A three ion channel genes-based signature predicts prognosis of primary glioblastoma patients and reveals a chemotherapy sensitive subtype

Abstract

Increasing evidence suggests that ion channels not only regulate electric signaling in excitable cells but also play important roles in the development of brain tumor. However, the roles of ion channels in glioma remain controversial. In the present study, we systematically analyzed the expression patterns of ion channel genes in a cohort of Chinese patients with glioma using RNAseq expression profiling. First, a molecular signature comprising three ion channel genes (KCNN4, KCNB1 and KCNJ10) was identified using Univariate Cox regression and two-tailed student's t test conducted in overall survival (OS) and gene expression. We assigned a risk score based on three ion channel genes to each primary Glioblastoma multiforme (pGBM) patient. We demonstrated that pGBM patients who had a high risk of unfavorable outcome were sensitive to chemotherapy. Next, we screened the three ion genes-based signature in different molecular glioma subtypes. The signature showed a Mesenchymal subtype and wild-type IDH1 preference. Gene ontology (GO) analysis for the functional annotation of the signature showed that patients with high-risk scores tended to exhibit the increased expression of proteins associated with apoptosis, immune response, cell adhesion and motion and vasculature development. Gene Set Enrichment Analysis (GSEA) results showed that pathways associated with negative regulation of programmed cell death, cell proliferation and locomotory behavior were highly expressed in the high-risk group. These results suggest that ion channel gene expression could improve the subtype classification in gliomas at the molecular level. The findings in the present study have been validated in two independent cohorts.

Keywords: alpha-fetoprotein; antigen epitope; functional peptide; heat shock protein 70; immunity.

Figure 1. Prognostic values of three ion channel genes-based signature for patients in training and validation datasets

Patients in low risk group showed a better prognosis than those in high risk group according to the signature risk score in the CGGA dataset (A–B), the TCGA data (C), and the Rembrandt data (D). L, low risk group; H, high risk group; pGBM, primary GBM.

Figure 2. Distributions of risk score of pGBM and OS of their patients in the three datasets

(A) Signature risk score distribution. (B) Patient overall survival duration. (C) Expression of the three ion channel genes-based signature along the risk score. Red indicates high expression and green indicate low expression.

Figure 3. The signature predicted the efficacy of radiotherapy with or without chemotherapy in pGBM patients

(A, D) pGBM patients in CGGA and TCGA treated with RT + CT showed a better prognosis than those with RT alone. (B, E) Benefit of RT + CT was observed in the high risk group with significantly improved OS (p < 0.05). (C, F) The addition of CT to RT did not improve OS of patients in the low risk group (p > 0.05). RT, radiotherapy; CT, chemotherapy; LR, low risk group; HR, high risk group.

Figure 4. Distribution of molecular and clinical pathological features for pGBM patients in three datasets

(A) The high risk score tumors displayed Mesenchymal subtype and wild-type IDH1 preference. (B) The differentially expressed genes were shown arranged from the low to high risk score. Pink represents the high expression of genes in the high risk group; blue represents the low expression of the genes in high risk group.

Figure 5. Functional annotation of the high risk group

(A) GO analysis revealed the significant association of the genes with increased expression in high risk group with four main pathways. Column height: gene counts; point height: enrichment p value. (B) Three representative plots of GSEA from enriched pathways in high risk group, analyzed by gene set enrichment analysis of CGGA and TCGA RNAseq data.