Abstract
Here, we utilized a database mining approach to unfold the prognostic and therapeutic potentials of Transmembrane EmP24 Trafficking Protein 4 (TMED4) and 9 (TMED) coding gene expressions in glioma. Both the genes were found to be overexpressed at the mRNA and protein level in low grade glioma (LGG) and glioblastoma multiforme (GBM) tissues including different glioma cell lines. Significant increase in the expression level of these genes with advancing glioma patients’ age, glioma grades and histological subtypes was observed. Differential and distinct promoter and coding sequence methylation pattern of TMED4 and TMED9 was observed in LGG and GBM tissues that may aid in methylation-sensitive diagnosis of glioma patients. The presence of multiple heterozygous genetic alterations (frequency: 0.4-1.1%) in those genes unveiled their potentials in high-throughput screening of glioma patients. The overexpression of TMED4 and TMED9 genes was associated with poor overall survival (OS) of LGG and GBM patients (HR:>1.6). Association of the expression levels of these genes with different immune cell infiltration levels i.e., B cell and T cell and modulators like CD274 and IL10RB was observed providing assurance in TMED-based diagnostic measure and therapeutic intervention discovery. Furthermore, functional enrichment analysis of the neighbor genes of TMED4 and TMED9 revealed that they are involved in metal ion binding, focal adhesion of cells and protein processing, and the deregulation of these activities are associated with gliomagenesis. Altogether, this study suggests that TMED4 and TMED9 are potential prognostic and therapeutic targets for glioma. However, further laboratory research is warranted.
1. Introduction
Brain tumor, the commonest solid tumor in children, refers to a diverse group of neoplasms emerging from distinct anatomic sites of the intracranial tissues and the meninges with varying degrees of malignancy ranging from benign to aggressive [1]. Malignant and non-malignant brain tumors are rare comprising only 2% of all cancers but are amongst the most fatal cancers accounting for substantial mortality and morbidity across the world [2]. Glioma, a broad category of diffusely infiltrative brain and spinal cord tumor arising from the glial cells, is the most common primary malignant brain tumor in adults with a very poor prognosis. About 33% of all brain tumors are gliomas, out of which approximately 70%-80% are malignant [3,4]. Global incidence of glioma varies greatly, displaying a higher incidence in men than women, and a higher occurrence across the Western population compared to the Asian or African population [5].
In accordance with the presumed cell of origin and histological features, gliomas can be broadly classified into astrocytomas including glioblastoma, oligodendrogliomas, ependymomas, and mixed gliomas [6-8]. Amongst these, glioblastoma multiforme (GBM), also commonly referred to as glioblastoma, accounts for approximately 60% to 70% of malignant gliomas and is the most frequently occurring primary astrocytomas [9]. World Health Organization (WHO) has categorized gliomas based on their histopathological characteristics such as cytological atypia, anaplasia, mitotic activity, microvascular proliferation, and necrosis [3]. According to this standard for defining the nomenclature, diagnosis, and malignancy of glioma, GBM is the most aggressive, invasive glioma subtype and by definition a high-grade tumor (HGG, WHO grade IV) [3,10]. Although the global incidence of GBM is less than 10 per 100,000 people, it’s extremely poor prognosis with a median survival time period of 14-15 months approximate from diagnosis makes it a serious health threat [11]. Despite several international therapeutics approaches, GBM treatment is still the most challenging task in clinical oncology with very limited success in the range of different treatments investigated [12]. Low-grade glioma (LGG, WHO grade I-IV), brain tumors arising from two different cell types namely astrocytes and oligodendrocytes, are another diverse group of primary brain tumors that often arise in young otherwise healthy people and is the slowest growing glioma types in adults [3]. LGGs account for 6.4% of all the primary CNS tumors, and although it rarely causes major neurological deficits besides seizures, recent data suggests that LGG can grow at a continuous rate reaching 4 to 5 mm per year [13,14]. Continuous growth of this untreated symptomatic LGG can eventually undergo malignant transformation triggering a more complicated disease course, reduced quality of life, and worse prognosis [14], raising serious complications and health concerns. Although LGG and GBM remains the most devastating form of brain cancer, the proper management and treatment of glioma patients is often interrupted by a lack of proper understanding of glioma pathogenesis and the protective structure of the CNS, as well as the unavailability of efficient diagnostic and therapeutic measures [15,16]. Therefore, there is an ever-increasing demand to discover an effective molecular target for appropriate and accurate stratification and therapeutic interventions in glioma patients.
