CTHRC1 Is a Prognostic Biomarker and Correlated With Immune Infiltrates in Kidney Renal Papillary Cell Carcinoma and Kidney Renal Clear Cell Carcinoma

Kidney renal clear cell carcinoma (KIRC) and kidney renal papillary cell carcinoma (KIRP) are the most common RCC types. RCC has high immune infiltration levels, and immunotherapy is currently one of the most promising treatments for RCC. Collagen triple helix repeat containing 1 (CTHRC1) is an extracellular matrix protein that regulates tumor invasion and modulates the tumor microenvironment. However, the association of CTHRC1 with the prognosis and tumor-infiltrating lymphocytes of KIRP and KIRC has not been reported. We examined the CTHRC1 expression differences in multiple tumor tissues and normal tissues via exploring TIMER, Oncomine, and UALCAN databases. Then, we searched the Kaplan-Meier plotter database to evaluate the correlation of CTHRC1 mRNA level with clinical outcomes. Subsequently, the TIMER platform and TISIDB website were chosen to assess the correlation of CTHRC1 with tumor immune cell infiltration level. We further explored the causes of aberrant CTHRC1 expression in tumorigenesis. We found that CTHRC1 level was significantly elevated in KIRP and KIRC tissues relative to normal tissues. CTHRC1 expression associates with tumor stage, histology, lymph node metastasis, and poor clinical prognosis in KIRP. The CTHRC1 level correlates to tumor grade, stage, nodal metastasis, and worse survival prognosis. Additionally, CTHRC1 is positively related to different tumor-infiltrating immune cells in KIRP and KIRC. Moreover, CTHRC1 was closely correlated with the gene markers of diverse immune cells. Also, high CTHRC1 expression predicted a worse prognosis in KIRP and KIRC based on immune cells. Copy number variations (CNV) and DNA methylation might contribute to the abnormal upregulation of CTHRC1 in KIRP and KIRC. In conclusion, CTHRC1 can serve as a biomarker to predict the prognosis and immune infiltration in KIRP and KIRC.


INTRODUCTION
Kidney cancer is among the top ten causes of cancer-related deaths. There are various subtypes of kidney cancer based on mixed histology, clinical course, and gene course. Renal cell carcinoma (RCC) is the most important type of kidney cancer (1). Besides, kidney renal clear cell carcinoma (KIRC) and kidney renal papillary cell carcinoma (KIRP) account for almost 95% of all renal cell cancers (2). The diagnosis, examination, surgery, and drug therapy of RCC have been advanced. However, its clinical outcome remains unsatisfactory (3,4). RCC is a heterogeneous tumor that requires useful molecular markers suitable for personalized therapy (5,6).
During cancer tumorigenesis and progression, tumor cells are affected by the tumor-infiltrating immune cells (7,8). The immune invasion of the tumor is closely associated with the clinical prognosis of RCC. Precious studies indicated that tumorinfiltrating macrophages, regulatory T Cells (Treg cells), and CD8 + T cells influence RCC treatment outcomes (9)(10)(11). Besides, M1 macrophages are associated with better prognosis, while M2 macrophages predict poor outcome in KIRP. Immunoregulatory molecules CTLA-4 and LAG-3 associate with a poor prognosis in KIRC, while IDO1 and PD-L2 correlate with a poor prognosis in KIRP (12). These findings demonstrate that tumor infiltration of immune cells may be a useful drug target that improving clinical outcomes.
Tumor microenvironment comprises infiltrating immune cells, stromal cells, extracellular matrix, and tumor cells. Several studies have reported that tumor-infiltrating lymphocytes have different vital roles in tumor development. For example, tumorassociated macrophages (TAM) promotes cancer metastasis (13). CTHRC1 is a 30 kDa secreted protein, which is highly expressed in cartilage, developing bones, and myofibroblasts during skin wound healing and solid tumors (14,15). Previous studies indicate that CTHRC1 promotes tumor cell progression via influencing specific pathways in various cancer types. CTHRC1 is elevated in cervical carcinoma and promotes metastasis through the Wnt/PCP pathway. In contrast, CTHRC1 modulates aggressiveness via GSK-3b/b-catenin pathway in human non-small cell lung cancer (16). Therefore, CTHRC1 is suggested to play an essential role in cancer progression. Current studies have found CTHRC1 function in modulating the tumor microenvironment via the E6/E7-p53-POU2F1 axis or focal adhesion kinase signal pathway (17,18). In endometrial cancer, CTHRC1 promotes M2-like macrophage recruitment and myometrial invasion via the integrin-Akt signaling pathway (19). Thus, CTHRC1 has multifaceted functions in the tumor microenvironment. However, the underlying mechanisms of CTHRC1 in KIRP and KIRC progression and tumor-infiltrating lymphocytes remains unclear.
In this study, we used Oncomine, TIMER, UALCAN datasets, and Kaplan-Meier plotter web to analyze CTHRC1 expression and its association with the prognosis. Furthermore, we used the TIMER web resource and TISIDB database to analyze the correlation between CTHRC1 and tumor-infiltrated immune cells in the tumor microenvironment. Besides, we further explored the molecular mechanisms of CTHRC1 dysregulation, such as analysis of the CNV, DNA methylation, and somatic cell mutations. Our findings underline the vital role of CTHRC1 in KIRP and KIRC prognosis. Also, we provide an underlying mechanism of CTHRC1 expression in potentially regulating the infiltration of immune cells, partly affecting the prognosis of KIRP and KIRC.

