Identification of an EMT-related lncRNA signature in glioma
Abstract Background Epithelial-mesenchymal transition (EMT) has been implicated in the invasion and progression of gliomas, the present study investigated EMT-related long noncoding (lnc)RNAs in gliomas.Methods Candidate lncRNAs screened from a glioma dataset in The Cancer Genome Atlas were used to construct EMT-related lncRNA coexpression networks in order to identify EMT-related lncRNAs; these were validated using the Chinese Glioma Genome Atlas (CGGA). Gene set enrichment analysis and principal component analysis (PCA) were used to annotate lncRNA functions. We established an EMT-related lncRNA signature composed of 9 lncRNAs (HAR1A, LINC00641, LINC00900, MIR210HG, MIR22HG, PVT1, SLC25A21-AS1, SNAI3-AS1, and SNHG18) that could distinguish between low- and high-risk glioma patients. The functions of the identified lncRNAs were evaluated by constructing a competing endogenous RNA network.Results The low-risk group had longer overall survival (OS) than the high-risk group (P<0.0001). High-risk patients also had no deletion on chromosome arms 1p and/or 19q, expressed wild-type isocitrate dehydrogenase, and had higher World Health Organization (WHO) tumor grade. The signature was an independent factor significantly associated with OS (P=0.041, hazard ratio=1.806). These findings were validated in the CGGA dataset. PCA also revealed differences in EMT status between low- and high-risk patients. Competing endogenous RNA network showed that complex lncRNA–microRNA–mRNA interactions potentially contributing to EMT progression in glioma. Conclusion Our results demonstrate that the EMT-related 9-lncRNA signature has prognostic value for glioma patients.