The Transmembrane EmP24 Domain-Containing Proteins (TMED), also known as p24 proteins, are members of a family of sorting receptors abundantly present in the cellular subcompartments of the early secretory pathway including the endoplasmic reticulum (ER), Golgi body, and the intermediate compartment of all representatives of the domain Eukarya [15-17]. This family of proteins has an average molecular mass of 24 kDa and are central regulators in cell signaling, the secretion of biomolecules, and intracellular protein transport [17]. Due to their active influence in determining the composition, structure, and function of the ER and the Golgi, as well as their role in maintaining cellular traffic, they have been associated with a variety of human diseases, including carcinomas. For example, TMED3 promotes hepatocellular carcinoma cell proliferation in by regulating STAT-signaling pathway [18, 19]. TMED2 is associated with poor prognosis of liver hepatocellular carcinoma, lung adenocarcinoma and head-neck squamous cell carcinoma. Again, TMED3 overexpression is associated with poor clinical outcome in lung squamous cell carcinoma and colorectal cancer patients’ survival [20-22]. Considering the significant role of TMED proteins in cellular proliferation and differentiation, all of the ten known TMED proteins have been studied in tumour-associated contexts [23-25], and have displayed a promising role as biomarkers of different cancer types. Amongst these proteins, TMED4 and TMED9 have been less investigated for their role in brain tumor pathomechanism. However, a recent study screened TMED9 proteins for unleashing their link to vascular invasion and poor prognosis with hepatocellular carcinoma, displaying promising results [26]. Moreover, the prognostic potential of TMED9 protein has been investigated in another recent study demonstrating an association between high expression of TMED9 and breast carcinoma cell proliferation and migration [29]. Another study reported significant upregulation of TMED4 proteins in longer-surviving patients with pancreatic ductal adenocarcinoma (PDAC) [30].
In this study, we have investigated the differential expression of TMED4 and TMED9 genes within different subgroups of glioma patients and associated different molecular aspects of these proteins with the glioma patients’ clinical outcomes. We have aimed to unleash the role of TMED4 and TMED9 genes and their post-transcriptional products as prognostic markers and therapeutic targets in brain cancer gliomas, more specifically LGGs and GBM. The scientific findings of this study should assist in TMED-based therapeutic and diagnostic measure formulation in glioma.
2. Materials and Methods
2.1. Differential Expression Analysis on TMED4 and TMED9 Genes in LGG and GBM Tissues
The mRNA level expression pattern of TMED4 and TMED9 genes in LGG and GBM tissues was determined using two servers i.e., GEPIA 2 (http://gepia2.cancer-pku.cn/) and OncoDB server (http://oncodb.org/) to increase the fidelity of the findings. GEPIA 2 is an online platform that allows the RNA sequencing data analysis on cancerous and normal tissues from integrated The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) samples [31]. OncoDB also assists in differential expression analysis of different genes in cancerous tissues and correlating gene expression with cancer patient’s clinical outcome [32]. The expression pattern of the genes in different glioma cell line was analyzed using the Expression Atlas server (https://www.ebi.ac.uk/gxa/home) [33]. Finally, the protein level expression of TMED4 and TMED9 genes was determined from the Human Protein Atlas (HPA) (https://www.proteinatlas.org/) server by exploring the tissue and pathology modules containing the immunohistochemistry images of gene expression [34].
2.2. Expression Analysis of TMED4 and TMED9 Genes in Accordance with Glioma Patients’ Demographic and Clinical Features
The expression level of our genes of interest in relation to glioma patients’ age, glioma grades (WHO), Isocitrate Dehydrogenase (IDH) mutation status and histological subtypes of glioma was determined using the online platform Chinese Glioma Genome Atlas (CGGA) (http://www.cgga.org.cn/). This is a web-based user-friendly database enabling users to compare the expression of mRNA, DNA methylation and miRNA levels between normal and brain tumor samples [35]. This server also allows the analysis of correlation between different gene/mRNA level features with brain cancer patients’ clinical manifestation by performing Analysis of Variance (ANOVA) test in between the control and test cohorts. The result was considered significant based on the p value cutoff of <0.5. Finally, the co-expression pattern of TMED4 and TMED9 in glioma tissues was also evaluated from this server.
