APOC1 is a Potential Prognostic Biomarker and Correlated With Immune Inltration in ESCC

Purpose Esophageal cancer (EC) is the sixth leading cause of cancer death worldwide. Esophageal squamous cell carcinoma (ESCC) is a predominant subtype of EC. Identifying diagnostic biomarkers for ESCC is necessary for cancer practice. Increasing evidence illustrates that apolipoprotein C-1 (APOC1) participates in the carcinogenesis. However, the biological function of APOC1 in ESCC remains unclear. Patients and methods We investigated the expression level of APOC1 using TIMER2.0 and GEO databases, the prognostic value of APOC1 in ESCC using Kaplan-Meier plotter and TCGA databases. We used LinkedOmics to identify co-expressed genes with APOC1 and perform GO and KEGG pathway analysis. The target networks of kinases, miRNAs and transcription factors were predicted by gene set enrichment analysis (GSEA). The correlations between APOC1 and immune inltration were calculated using TIMER2.0 and CIBERSORT databases. We further performed the prognostic analysis based on APOC1 expression levels in related immune cells subgroups via Kaplan-Meier plotter database. marker in ESCC.


Introduction
As one of the most signi cant diagnosed carcinomas, esophageal cancer (EC) ranks the eighth most prevalent cancer type and sixth leading cause of cancer death, with approximately 482,300 new cases and more than 400,000 deaths annually. 1,2 Esophageal squamous cell carcinoma (ESCC) is a predominant pathological type of EC, accounting for 80% of the world's total EC cancer and up to about 95% in China. 3,4 The warning symptoms of ESCC often become noticeable when the cancer has developed into the advanced or even metastatic stage because of the muscular and expansive nature of the esophageal. 5 Despite the development in the diagnosis and comprehensive therapy, the prognosis of EC patients is still poor with a 5-year survival rate of approximately 20%. 1,6,7 Therefore, it is essential to explore the molecular mechanisms and potential therapeutic targets involved in the occurrence and progression of ESCC.
Apolipoproteins are important lipoproteins that play a crucial role in maintaining the structure of lipoprotein, regulating the interaction between lipoproteins and cellular receptors and modulating the activities of enzymes associated with intravascular lipoproteins metabolism. 8 Apolipoprotein C-1(APOC1), located on chromosome 19, is the smallest of all apolipoproteins (only 6.6 kDa) which can be found within HDL or VLDL. 9 Previous studies have shown that APOC1 contributes to lipid transport and metabolism in serum and plasma by binding to free lipids to form lipoprotein and transport lipids through the lymphatic and circulatory systems. 10 Interestingly, recent studies have shown that the deregulation of APOC1 are detected in some malignancies including breast, gastric and prostate cancer. [10][11][12] Su et al. reported that APOC1 overexpressed in prostate cancer (PCa) and APOC1 silencing inhibited cell proliferation, colony formation, cell cycle progression and promoted apoptosis. 12 Furthermore, APOC1 regulates the cell cycle and mediate the proliferation and apoptosis of the cancer cells by controlling the signal pathways for survivin, p21 and capase-3. 13 Though these ndings suggest that APOC1 may be signi cant in the occurrence and progression of tumor, the expression level and pattern of APOC1 in ESCC are still unknown.
In the present study, we rst reviewed the APOC1 expression in different tumors in TIMER2.0 databases and observed that APOC1 overexpressed in tumor tissues in multiple ESCC datasets. We used the Kaplan-Meier plotter and TCGA databases to analyze the prognostic relationship between APOC1 and ESCC. We also explored the biological functions and signal pathways of APOC1 in ESCC and predicted the underlying targets of APOC1, including kinases, miRNAs and transcription factors. In addition, we investigated the correlation between APOC1 and immune in ltration in ESCC using TIMER2.0 and CIBERSORT.

Data acquisition
The mRNA expression pro les and clinical information of ESCC patients were downloaded form TCGA database (https://cancergenome.nih.gov/). 14 17 We searched for APOC1 co-expressed genes in the ESCC dataset (n = 96) on the basis of the Pearson correlation coe cient. The analysis results were visualized by volcano plots, heat maps, or scatter plots. Gene Set Enrichment Analysis (GSEA) of the LinkInterpreter module was used to perform the analysis of biological process (BP), cell component (CC), molecular function (MF), KEGG pathways, kinase target networks, miRNA target networks and transcription factor target networks. The GSEA program was performed with 500 simulations. If p-value and false discovery rate (FDR) were both less than 0.05, the gene set was considered signi cantly enriched. The relationship between kinase targets and clinical information in ESCC was analyzed using the LinkFinder module. P-values less than 0.05 were considered statistically signi cant.
TIMER2.0 database analysis TIMER2.0 (http://timer.cistrome.org/) is a comprehensive database for systematic analysis of immune in ltration in 32 different cancer types from TCGA. 18 We used TIMER2.0 to analyze the expression levels of APOC1 in different cancers and the correlation between APOC1 expression and immune in ltration (B cells, CD4 + T cells, CD8 + T cells, neutrophils, macrophages and dendritic cells) in esophageal cancer.
CIBERSORT database analysis CIBERSORT (https://cibersort.stanford.edu/) is a deconvolution algorithm on the basis of gene expression which infers 22 human immune cells and evaluates the relative scores for each cell type. 19 In order to explore the relationship between APOC1 expression and immune in ltration, we analyzed 106 samples from the GEO cohort (GSE23400), and 53 tumors and 44 normal samples were eligible after ltering with CIBERSORT p < 0.05.

