Increased Expression of TSPEAR in Colorectal Cancer Predicts Poor Prognosis


 Background: Colorectal cancer (CRC) is, at present, the most frequently and most lethal gastrointestinal tract cancer, leading to significant mortality and a large global cancer burden. TSPEAR has been reported to associate with the development of human tooth and hair follicle morphogenesis. However, there is no study on TSPEAR in any cancers have been reported. Methods: The information of TSPEAR gene expression in CRC was collected from TCGA database and GTEx. The Rt-qPCR and WB were used to compare the TSPEAR expression level between normal colon cells NCM-460 and human CRC cell lines, included SW-480, SW-620, HCT-116, Ht-29, and LoVo. Then, we performed the survival, gene function/pathway enrichment, and tumor immune infiltration analyses via using R packages and TIMER database.Results: It was demonstrated that TSPEAR gene showed high expression in CRC tissue and cell lines and up-regulated expression correlating with a poor overall prognosis. Besides, the up-regulated TSPEAR was correlated with the M1 stage, CEA level, and residual tumor. Multivariate analysis further confirmed that TSPEAR remained an independent CRC patient-related risk factors, whereas gene function/pathway enrichment analysis indicated that it may can regulate Wnt, Hippo signaling pathway, as well as many cell metabolism activities. For immune infiltration, the expression level of TSPEAR was positively correlated with immune cells and showed a strong correlation with diverse immune marker sets, especially in COAD.Conclusions: The results demonstrated that TSPEAR is possibly a clinically valuable predictive biomarkers for CRC patients, as well as associated with tumor infiltrating immune cells.


Background
Colorectal cancer (CRC), as the most frequently digestive tract malignant diseases, is the second leading cause of cancer related deaths globally, with an estimated 1.9 million new cases and over 935,000 deaths recorded alone in 2020 (1). The primary treatment for CRC includes cancer-directed surgery, adjuvant or neoadjuvant therapies, systemic chemotherapy, radiation therapy, biologic therapy, and their combinations, which have signi cantly improved the therapeutic effects (2,3). To date, the pathogenesis of CRC remains incompletely understood, and more effective treatment is also lacking. Therefore, the underlying molecular mechanisms of CRC development are urgently needed to be further investigated, and novel candidate biomarkers as well as new therapeutic targets for prognostic evaluation should be identified. Although some new biomarkers have been proposed as prognostic indicators, their clinical application is still limited due to tumor heterogeneity. Therefore, genetic and molecular profiling were used to identify new effective prognostic and predictive biomarkers for enhancing prognosis and individualized treatment.
Thrombospondin-type laminin G domain and EAR repeats (TSPEAR), locates on the chromosome 22 (22q22.3), whose function is poorly understood, especially in cancer eld. Up to now, the study about how TSPEAR involves in occurrence and progression of cancer is lacking. However, previous studies have shown that TSPEAR was capable of regulating Notch signaling pathway, thus in uence the development of human tooth and hair follicle morphogenesis, as well as dysmorphic features(4-6). The Notch signaling pathway is oncogenic, involved in regulating many cellular activities, including cell division, differentiation, apoptosis, as well as disease progression (7). The deregulation or mutation of Notch signaling could impact the tumorigenicity and the most hallmarks of cancer, especially, which have been highly noted in CRC(8, 9).
Herein, we evaluate the expression patterns, potential functions, and prognostic value of TSPEAR expression in CRC, with reference to the bioinformatics analysis of gene expressions and relevant clinicopathologic parameters published online.

Patient data sets
The relevant gene expression data and clinical information (698 samples, HTSeq-FPKM) were retrieved from by the Colorectal Adenocarcinoma projects (COADREAD) in TCGA database (https://cancergenome.nih.gov), which is an available largescale publicly date portal of genome project, including 33 types of tumor sample information. The samples with insu cient survival information are not counted in this measurement. A total of 619 patients with CRC with the clinicopathological details and general features were acquired and further analyzed in this study.

Functional enrichment analysis
The function "cor.test" implemented in R with the Spearman method was utilized to filter the most positive or negative correlation coefficients with TSPEAR. Then, we perform GO and KEGG analysis by using "enrichGo" and "enrichKEGG" functions of the "clusterprofile.package". GO enrichment analysis was performed to predict the functional roles of TSPEAR gene, included BP, CO, MF. Similarly, KEGG pathways analysis was utilized to determine the pathways related toTSPEAR alternative functions and the frequently altered neighbor genes; Immune cell in ltration TIMER, a public available website (http:// cistrome.org/TIMER/), inclusion of 32 cancer types and 10,897 samples derived from the TCGA database (11). Herein, we mainly focused on the abundance of six types of in ltrating immune cells in CRC patients, included B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages, and dendritic cells. Furthermore, we performed the correlation module analysis to explore relationships between TSPEAR expression and possibly gene markers of tumor-infiltrating immune cells, included CD8 + T cells, B cells, T-helper 1 (Th1) cells, -helper 2 (Th2) cells, T-helper 22 (Th22) cells, exhausted T cells, macrophages, TAMs, natural killer (NK) cells, and dendritic cells. Prior studies provide a reference for these gene markers (12,13).

