Multiomics Integrative Analysis Identifying EPC1 as a Prognostic Biomarker in Head and Neck Squamous Cell Carcinoma

Background Biomarker research in head and neck squamous cell carcinoma (HNSCC) is constantly revealing promising findings. An enhancer of polycomb homolog 1 (EPC1) was found to play a procancer role in nasopharyngeal carcinoma (NPC), but its role in HNSCC with strong heterogeneity is still unclear. Herein, we investigated the prognostic significance and related mechanisms of EPC1 in HNSCC. Methods The Kaplan-Meier plotter was used to evaluate the prognostic significance of EPC1. Based on a range of published public databases, the multiomics expression of EPC1 in HNSCC was explored to investigate the mechanisms affecting prognosis. Results According to the clinical data, high EPC1 expression in HNSCC was a predictor of patient prognosis (hazard ratio (HR) = 0.64; 95% confidence interval (CI) 0.49-0.83; P < 0.01). EPC1 expression varied among clinical subtypes and was related to key factors, such as TP53 and human papillomavirus (HPV) (P < 0.05). At the genetic level, EPC1 expression level may be associated with protein phosphorylation, cell adhesion, cancer-related pathways, etc. For the noncoding region, a competing endogenous RNA network was constructed, and 6 microRNAs and 12 long noncoding RNAs were identified. At the protein level, a protein-protein interaction (PPI) network related to EPC1 expression was constructed and found to be involved in HPV infection, endocrine resistance, and multiple cancer pathways. At the immune level, EPC1 expression was correlated with a variety of immune cells and immune molecules, which together constituted the immune microenvironments of tumors. Conclusion High EPC1 expression may predict a better prognosis in HNSCC, as it is more frequently found in HNSCC with HPV infection. EPC1 may participate in the genomics, transcriptomics, proteomics, and immunomics of HNSCC, and the results can provide a reference for the development of targeted drugs and evaluation of patient prognosis.


Background
Due to increases in tobacco use and the human papillomavirus (HPV) infection rate, the number of patients with head and neck squamous cell carcinoma (HNSCC) is increasing, which is one of the most common cancers and accounts for approximately 5% of all malignancies [1,2]. The successful development of targeted therapies in biomarker-selected patients for personalized medicine has shifted expectations in cancer research [3,4], and the lack of targetable genomic abnormalities in HNSCC has limited the development of targeted therapies in the past [5]. Thus, identifying a reliable molecular biomarker to predict the prognosis of patients with HNSCC is an urgent task. To better treat HNSCC patients, many studies have focused on identifying relevant biomarkers to predict patient prognosis [6][7][8]. However, because the human body is a complex organism and the occurrence and development of cancer may involve many aspects, the limitations of mining disease-related factors based on the one-omics perspective have become    3 BioMed Research International increasingly apparent in recent years. Additionally, constantly improving databases provide technical support, enabling multiomics studies involving genomics, transcriptomics, and proteomics. Therefore, the application of multiomics data systems to explore cancer biomarkers has become an important trend in precision medicine, allowing joint research on macro and micro aspects.
The enhancer of polycomb homolog 1 (EPC1) has a protective function against DNA damage. Epigenetic factor EPC1 is a master regulator of the DNA damage response by interacting with E2F1 to silence cell death and activate metastasis-related gene signatures [9]. Pathways known to be associated with this gene include chromatin-modifying enzymes, chromatin organization, and histone acetyl transferases (HATs) [10]. Sophisticated studies have demonstrated that EPC1 is involved in the NuA4 HAT complex, and the crystal structure and molecular basis for EPC1 bound to MBTD1 were determined [11]. Additionally, hsa_circ_0007919 knockdown resulted in hsa-let-7a downregulating EPC1 mRNA [12]. According to previous reports, abnormal EPC1 was present in both endometrial stromal sarcoma [13,14] and ossifying fibromyxoid tumors [15], whereas EPC1 silencing inhibited lung cancer cell proliferation and tumor growth [16]. Additionally, EPC1 has been correlated with patient prognosis in microarray screenings of nasopharyngeal cancer [17]. These data suggest that EPC1 may be a prognostic biomarker that is worth studying. Therefore, we aimed to provide further insight into the prognostic significance of EPC1 in patients with HNSCC and comprehensively analyze EPC1 from a multiomics perspective to explore its mechanisms.

Patients and Transcriptional Expression Profile.
The clinical data and gene expression profiles of patients with HNSCC were downloaded from the Genomic Data Commons. Clinical data were mainly used for survival analysis, and gene expression profiles were used for subsequent multiomics analyses. This study was conducted in accordance with TCGA publication guidelines.

