EXOC3L1: A Novel Prognostic Biomarker Correlated with Immune Infiltration in Esophageal Squamous Cell Carcinoma

Background

Esophageal carcinoma (EC) is the eighth most common cancer and the sixth most common cause of cancer-related mortality in the world [1], including esophageal adenocarcinoma (EAC) and esophageal squamous cell carcinoma (ESCC) [2,3]. China accounts for more than half of the world’s EC patients and more than 90% of those are ESCC [4,5]. Despite the developments in multi-disciplinary team therapy strategies, such as surgery, chemotherapy, radiotherapy, and immunotherapy, the 5-year overall survival (OS) rate remains unsatisfactory, being less than 30% [6–8], even in developed countries [9]. The low survival rate of ESCC is partly due to the lack of effective therapeutic targets [10]. Although many biomarkers have been explored in ESCC, such as epidermal growth factor receptor (EGFR), vascular endothelial growth factor receptor (VEGFR), human epidermal growth factor receptor-2 (HER2), and others, they failed to be effective therapeutic targets for ESCC [11,12]. Since the pathogenesis of ESCC is still unclear, it is necessary to identify novel biomarkers for early diagnosis, oncology therapy, and prognosis prediction.

Exocyst complex component 3-like 1 (EXOC3L1) is a protein-coding RNA located on chromosome 16 (16q22.1), also known as EXOC3L, which was originally isolated and reported as an isoform of Sec6 to regulate insulin secretion in pancreatic β cells. Previous studies suggested that EXOC3L1 is associated with insulin secretion, high-density lipoprotein (HDL) concentration [13], and spontaneously-induced apoptosis [14]. However, few studies on EXOC3L1 have been reported, EXOC3L1 has not been studied in any cancers so far, and the association between EXOC3L1 and ESCC has not yet been characterized, and these topics warrant further investigation. The purpose of the present study was to investigate the relationship between EXOC3L1 and ESCC.

Material and Methods

DATA SOURCE:

The RNA-seq data were downloaded from the University of California Santa Cruz (UCSC) Xena (https://xenabrowser.net/datapages/), including the Genotype-Tissue Expression (GTEx) data and The Cancer Genome Atlas (TCGA) data, a total of 652 normal samples and 82 ESCC samples were obtained. In addition, GSE53625 and GSE161533 datasets were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/gds) to further confirm the expression of EXOC3L1 in ESCC. The use of the data was in compliance with the published guidelines of UCSC Xena, TCGA, and GEO databases; therefore, ethics approval and informed consent of the patients were not required. This study was performed in accordance with the principles of the Declaration of Helsinki.

DIFFERENTIALLY EXPRESSED GENES (DEGS) ANALYSIS:

EXOC3L1-related DEGs were identified in 82 ESCC samples by the DESeq2 R package [15] using Wilcoxon rank-sum test, and the cut-off value was 50%. The DEGs met the requirements of |logFC| ≥1.5, and P values <0.05 were considered as significantly different DEGs. The relationship between EXOC3L1 and the top 10 upregulated genes were displayed by a heatmap, and the significantly different DEGs were used for subsequent enrichment analysis and protein–protein interaction (PPI) network analysis.

FUNCTIONAL ENRICHMENT ANALYSIS:

In this study, the significantly different DEGs were chosen to further perform functional enrichment analysis by using clusterProfiler R package [16], which included Gene Ontology (GO) functional analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and gene set enrichment analysis (GSEA).

PPI NETWORK ANALYSIS:

The PPI network of EXOC3L1-related DEGs was constructed using the search tool for the retrieval of interacting genes (STRING) online database (https://cn.string-db.org/) with an interaction threshold of 0.4, the interaction network was displayed by Cytoscape software (version 3.8.2) [17], and the most significant module in the PPI network was extracted by the MCODE plug-in.

IMMUNE INFILTRATION ANALYSIS:

The immune infiltration analysis of EXOC3L1 in ESCC was performed using the GSVA R package [18] using the single-sample GSEA (ssGSEA) method, the infiltration abundance of 24 types immune cells previously described [19] were quantified, and the immune infiltration difference between the high EXOC3L1 expression and low EXOC3L1 expression groups was taken into consideration as well.

STATISTICAL ANALYSIS:

All statistical analyses were performed using R software (v.3.6.3). The Wilcoxon rank-sum test was used to analyze the expression of EXOC3L1. Survival analysis was performed with the Kaplan-Meier method, and the difference between survival curves was compared with a log-rank test. Univariate and multivariate Cox regression analyses were performed to estimate the factors that contributed to the prognosis of ESCC. The Spearman test was applied to analyze the correlation between EXOC3L1 expression and immune cell infiltration. All statistical tests were 2-tailed and P values <0.05 were considered statistically significant.

