Expression Patterns and Prognostic Values of ORMDL1 in Different Cancers

The mammalian orosomucoid-like gene family (ORMDL), containing ORMDL1, ORMDL2, and ORMDL3, is the important regulator of sphingolipid metabolism, which is relevant to cell growth, proliferation, migration, and invasion. Since the role of ORMDL1 in cancers remained unclear, the main purpose of our study was to explore the expression patterns and prognostic values of ORMDL1 in different tumors, especially in cholangiocarcinoma (CHOL), lymphoid neoplasm diffuse large B cell lymphoma (DLBCL), acute myeloid leukemia (LAML), and thymoma (THYM). Bioinformatics tools including GEPIA, CCLE, LinkedOmics, cBioPortal, and TIMER databases were used. As a result, the expression levels of ORMDL1 in tumor tissues and normal tissues varied in different cancers, especially significantly upregulated in CHOL, DLBCL, LAML, and THYM. Moreover, ORMDL1 mRNA was also highly expressed in cell lines of DLBCL and LAML. Further studies showed that ORMDL1 overexpression was associated with poor prognosis in DLBCL, but not significant in CHOL, LAML, and THYM. Consistently, there were genetic alterations of ORMDL1 in DLBCL, and patients with genetic alterations indicated worse survival. Coexpressed genes and related biological events with ORMDL1 in DLBCL were found via LinkedOmics, Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. The relationship between ORMDL1 and cancer immune cells was investigated, and ORMDL1 expression was positively correlated with infiltrating levels of B cells. In conclusion, ORMDL1 is suggested to be a tumorigenic factor and considered as the potential therapeutic target and prognostic biomarker in DLBCL.

As an important component of the cell membrane, sphingolipids can regulate cell growth, proliferation, migration, invasion, and metastasis through cancer signaling pathways, in addition to exerting barrier function and maintaining membrane fluidity [16][17][18]. Since ORMDL1 was a regulator of sphingolipid levels in cells, we hypothesized that ORMDL1 might play a role in the pathogenesis and progression of tumors. Thus, our study shed light on the specific role of ORMDL1 in different tumors via bioinformatics analysis, particularly in cholangiocarcinoma (CHOL), lymphoid neoplasm diffuse large B cell lymphoma (DLBCL), acute myeloid leukemia (LAML), and thymoma (THYM).  BLCA  BRCA  CESC  CHOL  COAD  DLBCL  ESCA  GBM  HNSC  KICH  KIRC  KIRP  LAML  LGG  LIHC  LUAD  LUSC  MESO  OV  PAAD  PCPG  PRAD  READ  SARC  SKCM  STAD  TGCT  THCA  THYM  UCEC  UCS      3 BioMed Research International GEPIA provides 9,736 tumors and 8,587 normal samples from The Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) projects for analyzing the RNA sequencing expression data [19]. GEPIA was used to analyze the differential expression of ORMDL1 between normal tissues and tumor tissues in 33 different types of cancer. We compared the expression patterns of ORMDL1 in four cancer types, including CHOL, DLBCL, LAML, and THYM. Moreover, GEPIA also provided the function for prognostic curve analysis and pathological stage evaluation.

CCLE Dataset
Analysis. The Cancer Cell Line Encyclopedia (CCLE, http://www.broadinstitute.org/ccle/home) project is a collaboration between the Broad Institute and the Novartis Institutes for Biomedical Research and its Genomics Institute of the Novartis Research Foundation. It can be applied to conduct a detailed genetic and pharmacologic characterization of a large panel of human cancer models, develop integrated computational analyses that link distinct pharmacologic vulnerabilities to genomic patterns, and translate cell line integrative genomics into cancer patient stratification. CCLE provides public access to genomic data, analysis, and visualization for about 1,000 cell lines [20]. The expression of ORMDL1 was verified by the CCLE dataset.
2.3. LinkedOmics Dataset Analysis. LinkedOmics (http:// www.linkedomics.org/login.php) provides a newly developed platform for analyzing large-scale cancer omics data from TCGA and Clinical Proteomic Tumor Analysis Consortium (CPTAC) [21]. We used LinkedOmics to inquire into the prognostic values of ORMDL1 expression in the four cancer types, including CHOL, DLBCL, LAML, and THYM. The survival differences were visualized by Kaplan-Meier plots. Furthermore, the correlation coefficient and coexpressed gene patterns were calculated according to the online instruction.