The Associations between Gene Expression and Key
Prognostic Factors. To analyze possible associations between clinical parameters and EPC1 expression, the TISIDB database (http://cis.hku.hk/TISIDB/, last accessed on 31 July 2022), which integrated clinical data from TCGA, was used to identify differences in EPC1 expression between different HNSCC subtypes [18]. Survival curves associated with EPC1 expression in HNSCC were plotted using the Kaplan-Meier plotter (http://kmplot.com/analysis/, last accessed on 31 July 2022) [19]. Overall survival (OS) was selected for the HNSCC clinical endpoint analysis. OS was defined as the period from the date of diagnosis to the date of death from any cause [20]. UALCAN (http://ualcan.path.uab.edu/, last accessed on 31 July 2022) is a web platform based on Java-Script, CSS, and PERL-CGI. Differences in EPC1 expression in HNSCC can be visualized using UALCAN according to TP53 mutation status or the presence of HPV [21]. TIMER 2.0 (http://timer.comp-genomics.org/, last accessed on 31 July 2022) was used to analyze three sample types: total HNSCC samples, HPV-positive HNSCC samples, and HPV-negative HNSCC samples. Therefore, the association between EPC1 and clinical outcomes of patients with HNSCC was investigated based on the presence of HPV [22]. At the same time, the Wilcoxon rank sum test was used to analyze the relationship between the EPC1 and TP53 mutation status.

Protein-Protein Interaction (PPI) Networks and KEGG Pathway Enrichment.
To reveal the role of EPC1 in proteomics, EPC1-related proteins in HNSCC were screened using LinkedOmics (set at P < 0:05) and used to construct EPC1related PPI networks using the STRING database [28] (https://string-db.org/, last accessed on 6 October 2020). In 8 BioMed Research International addition, KEGG pathway enrichment was used to predict the role of EPC1 at the protein level in HNSCC, which was visualized using the "ggplot2" package of R software 3.6.3.
2.6. EPC1-Related Immune Cells and Immunoreactive Substances in HNSCC Samples. TIMER 2.0 was used to sys-tematically analyze immune infiltration. The tool integrated CIBERSORT-ABS with published existing data [29]. The associations between immune infiltration and EPC1 gene expression in HNSCC patients (HNSCC, HPV-positive HNSCC, and HPV-negative HNSCC) were explored.

Statistical Analysis.
Kaplan-Meier curves were used to compare the differences in survival time. OS was selected for the HNSCC clinical endpoint analysis. Hazard ratios (HRs) and their corresponding 95% confidence intervals (CIs) were calculated to assess the role of the EPC1. A logrank test (P < 0:05) indicated a significant survival time difference. The results were verified using another database [30]. We also performed GSEA. And the rank criteria were a P value <0.05, a minimum number of genes size=5, stimu-lations=500, and aFDR ≤ 0:05. Additionally, the spearman correlation analysis and the wilcoxon rank-sum test were applied to show the correlations between the EPC1 gene and other factors.

Effect of the Differential Expression of EPC1 on the Prognosis and Clinical Outcomes of Patients with HNSCC.
Using the TISIDB platform, spearman correlation analysis was performed to study the association between EPC1 expression and HNSCC subtypes. EPC1 expression levels were not equal among the different subtypes (Figure 1(a)). The Kaplan-Meier plotter platform was used to analyze the association between survival and EPC1 expression (Figure 1(b)). The median survival time for the low-EPC1 expression group was 33.10 months, and the median survival time for the high-EPC1 expression group was 61.27 months; the difference was statistically significant (HR < 1, P < 0:01), suggesting that HNSCC patients with high EPC1 expression have a better prognosis. Based on TCGA samples and the UALCAN website, EPC1 expression in HPV positive HNSCC tumors was not only significantly higher than that in paracancerous tissues (P < 0:01) but also significantly higher than that in HPV negative HNSCC samples (P < 0:01) (Figure 1(c)). Using HPV positive HNSCC samples, we explored the relationship between EPC1 expression and patient prognosis. The results showed that patient prognosis was significantly better with higher EPC1 expression (Figure 1(d)) (P < 0:01). However, no significant effect of EPC1 expression on patient prognosis was found when HPV negative HNSCC samples were analyzed. In addition, compared with that in TP53-mutated HNSCC, EPC1 expression in wild-type TP53 HNSCC was significantly higher (P < 0:05) (Figure 1(e)). The expression of wild-type TP53 EPC1 was relatively high in HPV positive HNSCC samples. In HPV negative HNSCC samples, no significant difference in EPC1 expression was identified between the wild-type TP53 and TP53-mutated samples (Figure 1(f)). Therefore, we hypothesized that high EPC1 expression was suggestive of better prognosis in patients with HNSCC, because it tended to be present more often in HNSCC patients with HPV positive, who were more sensitive to radiotherapy and had a greater prognosis compared with HNSCC patients without HPV.

