Background
Recent studies indicate that cell mechanics are associated with malignancy through its impact on cell migration and adhesion. Gliomas are the most common primary malignant brain tumors. Low-grade gliomas (LGGs) include diffuse LGGs (WHO grade II) and intermediate-grade gliomas (WHO grade III). Few studies have focused on membrane tension in LGGs. Herein, we assessed the prognostic value of plasma membrane tension-related genes (MTRGs) in LGGs.
Conclusion
Our plasma membrane tension-related gene prognostic signature is a prospective tool that can differentiate between prognosis, gene mutation landscape, immune microenvironment, immune infiltration and immunotherapeutic efficacy in LGG.
Methods
We selected plasma MTRGs identified in previous studies for analysis. Based on LGG RNA sequencing (RNA-seq) data in The Cancer Genome Atlas, a prognostic signature containing four genes was constructed via log-rank testing, LASSO regression and stepwise multivariate Cox regression and was validated with other datasets. Additionally, functional annotation, pathway enrichment and immune and molecular characteristics of the prognostic model defined subgroups were analyzed. Thereafter, a predictive nomogram that integrated baseline characteristics was constructed to determine the 3, 5, and 10-year overall survival (OS) of patients with LGG. Differentially expressed genes were confirmed via quantitative reverse-transcription polymerase chain reaction (qRT-PCR) and immunohistochemistry (IHC).
Results
Our MTRG prognostic signature was based on ARFIP2, PICK1, SH3GL2, and SRGAP3 expression levels. The high-risk group was more positively associated with apoptosis and cell adhesion pathways and exhibited a low IDH1 mutation rate, high TP53 mutation rate and a low 1p19q co-deletion rate. The high-risk group also exhibited incremental infiltration of immune cells, more forceful immune activities and high expression of immune checkpoints as well as benefited less from immune therapy compared with the low-risk group. Our prognostic model had better forecasting ability than other scoring systems. We found that the nomogram was a better tool for predicting outcomes for patients with LGG. Finally, qRT-PCR confirmed that SH3GL2 and SRGAP3 expression levels in glioma tissues were significantly lower than those in normal brain tissues. The results of IHC analysis confirmed that SH3GL2 protein expression was higher in patients with longer survival.