Oncomine Database Analysis
Oncomine database (https://www.oncomine.org/resource/main. html) integrates literature and databases of tumor microarray results and is mainly used for gene expression analysis, coexpression analysis, enrichment analysis, interaction networks (20). We used the Oncomine database to analyze CTHRC1 expression in various cancer types.

TIMER Database Analysis
TIMER web server (https://cistrome.shinyapps.io/timer/) is a website for comprehensive analysis of gene expression and tumor-infiltrating immune cells of diverse cancer types. This web assesses the abundances of six tumor-infiltrating cells (B cells, CD4 + T cells, CD8 + T cells, neutrophils, macrophages, and dendritic cells), using the TIMER algorithm (21). TIMER website also enables the user to explore gene expression in tumor tissues and normal tissues in multiple cancers. We used the TIMER website to analyze the differential expression of CTHRC1 in tumor and normal tissues in various cancers. We evaluated the correlation of CTHRC1 with 6 tumor immune infiltrating cells and molecular markers of 16 immune cells. We also used this web to explore the relationship between immune infiltrating cells and gene expression that affects clinical prognosis in KIRP and KIRC. The levels of gene expression were expressed as log2 RSEM.

UALCAN Database Analysis
UALCAN database (http://ualcan.path.uab.edu/index.html) is available for online analysis of differential gene expression in cancer and normal tissue from the TCGA RNA sequencing data and clinical data of 31 malignancies (22). Besides, this website provides survival prognosis data based on gene expression differences in 31 cancer types. This study used the UALCAN database to validate the analysis results of the Oncomine database, and furtherly determined the correlation between CTHRC1 gene expression and clinical features. Differences at p<0.05 were considered statistically significant.

Kaplan-Meier Plotter Database Analysis
Kaplan-Meier plotter (http://kmplot.com/analysis/) (23) is an open, intuitive portal tool for prognostic analysis. It contains 54,675 genes survival data from 10,461 cancer samples. Kaplan-Meier plotter database was used to assess the relationship between clinic outcomes and CTHRC1 expression in different cancers. We performed a prognostic analysis based on CTHRC1 expression levels in relevant immune cell subgroups using this web. We calculated hazard ratios (HRs) of 95% confidence intervals (CIs) and the log-rank p-value.

TISIDB
TISIDB database (http://cis.Hku.hk/TISIDB/) is a portal for analyzing tumor and immune cell interactions that integrates multiple heterogeneous data types (24). We analyze the correlation between CTHRC1 expression and tumor-infiltrating lymphocytes via this platform.

