Cellular senescence escape and antiviral response discriminate glioblastoma from lower-grade gliomas

Abstract Background Cellular senescence is a negative prognostic indicator in glioblastoma (GB). However, the specific cell types exhibiting senescence and the molecular mechanisms by which senescent cells contribute to GB pathogenesis remain unknown. Methods We performed multi-omics integration of publicly available GB patient-derived datasets to identify senescent cells and their functional impact on GB pathogenesis. We created a transcriptomic definition of senescent cells to verify their presence in GB datasets. Next, we analyzed transcriptomic profiles of GB and low-grade gliomas to reveal key features of GB aggressiveness. The GB patients’ phosphoproteome was analyzed with a focus on key regulators of senescence. To assess chemotherapy-evoked secondary senescence, we performed a proteomic analysis of temozolomide-induced senescence in human GB spheroids. Results We identified GB-associated senescent cells in clusters radial glia, endothelial cells, and immature astrocytes localized primarily in the hypoxic zones. Notably, we identified senescence-escape features and tumor antiviral responses as processes distinguishing GB from low-grade gliomas. In GB samples, we detected inhibitory phosphorylation of p21 and p27 proteins and active PI3K signaling, which can lead to senescence escape and belong to the typical manipulation arsenal of herpesviruses. Our proteomic analysis of temozolomide-induced senescence GB cells reveals that primary (pre-radiochemotherapy) and secondary senescence in GB share similar phenotypic features. Pathways associated with GB aggressiveness are upregulated after therapy, which can promote more aggressive behavior of recurrent tumors. Conclusions Our data indicate that senescence and viral reactivation may fuel GB progression, including recurrence, suggesting that senolytics and antiviral drugs are potential therapeutic avenues.