2.3. Analysis of the Promoter and Coding Sequence Methylation Pattern in TMED4 and TMED9 Genes in LGG and GBM Tissues
The genomic profiles of TMED4 and TMED9 in LGG and GBM tissues delineating their methylation status was retrieved from the OncoDB server. The result displayed significant methylation position across the coding sequence of the selected genes. Thereafter, the methylation pattern of TMED4 and TMED9 gene was compared to the sample type and histological type of TCGA LGG (n=530) and TCGA GBM samples (n=631), respectively from the UCSC Xena Browser. UCSC Xena browser in an online data repository that allows the users to explore genomics data and compare genotypic and phenotype variables [36]. Finally, the association between TMED4 and TMED9 methylation and glioma patients’ survival rate was discovered utilizing the Gene Set Cancer Analysis (GSCA) web-based tool (http://bioinfo.life.hust.edu.cn/GSCA). GSCA is an integrated tool that helps the users in identifying correlation between specific gene expression and genomic variation, patients’ clinical features, immune infiltration level and so on [37]. The results obtained from this served were considered significant based on a p-value cutoff of >0.05.
2.4. Analysis of the Frequency of Mutation and Copy Number Alterations in TMED4 and TMED9 Genes in Glioma Tissues
The mutation and Copy Number Alteration (CNA) frequency in TMED4 and TMED9 in LGG and GBM tissues was identified from the cBioPortal Server (https://www.cbioportal.org/) [38]. In the server, the genomic data deposited for LGG and GBM by TCGA, CPTAC, MSK, UCSF and a few other organizations spanning a total of 7 studies incorporating 2,404 samples were selected for the analysis. Thereafter, the OncoPrint and cancer types summary were investigated to observe the mutation and CNA events in those selected genes in glioma tissues. Finally, the association of CAN events in TMED4 and TMED9 genes with the glioma patients’ Overall Survival (OS), Disease Specific Survival (DSS), Relapse-free Survival (RFS) and Progression-free Survival (PFS) was evaluated from the GSCA server. The correlation in this step was analyzed based on hazard ratio and considered significant for p<0.05.
2.5. Survival Analysis on Glioma Patients in Relation to TMED4 and TMED9 Expression
Initially, the OS of LGG and GBM patients in accordance with TMED4 and TMED9 expression was compared in the PreCog using the individual cohort of this server (https://precog.stanford.edu/) [39]. This server enables the users to assess genomic data and find out correlations with cancer patients’ clinical sequalae. In other words, this server allows to define whether upregulation or downregulation of a particular gene is favorable or unfavorable for cancer patients’ OS and DSS. To further increase the credibility of survival analysis, the association between the expression level of our gene of interest and glioma patients’ OS was established from the OncoLnc server (http://www.oncolnc.org/) that helps linking the survival information on cancer patients with mRNA, miRNA and lncRNA expression levels from TCGA cohorts [40]. In either case, the survival data were analyzed based on HR and considered significant based on p values cutoff of >0.05.
2.6. Analysis on the Association between Immune Infiltration Level and TMED4 and TMED9 Expressions in Glioma Patients
Initially, the association between the selected gene expression and different immune cell infiltration levels i.e., B cell, T cell, Macrophage, Natural Killer (NK) cells in glioma tissues was determined from the immune module in TIMER 2 web-based tool (http://timer.cistrome.org/). TIMER 2 is a publicly available database that helps users in identifying specific gene expression with immune cell abundance levels based on multiple immune deconvolution strategies [41]. In the next step, the correlation between TMED4 and TMED9 expression and different immune modulator expression in LGG and GBM tissues was identified from the TISIDB online portal (http://cis.hku.hk/TISIDB/). This platform allows the interaction analysis between tumor and immune system b enabling the users to explore the association between different gene expression immunomodulators expression like chemokine, immune-receptors, immune-inhibitors [42]. In both cases, the results were analyzed based on spearman correlation coefficient and considered significant on the basis of p value.