Statistical analysis
The R-3.6.3 was utilized to acquire and analyze statistic. Kaplan-Meier survival curves was used to estimate prognostic value of TSPEAR expression. The TSPEAR gene expression level in CRC patients was evaluated by box plots. The logstic regression was used to analyze accordingly correlation between clinicopathologic variables and TSPEAR expression in CRC. Besides, using a univariate Cox regression approach, with a multivariate Cox analysis, the influence of TSPEAR on survival and other clinicopathologic variables was estimated. A p-value less than 0.05 was considered statistically signi cant.

Baseline characteristics of patients
As shown in Table 1, a total of 619 primary tumor cases with the corresponding clinical details and expression data were retained from TCGA data portal. Among the 619 patients, male patients were 343 (53.3%), with female 289 (46.7%). The median age was 68 years. In terms of pathologic stage, 105 patients were stage I (17.5%), stage II in 227 (37.9%), stage III in 179 (29.9%), and stage IV in 88 (14.7%). The neoplasm type included 454 patients with colon adenocarcinoma (73.3%) and 165 rectum adenocarcinoma patients (74.0%). Primary therapy outcomes of CRC included 12.8% in stable disease (SD) and progressive disease (PD), most patients were in partial remission (PR) and complete remission (CR)(87.3%).

High TSPEAR expression in CRC
To assess the expression level of TSPEAR, the CRC samples from TCGA were compared normal tissues and GTEx databases. The results showed that TSPEAR gene expression level was signi cantly higher in CRC tissues (p<0.001) than that in normal tissues ( Figure 1A). The results were veri ed in CRC tissues and paired normal tissues (p<0.001) ( Figure 1B). Besides, TIMER database further veri ed the high expression of TSPEAR both in COAD and READ ( Figure 1F). In term of cell lines, WB and Rt-qPCR both showed that TSPEAR was up-regulated in CRC cell lines ( Figure 1C-1E).

Correlation and enrichment analyses
TCGA data was used to perform an analysis of correlated genes between TSPEAR and other genes in CRC. Firstly, the top 50 genes most positively-or negatively-correlated with TSPEAR were shown in the heatmap ( Figure 5E, 5F). For gene function enrichment analysis, the top 300 genes correlated most positively with TSPEAR were selected. GO analysis shows that TSPEAR is signi cantly associated with positive regulation of peptide hormone secretion, left/right axis speci cation, phospholipase A2 activity (consuming 1,2−dipalmitoylphosphatidylcholine), integral component of presynaptic membrane, etc. (Figure 5A-5C). In addition, KEGG pathway analysis indicated the enrichments in the Wnt signaling pathway, Hippo signaling pathway, Arachidonic acid metabolism, and alpha-Linolenic acid metabolism ( Figure 5D).

Correlation between immune in ltration and TSPEAR
Previous studies have shown that tumor-infiltrating lymphocytes serve as an indicator of TNM stage and survival in CRC (14). Therefore, we assessed the relationship between TSPEAR expression and immune infiltration levels in colon adenocarcinoma (COAD) and rectal adenocarcinoma (READ) from TIMER. As shown in gure 6A, the results displayed that TSPEAR expression has negative correlation with in ltrating levels of B cells (r = -0.108, P = 3.00e-02), CD8+ T cells (r = -0.201, P = 4.34e-05), neutrophils (r = -0.123, P = 1.35e-02), and dendritic cells (r = -0.117, P = 1.88e-02), while positively with CD4+ T cells (r = 0.12, P = 1.61e-02) in COAD. However, in terms of READ, only neutrophils (r = -0.176, P = 3.94e-02) is negative with the expression of TSPEAR. In addition, TSPEAR expression has no signi cant association with macrophages in COAD and no correlations with tumor purity, B cells, CD8 + T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells in READ. P < 0.05 was considered as the difference is of signi cance.
Above results show that TSPEAR is involved in tumor immune regulation in CRC, especially in COAD. To further exploration, the TIMER databases were used to execute the association between TSPEAR and speci c markers of immune cell both in COAD and READ. As shown in table 2, the results demonstrated that the TSPEAR expression level was correlated with most immune marker genes of immune cells in COAD, included CD8 + T cells, different functional T cells, natural killer cell, as well as dendritic cell ( Figure 6B-6G). However, the TSPEAR expression level was only correlated with 4 immune marker sets in READ (Table 2). These ndings further veri ed that TSPEAR is associated with immune infiltrating in COAD.

