Deciphering ECM Remodelling in Retinoblastoma through Integrative Transcriptomic and Computational Approaches: Identification of Key Regulators and Natural Inhibitors
Background: Retinoblastoma (RB) is the most common intraocular malignancy in children, primarily associated with inactivation of the RB1 gene. Besides genetic alterations, Extracellular Matrix (ECM) remodelling and changes in the tumour microenvironment play critical roles in RB invasion, angiogenesis, and drug resistance. Aim: This study aimed to identify key ECM remodelling regulators in RB and evaluate potential natural inhibitors using integrated in silico approaches. Methods: Transcriptomic profiling from the GEO dataset GSE111168 was analysed to identify differentially expressed ECM-related genes. Machine learning and systems biology approaches were applied to prioritise significant targets. Molecular docking and Molecular Dynamics (MD) simulations were performed to assess the interaction and stability of natural compounds with the top-ranked target, PLOD1. Pharmacokinetic, ADMET, and oral bioavailability analyses were further conducted to evaluate drug-likeness. Results: PLOD1 emerged as a major ECM remodelling gene overexpressed in RB. Among the screened phytochemicals, the natural flavonoid luteolin demonstrated the highest docking affinity (score: 119.15) compared to carboplatin (65.96). MD simulations confirmed the stable binding of luteolin in the catalytic pocket of PLOD1, supported by favourable pharmacokinetic and safety profiles. Conclusion: The study identifies PLOD1 as a candidate promising therapeutic target in retinoblastoma and luteolin as a potential inhibitor. These findings provide a computational framework for future experimental validation and the development of phytochemical-based therapies in pediatric ocular oncology. Major Findings: PLOD1 was identified as a key ECM-associated candidate biomarker in retinoblastoma through integrative transcriptomic and systems-level analysis. Among screened phytochemicals, luteolin demonstrated the strongest predicted binding affinity and stable interaction with PLOD1 in docking and mo