2.7. Examining the Co-expressed Neighbor Genes of TMED4 and TMED9 in Glioma Tissues and Their Functional Enrichment Analysis
The co-expressed neighbor genes of TMED4 and TMED9 in glioma tissues was evaluated from the LinkedOmics server (http://linkedomics.org/login.php) using the TCGA LGG and TCGA GBM cohorts [43]. Initially, top positively correlated genes were inspected on the basis of pearson correlation coefficient. Thentop 10 highest positively co-expressed genes were selected from each cohort and utilized in the subsequent functional enrichment analysis using the Enrichr server [44]. Eventually, the co-expression mode of the top selected genes in glioma tissues was inspected using the CGGA server.
3. Results
3.1. mRNA and Protein Level Expression of TMED4 and TMED9 in LGG and GBM
The selected genes were investigated to understand their expression level at the mRNA and protein levels. The GEPIA 2.0 analyses revealed that TMED4mRNA is highly expressed both in LGG tissues (Tumor:518, Normal:207) and GBM tissues (Tumor:163, Normal:207) than their respective normal counterparts (Figure 1a). The similar inspection for TMED9 gene also revealed that its transcriptional product is highly expressed both in LGG and GBM tissues compared to the normal brain tissues (Figure 1b). Thereafter, the TMED4 and TMED9 genes were also analyzed in the OncoDB database to observe their expression patterns in two forms of cancerous samples. Similar to the previous step, it was observed that TMED4 mRNA is highly expressed in LGG tissues (p=3.31e-21) and GBM (p=2.1e-30) tissues when compared to the normal samples (Figure 1c and 1d). For TMED9, LGG tissues (p=3.9e-129) and GBM tissues (1.3e-72) also showed higher expression in comparison with normal tissues (Figure 1e and 1f). Interestingly, most noticeable differences between LGG and GBM cohorts were observed for TMED9 expression along with the highest fold change in GBM tissues as observed from the results obtained from both servers. Afterward, the expression pattern of the genes was analyzed across different glioma cell lines. As in par with the previous results, TMED9 showed higher expression in most of cell lines than TMED4 (Figure 1f). Finally, the protein level expression of the TMED4 and TMED9 genes was inspected by analyzing the immunohistochemistry (IHC) images from the HPA server. Herein, the normal central nervous system (CNS0 tissues showed low staining for the IHC antibody administered against the TMED4 and TMED9 proteins, whereas, cancerous tissues showed medium to high IHC intensity reflecting their higher protein level expression (Figure 2).
3.2. Expression Pattern of TMED4 and TMED9 across Different Grades, Subtypes and Demographic Conditions of Glioma Patients
In this step, the TMED4 and TMED9 genes were enquired in the CGGA server to discover their expression across different clinical and demographic conditions of glioma patients. Significant association between the expression levels and patients’ age was observed (Figure 3). Both the TMED4 (p=2.5e-07) and TMED9 (p-1.2e-06) were found be expressed at a higher level in >42 years age group compared to the <42 years group. Moreover, upward trend in TMED4 (p=1.3e-11) and TMED9 (p=1.9e-26) expression was also observed across World Health Organization (WHO) grade II, III and IV gliomas (Figure 3). Considering the prevalence of Isocitrate Dehydrogenase (IDH) mutation in glioma patients, the expression level of the selected genes in this study was also analyzed in relation to IDH mutation status. Interestingly, wildtype glioma tissues showed significantly high level of expression of both TMED4 (p=5.9e-11) and TMED9 (p=2.8e-22) genes (Figure 3). Thereafter, their differential expression pattern across different histological subtypes was determined. Similar pattern of increment in those gene expression in accordance with advancing subtype was observed. In both the cases, TMED4 (p=1.1e-08) and TMED9 (p=1.1e-08) showed least expression in Oligoastrocytoma tissues followed by a gradual rise in expression level with a few events of downward trends and finally achieving maximum expression level in Glioblastoma (Figure 3). Finally, the co-expression pattern of TMED4 and TMED9 was analyzed and it was found that both the genes are highly co-expressed in primary (Cor: 0.609, p: 2.65e-24) and recurrent glioma (Cor: 0.577, p: 9.36e-07) tissues (Supplementary Figure S1).