Statistical analysis
Student's t-test was employed to perform the different expression analysis between ESCC and normal samples. Receiver operating characteristic (ROC) curve was performed to judge the diagnostic value of APOC1 in ESCC. The ESCC patients were divided into two groups using the optimal threshold evaluated by X-tile 3.6.1 on the basis of APOC1 expression levels. 20 The relationship between APOC1 expression and clinicopathlogical parameters was analyzed using Chi-square test, Wilcoxon signed-rank test and Kruskal-Wallis test. The overall survival (OS) between APOC1-high and APOC1-low expression groups was compared by KM analysis and log-rank test. Cox proportional hazards model was used to identify the independent indicators associated with OS. The correlation heatmap of 21 types of in ltrating immune cells was performed by "corrplot" package. The violin plot of the 22 immune cells to illustrate the difference between the ESCC and normal tissues was performed by "vioplot" package. The correlation between APOC1 and in ltrating immune cells was evaluated by Spearman correlation. The GraphPad Prism 8, SPSS 26.0 and R version 4.0 were used for statistical analyses. P < 0.05 was considered statistically signi cant.

Results
The mRNA expressions of APOC1 In order to determine the differences of APOC1 expression in tumor and normal tissues, we reviewed the expression levels of APOC1 in different tumors and normal tissues of multiple cancer types via the TIMER2.0 database. The analysis showed that APOC1 expression was signi cantly higher in most common tumor tissues, such as BLCA ( (Fig. 1A). Then, we further studied the differences of APOC1 expression levels between esophageal squamous cell carcinoma and adjacent normal samples by analyzing four cohorts (GSE23400, GSE23407, GSE45186, GSE70409). The result demonstrated that APOC1 expression was signi cantly increased within paired tumor samples compared to normal esophageal samples ( Fig. 1B-E). We also performed ROC curves to evaluate the diagnostic value of APOC1 upregulation for ESCC using data of GSE23400 (AUC = 0.941, P < 0.001) (Fig. 1F).

Association between APOC1 expression and clinicopathological characteristics
The clinicopathological data of 94 ESCC patients obtained from TCGA database were segmented into low or high APOC1 group based on the optimal threshold of the OS. According to Table 1, there was no signi cant difference in age, BMI, T classi cation, N classi cation, cancer status and alcohol history between two groups (all p > 0.05). However, high APOC1 expression was signi cantly associated with gender, M classi cation, TNM stage (all p < 0.05). Besides, we further studied the relationship between APOC1 expression and these different clinicopathological groups by Wilcoxon signed-rank test or   The GO analysis results illustrated that APOC1 co-expressed genes were primarily related to adaptive immune response, cellular defense response and lymphocyte mediated immunity in the biological process(BP), MHC protein complex, immunological synapse and protein complex involved in cell adhesion in the cellular component(CC), antigen binding, immunoglobulin binding and cytokine receptor activity in the molecular function(MF). The KEGG pathway analysis showed enrichment in the natural killer cell mediated cytotoxicity, phagosome, cell adhesion molecules (CAMs), AMPK signaling pathway, axon guidance and hippo signaling pathway, ect. These above results suggest that a widespread impact of APOC1 on the immune response.

Kinase, miRNA and transcription factor targets of APOC1 in ESCC
To further explore the underlying targets of APOC1 in ESCC, we predicted the kinase, miRNA, transcription factor targets of APOC1 co-expressed genes by GSEA. The top 5 most signi cant targets of kinases, miRNAs and transcription factors were shown in Table 3.

Correlation between APOC1 expression and immune in ltration
Previous analyses showed that tumor in ltration is associated with the prognoses in various cancers. Therefore, we tried to explore the relationship between APOC1 expression and immune in ltration. We  Table 4).  (Fig. 8). These above analyses suggest that high APOC1 may affect prognoses in part due to immune in ltration.