UCSC Xena
UCSC Xena database (http://xena.ucsc.edu/) is a genome-related database, which brings approximately 200 public databases together, including TCGA, ICGC, TARGET, GTEx, CCL, etc. (25). The database is available to examine copy number and methylation, somatic mutation, gene expression, protein expression. This web also provides clinical information such as patient treatment and survival.

Statistical Analysis
The CTHRC1 expression was analyzed via the Oncomine, TIMER, and UALCAN database. Survival curves were generated using the Kaplan-Meier plotter database and R project using "survival" packages. We used Spearman's correlation analysis to evaluate the correlation of gene expression in the TIMER. p<0.05 were considered statistically significant.

The Collagen Triple Helix Repeat Containing 1 mRNA Expression in Different Cancers
We analyzed the mRNA expression of CTHRC1 using the Oncomine database. The results showed that CTHRC1 was significantly high in various cancer tissues, compared to normal tissues ( Figure 1A). Then, the mRNA level in the TIMER database was determined. We found that CTHRC1 mRNA expression was significantly high in most human tumors, especially in KIRP and KIRC, compared with the corresponding normal tissues ( Figure 1B). These results showed that CTHRC1 was highly expressed in various cancers. Besides, we used the UALCAN database to validate the findings in Oncomine and TIMER web and reported higher expression of CTHRC1 in KIRP and KIRC tissues than in normal tissues ( Figures 1C, G). Notably, CTHRC1 expression was associated with tumor histology, stage, lymph node metastasis in KIRP ( Figures 1D-F). Meanwhile, the high CTHRC1 level in KIRC was related to lymph node metastasis high grade and stage ( Figures 1H-J).

Prognostic Significance of Collagen Triple Helix Repeat Containing 1 Expression in Human Cancers
We investigated the Kaplan-Meier plotter database for the prognostic significance of CTHRC1 expression in human cancers. High levels of CTHRC1 predicted poor prognostic in KIRP (Figures 2A, B), KIRC ( Figures 2C, D), THCA ( Figures  2E, F), and LUAD ( Figures 2G, H). As Kaplan-Meier plotter analyzes only OS and RFS value, we assessed the multiple clinical prognostic value of CTHRC1 in a variety of cancers by R project using "survival" packages. Forest plot showed CTHRC1 as a risk factor of different prognosis in KIRP and KIRC ( Figure 3). Besides, we generated the Kaplan-Meier plot, which showed that high expression of CTHRC1 had a poor prognosis in KIRP (Supplementary Figures S1A-D) and KIRC (Supplementary Figures S1E-H). These findings indicated that CTHRC1 is a hazard for predicting worse prognostic in KIRP and KIRC.

Correlation of Collagen Triple Helix Repeat Containing 1 Expression With Clinical Characteristics of Kidney Renal Papillary Cell Carcinoma and Kidney Renal Clear Cell Carcinoma Patients
Then, we investigated the association of CTHRC1 expression with different clinical characteristics of KIRP and KIRC using the Kaplan-Meier Plotter database ( Table 1)

Collagen Triple Helix Repeat Containing 1 Expression is Correlated With Immune Infiltration in Kidney Renal Papillary Cell Carcinoma and Kidney Renal Clear Cell Carcinoma
Tumor-infiltrating lymphocytes can independently be used to predict sentinel lymph node status and prognosis in cancers    used TISIDB database to furtherly explore the relationship between CTHRC1 level and 28 tumor immune infiltrating cell subtypes. These results showed that CTHRC1 is associated with twenty-seven immune cell subtypes in KIRP ( Figure 4D,  Figure  4E). In addition, CTHRC1 showed a positive correlation with 24 immune cell ( Figure 4D,  Figure 4F). These results strongly implicate that CTHRC1 could serve as a major tumor immune infiltration regulator in KIRP and KIRC.

