Expression Profiling of Glioblastoma Cell Lines Reveals Novel Extracellular Matrix-Receptor Genes Correlated With the Responsiveness of Glioma Patients to Ionizing Radiation

Abstract

Glioblastoma (GBM) is the most lethal and frequent type of brain tumor, leading patients to death in approximately 14 months after diagnosis. GBM treatment consists in surgical removal followed by radio and chemotherapy. However, tumors commonly relapse and the treatment promotes only a slight increase in patient survival. Thus, uncovering the cellular mechanisms involved in GBM resistance is of utmost interest, and the use of cell lines has been shown to be an extremely important tool. In this work, the exploration of RNAseq data from different GBM cell lines revealed different expression signatures, distinctly correlated with the behavior of GBM cell lines regarding proliferation indexes and radio-resistance. U87MG and U138MG cells, which presented expressively reduced proliferation and increased radio-resistance, showed a particular expression signature encompassing enrichment in many extracellular matrix (ECM) and receptor genes. Contrasting, U251MG and T98G cells, that presented higher proliferation and sensibility to radiation, exhibited distinct signatures revealing consistent enrichments for DNA repair processes and although several genes from the ECM-receptor pathway showed up-regulation, enrichments for this pathway were not detected. The ECM-receptor is a master regulatory pathway that is known to impact several cellular processes including: survival, proliferation, migration, invasion, and DNA damage signaling and repair, corroborating the associations we found. Furthermore, searches to The Cancer Genome Atlas (TCGA) repository revealed prognostic correlations with glioma patients for the majority of genes highlighted in the signatures and led to the identification of 31 ECM-receptor genes individually correlated with radiation responsiveness. Interestingly, we observed an association between the number of upregulated genes and survivability greater than 5 years after diagnosis, where almost all the patients that presented 21 or more upregulated genes were deceased before 5 years. Altogether our findings suggest the clinical relevance of ECM-receptor genes signature found here for radiotherapy decision and as biomarkers of glioma prognosis.