3.3. The Promoter and Coding Sequence Methylation Pattern of TMED4 and TMED9 Genes in LGG and GBM Tissues
The promoter and coding sequence methylation status of TMED4 and TMED9 genes was retrieved from the OncoDB server. TMED4 gene’s promoter region was found to be less methylated in LGG tissues whereas the 3’ end of the coding region showed highest methylationlevel (Figure 4a). Similar pattern of methylation was also observed for TMED4 in GBM tissues, however, a delicate difference both in the promoter region and downstream landscape to promoter was observed (Figure 4b). In case of TMED9, a completely different pattern of methylation was observed when compared to TMED4 gene methylation. The result suggested that the promoter region of TMED9 is less methylated than TMED4 in LGG tissues (Figure 4c). However, unlikely to TMED4, no observable difference in methylation pattern of TMED9 in GBM tissues in contrast to the LGG tissues was observed (Figure 4d). Thereafter, their coding sequence methylation pattern was analyzed from the UCSC Xena browser. The analysis result from this server supported the findings of previous steps by revealing that most of CpG islands might be present at the 3’ end of the coding sequence as observed in the red colored landscape (Supplementary Figure S2). The result also indicated that both the genes have distinct methylation pattern which might be maintained through the transition between LGG and GBM states. Finally, the relation between TMED4 and TMED9 methylation and glioma patients’ survival rate was observed. Significant association between TMED9 hypomethylation and GBM patients’ poor overall survival (OS) (p=0.0046) and disease specific survival (DSS) (p=0.0095) was observed (Figure 4e and 4f).
3.4. The Frequency of Mutation and Copy Number Alterations in TMED4 and TMED9 Genes in LGG and GBM Tissues
The mutation and copy number alteration (CNA) frequency in TMED4 and TMED9 genes in glioma tissues was determined using the cBioPortal server. Overall, TMED4 (frequency:1.1%) was found to undergo more CNA events than the TMED9 (frequency: 0.4%) gene in glioma tissues (Figure 5a). Moreover, amplification was the only CNA event in TMED4, whereas, TMED9 experienced deep deletion along with amplification in different selected glioma studies (Figure 5a and 5b). However, no synonymous or nonsynonymous mutation was recorded for the selected genes in glioma tissues.Thereafter, the type of amplification and deletion in TMED4 and TMED9 genes in glioma tissues was determined from the GSCA server. It was found that, TMED4 experienced more amplification events in the GBM tissues compared to the LGG tissues. Furthermore, majority of the amplification was heterozygous while homozygous amplification contributed to a minor portion (Figure 5c). On the contrary, TMED9 gene underwent more heterozygous deletion than amplification both in the LGG and GBM tissues. Finally, the association between CNA in those genes and glioma patients’ survival was evaluated from the GSCA server. Although TMED9 did not show any significant correlation, TMED4 CNA was discovered to be negatively associated with both the LGG and GBM patients’ OS, DSS and progression free survival (PFS) (p<0.001) (Figure 5d) (Supplementary Table S1).
3.5. Association between TMED4 and TMED9 Expression and Glioma Patients’ Survival Rate
In this step, the TMED4 and TMED9 genes were explored to examine the association between their expression and LGG and GBM patients’ OS rate in the PreCog server. TMED4 gene expression was found to negatively affect the OS of LGG patients (Hazard Ratio (HR): 4.59, p<0.0001) (Figure 6a). Similar to the LGG patients, TMED4 expression was also discovered to be responsible for poor OS of GBM patients (HR: 1.12) (Figure 6b). However, the association was not found to be statistically significant as evidenced by higher p value of 0.12. In case of TMED9 in LGG patients, high expression was again significantly accounted for the worse OS of the patients (HR: 3.47, p<0.0001) (Figure 6c). Finally, high expression of TMED9 was found to be negatively correlated with the OS of GMB patients too (HR: 1.86, p=0.048) (Figure 6d). To further validate the findings of survival analysis from PreCog server we carried out the OS analysis again with the help of OncoLnc server. As in par with the previous findings, TMED4 overexpression was revealed to be the significant cause of poor prognosis in LGG (p=4.55e-07) and GBM patients (p=0.003) (Supplementary Figure S3a and S3b). Similarly, significant negative association of TMED9 overexpression in LGG (p=3.85e-06) and GBM patients’ (p=0.018) OS was also reported (Supplementary Figure S3c and S3d).