Collagen Triple Helix Repeat Containing 1 Expressions Were Correlated With Immune Cell Type Markers
We assessed the correlation between CTHRC1 expression and tumor-infiltrating immune cell gene marker levels in KIRP, KIRC, and CESC tissues by exploring the TIMER database.
Our results showed that the CTHRC1 level in KIRP tissues was strongly associated with immune markers of B cells, monocytes, and dendritic neutrophils, DCs, CD8 + T cells, T helper, Tregs, and T exhaustion cells. Moreover, CTHRC1 in KIRC positively correlates with the marker genes of B cells, macrophages, and neutrophils. However, there were only ten significant markers associated with CTHRC1 in CESC.

Prognostic Potential of Collagen Triple Helix Repeat Containing 1 Expressions in Different Tumors Based on Immune Cells
This study showed that the CTHRC1 level was associated with the immune infiltration of KIRP and KIRC. Also, upregulated CTHRC1 has a worse prognosis in KIRP and KIRC patients. Thus, we propose a hypothesis that CTHRC1 may affect the prognosis of KIRP and KIRC patients partly through immune infiltration.
We perform Kaplan-Meier plotter analyses of CTHRC1 expression in KIRP and KIRC following B cells, CD4+ memory T cells, CD8 + T cells, macrophages, NK T cells, Treg  (Figures 5L, N). The above analysis suggested that immune infiltration may, in part, affect high CTHRC1 expression prognosis of KIRC and KIRP patients.

Mutation, Copy Number Variation, and Methylation Analysis of Collagen Triple
Helix Repeat Containing 1 CTHRC1 expression was significantly elevated in KIRP and KIRC. We assessed the cause of elevated CTHRC1 levels. DNA methylation, gene mutation, CNV was critically involved in genetic and epigenetic regulation and were highly associated with the process of cancers. We verified the DNA methylation, gene mutation, CNV levels of the CTHRC1 in KIRP and KIRC via the UCSC Xena database. The heatmap indicates that the expression of CTHRC1 mRNA was correlated with CNV and DNA methylation, but not with a somatic mutation in KIRP ( Figure 6A) and KIRC ( Figure 6B). The human disease methylation database was used to further validate the lower methylation level in KIRP (Supplementary Figures S2A, B) and KIRC (Supplementary Figures S2C, D), compared to normal tissues. Therefore, we suggested that CNV and DNA methylation might contribute to the elevated level of CTHRC1 in KIRP and KIRC, respectively.