Screening and Functional Prediction of Genes Associated
with the Differential Expression of EPC1 in HNSCC. Linke-dOmics was used to screen genes that were significantly positively correlated with the EPC1 and genes that were significantly negatively correlated with EPC1. A total of 20,164 related genes were obtained, including 8208 genes with negative correlations and 11,956 genes with positive correlations, and volcano plots were constructed (Figure 2(a)). Notable positively correlated genes included ZNF41, NR2C2, and CEP350. Notable negatively correlated genes included MRPL28, C14orf156, and TMEM280. After obtaining the gene dataset, we performed GSEA. The rank criteria were a P value < 0.05, FDR ≤ 0:05, minimum number of genes ðsizeÞ = 5, and simulations = 500. KEGG pathway analysis was also performed. We selected "redundancy reduction: weighted set cover" and screened 5 positively correlated KEGG pathways (labeled blue, Figure 2(b)): phosphatidylinositol signaling system, cell adhesion molecules, transcriptional misregulation in cancer, pathways in cancer,

11
BioMed Research International and neuroactive ligand-receptor interaction. We also screened five negatively correlated KEGG pathways (labeled orange, Figure 2(b)): purine metabolism, thermogenesis, spliceosome, proteasome, and ribosome. Using pathways in cancer as an example, 190 genes were enriched (enrichment score = 0:48; normalized enrichment score = 1:55; P < 0:01), and the difference was statistically significant (Figure 2(c)). The above steps were repeated for the GO analysis (biological processes). Five positively correlated biological processes were screened (labelled blue, Figure 2(d)): protein autophosphorylation, covalent chromatin modification, positive regulation of cell motility, regulation of GTPase activity, and cell-cell adhesion via plasma-membrane adhesion molecules. Additionally, five negatively correlated biological processes were screened (labeled orange, Figure 2(d)): protein targeting, protein folding, nucleoside triphosphate metabolism, ribonucleoprotein complex biogenesis, and mitochondrial gene expression. The above analyses showed that the differential expression of EPC1 at the gene level was related to cancer pathways. In addition, functional enrichment results showed that EPC1 expression was mainly related to protein phosphorylation and cell adhesion.

Screening and Pathway Enrichment of Proteins Related to Differential EPC1 Expression in HNSCC.
LinkedOmics was used to screen 13 proteins that were positively related to EPC1 gene expression (Figure 4(a)) and 7 proteins that were negatively related to EPC1 expression (Figure 4(b)), all of which met the criteria ofP < 0:05. The corresponding heat maps were drawn. Using the STRING database, an interaction network consisting of 21 proteins was constructed (Figure 4(c)) and protein enrichment analysis was used to obtain the top 10 related pathways based on gene ratios (Figure 4(d)). The PPI network suggested that proteins coexpressed with EPC1 might be involved in various cancer-related signaling pathways, such as HPV infection, endocrine resistance, cell cycle disruption, plaque adhesion, breast cancer, gastric cancer, hepatocellular carcinoma, pancreatic cancer, and small-cell lung cancer.
3.5. Differential Expression of EPC1 among All HNSCC Samples, HPV-Positive HNSCC Samples, and HPV-Negative HNSCC Samples and the Association with Immunity. At the cellular level, HNSCC, HPV positive HNSCC, and HPV negative HNSCC samples were subjected to immunological analysis using the TIMER 2.0 platform. We used CIBERSORT-ABS algorithm for immune scoring analysis. After adjusting for tumor purity, 6 kinds of EPC1-related immune cells were screened out from the HNSCC samples under the restriction condition that one of three sample types reached jrhoj > 0:3 and P < 0:05. Except for alternatively activated macrophages (M2), we found EPC1 gene expression in HNSCC patients with HPV infection showed a higher infiltration levels of most immune cell (Table 1).
To further explore the correlation between macrophages and EPC1, we focused on classically activated macrophage (M1), M2, and their related molecules. Among HPV positive HNSCC patients, EPC1 gene expression was positively associated with the infiltration level of M1 ( Figure 5(a)). Also, EPC1 gene expression level was correlated with IL2 ( Figure 5(b)) and TNFSF15 ( Figure 5(c)). However, M2 macrophage infiltration level was positively associated with EPC1 gene in HNSCC patients without HPV infection ( Figure 5(d)), compared with HNSCC patients with HPV infection. And, EPC1 gene expression in HPV HNSCC patients was associated with IL10 ( Figure 5(e)) and MRC1 ( Figure 5(f)). These results suggested that EPC1 may alter the tumor microenvironment of HNSCC by affecting the immune cells and immune-related molecules.