3.6. Association between TMED4 and TMED9 Expression and the Abundance of Tumor-infiltrating Immune Cells in LGG and GBM Patients
The association between the selected gene expression and tumor-infiltrating immune cell abundance was identified from the TIMER 2 server. Significant positive correlation between TMED4 expression and B cell infiltration was observed in LGG patients (Cor:0.09, p=3.20e-02) (Figure 7a). Moreover, TMED4 expression was significantly and positively associated with the CD8+ T cell (Cor:0.13, p=4.21e-03) infiltration level in LGG patients, whereas, a negative correlation was identified for Natural Killer (NK) cell (Cor: -0.389, p=1.07e-18). In GBM patients, TMED4 did not show any significant positive or negative correlation with any of the immune cell infiltration level (Figure 7b). On the contrary, TMED9 expression was negatively correlated with the B cell (Cor: -0.39, p=4.27e-19) and CD8+ T cell (Cor: 0.44, p=1.78-24) abundance in LGG patients (Figure 7c). Significant positive correlation between TMED9 expression and CD4+ T cell (Cor: -0.39, p=4.27e-19), Macrophage (Cor: -0.39, p=4.27e-19) and NK cell (Cor: -0.39, p=4.27e-19) abundance was observed in LGG patients (Figure 7c). Interestingly, TMED9 expression was negatively correlated with B cell (Cor: -0.27, p=1.42e-03), CD8+ T cell (Cor: -0.38, p=4.77e-06), CD4+ T cell (Cor: -0.12, p=1.51e-02), and NK cell (Cor: -0.26, p=2.18-103) infiltration level in GBM patients (Figure 7c). Moreover, significant positive correlation for TMED9 expression was recorded only for Macrophage abundance in GBM patients (Cor: 0.20, p=1.48e-02). Finally, the expression level of TMED4 and TMED9 was explored to identify the association with different immunoinhibitor i.e., IL10RB, CD274 expression in glioma cells from TISIDB server. Significant association between TMED4 expression and IL10RB (Cor: 0.121, p=0.005) and CD274 (Cor: 0.157, p=3e-03) abundance in LGG tissues was observed (Figure 8a and 8b). IL10RB expression also showed positive correlation with TMED4 gene expression in GBM tissues (Cor: 0.548, p<2.2e-16) (Figure 8c). For TMED9 expression, significant association with IL10RB (Cor: 0.204, p=0.008) and CD274 (Cor: 0.161, p=0.038) expression in LGG tissues was reported (Figure 8e and 8f). Additionally, IL10RB (Cor: 0.454, p=1.17e-09) and CD274 (Cor: 0.298, p=1e-02) expression in GBM tissues were significantly and positively correlated with TMED9 expression levels (Figure 8g and 8h).
3.7. The Positively Co-expressed Neighbor Genes of TMED4 and TMED9 in Glioma Patients and Their Functional Enrichment Analysis
The neighbor genes of TMED4 and TMED9 in LGG and GBM patients were identified using the LinkedOmicsserver (Supplementary Figure S4). TMED4 was found to be highly and positively co-expressed with TRNA-YW Synthesizing Protein 1 Homolog (TYW1) (Cor: 0.73, p=1.84e-90) in LGG patients whereas it showed highest co-expression level with NudC Domain Containing 3 (NUDCD3) (Cor: 0.67, p=5.84e-21) genein GBM patients (Figure 9). On the other hand, TMED9 showed highest co-expression level with Lectin, Mannose Binding 2 (LMAN2) both in the LGG (Cor: 0. 80, p=7.41e-117)and GBM (Cor: 0.71, p=1.73e-24)patients (Figure 9). Thereafter, the top 10 positively co-expressed genes of TMED4 and TMED9 in two different glioma tissues were selected for their functional enrichment analysis. The biological process analysis of the neighbor genes revealed that they were predominantly involved in collagen fibril organization, negative regulation of trophoblast and platelet degranulation (Figure 10a). Among the top selected molecular functions of the neighbor genes hexosyl transferase activity, transition metal ion binding and zinc ion binding were most significant (Figure 10b). The cellular component analysis revealed that the gene cluster was mostly operating in the focal adhesion point, cell-substrate junction and intracellular organelle lumen (Figure 10c). Different pathway analysis on the neighbor genes revealed that, most of them are involved in post-translational modification of different proteins, modulating protein functions in endoplasmic reticulum (ER), extracellular matrix organization, cell communication and so on (Figure 10d, 10e and10f).