BioMed Research International
and abundance of immune infiltrates [23]. In this study, the relationship between ORMDL1 expression and the six immune cells was estimated by TIMER in DLBCL.
2.6. Statistical Analysis. The difference in ORMDL1 expression between tumor tissues and normal tissues was compared with an independent t-test. ORMDL1 expression in different clinical stages was evaluated using one-way ANOVA. The relationship between ORMDL1 expression and patient prognosis was detected using the Kaplan-Meier survival analysis and log-rank test. The correlation between ORMDL1 and related genes was analyzed using the Pearson correlation test. P < 0:05 indicated statistical significance.

ORMDL1 mRNA Expression in Different Kinds of Cancer
Cell Lines. By collecting genetic information from CCLE, investigation of ORMDL1 expression was extended to various cancer cell lines. As a result, ORMDL1 mRNA was found to be highly expressed in cell lines of LAML and DLBCL, which ranked 1 st and 11 th among 40 kinds of cancers ( Figure 2).

The Prognostic
Influence of ORMDL1 on CHOL, DLBCL, LAML, and THYM. We further explored the influence of ORMDL1 expression levels on the survival of patients in four cancer types, including CHOL, DLBCL, LAML, and THYM. The Kaplan-Meier curves and log-rank test analysis revealed that increased ORMDL1 was associated with poor overall survival (OS) in DLBCL significantly, but not in CHOL, LAML, and THYM by LinkedOmics (Figures 3(a)-3(d)).
Also, similar results were predicted in GSE10846 and GSE53786 using shinyGEO online tool, which could analyze patient survival from the GEO database, suggesting that DLBCL patients with higher ORMDL1 levels tended to have lower OS (Figures 3(e) and 3(f)).

Genetic Alterations of ORMDL1 in DLBCL.
Since ORMDL1 might play a role in DLBCL, genetic alterations of ORMDL1 in DLBCL were determined using cBioPortal database analysis. ORMDL1 mutations included gene gain and shallow deletion from 48 DLBCL patients (TCGA, Provisional) (Figure 4(a)). The relationship between ORMDL1 genetic alterations and DLBCL patient survival was further evaluated. The Kaplan-Meier survival analysis showed that cases with genetic alterations were associated with worse prognosis (Figure 4(b)).

Coexpressed Genes and Functional Analysis of ORMDL1 in DLBCL.
To figure out the potential interaction of ORMDL1 with other genes in DLBCL, correlation analysis between ORMDL1 and various genes and markers was performed via LinkedOmics. As shown in Figure 5, the top 50 significant genes positively and negatively correlated with ORMDL1 were shown in the heat map. A detailed description of the coexpression genes is listed in Table 1. Furthermore, Gene Ontology (GO) analysis in biological process by GSEA indicated that ORMDL1 coexpressed genes mainly participated in DNA damage response, nucleus localization, rRNA metabolic process, and cell cycle checkpoint (Figure 6(a)).

ORMDL1 Is Correlated with Immune Infiltration Level in DLBCL.
To understand the relationship between ORMDL1 expression and immune signatures, we analyzed the six immune marker genes in DLBCL, including B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells. The results revealed that the expression level of ORMDL1 was significantly correlated with the infiltrating level of B cells in DLBCL (Figure 7(a)). Moreover, the gene gain mutation of ORMDL1 promoted the B cell infiltration in DLBCL (Figure 7(b)).