Results

ABNORMALLY HIGH EXPRESSION OF EXOC3L1 IN ESCC:

The expression difference of EXOC3L1 was assessed in 652 normal samples from GTEx and 82 ESCC samples from TCGA. The results showed that EXOC3L1 was highly expressed in ESCC (Figure 1A). A receiver operating characteristic (ROC) curve was used to evaluate the discriminative power of EXOC3L1 expression to distinguish tumor tissues from normal tissues, with area under curve (AUC) of EXOC3L1 0.812 (Figure 1B), suggesting that EXOC3L1 could serve as an ideal biomarker for the diagnosis of ESCC. In addition, the expression of EXOC3L1 was assessed in the GEO database, showing that EXOC3L1 was significantly overexpressed in tumor tissues in the GSE53625 dataset (Figure 1C) and GSE161533 dataset (Figure 1D).

EXOC3L1 EXPRESSION WAS CORRELATED WITH THE PROGNOSIS OF ESCC:

The associations between EXOC3L1 expression and clinicopathological characteristics were investigated (Table 1). The results demonstrated that the expression of EXOC3L1 was correlated with the N stage (Figure 2A) and the pathologic stage (Figure 2B). To explore whether the expression of EXOC3L1 has an effect on the prognosis of the patients with ESCC, the associations between EXOC3L1 expression and OS, disease-specific survival (DSS), and progression-free interval (PFI) were evaluated by Kaplan-Meier method with log-rank test, showing that ESCC patients with high expression of EXOC3L1 had a worse prognosis than patients with low expression (Figure 2C). Next, univariate and multivariate Cox proportional hazards regression analyses were used to evaluate the factors influencing the OS, showing that pathologic stage and EXOC3L1 expression were independent predictive factors for OS in ESCC (Table 2).

IDENTIFICATION OF DEGS IN ESCC:

The DEGs between EXOC3L1 high-expression and low-expression groups were identified by the DSEeq2 R package in the RNA-sequencing data of ESCC in TCGA database (Figure 3A). The DEGs met the requirements of |logFC| ≥1.5, and P values <0.05 were considered as significantly different DEGs; 556 genes were obtained after filtering, covering 290 upregulated and 266 downregulated genes. The relationship between EXOC3L1 and the top 10 upregulated genes were displayed by a heatmap (Figure 3B), and the significantly different DEGs were used in functional enrichment analysis and then in the PPI network.

FUNCTIONAL ENRICHMENT ANALYSIS AND PPI NETWORK ANALYSIS BASED ON DEGS:

To explore the functions of the statistically significant DEGs in ESCC, conducted GO and KEGG functional enrichment analyses was performed using the clusterProfiler R package. GO analysis included biological process (BP), cellular composition (CC), and molecular function (MF). The results indicated that BP might be associated with “humoral immune response”, CC might be associated with “immunoglobulin complex”, and MF might be associated with “immunoglobulin receptor binding” (Figure 3C). KEGG pathway analysis showed that the “cytokine-cytokine receptor interaction” pathway might be involved in the biological process of ESCC (Figure 3D).

To obtain the interactions of the statistically significant DEGs, a PPI network was constructed using the STRING online database, and the interaction threshold was 0.40. Subsequently, the interaction was displayed with Cytoscape software (Figure 4A), and the most significant module in the PPI network was extracted by the MCODE plug-in (Figure 4B).

Additionally, GSEA was also performed to reveal the EXOC3L1-related pathways in ESCC, and the pathways met the requirements of P value <0.05, FDR <0.25, and |NES| ≥1 and were considered as significant different pathways. Among the significant different pathways, several were selected to exhibit, including intestinal immune network, T cell receptor signaling, Natural Killer T cell pathway, and cancer immunotherapy by PD1 blockade (Figure 5A–5D).

CORRELATION BETWEEN EXOC3L1 EXPRESSION AND IMMUNE INFILTRATION:

Finally, the correlations between the expression of EXOC3L1 and immune cell infiltrations were assessed by Spearman correlation test using ssGSEA method (Figure 6A). The findings revealed that the expression of EXOC3L1 was positively correlated with the abundance of active dendritic cell (aDC), CD8+ T cells, cytotoxic cells, eosinophils, immature dendritic cell (iDC), CD56dim Natural killer cells, T cells, Th1 cells, and regulatory T cells (Figure 6B, 6C).

Discussion

ESCC is one of the most aggressive cancers all over the world and the incidence continues to rise, despite the advances in treatment technology, the prognosis is still gloomy, and the treatment of ESCC is a worldwide challenge. Although numerous biomarkers have been investigated in ESCC, their reliability remains controversial, thus, it is essential to discover new biomarkers to perform early diagnosis, oncology therapy, and prognosis prediction of ESCC. EXOC3L1 has not yet been investigated in ESCC.