4. Discussion
In this study, a database mining approach was utilized to evaluate and establish the prognostic value of TMED4 and TMED9 gene expression in glioma patients. Differential gene expression analysis can aid in understanding the roles of specific gene overexpression or under-expression in the oncogenic development of a cell and its subsequent growth [45]. Initially, TMED4 and TMED9 genes were found to be differentially expressed (upregulated) in LGG and GBM tissues at the mRNA and protein levels suggesting their possible oncogenic functions in glioma growth and progression (Figure 1 and 2). Nevertheless, they showed higher level of expression across different glioma cell lines. Since cancer development is a very complicated and multistep process varying across different cancer subtype, grade and patients’ demographic conditions [46], a comprehensive understanding on the expression level of a particular gene across different variables is required to understand its role in the oncogenic processes. In this study, our genes of interest showed increment in their expression with advancing age of glioma patients, glioma grades and histological subtypes which indicates that their expression might have significant association with glioma exacerbation and poor prognosis (Figure 3).
DNA methylation is one of the major epigenetic drivers in cancer development and growth. Specifically, promoter methylation controls the gene activity of a particular gene by silencing or activating the transcription of that gene and hypomethylation is associated with upregulation of gene activity [47]. Coding sequence methylation also regulates the gene expression by altering the genome sequence structure in the chromatin architecture [48]. In this study, both the TMED4 and TMED9 promoter regions were found to be less methylated compared to the coding sequence regions, might be accounting for their overexpression in glioma tissues (Figure 4). Moreover, the differential DNA methylation pattern of specific genes can serve in formulating methylation-sensitive diagnosis of glioma cells and making epigenetic clinical decision [49]. We also found that TMED4 showed slight difference in methylation pattern between LGG and GBM tissues, whereas, TMED9 maintains its distinct methylation feature even after transition from LGG to GBM state (Figure 4). Therefore, TMED9 may aid in the methylation-based diagnosis of glioma patients, whereas, TMED4 is supposed to offer further stratification of LGG and GBM patients. Additionally, the association between TMED9 hypomethylation and GBM patients’ poor OS and PFS signifies that it could also aid in the discovery of TMED9-based epigenetic glioma treatment measures (Figure 4). However, such assumptions require further laboratory investigations.
Somatic driver mutations are the predominant causes of tumorigenic transformation of healthy cells inside human body [50,51], and somatic CNA contributes to a larger fraction in the cancer development than any other type of mutations [52]. These genetic changes influence the cancer initiation process by regulating oncogenes or tumor-suppressor genes [53]. Moreover, heterozygous amplification and deletion in specific genes like EGFR and MDM2 can also serve as the molecular diagnostic target for high-throughput heterozygosity mapping in glioma patients [54,55]. Both TMED4 and TMED9 genes in glioma tissues were discovered to undergo multiple heterozygous CNA events with more amplification incidents in TMED4 and deletion in TMED9 which supports their potentiality as effective diagnostic candidates for glioma patients (Figure 5). Moreover, the association between TMED9 expression and different poor survival rates of glioma patients indicates that the CNA present in those genes might have critical underlying mechanisms in glioma development and prognosis that requires further investigation.
Thereafter, the survival analysis revealed that the higher expression of both the genes was negatively correlated with the OS of both LGG and GBM patients (Figure 6). This suggests that the overexpression of TMED4 and TMED9 might have negative effects on glioma patients throughout different disease stages. This along with their higher expression in glioma tissues irrespective of patients’ demographic and clinicopathological conditions emphasizes that both the genes should aid in TMED-based tracking of glioma patients throughout the clinical courses.
Different immune cells along with adjacent intrinsic cell of the CNS forms the tumor microenvironment in gliomas which determines the cancer growth, metastasis and response to therapy [56,57]. Thus, different immune cells are prominent measures for formulating immunotherapy against glioma considering their ability to cross the blood brain barrier [58]. Different cytokine and chemokine secreting immune cells after getting produced in higher titers in the glioma microenvironment control the immunity against the glioma cells [59]. Abundance of numerous immunoinhibitors along with immune cells can assist in the diagnosis of glioma patients and keeping track on the patients throughout the disease state. For example, high abundance of CD8+ T cells in glioma patients predicts better survival [60]. In this study, TMED4 and TMED9 expression was found to be significantly correlated to the abundance of different immune cells (i.e., B cells and T cells) in LGG and GBM tissues that may aid in administering combinatorial immunotherapy along with TMED-based glioma treatment method or help in dual diagnosis (Figure 7). The expression level of our genes of interest also showed significant association with the expression level of different immunoinhibitors i.e., PDC274, IL10RB (Figure 8). Previously, CD274 infiltration level in glioma cells has been proposed to predict the favorable survival of glioma patients [61], as well as, IL10RB upregulation has been linked to the unfavorable survival of glioma patients [62].