DISCUSSION
CTHRC1 is an extracellular matrix protein that regulates tumor metastasis and the extracellular microenvironment. In this study, we analyzed CTHRC1 expression, prognostic value, genetic variations, and correlation with tumor immune cell infiltration in KIRP and KIRC for the first time.
In this study, we found that CTHRC1 expression was highly elevated in KIRP and KIRC, compared to normal tissues. Moreover, CTHRC1 expression has associations with tumor histology, stage, lymph node metastasis in KIRP (Figures 1D-F). Meanwhile, the high CTHRC1 level was related to lymph node metastasis, high grade, and stage ( Figures 1H-J). These results suggest that CTHRC1 plays an important role in the progression and metastasis of KIRP and KIRC. Our findings are consistent with previous researches. CTHRC1 was elevated in some tumor tissues and associated with clinicopathological features, including late T stage, lymph nodal metastasis, and TNM staging (17,29).
Results from survival analysis showed that high CTHRC1 expression was associated with poor OS, DSS, DFI, and PFI ( Figure S1) in KIRP and KIRC, consistent with previous findings CTHRC1 affects tumor growth and invasion and leads to a poor prognosis (30,31). Ni et al. (32) reported that CTHRC1 promotes metastasis through an epithelial-mesenchymal transformation in colorectal cancer, resulting in a poor prognosis. Our results strongly indicate that CTHRC1 can be used as a prognostic biomarker for KIRP and KIRC.
CTHRC1 was previously reported to regulate tumor microenvironment. Through correlation analysis, we reported that CTHRC1 expression is associated with several immune infiltrating cells in KIRP and KIRC ( Figure 4, Table 3). These results suggest that CTHRC1 is involved in the regulation of tumor immune cells. Lee et al (33). study revealed that CTHRC1 recruits Tie2-expressing monocytes into tumor tissues by activating ERK-dependent AP-1 to promote angiogenesis. Besides, the use of CTHRC1 antibodies reduced tumor burden and TEMs infiltration in tumor tissue in xenograft mouse models. Another study demonstrated that CTHRC1 expression has a role in tumor-associated macrophages infiltration by upregulating fractalkine chemokine receptor (CX3CR1) expression (19). Our analysis had the effect of mutual authentication with the results of this researches.
We further analyzed the immunotype markers in KIRP and KIRC. After cell purity correction, CTHRC1 was positively correlated with many immune cell makers in KIRP and KIRC ( Table 4). The results further imply that CTHRC1 is associated with immune infiltration in KIRP and KIRC. Besides, our results suggest that CTHRC1 can potentially modulate Tregs and results in T cell exhaustion. Notably, increased CTHRC1 level was positively associates with Treg and T cells exhaustion markers, such as FOXP3. FOXP3 is a valid target for identifying Treg in the tumor microenvironment and contributes significantly to Treg cells differentiation and mediated tumor immune escape (34). There was a significant correlation between CTHRC1 level and several T helper cells (Th1, Th2, Tfh, and Th17) markers in KIRP. These connections may indicate the underlying mechanisms for CTHRC1 regulation of T cell function in KIRP. Therefore, it was potentially related to the poor prognosis of KIRP and KIRC by recruiting and regulating immune cells.
Through the Kaplan Meier-Plotter database analysis, high expression levels of CTHRC1 enriched in a variety of immune cells cohort of KIRP and KIRC had a worse prognosis ( Figure 5). Tregs can suppress anti-tumor responses, leading to tumor immune escape (35). DC can promote tumor metastasis by increasing Treg cells and decreasing the cytotoxicity of CD8 + T cells (36). Myeloid origin suppressor cells (MDSC) contact with T cells, which rapidly depletes arginine from the microenvironment and leads to tumor-mediated immune escape (37). Previous studies also have proven that the proportion of macrophages, CD8 + T cells, Tregs, and MDSC in RCC patients correlates with poor prognosis (9-11, 38, 39). These results may explain that high expression of CTHRC1 partly affects the prognosis of KIRP and KIRC patients through immune infiltration.
Genetic and epigenetic phenomena play an essential role in regulating gene expression (40). In this study, we found that CTHRC1 expression was strongly correlated with DNA methylation and CNV and not with somatic mutations. DNA methylation is the most common epigenetic phenotype that usually acts as a transcriptional repressor and plays an essential role in tumor progression (41,42). Besides, CTHRC1 is reported to be upregulated by promoter demethylation in gastric cancer and down-regulated by hypermethylation in hepatocellular carcinoma (43). DNA copy number variation, including gene amplification, gain, loss, and deletion. And CNV influences the gene expression in carcinogenesis (44). Wang et al. (45) presented the evidence that 5-aza-2'deoxycytidine (the demethylating agent) can restore CTHRC1 expression, and TGF-b1 led to an increase in levels of CTHRC1 mRNA and protein. Therefore, DNA hypomethylation and CNV may be a cause for CTHRC1 upregulated in KIRP and KIRC.
In conclusion, the upregulated CTHRC1 is strongly associated with clinicopathological features, poor prognosis, and immune cell infiltration. DNA methylation and copy number variation may attribute to CTHRC1 upregulated. Furthermore, our study provides a new mechanism that CTHRC1 may affect the prognosis of KIRP and KIRC through tumor immune infiltration. Therefore, this study offers insights for further studies on tumor immunotherapy of KIRP and KIRC. The current study is the preliminary part of a larger study, including validation in a study population prospectively enrolled. Undoubtedly, we will make further validation when there are available independent datasets and perform experiments in the future.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article.