MRGBP: A New Factor for Diagnosis and Prediction of Head and Neck Squamous Cell Carcinoma

MRG-binding protein (MRGBP) is a transcription factor widely involved in physiological and pathological processes. Many studies have discussed the relationship between the expression level of MRGBP and the prognosis of various malignant tumours. However, the role and clinicopathological significance of MRGBP in head and neck squamous cell carcinoma (HNSC) are unclear. In this study, the Wilcoxon signed-rank test and logistic regression were used to analyze the relationship between clinical characteristics and MRGBP expression in HNSC. The Kaplan-Meier plotter analysis and Cox regression analysis were established to evaluate the effect of MRGBP on prognosis, and the receiver operating characteristic (ROC) curve and nomogram was constructed. Gene set enrichment analysis (GSEA) and single-sample gene set enrichment analysis (ssGSEA) were used to analyze the correlation between MRGBP and immune infiltration. The results showed that the expression of MRGBP in HNSC tissues was significantly higher than that in normal tissues. The KM plotter analysis showed that the OS of HNSC patients was shorter. The multivariate Cox analysis further confirmed that increased expression of MRGBP was an independent risk factor for OS in HNSC patients. In addition, ROC analysis confirmed its diagnostic value and constructed prognostic nomograms, including age, T, M, N classification, pathological stage, and MRGBP. GSEA showed that MRGBP was associated with high expression of GPCR ligand binding, interleukin receptor binding, and neutrophil degranulation, and ssGSEA showed that MRGBP was associated with T cells and mast cells. In conclusion, MRGBP can serve as an independent prognostic biomarker related to immune invasion of head and neck squamous cell carcinoma.


Introduction
In recent years, the number of patients with head and neck squamous cell carcinoma (HNSC) has been on the rise. For example, thyroid tumours, as a type of HNSC, account for up to 4.9% of female tumours (https://www.uicc.org/news/ globocan-2020-new-global-cancer-data). The mechanism of HNSC development is complex, involving changes in multiple genes and multiple signaling pathways [1,2]. In this process, viral, environmental, and genetic factors affect the regulation and abnormal expression of tumour-related genes [3,4]. Due to the hidden physiological location of HNSC, early diagnosis is complex, and lymph node metastasis and distant metastasis are prone to occur. Therefore, screening biomarkers are helpful to the diagnosis and pathological indicators of tumour development. MORF4-related gene-binding protein(MRGBP), a protein encoded by FLJ10914 ORF, was first reported in 2003 and is associated with the structure and function of the mammalian TRRAP/TIP60 histone acetyltransferase complex [5].
More and more researchers have begun to explore the relationship between MRGBP and tumours in recent years. Studies have shown that MRG-binding protein (MRGBP) is upregulated in most colorectal tumours [6]. However, the potential role of MRGBP in the development or metastasis of HNSC remains unclear.
In the tumour microenvironment, immune and stromal cells are two main nontumour components. The degree of tumour immune invasion and mesenchymal cells is essential for tumour diagnosis and prognosis evaluation [7]. This study comprehensively analyzed the expression of MRGBP in the cancer database and its relationship with the prognosis of cancer patients. We then used clinical specimens in the TCGA database for validation of gene expression. In addition, we investigated the relationship between MRGBP and tumourfiltering immune cells in HNSC using tumour immune evaluation resources. This study comprehensively analyzed the expression of MRGBP in the cancer database and its relationship with the prognosis of HNSC patients.

Gene
Set Enrichment Analysis (GSEA). GSEA, a method to reveal genomic expression data through basic knowledge, is performed to identify high-risk populations' potential biological processes and pathways in a "cluster analysis" R package [9,10]. KEGG (Kyoto Encyclopedia of Genes and Genomes) and GO (Gene Ontology) data were downloaded from the Molecular Signature Database (MSigDB). Gene sets with jNESj>1, NOM P < 0:05, and FDR < 0:05 were considered statistically significant.
2.3. Immune Cell Infiltration of ssGSEA. Single-sample gene set enrichment analysis (GSA) in the "GSVA" R package [11] was used to analyze the immune infiltration of HNSCS, and the infiltration levels of 24 resistant cell types [12] were quantitatively analyzed according to the gene expression profile. The Spearman correlation was used to analyze the relationship between immune cell infiltration and MRGBP

The Relationship between High Expression of MRGBP
and Clinicopathological Features in HNSC Patients. Comparing the expression of MRGBP in 502 HNSC tissues and 44 paracancer tissues, it was found that the expression level of MRGBP in HNSC tissues was significantly higher than that in paracancer tissues (P = 2:6E − 39) (Figure 1(a)).

BioMed Research International
Meanwhile, MRGBP expression increased in tumour tissues of 43 HNSC samples after reservation paired sample screening (P = 2E − 15) ( Figure 1(b)). In addition, the effectiveness of MRGBP expression in normal GTEx combined with adjacent HNSC tissues and HNSC samples was analyzed using receiver operating characteristic (ROC) curves. The area under the MRGBP curve is 0.982, which has a high diagnostic value (Figure 1(c)).