The co-expression analysis revealed that TMED4 gene is highly co-expressed with TYW1 gene in LGG tissues and NUDCD3 gene in GBM tissues (Figure 9). TYW1 overexpression and mutations are associated with leukemia and breast cancer prognosis [63,64]. NUDCD3 overexpression is also linked to the poor prognosis of cervical cancer patients [65]. TMED9 gene was found to be highly co-expressed with LMAN2 gene both in LGG and GBM tissues (Figure 9). Recent laboratory research has grounded that this particular gene is differentially methylated in brain white matter tissues and it is overexpressed in brain metastatic breast cancer tissues [66,67].
Pathway analysis of the co-expressed neighbor genes in glioma tissues indicated that most of the genes are involved in metal ion such as Zinc binding (Figure 10). Previous study suggests that deficiency in Zinc concentration, imbalance and deregulation is associated with different cancers including childhood brain tumor development [68]. Maintaining focal adhesion was among the top selected molecular functions of neighbor genes and suppression of such activity reduces brain cancer growth [69,70]. Cellular component analysis on the neighbor genes of TMED4 and TMED9 revealed that they are predominantly involved in protein processing in ER and any unfolded protein response in ER can influence the tumorigenesis in GBM tissues [71]. These evidences suggest that the neighbor genes of TMED4 and TMED9 in LGG and GBM tissues can also be investigated in glioma diagnosis and therapeutic measure discovery.
Overall, the reports in this study obtained from differential expression analysis, methylation and mutation frequency of TMED4 and TMED9 genes in LGG and GBM tissues indicate that these genes might have underlying mechanism in the glioma development and progression. Their overexpression pattern and effects on survival pattern signifies that the selected genes could be potential diagnostic and therapeutic target for glioma diagnosis and prognosis. Finally, this study recommends thar TMED4 and TMED9 are potential prognostic and therapeutic targets for glioma. However, further laboratory research is warranted to validate the significance of this study which is currently underway.
5. Conclusion
In summary, this study demonstrated the differential expression and variation in the methylation pattern of the promoter and coding sequences of TMED4 and TMED9 in LGG and GBM tissues. A number alteration events were reported in the coding regions of those genes in glioma tissues. Significant association between TMED4 and TMED9 overexpression and glioma patients OS was observed. These evidences suggest that TMED4 and TMED9 could be potential target for glioma diagnosis and treatment. Moreover, expression levels of multiple immune cells and immunoinhibitors were found to be associated with TMED4 and TMED9 expression in LGG and GBM tissues that may aid in formulating TMED-based diagnostic method and therapeutic interventions. The functional enrichment analysis revealed that their co-expressed genes are involved in functions and deregulation in those activity can promote gliomageneis. Thus, the neighbor genes of TMED4 and TMED9 could also be investigated further while extending laboratory work on making TMED-based diagnostic and therapeutic measures for glioma patients.
Author’s Contribution Statement
MU conceived and designed the study. MU and AM carried out the experiment. MU and AM illustrated the graphs and figures. MU, TT and MF wrote the initial draft. MU, TT, MF, UZ and MR edited and revised the paper. MR supervised the study. All authors approved the final manuscript for publication.
Ethics Approval and Consent to Participate
Not Applicable
Consent for Publication
Not Applicable
Availability of Data and Material
All the data are provided within the manuscript and the supplementary material.
Competing Interest
All the authors declare that they have no conflict of interest regarding the publication of the paper.
Funding
No specific grant was received for this study.
Acknowledgement
Authors are thankful to the members of Bio-resources Technology and Industrial Biotechnology Laboratory, Jahangirnagar University, Dhaka, Bangladesh and Swift Integrity Computational Lab, Dhaka, Bangladesh for their supports during the preparation of the manuscript.