IGF2 associated PI3K signaling shapes temozolomide sensitivity in glioblastoma

Resistance to temozolomide (TMZ) remains a major obstacle in glioblastoma (GBM) therapy and cannot be fully explained by canonical DNA repair mechanisms alone. Here, by integrating large scale transcriptomic datasets with machine learning based prioritization and functional validation, we identify insulin-like growth factor 2 (IGF2) as a key molecular hub linking tumor metabolic reprogramming to chemoresistance in GBM. Differential expression analysis of TCGA-GBM and GTEx normal brain tissues, followed by intersection with chemoresistance- and metabolism-related gene sets, yielded candidate cross-hub genes that were further refined using LASSO regression and XGBoost modeling. Among these, IGF2 emerged as the gene consistently associated with poor overall survival across both TCGA and independent CGGA cohorts, which higher expression was validated by human GBM tissues. Functionally, IGF2 silencing suppressed GBM cell viability, promoted apoptosis, and significantly enhanced TMZ sensitivity. Transcriptomic profiling revealed that IGF2 knockdown broadly downregulated resistance associated gene programs and disrupted pathways involved in cellular metabolism, DNA damage response, and cell cycle regulation. Notably, both patient stratification analyses and IGF2 knockdown induced transcriptional changes converged on the PI3K signaling axis. Pharmacological inhibition of PI3K using Paxalisib further potentiated TMZ efficacy, with combination treatment producing the strongest anti-proliferative effects. Together, these findings indicate a link between IGF2-associated PI3K signaling, metabolic reprogramming, and TMZ resistance in GBM, providing a rationale for further exploration of this pathway as a therapeutic target