Molecular and clinical characterization of TMEM71 expression at the transcriptional level in glioma

Abstract Background Glioma is the most common and aggressive type of primary brain tumor in adults. Although radiotherapy and chemotherapy are used in the treatment of glioma, survival remains unsatisfactory. Chemoresistance is one of the primary reasons for the poor prognosis of glioma. Several studies have demonstrated that glioma stem cells (GSC) may be one of the reasons for chemoresistance. In this article, we attempt to search for a new biomarker related to GSC and chemoresistance in glioma. Methods We used three datasets (GSE23806, COSMIC, and CGGA) to search for the genes related to GSC, temozolomide (TMZ) resistance, and overall survival. The selected gene was investigated with respect to the relationship between mRNA levels and clinical characteristics in the CGGA and TCGA dataset. Gene ontology (GO) analysis was used for bioinformatics analysis. Kaplan‐Meier survival analysis and Cox regression analysis were used for survival analysis. Results The transmembrane protein 71 (TMEM71) gene was selected for further research. TMEM71 was highly expressed in GSCs and TMZ‐resistant cells. The TMEM71 mRNA levels increased with increasing grades of glioma. In IDH‐wild‐type and MGMT‐unmethylated samples, TMEM71 was overexpressed. The TMEM71 transcript levels were also increased significantly in mesenchymal subtype gliomas. GO analysis demonstrated that TMEM71 was related to the immune and inflammatory responses, cell proliferation, cell migration, chemotaxis, and the response to drugs. Specifically, PD‐1, PD‐L1, TIM‐3, and B7‐H3 were tightly associated with TMEM71 expression. This result indicates that TMEM71 may play an important role in the immune response. More importantly, high expression of TMEM71 was correlated with short survival time in both glioma and glioblastoma patients. Conclusion In summary, TMEM71 expression was increased in GBM and associated with immune response. Our study suggests that TMEM71 may function as an oncogene and serve as a new effective therapeutic target for the treatment of glioma.


| INTRODUC TI ON
Glioma is the most common and aggressive type of primary brain tumor in adults. 1,2 Current therapies including surgery, chemotherapy, and radiotherapy are partly effective for glioma. Therefore, treatment for glioma remains a serious clinical and scientific problem. The median survival in patients with glioma remains unsatisfactory, especially in glioblastoma multiforme (GBM), which is the most malignant type. 1,3 For further understanding the molecular mechanisms behind the disease manifestation and progression, a new prognostic biomarker and effective therapeutic targets for GBM still need to be identified.
Recently, the most common drug used in chemotherapy for glioma is temozolomide (TMZ), which is a standard treatment. 1 Although TMZ has improved both the overall survival (OS) and the progression-free survival (PFS), patients have often suffered from chemoresistance after using TMZ. The mechanism of the chemoresistance is still unclear. Several studies focus on the glioma stem cells (GSC), which may be related to chemoresistance. 4 Cancer stem cells (CSC) are a subpopulation of tumor tissue and are postulated as a mediator of chemoresistance. After chemotherapy, surviving CSC may later differentiate into rapidly proliferating tumor cells and lead to tumor recurrence. 5 In recent years, multiple genes have been discovered as signatures of CSC.
CD133 is one of the most studied CSC biomarkers. 6 CD133positive glioma cells can grow spheres in serum-free medium, whereas CD133-negative cells cannot grow spheres. 7,8 Several studies have demonstrated that knocking down CD133 could weaken tumorigenicity. 8 SOX2 is another signature of CSC. SOX2 is highly expressed in CSC, and it can promote the self-renewal of CSC. 9 Moreover, other CSC-relevant molecular markers, such as CD15, CD36, A2B5, L1CAM, and CD44, have been reported by researchers. [10][11][12][13] Recently, there has been insufficient research on utilizing GSC-related genes to predict prognosis and chemoresistance. For this reason, we focused on these genes to search for a new biomarker in glioma. In the present study, we discovered that transmembrane protein 71 (TMEM71) mRNA expression was related to TMZ chemoresistance, and high expression of this protein was associated with GSC. Further research revealed that TMEM71 expression was significantly higher in GBM than in lower-grade glioma tissues. The TMEM71 expression level was also positively related to the histological grades of gliomas. The patients with high TMEM71 expression level tend to enrich in IDH wild-type and MGMT unmethylated samples. Bioinformatics analysis indicates that overexpression of TMEM71 was involved in multiple crucial biological processes including immune and inflammatory response, cell proliferation, cell migration, response to drug, and so on. Collectively, these results suggest that TMEM71 is a potential oncogene in GBM and that it may serve as a therapeutic target for this disease.

