Glioblastoma (GBM) is the most common primary brain cancer, and the prognosis of traditional treatment methods is not favorable. Therefore, we aimed to identify new molecular targets and construct a risk model for GBM diagnosis and prognosis prediction on the basis of neutrophil extracellular trap (NET)-related genes. Gene expression and clinical data were obtained from the TCGA. We constructed a risk model based on four NET-related genes, namely, MAPK1, P2RX1, PARVB and STAT3, using univariate, LASSO and multivariate Cox analyses. The receiver operating characteristic curve analysis with the pROC package revealed that the area under the curve (AUC) values for 1-, 2- and 3-year survival were 0.73, 0.81 and 0.83, respectively. In addition, correlation analysis revealed that MAPK1 expression was negatively correlated with IL15, whereas PARVB, P2RX1 and STAT3 expression levels were positively correlated with IL15, PADI4, CXCL1 and CSF1. In vitro, overexpression of PARVB and P2RX1 STAT3 and the knockdown of MAPK1 promoted the malignant behaviors of GBM cells and activated EMT pathways. Drug sensitivity analysis revealed that STAT3 expression was positively associated with sensitivity to most drugs, whereas MAPK1, PARVB and P2RX1 expression levels were negatively correlated with most drug sensitivity. In conclusion, we identified four NET-related hub genes related to the prognosis of GBM and constructed a risk model with good performance.
Glioblastomas (GBM) is a kind of malignant brain tumor with poor prognosis. Identifying new biomarkers is promising for the treatment of GBM. The mRNA-seq and clinical data were obtained from The Cancer Genome Atlas and the Chinese Glioma Genome Atlas databases. The differentially expressed genes were identified using limma R package. The prognosis-related genes were screened out and a risk model was constructed using univariate, least absolute shrinkage and selection operator, and multivariate Cox analysis. Receiver operating characteristic curve was used to assess the efficiency of model. Kaplan–Meier survival curve was applied for the survival analysis. Mutation analysis was conducted using maftools package. The effect of immunotherapy was analyzed according to TIDE score, and the drug sensitivity analysis was performed. The Gene Ontology, Kyoto Encyclopedia of Genes and Genomes, and Gene Set Enrichment Analysis enrichment analyses were performed for the functional analysis. The regulatory network was constructed by STRING and Cytoscape software. RT-qPCR was performed to validate the expression of 3 hub genes in vitro. A risk model was constructed based on 3 ion channels related genes (gap junction protein beta 2 [GJB2], potassium voltage-gated channel subfamily h member 6 [KCNH6], and potassium calcium-activated channel subfamily n member 4 [KCNN4]). The risk score and hub genes were positively correlated with the calcium signaling pathway. Patients were divided into 2 groups based on the risk score calculated by 3 signatures. The infiltration levels of T cell, B lineage, monocytic lineage, and neutrophils were increased in high risk group, while TIDE score was decreased. IC50 of potential drugs for GBM treatment was elevated in the high risk group. Furthermore, GJB2, KCNH6, and KCNN4 were oncogenic, and GJB2 and KCNN4 were upregulated, while KCNH6 was downregulated in high risk group and GBM cells. The regulatory network showed that KCNH6 was targeted by more miRNA and transcription factors and KCNN4 interacted with more drugs. We constructed a three-signature risk model, which could effectively predict the prognosis of GBM development. Besides, KCNH6 and KCNN4 were respectively considered as the targets of molecular targeted treatment and chemotherapy.