Discussion
APOC1, the smallest apolipoprotein (only 6.6 KDa), is a component of both triglyceride-rich lipoproteins and high-density lipoproteins. 9 Previous studies have shown that APOC1 contributes to lipid transport and metabolism in serum and plasma. 10,13 Interesting, emerging researches highlight the relationship between APOC1 and cancers. The upregulation of APOC1 was previously reported to associated with advanced tumor progression and dismal prognosis in patients with pancreatic cancer, colorectal cancer, lung cancer and papillary thyroid cancer. [21][22][23][24] Nevertheless, the expression level and pattern of APOC1 in ESCC has still not been investigated. In this study, we conducted bioinformatic analysis using public data to explore the potential functions of APOC1 in ESCC.
We initially used TIMER2.0 to review the differential expression of APOC1 between tumor and normal tissues in various types of cancers and observed increased expression of APOC1 in esophageal cancer.
Then, we further found that APOC1 overexpression occurred in multiple cohorts of ESCC and the mRNA expression level of APOC1 displayed favorable diagnostic value for ESCC accessed by ROC curves. In addition, we evaluated the clinical implication of APOC1 overexpression in ESCC and observed that APOC1 overexpression was correlated with gender and TNM stage. Meanwhile, Kaplan-Meier plotter analysis illustrated that ESCC patients with high APOC1 expression had unfavorable OS and RFS (all P < 0.05). We further con rmed the prognostic value of APOC1 by analyzing clinical information of 94 ESCC patients from TCGA. The KM analysis showed that increased expressions of APOC1 were associated signi cantly with poor OS (p < 0.05) and the multivariate analysis indicated that APOC1 overexpression was an independent prognostic factor of unfavorable OS in ESCC. These above results suggest that APOC1 may serve as a valuable diagnostic and prognostic markers of ESCC.
To probe the potential mechanisms of APOC1 upregulation in ESCC, we analyzed APOC1 co-expressed networks using GSEA. GO enrichment analysis revealed that APOC1 was mainly associated with adaptive immune response, cellular defense response and lymphocyte mediated immunity. These results suggest that APOC1 may participate in the immune response of ESCC. Moreover, the KEGG pathway analysis primarily included natural killer cell mediated cytotoxicity, phagosome, AMPK signaling pathway and hippo signaling pathway. AMPK is a crucial energy-sensing enzyme maintaining cellular energy homeostasis. It can be activated by cellular stress which increases the AMP/ATP ratio and leads to the production of metabolic poisons, the progression of hypoxia, glucose starvation, etc. 25 A research reported that metformin exerted its anti-proliferative effects on ESCC cells through inducing the activation of AMPK pathway, suggesting that AMPK pathway may play a crucial role in the inhibition of the growth of ESCC. 26 The primary functions of Hippo pathway are restricting tissue growth and modulating cell proliferation, differentiation and migration. Recent studies suggest that deregulation of hippo pathway plays an important role in cancer initiation and progression. 27 Gao et al reported that miR-31 could promote ESCC tumorigenesis by inhibiting LATS2 expression via the hippo pathway. 28 These above researches are consistent with our results, suggesting that APCO1 may participate in tumorigenesis and progression of ESCC through these signaling pathways.
To further gain insight into the regulators potentially responsible for APOC1 overexpression, we investigated networks of kinases, miRNAs and transcription factors. We observed that APOC1 in ESCC was related to the network of kinases including LCK, SYK, LYN, ITK and HCK. Additionally, LYN and HCK showed signi cantly overexpressed in ESCC tissues in the GSE23400 dataset (all p < 0.05). And HCK overexpression was associated with the poor prognosis and pathological stage in ESCC patients (all p < 0.05). In fact, HCK is the important paralog of LYN. HCK plays an important role in the regulation of innate immune responses, including neutrophil, monocyte, macrophage and mast cell functions, phagocytosis, cell survival and proliferation, cell adhesion and migration. Excessive HCK activation enhances cell proliferation and survival by associating with oncogenic fusion proteins and functionally interacting with receptor tyrosine kinases. 29,30 Consistent with previous reports, we speculate that APOC1 may regulate immune system process via HCK kinase in ESCC and further studies are needed to verify this hypothesis.
IRF family, a broad class of cytokines elicited on challenge to the host defense, are important for mobilizing immune responses to pathogens. 31 Moreover, statistical data indicated that IRF-2 expression was tightly correlated with progression of ESCC. 32 Reduction of the ratio of IRF-1/IRF-2 might lead to the enhancement of tumorigenicity of ESCC cells. 33 STAT3 plays important roles in the progression of various cancers by regulating the proliferation, invasion, angiogenesis and immune surveillance evasion. 34,35 Previous researches have shown that the STAT3 pathway was activated in some ESCC cells and STAT3 overexpression indicated the poor prognosis of ESCC patients who had undergone curative resection. 36 Our study suggests that IRF and STAT3 are potential regulators of APOC1 and that APOC1 may act through these factors to regulate the immune response of ESCC.
Previous research illustrated that miRNA which normally involve in post-transcriptional regulation of gene expression can contribute to human carcinogenesis. 37 Thus, we tried to explore the potential regulatory miRNAs of APOC1. The miRNAs associated with APOC1 in our study participate in the process of carcinogenesis. In the current study, miR-26a and miR-26b inhibit ESCC cell proliferation through suppression of c-MYC pathway. 38 Previous studies showed that the upregulation of miR-519 enhanced radiosensitivity of ESCC cells and facilitated ESCC cell apoptosis via targeting PI3K/AKT/mTOR signaling pathway. The low expression level of miR-519 indicated poor prognosis of ESCC patients. 39 The expression of miR-202 was reported to aberrantly decreased in ESCC and the down-regulation of miR-202 was associated with the metastasis of tumor. 40 Consistent with above literature evidence, deregulation of these miRNAs may partly contribute to APOC1 overexpression in ESCC.
Tumor-in ltrating immune cells (TIICs) are crucial parts of the tumor microenvironment, which are associated with patient outcome and tumor behavior. 41  Analyses on the role of immune in ltration in human cancers typically focus on T cells, and the CD8 + T cells are one of the major anti-tumor immune cells in the tumor microenvironment. It has been reported that CD8 + T cells cooperate with CD4 + T cells to promote better prognosis of ESCC patients. 42 Tregs are heterogeneous and have multiple context-dependent functions that are still not well-characterized. It may play a dual role in carcinogenesis, initially inhibiting in ammation that leads to tumorigenesis, but later suppressing anti-tumor immune responses. 43 Besides T cells, the innate immune system including NK cells and macrophages also role importantly in the tumor microenvironment. Several studies have shown that the in ltration of NK cells signi cantly increases in ESCC. And IL-6 or IL-8 secreted by primary ESCC cells impairs the function of NK cells via the STAT3 signaling pathway. 44 Tumor-associated macrophages (TAMs) have been reported to play a crucial role in in ammatory tumor microenvironment.
TAM in ltration is related to poor responses to chemotherapy and unfavorable prognosis in ESCC. 45 The above research ndings combined with our study indicate that the in ltrating immune cells play crucial roles in ESCC and should be the highlight of future studies. Moreover, we analyzed the correlation between the immune cells in ESCC. Our results showed that M1 macrophages cells were positively correlated with activated NK cells, and M0 macrophages cells were negatively related to CD8 T cells and neutrophils. The potential mechanisms of these correlations require further studies.
To uncover the association between APOC1 overexpression and the immune in ltration in ESCC, we preformed the correlation analysis. Our results indicated that APOC1 was signi cantly positively Furthermore, we performed the prognostic analysis of APOC1 expression levels in ESCC based on immune cells. The results showed that high APOC1 expression level had an unfavorable prognosis in the decreased B-cells, decreased CD4 + memory T-cells, decreased CD8 + T-cells, and enriched mesenchymal stem cells subgroups. These analyses suggest that the overexpression of APOC1 may partly affect the prognoses of ESCC patients due to immune in ltration, especially the decrease of CD8 + T cells.
In summary, this present study promotes our understanding of the association between APOC1 and ESCC and suggests that the upregulation of APOC1 may have an impact on the unfavorable prognosis of ESCC through the mechanism of immune in ltration. But this study also has some limitations. The study was preformed based on the data downloaded from public databases and there were few survival data of ESCC patients. It is essential to conduct additional experiments to verify our results and further clarify the role of APOC1 in ESCC.