ESCC is rich in lymphatic vessels and is very susceptible to lymph node metastasis [20], which is the main mode of metastasis in ESCC, and more than half of patients already have multiple lymph node metastasis at the time of initial diagnosis [21]. The presence of lymph node metastasis means a higher stage and even loss of the opportunity for radical surgery, with a gloomy prognosis [22]. Therefore, understanding the mechanism of ESCC lymph node metastasis is a vital step in research on ESCC. The present study is the first to show that EXOC3L1 expression is correlated with the N stage in ESCC, suggesting that EXOC3L1 is closely related to lymph node metastasis and plays a crucial role in the mechanism of lymph node metastasis in ESCC. This important discovery helps to clarify the mechanism of lymph node metastasis in ESCC, and may provide a new target for the treatment of ESCC.

In addition, the relationship between EXOC3L1 and ESCC prognosis has not been reported previously, and the present study demonstrated that EXOC3L1 was significantly elevated in ESCC samples, which indicated poorer clinical outcomes, and the OS, DSS, and PFI were significantly worse among patients with high EXOC3L1 expression than in those with low expression. Multivariate Cox regression analysis further demonstrated that EXOC3L1 could act as an independent prognostic indicator of ESCC. There have been few recent breakthroughs in targeted therapy of ESCC [7,8,10], and the lack of effective targeted drugs is one of the main reasons for the poor prognosis of ESCC [23]. The present study found that EXOC3L1 was significantly increased in ESCC and was significantly related to the prognosis of ESCC patients, which means that EXOC3L1 has the potential to be a novel target for ESCC. Further in-depth research on the pathogenesis of EXOC3L1 in ESCC is expected to provide new insights for the treatment of ESCC.

It is well known that the immune microenvironment plays a crucial role in the occurrence and development of tumors [24,25], the changes in the immune microenvironment may be the initiating factors for many pathophysiological, and the mechanism of immune microenvironment is extremely complicated [26,27]. In the present study, GO analysis, KEGG analysis, and GSEA analysis all suggested that EXOC3L1 is related to the immune function of ESCC, which was consistent with the immune infiltration analysis, showing that high expression of EXOC3L1 was positively correlated with an increase in immune infiltration levels in activated dendritic cells, CD8+ T cells, cytotoxic cells, eosinophils, immature dendritic cells, natural killer CD56dim cells, T cells, Th1 cells, and regulatory T cells. Although some previous studies reported that most immune cells have a positive impact on clinical outcomes [28–32], this is not totally consistent with the results of the present study. A reasonable explanation for the disagreement is that the changing immune microenvironment is very complex, immune cells in the tumor microenvironment have dual effects of anti-tumor and tumor-promoting [33,34], and immune cells can perform different functions in different tumors or at different stages of development in the same type of tumor. To date, clinical trials of targeted research in ESCC have not achieved ideal results, while the clinical trials of immunotherapy in ESCC show promise, and immunotherapy has shown great potential in the treatment of ESCC [35–38], suggesting that the therapeutic strategies targeting changes in the immune microenvironment are likely to benefit patients with ESCC; thus, the present study provides promising insights for the immunotherapy of ESCC by manipulating the immune microenvironment.

The highlights of the present study are as follows. Above all, EXOC3L1 has not been investigated in any tumors so far and this was the first study to reveal the relationship between EXOC3L1 and ESCC. The function of EXOC3L1 as an oncogene in ESCC was initially clarified, which could affect the metastasis and prognosis of ESCC patients and was related to immune infiltration, laying a foundation for further study of the mechanism of EXOC3L1 in ESCC in the future. Another important aspect of this study was that the expression of EXOC3L1 in ESCC was investigated in GTEx and TCGA, and was confirmed by GEO database and paired fresh tissues, which made the conclusion more reliable.

Although this study improves understanding of the relationship between EXOC3L1 and ESCC, there were still some unavoidable limitations. Firstly, the research was performed mainly based on bioinformatics analysis, so further verification should be carried out to reveal the biological impacts of EXOC3L1 in ESCC by in vitro and in vivo experiments. Secondly, a retrospective study has its own disadvantages, especially in that the interventions were not uniform and some of the information was missing; thus, a prospective study should be performed in the future to avoid analysis bias arising from the retrospective design. The third shortcoming of this study was that the sample size was small, and to enhance the credibility of the results, the sample size should be expanded.

Conclusions

In summary, the present study assessed the prognostic value of EXOC3L1 in ESCC for the first time, suggesting that EXOC3L1 offers a promising biomarker for predicting the prognosis and is a potential therapeutic target for ESCC patients. However, to elucidate the biological impact and underlying mechanism of EXOC3L1 in ESCC, further experimental studies are warranted.