mol). Residue-level interaction analysis, however, indicated that diosquinone does not have the classical binding interactions for the hinge-region of the ATP binding site characteristic of erlotinib. Rather, alternative non-covalent interactions, such as cation π interactions, seemed to be responsible for ligand stabilisation. Pharmacokinetic assessment showed good drug-likeness properties and low predicted penetration into the CNS, suggesting low exposure to the CNS even though there is high affinity for the receptor. In summary, the results show that docking affinity is not necessarily a good indicator of biologically relevant EGFR inhibition and underscore the need to incorporate docking geometry and pharmacokinetic analysis into computational drug discovery pipelines. The scaffold of diosquinone could thus serve as a structurally relevant scaffold for future optimization studies aimed at improving the engagement of the hinge region and CNS penetrance for therapeutic development in GBM. Keywords— EGFR, Glioblastoma, Diosquinone, Erlotinib, Molecular Docking, ADME, in silico analysis. Result: The docking analysis revealed that diosquinone had the highest predicted binding affinity towards EGFR (−10.1 kcal/mol), which is higher than the reference inhibitor erlotinib (−7.2 kcal/mol). Interactions analysis showed a non-classical binding mode for diosquinone, vi

and ADME analysis showed good drug-likeness and limited blood–brain barrier permeability. Conclusion: The study revealed that diosquinone is a potential natural compound with high predicted receptor affinity and is a promising compound for targeting EGFR. The results, however, indicate that the binding affinity is not enough to predict the inhibitory potential and thus, the use of docking, interaction profiling and pharmacokinetic evaluation should be considered in computational drug discovery for glioblastoma.