| Samples
In this article, eight datasets were used for the analysis.

| Bioinformatics analysis
Gene ontology (GO) was used to analyze biological function. 14 The correlation between TMEM71 mRNA expression and other genes was analyzed by the Pearson correlation analysis using the R programming language. The positive correlative gene (r > 0.4, P < 0.05) and the negative correlative gene (r < −0.4, P < 0.05) were chosen for analysis by DAVID (http://david.abcc.ncifc rf.gov/home.jsp) to detect the biological processes that were correlated with TMEM71 expression. The results were showed by heatmap using R programming language.

| Statistical analysis
In this study, SPSS 16.0 (Armonk, NY, USA), the R programming for the survival analysis. The pheatmap, pROC, circcos, and corrgram 15,16 were generated by several R packages. P value < 0.05 was considered statistically significant.

| Gene selection
Student's t test was conducted in the GSE23806 dataset to identify genes with different expressions between conventional glioma cell lines (n = 36) and glioblastoma stem-like cell lines (n = 27) (P < 0.05). The detailed information of the cell lines was given by the dataset to choose which type the cell line was. In the glioblastoma stem-like cell line, 6752 genes exhibited high expression, and they were selected for further analysis ( Figure 1A). In COSMIC dataset, the half maximal inhibitory concentration (IC50) of TMZ was given by the dataset and the data of glioma cell lines was chosen to do the analysis. According to the median IC50 of TMZ to the glioma cell lines, the cell lines were divided into two groups.
Student's t test was also used in the COSMIC dataset to identify genes with different expression between different sensitivities to TMZ according to the IC50. In the high-IC50 group, 447 genes were highly expressed and were selected for further analysis F I G U R E 1 (A) The different genes between GSCs and conventional GBM cell lines. (B) The different genes between TMZ-sensitive cell lines and TMZ-resistant cell lines F I G U R E 2 (A-B) TMEM71 expression in different grades according to IDH status from the CGGA and TCGA databases. (C-D) TMEM71 expression according to MGMT status from the CGGA and TCGA databases. ***P < 0.001, **P < 0.01, *P < 0.05 (P < 0.05, Figure 1B). Next, we used univariate Cox regression in the CGGA RNA sequencing dataset to analyze the genes related to survival (P < 0.05) in GBM samples (n = 144). The genes significantly related to survival were retained (n = 1302). Only three genes were left in all three datasets: TMEM71, CA12, and FUT7 (Supplementary Figure S1A). We focus on TMEM71, which has been hardly researched in glioma, even in tumors (Supplementary Figure S1B, C).

| Heterogeneity of TMEM71 expression in glioma
By examining the RNA sequencing data of glioma from the CGGA database, we found that the expression level of TMEM71 was posi-

| TMEM71 expression level shows a subtype preference
To further explore the molecular relevance between TMEM71 and glioma, we analyzed TMEM71 expression in different molecular subtypes. The results showed that the TMEM71 expression levels in the four molecular subtypes were considerably different in the CGGA and TCGA datasets. Compared with the other three subtypes, TMEM71 was significantly upregulated in the mesenchymal subtype, which is typically associated with poor outcomes ( Figure 3A, B and Supplementary Figure S2C, D). Moreover, ROC curves were generated for TMEM71 expression and the mesenchymal subtype. As shown in Figure 3C and D, the area under the curve (AUC) was up to 86.4% and 86.6% in the CGGA and TCGA sequencing datasets, respectively. These results enlightened us that TMEM71 was a potential biomarker for mesenchymal subtype glioma.
F I G U R E 3 (A-B) TMEM71 expression in molecular subtypes from the CGGA and TCGA datasets. (C-D) TMEM71 expression was used to predict mesenchymal subtype or not in both CGGA and TCGA subtype. The AUC was more than 80% in both datasets. ***P < 0.001, **P < 0.01, *P < 0.05