Discussion
TSPEAR is a gene never reported in cancer. It has been shown that TSPEAR is expressed in many tissues such as the liver, intestine, lung, kidney, and testis (15). However, the function of TSPEAR is still poorly understood. One study proposed that TSPEAR gene was capable of encoding a member of the Epilepsy-Associated Repeat protein family, which play an important role during the development of the auditory system (15). In addition, the wild-type TSPEAR was detected at the plasma membrane, which indicates that TSPEAR is an extracellular protein and the presence of a ubiquitous TSPEAR receptor at the cell surface. But how the TSPEAR regulating normal cell and tumor cell growth and differentiation is still unclear. Therefore, the function role of FCGBP in cancer development and treatment is urgent to be clari ed.
Herein, we proposed that TSPEAR gene expression level was associated with prognosis in CRC. An upregulated TSPEAR expression can be reckoned an independent prognostic biomarker for poor prognosis. Furthermore, more analysis results demonstrated that TSPEAR expression was correlated with several clinicopathologic features, included M stage, CEA level, as well as residual tumor. Furthermore, according to GO analyses, we found that TSPEAR expression phenotype was associated with the hormone regulation, component of presynaptic membrane and phospholipase A2 activity, however, all of which were the first to be studies, the speci c mechanism needed to be further elucidated. The KEGG results show that TSPEAR expression was associated with many cell metabolism activities, such as Arachidonic acid metabolism, alpha−Linolenic acid metabolism, and Linoleic acid metabolism, and many oncogenic signaling pathways, such as Wnt signaling pathway and Hippo signaling pathway. The Wnt is a core cancer signaling pathway, which are often aberrantly regulated in a large variety of cancers(16). Especially, an increasing studies have demonstrated that Wnt signaling pathway was very essential during the development of human CRC (17).
Besides, previous study has reported that Hippo signaling pathway was involved in diverse physiological processes, spanning from cell growth, proliferation, metabolism and tissue homeostasis to tumor formation and progression(18). These ndings indicate that TSPEAR may in uence tumor development through regulating canonical Wnt and Hippo signaling pathway. However, More speci c studies are is needed for con rmation.
Tumor-infiltrating lymphocytes and tumor microenvironment immune cells were the key factors in tumor development, such as tumorigenesis, angiogenesis, tumor cell growth, and tumor metastasis (19).
Immunotherapeutic strategies were reckoned as a potential direction for predicting survival status and therapeutic e cacy for tumor patients (20). Anti-programmed death-1 (PD-1) and anti-programmed death ligand-1 (PD-L1) have showed signi cantly treatment potential in advanced CRC especially (21,22). Herein, our analysis reveals that different types of immune-infiltrating cells and speci c markers of immune cell are correlated with TSPEAR expression in CRC, mainly in COAD, the mechanism of which is still unclear. Firstly, most of the T cells gene markers, such as CD8A of CD8+ T cells, TBX21, STAT1 and IFNG of Th1, STAT6 and STAT5A of Th2, CCR10, AHR of Th22, were inversely correlated with TSPEAR expression in COAD. These results reveal a potential mechanism that TSPEAR may in uence COAD progression by regulating T cell functions. In addition, gene markers of NK cell markers, such as KIR2DL1, KIR2DL4, KIR3DL2, KIR3DL2, and CD7, showed moderate and strong correlations with TSPEAR expression, which indicate that TSPEAR is a significant factor that relate to immune infiltration of NK cells. Furthermore, we observed significant correlations between TSPEAR and gene markers of Dendritic cell monocytes, such as HLA-DRA and HLA-DPA1, which further assess the potential regulating role of TSPEAR in polarization of tumor-associated macrophages (TAM), indicating that TSPEAR may be involved in regulating immune infiltration role in COAD microenvironment. Based on above ndings, what we can be quite certain of, however, is that TSPEAR is capable of in uencing immune cell in ltration and can be reckoned as a potential cancer biomarker.

Conclusion
All these ndings identified that TSPEAR expression was up-regulated in CRC tissues and cell lines. Moreover, TSPEAR can be seen as an independent promising prognostic biomarker in patients with COAD. In addition, TSPEAR is certainly involved in immune infiltration to regulate CRC development. Therefore, elucidating a more speci c molecular mechanism is of necessity, which can provide novel insights and therapeutic strategies to improve the long-term CRC treatments. Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.
Availability of data and materials Not applicable. * P < 0.01; ** P < 0.001; *** P < 0.0001.    Forest plot of the multivariate Cox regression analysis in colorectal adenocarcinoma.