Conclusion
In summary, APOC1 expression increased signi cantly in ESCC compared to normal esophageal tissues.
The APOC1 overexpression was related to dismal prognosis and in ltrating immune cells. In addition, the prognostic analysis of APOC1 expression levels in ESCC based on immune cells indicated that elevated APOC1 may partly affect the prognoses of ESCC patients due to immune in ltration. Therefore, APOC1 may serve as a potential diagnostic and prognostic biomarker for esophageal squamous cell carcinoma. Further pathogenesis mechanisms and clinical studies are needed to prove the absoluteness of these ndings.

Declarations
Ethics approval and consent to participate Not applicable

Consent for publication
Not applicable Availability of data and materials The datasets generated during and/or analysed during the current study are available in the TCGA repository, https://cancergenome.nih.gov/ and GEO repository, https://www.ncbi.nlm.nih.gov/geo/.

Competing interests
The authors declare that they have no competing interests.

Funding
The present study was supported by the grants from the Guangdong Provincial Science and Technology Research Program (2019A141405016 and 2017B020209003).
Authors' contributions JX and ML designed this study. JX, ZW, ZL, XR and LX performed data collection and analysis. JX, YL, RH and AL contributed to manuscript preparation. SL, XL, ZH revised the manuscript. All authors read and approved the nal manuscript.