Discussion
Actually, the ORMDL gene family is a group of evolutionary conserved gene sequence found in Drosophila, yeast, and mammals. Among them, Drosophila only has a single-copy gene, yeast has two homologous genes of Orm1 and Orm2, and mammalian cells contain three homologous genes with ORMDL1, ORMDL2, and ORMDL3 [1]. The three human ORMDL isoforms are located on chromosomes 2q32, 12q13, and 17q21, respectively, with approximately 80% identical amino acid, proving that they may have some common biological functions [1,12,24]. The most important function of ORMDLs is to regulate sphingolipid biosynthesis and maintain ceramide homeostasis [3][4][5][6][7][8][9][10][25][26][27][28], where SPT is the rate-limiting enzyme. In yeast, the Orm/SPT compound regulates sphingolipid expression levels by a negative feedback response. When the sphingolipid concentration is high, the Orm protein binds to SPT to inhibit SPT activity and reduce the further synthesis of sphingolipid. When the sphingolipid concentration is low, the N-terminal region phosphorylation of Orm protein causes its separation from SPT, which eliminates the repression of SPT and promotes sphingolipid biosynthesis [3,4,24,25]. However, the regulation mechanism similar to that of yeast cannot be found in mammalian ORMDLs since human ORMDL proteins lack N-terminal phosphorylation sites [1,4]. The study by Siow and Wattenberg demonstrated the feedback response of ORMDL-mediated sphingolipid synthesis [10]. The phenomenon that inhibition of SPT activity caused by permeable cells treated with C6 ceramide suggested ORMDLs might have a structural domain interacting with C6 ceramide, which was further involved in the regulation of ORMDLdependent SPT activity. In addition, Wang et al. explored the relationship between ORMDL1, SPT, and sphingomyelin based on a free cholesterol-(FC-) loading microenvironment in human atherosclerotic macrophages [7,9]. According to their research, the induction of endoplasmic reticulum (ER) stress and autophagy in FC-loaded macrophages led to ORMDL1 shifting from ER to autophagosome, followed by the dissociation of SPT, which was originally bound to ORMDL1. Then, the activation of SPT resulted in increased sphingomyelin synthesis, excessive FC buffering, and reduced cytotoxicity. Dysregulation of sphingolipid metabolism in cancers has been described in several studies [29][30][31][32][33][34][35]. Typical sphingolipid metabolites such as ceramide and sphingosine could be used as bioactive signaling molecules, suppressing cell growth and promoting apoptosis [31]. Phosphorylated metabolites such as sphingosine-1-phosphate (S1P) are related with survival, proliferation, and migration of cancer cells [36,37]. The metabolism of bioactive sphingolipids in mammals is regulated by around 40 enzymes, which play key roles in cancer signaling pathways and therapeutic targets [29,31,32]. Consistently, sphingolipid enzymes and metabolites were abnormally expressed in a variety of cancers. For example, ceramide levels were upregulated in head and neck cancer and breast cancer [38,39], while they were downregulated in ovarian cancer and colon cancer [40,41]. Moreover, sphingosine was highly expressed in endometrial cancer [42]; S1P was overexpressed in glioblastoma [43], and SPT was lowly expressed in colon cancer [44]. Therefore, it was reconfirmed that metabolic disorders of sphingolipids interacted closely with tumorigenesis, tumor development, and chemoresistance of cancer patients.
As described above, ORMDLs act as critical factors in maintaining the balance of cellular sphingolipid levels. However, it remains unclear whether ORMDLs are involved in cancer networks associated with sphingolipid metabolism. In our study, we first preliminarily analyzed ORMDL1 expression in tumor tissues and normal tissues using the GEPIA database and found that ORMDL1 was expressed differently in diverse cancer tissues and adjacent tissues, especially highly expressed in CHOL, DLBCL, LAML, and THYM. Second, high expression of ORMDL1 in cell lines of DLBCL and LAML was verified by the CCLE database, which was consistent with the results in the corresponding tumor samples. To further elucidate the prognostic effect of ORMDL1 expression on CHOL, DLBCL, LAML, and THYM, the Kaplan-Meier survival curves were generated by GEPIA. It revealed that ORMDL1 overexpression was significantly

11
BioMed Research International associated with poor survival of DLBCL, indicating ORMDL1 might facilitate tumorigenesis and recurrence in DLBCL. The results of GSE10846 and GSE53786 further confirmed the results that high expression of ORMDL1 indicated poor prognosis in DLBCL patients. In addition, the cBioPortal database was used as a powerful tool for discovering ORMDL1 genetic alterations in DLBCL, since genetic alteration was considered as an important factor in cancer development [45,46]. Expectedly, increased gene copies and slight gene deletion existed in DLBCL, which partly explained the higher expression of ORMDL1 in DLBCL compared with normal samples. The concomitant result also showed cases with ORMDL1 genetic alterations had worse prognosis. Finally, through LinkedOmics and Pearson correlation test, genes that positively and negatively interacted with ORMDL1 were found and functional analysis in GO and KEGG pathways was further explored, which might be jointly involved in the ORMDL1-related cancer signaling pathways.
There were still some limitations to be solved. Firstly, differences of sample sizes among multidatabases might cause some bias. Secondly, this study only analyzed transcriptional levels of ORMDL1 in cancers, without its posttranslational levels. Finally, molecular mechanism investigation should be carried out to further explore can-cer pathways associated with ORMDL1 and sphingolipid metabolism.

Conclusions
This was the initial study comprehensively analyzing the expression patterns and prognostic values of ORMDL1 in different tumors. ORMDL1 is promising to be the potential therapeutic target and prognostic marker in DLBCL.

Data Availability
Data can be available upon request.

Conflicts of Interest
All authors have no conflicts of interest to declare.

Authors' Contributions
Dr. Tengjiao Zhu and Dr. Yingtong Chen contributed equally to this article in the aspects of drafting the work as well as acquiring and interpreting the data for the work. Dr. Shuyuan Min was accountable for the statistics work of the article. Dr. Li