Bioinformatics Analysis and Experimental Validation for Exploring Key Molecular Markers for Glioblastoma.

Abstract

Glioblastoma (GBM) is the most common primary intracranial malignancy with a very low survival rate. Exploring key molecular markers for GBM can help with early diagnosis, prognostic prediction, and recurrence monitoring. This study aims to explore novel biomarkers for GBM via bioinformatics analysis and experimental verification. Dataset GSE103229 was obtained from the GEO database to search differentially expressed lncRNA (DELs), mRNAs (DEMs), and miRNAs (DEMis). Hub genes were selected to establish competing endogenous RNA (ceRNA) networks. The GEPIA database was employed for the survival analysis and expression detection of hub genes. Hub gene expression in GBM tissue samples and cell lines was validated using RT-qPCR. Western blotting was employed for protein expression evaluation. SYT1 overexpression vector was transfected in GBM cells. CCK-8 assay and flow cytometry were performed to detect the malignant phenotypes of GBM cells. There were 901 upregulated and 1086 downregulated DEMs identified, which were prominently enriched in various malignancy-related functions and pathways. Twenty-two hub genes were selected from PPI networks. Survival analysis and experimental validation revealed that four hub genes were tightly associated with GBM prognosis and progression, including SYT1, GRIN2A, KCNA1, and SYNPR. The four genes were significantly downregulated in GBM tissues and cell lines. Overexpressing SYT1 alleviated the proliferation and promoted the apoptosis of GBM cells in vitro. We identify four genes that may be potential molecular markers of GBM, which may provide new ideas for improving early diagnosis and prediction of the disease.