m <sup>6</sup>A Signature and Tumour Immune Microenvironment for Predicting Prognostic Value in Gliomas

Background: Gliomas comprise about 80 per cent of all malignant brain tumours. Current gliomas staging alone cannot adequately predict prognosis between Glioblastoma multiforme (GBM) and low-grade gliomas (LGG). This study aims to explore effective diagnostic and prognostic biomarkers for gliomas. Methods: We performed consensus clustering to m6A RNA methylation regulators in TCGA-GBM/LGG datasets. The tumour-infiltrating immune cells of The Cancer Genome Atlas (TCGA) gliomas cohort was analyzed by single-sample gene set enrichment analysis (ssGSEA). Four hub gene were identified via weighted gene co-expression network analysis (WGCNA) and least absolute shrinkage, and selection operator (LASSO). The diagnostic and prognostic value were validated by risk score analysis, Kaplan-Meier and ROC. Finding: We identified two glioma subgroups (cluster 1/2) by applying consensus clustering to m6A RNA methylation regulators in 665 gliomas (TCGA-GBM and TCGA-LGG). The cluster 1 subgroup was associated with a poorer prognosis and a higher WHO classification than the cluster 2 subgroup. We performed chi-square test between m6A cluster subgroup and immune infiltration subgroup, the results indicated that the characteristics of the tumor-associated immune infiltration are significantly enriched in the cluster 1 subgroup. These findings suggest that m6A RNA methylation regulators are related to immune infiltration. In order to explore potential hub genes, we performed WGCNA between high/low immune infiltration and m6A cluster 1/2 groups. LASSO was utilized for the hub genes, and four immune and m6A related genes TAGLN2, PDPN, TIMP1, EMP3 were identified as biomarkers for glioma prognosis. Moreover, the 4-gene signature achieved AUC values of 0.80, 0.79, 0.67 and 0.72 in TCGA, Rembrandt, GSE16011 and Chinese Glioma Genome Atlas (CGGA), respectively. Interpretation: These findings provide new insight into gliomas prognosis and identify four genes (TAGLN2, PDPN, TIMP1 and EMP3) as candidate biomarkers for diagnosis and prognosis of glioma. Funding Statement: This work was supported by National Natural Science Foundation of China (no.81701359), Natural science foundation of Shanghai (18ZR1430400), and the Cross Research Fund of Medicine and Engineering of Shanghai Jiaotong University (No. YG2016QN32, YG2019QNA67). Declaration of Interests: The authors have declared that no conflict of interest exists. Ethics Approval Statement: Not required.