Discussion
EPC1 is a multicomb homolog 1 (Drosophila) enhancer involved in the regulation of cell growth and transcription [31]. EPC1 anomalies may be involved in ossifying fibromyxoid tumors [15], endometrial stromal sarcoma [32], pancreatic cancer [33], and nasopharyngeal carcinoma (NPC) [34]. However, the effect of this gene on the survival prognosis of HNSCC patients has not been studied. Therefore, we tried to identify the function of EPC1 as a prognostic biomarker in HNSCC by multiomics integrative analysis. We found that EPC1 might indicate a better overall survival for patients with HNSCC, which was inconsistent with the results of NPC from our study [34]. We considered that one of the reasons might be that the expression data of NPC was not available in TCGA, due to the rarity of NPC in America. In addition, we found that high EPC1 expression was correlated with HPV-positive and non-tp53 mutations, both of which were favorable prognostic factors in HNSCC [35]. In addition, coexpressed genes, competing endogenous RNAs, protein interaction networks, immune cells, and molecules may also participate in the prognosis of HNSCC patients.
The survival time of HNSCC patients with high EPC1 expression is longer, which was not noted in previous studies. EPC1 expression varied in different HNSCC subtypes. And EPC1 expression was relatively high in HPV positive HNSCC samples. In addition, among the EPC1-related protein enrichment pathways, HPV infection-related pathways predominated, suggesting that EPC1 and HPV may have a certain correlation that affects patient prognosis. In addition, TP53 mutations often indicate poor prognosis, and EPC1 expression is relatively low in HNSCC with TP53 mutations. However, EPC1 expression was relatively high in wild-type TP53 samples, suggesting that high EPC1 expression may indicate the better prognosis of patients with HNSCC.
To further investigate the possible mechanism by which EPC1 affects HNSCC prognosis, we performed a multiomics analysis of EPC1. At the gene level, EPC1-related gene enrichment results indicated that EPC1 participates in some important biological processes, such as protein phosphorylation, cancer-related pathways, and cell mobility. A study revealed that oxidative phosphorylation was significantly enriched in HPV + HNSCC, which was one of the differences from HPV -HNSCC [36]. Therefore, we can speculate that HPV may be an important reason for the close relationship between EPC1 gene expression and protein autophosphorylation. In transcriptomics studies, lncRNA H19 was an important link in the EPC1-related lncRNA-miRNA-mRNA network. High lncRNA H19 expression was positively correlated with the growth, migration, and invasion of lung tumor cells, but low H19 expression is associated with poor prognosis for patients with microinvasive follicular thyroid carcinoma and can be used to predict distant metastasis [37]. These results suggest that EPC1 may affect the prognosis of patients with HNSCC through lncRNA H19, leading to different prognoses in patients with different cancers. Besides, H19 [38] and MALAT1 [39] were upregulated in HPV + cancers than those in HPVcancers, which indicated that HPV might participate in the regulation of the ceRNA network. Proteomics studies have shown that EPC1-related proteins were mainly involved in HPV infection, endocrine resistance, cell cycle, and cancer pathways. In this study, high EPC1 expression in HPV positive HNSCC samples had a significant positive effect on prognosis, suggesting that EPC1 and HPV may be associated with patient survival prognosis. At the immunomics level, intratumoral immune status was a key factor affecting patient survival and response to immunotherapy. The tumor microenvironment has clinical significance in predicting therapeutic effects [40]. EPC1-related immune cells, such as macrophage, B cells, and T cells, may play a role in controlling tumor growth. Furthermore, M1 macrophage was considered to have greater antitumor activity [41], which was positively correlated with EPC1 expression in HPV + tumors. IL2 served as an M1 marker and TNFSF15 was shown to promote M1 production, both of which showed a correlation with EPC1 expression. Besides, EPC1 expression was more strongly correlated with M2 and its markers in HPVtumors than in HPV + tumors. M2 was regarded as a tumor promoting factor in previous studies [42,43]. Taken together, it may partly reveal why high EPC1 expression indicates a better prognosis in HNSCC, especially in HPV + tumors.

Conclusions
This study did not investigate the role of EPC1 alone but rather investigated the differentially expressed genes, ceRNA networks, interacting proteins, and immune infiltration levels associated with EPC1 in HNSCC samples. The data used in this study were experimentally validated or geneti-cally sequenced, based on real-world data. Using these databases, we further confirmed that high EPC1 expression was a favorable factor for the prognosis of patients with HNSCC. EPC1 expression correlated with HPV infection, protein phosphorylation, and immune cell infiltration. These factors may explain the role of EPC1 for HNSCC from multiple perspectives.
This study had some limitations. Data were obtained from public databases. However, the results have not yet been validated by animal model. In addiction, the functional role of EPC1 in HNSC should be validated by overexpression and knockdown experiments in the future.

Data Availability
The datasets used for the current study are available upon reasonable request from the corresponding authors.

Ethical Approval
The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki in their investigations.