Over the past decade, primary GBM is known for its aggressiveness and resistance to treatment, and because of its high recurrence, patients often die from the disease. Many novel therapies have emerged with the development of medical technology, such as surgery, chemotherapy, radiotherapy, molecular targeted agents, immunotherapy and nanotechnology [3, 28]. As we known, GBM alters major immunogenic signaling pathways and altering cellular immunity inside and outside the brain [8].Tumor-related immune responses and immunotherapy have been used in the treatment of malignant cancer [29].With the advent of targeted therapy and immunotherapy, the discovery of new biomarkers is urgent. In addition, the advancement of high-throughput sequencing technology made it possible for more and more function of non-coding RNA to be explored. Especially the study of some specifically expressed genes as a biomarker for treatment, diagnostic and prognostic in all kinds of tumors [30–32]. However, there are still few biomarkers used for diagnosis and prognosis in GBM, and further needed to explore.
Many researches have reported that lncRNAs not only regulate gene expression in some tumors, but also can be used as biomarkers for diagnosis and prognosis. For example, Qi Sun etc. found that LOXL1-AS1 upregulate the expression of USF1 as a ceRNA via sponging miR-708-5p [33]. LncRNA RPPH1 was significantly upregulated and was associated with poor prognosis in colorectal cancer [34]. LncRNA HOTTIP could mediate HOXA9 to enhance the Wnt/β-catenin pathway by combining with WDR5 in pancreatic cancer stem cells, HOTTIP/WDR5/HOXA9/Wnt axis was expected to be potential therapeutic targets for pancreatic cancer [35]. These results have given us a newer understanding of lncRNA, but they are not sufficient. In our study, we used the sequencing data of GBM and normal groups in the TCGA database to obtain DE-lncRNAs, DE-mRNAs, and DE-miRNAs through gene differential expression analysis. These DEGs serve as the basis for subsequent research.
Recently, with the increasing development of research on immune infiltration in the tumor microenvironment, immune-related genes have immediately become the focus of attention. Just as, 63 immune-related genes were founded associated with overall survival of melanoma and showed a powerful predictive ability in study of Rongzhi Huang etc. [36]. Wen Wang etc. revealed 9 immune-related lncRNAs have prognostic value for anaplastic gliomas [37]. Meng Zhou etc. verified that six-lncRNA signature could be an independent prognostic factor in GBM multiforme, with significantly different survival in high-risk and low-risk groups [38]. Therefore, it is necessary to explore the characteristics of immune-related molecules and evaluate the function of immune genes in GBM, which not only revealing the immune mechanism but also finding new therapeutic targets in GBM. In this study, immune-related mRNAs were obtained from ImmPort database and overlap with DE-mRNAs, 369 immune-related genes were selected. Furthermore, we took the intersection of 925 in the ImmLnc database and 32 prognostic-value DE-lncRNAs, and obtained 9 immune-related lncRNAs related to the prognosis. Similarly, we identified 34 DE-miRNAs, all the above data were used in the following analysis. An article was illustrated AC064875.2 as potential prognostic biomarker may regulate neutrophil infiltration in glioma [39]. MIR210HG was reported up-regulation upon hypoxia exposure in glioma cells, and was considered a resistance in GBM therapy [40, 41]. Transcriptome sequencing and qRT-PCR confirmed that lncRNA SMIM25 was high expression in cerebral cavernous malformations on the before study [42]. LINC01268 could influence the emotional regulation, and was involved in gene regulation of potential immune response [43]. The alias of GS1-358P8.4 is PDK3, the study reported that high expressions of PDK3 were poor prognostic factors of acute myeloid leukemia [44]. Besides, RP11-429B14.4, CTB-31O20.2, RP11-436K8.1, and RP11-268J15.5 have not been reported, and further needs to explore in the future.
The mechanism of lncRNAs as ceRNAs in GBM have been attracted many researchers [45, 46]. Therefore, we constructed an immune ceRNA network by predicting the targeting relationship between lncRNA-miRNA and miRNA-mRNA. In this ceRNA network, two lncRNAs were enriched. LINC01268 could regulate the expression of SLC11A1, NCK1, TNFSF14 and NOX4 by adsorbing hsa-miR-23b-5p and hsa-miR-139-3p, thereby affecting related immune responses. CTB-31O20.2 can regulate NR5A2 by adsorbing hsa-miR-139-5p to affect related immune responses. Otherwise, LINC01268 and CTB-31O20.2 could directly affect dendritic cell in the GBM. CTB-31O20.2 is a novel LncRNA that haven’t been reported. LINC01268 and CTB-31O20.2 are novel LncRNA that have never been reported in GBM. Besides, the univariate, multivariate, and nomogram analysis indicated that these two lncRNAs are prognostic markers for GBM. The results above revealed the prognostic role of LINC01268 and CTB-31O20.2 in GBM via immune ceRNA network, which are helpful to have a better understanding of the underlying mechanism in immune responses and GBM.
In conclusion, our study screen out two innovative molecules, LINC01268 and CTB-31O20.2, which play a crucial role in GBM and have the potential to predict the prognosis of GBM. Additionally, LINC01268 and CTB-31O20.2 reduces cancer cell proliferation and metastasis in GBM. Our results also provide new insights into the discovery of biomarkers for the prognosis of GBM. However, in vivo experiments need to be further verified.