| TMEM71-related GSC markers
In our research, we found that TMEM71 was upregulated in GSCs. To conduct further research on TMEM71-related GSC markers, we chose several genes that were related to GSC according to previous reports. 17 Pearson correlation analysis was used to search for the GSC-related genes that correlated with TMEM71 expression. In both CGGA and TCGA datasets, it was found that IL6, STAT3, CD44, and FUT4 were positively related to TMEM71 expression ( Figure 4A, B). The similar results were also obtained in REMBRANDT and GSE16011 datasets ( Figure 4C, D

| TMEM71 related biological process
Based on the above findings, we inferred that TMEM71 may exhibit essential biologic functions in glioma. To further investigate the biological process tightly associated with TMEM71 expression, Pearson correlation analysis was performed to  Figure 5A and B, genes that positively correlated with TMEM71 expression were highly enriched in the immune and inflammatory response, cell proliferation and migration, T-cell activation, chemotaxis, and response to drug in GO terms.  associated with TMEM71 expression (Figure 6A, B). This result indicates that TMEM71 may play an important role in immune response.

| TMEM71-related inflammatory activities
To conduct further research on TMEM71-related inflammatory activities, we chose seven clusters that represented several types of inflammation and immune responses. These seven clusters were calculated by Gene Sets Variation Analysis (GSVA) to generate seven metagenes. 23 In both CGGA and TCGA datasets, it was revealed that TMEM71 expression was positively related to HCK, LCK, MHC-II, and STAT1 while negatively related to IgG and interferon ( Figure 7A, B). These results suggest that TMEM71 expression was upregulated with the activation of macrophages and T-cell signaling transduction. with higher TMEM71 expression in glioma is shorter than patients with lower TMEM71 expression ( Figure 8A, B). Considering the significant biological heterogeneity between GBM and lower-grade glioma, we additionally analyzed the prognostic value of TMEM71 in GBM patients from the two datasets. We found that patients with high TMEM71 expression exhibited a significantly shorter survival than those with lower TMEM71 expression ( Figure 8C, D). Similar results were observed in the TCGA microarray and REMBRANDT datasets (Supplementary Figure S4A, B). The ROC curve analysis was performed to validate the prognostic value of TMEM71.

| TMEM71 was determined to be an independent prognostic factor for GBM patients
Compared with age at diagnosis and tumor grade, the AUCs for TMEM71 expression level in the prediction of 3 years of survival in glioma were close to "grade" and higher than "age" (Supplementary Figure S4C, D). These results revealed that TMEM71 is a negative prognostic factor in glioma and GBM patients.
Univariate Cox regression analysis was performed to further estimate the prognostic value of TMEM71 in GBM patients in the CGGA dataset. The results showed that high TMEM71 expression, Karnofsky Performance Status (KPS) score, MGMT methylation state, radiotherapy, and chemotherapy after resection were shown to be significantly associated with OS of GBM patients (Table 1).
Thus, we performed a multivariate Cox regression analysis, and the result revealed that the TMEM71 expression is an independent prognostic biomarker for GBM patients (Table 1). Furthermore, we obtained similar results for the TCGA microarray and RNA sequencing datasets (Tables S1 and S2). Those results implied that TMEM71 might be a novel independent prognostic biomarker for GBM patients. Moreover, the expression of TMEM71 was positively related to chemotaxis and the response to drugs. This might be the reason that TMEM71 overexpression leads to chemoresistance to TMZ therapy.

| D ISCUSS I ON
Nevertheless, we know little regarding the function of TMEM71 in GBM. It is necessary to conduct further research.

| CON CLUS IONS
In conclusion, the results described above revealed that TMEM71 acts as an oncogene in GBM. It may be related to GSC, TMZ resistance, and immune response. The upregulation of TMEM71 may indicate poor sensitivity to TMZ therapy. The study also points to TMEM71 as a potential therapeutic target for the clinical management of GBM.

E TH I C A L A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
This study was approved by the Beijing Tiantan Hospital institutional review board (IRB), and informed consent was obtained from all individual participants who were included in this study.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.