Relationship between MRGBP Gene Expression and Clinical Features.
Clinical and gene expression profiles of 502 HNSC patients were extracted from TCGA database. HNSC patients were divided into high expression group and low expression group according to the median MRGBP expression level ( Table 1). The Wilcoxon sign rank and logistic regression were used to analyze the correlation between MRGBP expression and clinical features. High expression of MRGBP was significantly related to T-type (P = 9:8E − 04), N-type (P = 0:03), and pathological stage (P = 0:01). In addition, the high expression of MRGBP was significantly correlated with gender (P = 6:9E − 05), smoking (P = 4:2E − 03), and race (P = 0:02) ( Figure 2). After excluding unqualified cases, univariate logistic regression analysis showed that MRGBP expression in HNSC patients was correlated with the clinical characteristics of poor prognosis ( Table 2) (Table 3 and Supplementary1, 2). Then, the 1-year, 3-year, and 5-year OS of TCGA-HNSC was predicted according to age, T, M, N grade, pathological stage, and MRGBP ( Figure 4). Meanwhile, a forest map was drawn according to CGA-HNSC multifactor Cox regression analysis ( Figure 5).

Discussion
MRGBP, identified initially as a novel TRRAP/Tip60 HAT complex component, is encoded by FLJ10914 ORF [5]. The MRG protein family is associated with an eventual loss of cell senescence or proliferation [14]. In addition, the overexpression of MRGBP is involved in the occurrence and development of cancer. MRGBP expression is upregulated in colorectal cancer, and MRGBP plays an essential role in cancer cell proliferation by regulating BRD8 [15]. MRGBP promotes colorectal cancer by playing advantage in cell proliferation or cancer cell division [5]. MRGBP was found to be expressed in human prostate cancer model cells and promote replication and invasion of these cells [16]. MRGBP is frequently upregulated in pancreatic ductal adenocarcinoma tissues and cell lines, and the upregulation of MRGBP is positively correlated with TNM stage and poor prognosis [17]. Of interest, MRGBP promotes the activation of      9 BioMed Research International androgen receptor-associated enhancers and promoters through epigenetic mechanisms [18].
Further studies indicated that MRGBP might play an epigenetic regulatory role as a direct downstream target gene of miR-137 in pancreatic cancer [19]. The latest study suggests that MRGBP also plays a role in DNA repair [20]. In this study, we used the TCGA database to study the expression profile of MRGBP in various cancers. The results showed that MRGBP expression was higher in adrenal cortical carcinoma, urothelial bladder carcinoma, hepatocellular carcinoma, oesophagal carcinoma, and head and neck squamous cell carcinoma than in adjacent normal tissues. Chai et al. found that MRGBP is a subunit of the NuA4 histone acetyltransferase complex and participates in the transcriptional activation of particular genes mainly through the acetylation of nucleosome histone H4 and H2A. They explored the expression profile of MRGBP in 33 tumours [21]. In this study, however, we assessed the potential prognostic value of MRGBP in HNSC patients in more detail. Similar to previous studies, in the current study [22], we found that the expression level of MRGBP in HNSC tissues was significantly higher than that in adjacent tissues. The KM plotter analysis showed significant shortening of OS in HNSC patients (P < 0:05), and the same was found in the floor of mouth subtype (P < 0:05). In addition, the high expression of MRGBP was related to the clinicopathological characteristics of HNSC, including T-type, N-type, pathological stage, and lymph node neck dissection. The results showed that high expression of MRGBP was associated with advanced head and neck squamous cell carcinoma, suggesting that MRGBP may be a marker for early or advanced gas-tric cancer. More importantly, the ROC analysis also confirmed its diagnostic value. In recent years, HNSC combined with MRGBP expression has not been reported to predict nomograms. Therefore, we constructed prognostic nomograms including age, T, M, N classification, pathological stage, and MRGBP, which doctors can use to improve the accuracy of identifying high-risk patients.
Furthermore, mutations in DNase I hypersensitive sites (DHS) may alter lung cancer susceptibility by regulating the expression of surrounding genes, a process closely related to MRGBP [23]. For hepatocellular carcinoma (HCC), the high expression of MRGBP was significantly correlated with tumour T stage, pathological stage, histological grade, vascular invasion, tumour protein P53 status, and overall survival. MRGBP has high diagnostic accuracy, and the area under the subject operating characteristic curve is 0.980. GSEA revealed the abundance of tumour-related pathways, such as cell cycle and DNA replication pathways, in the highly expressed MRGBP phenotype. GSA showed that MRGBP expression was significantly correlated with    , γδT, cytotoxic cells, regulatory T cells, plasmacytoid DC, and immature DC were significantly decreased. MRGBP may be a new biomarker to predict the prognosis of liver cancer and a therapeutic target related to immune infiltration [24]. Our study has some limitations-lack of cellular validation as other hard evidence. In addition, due to data from public databases, there may be some bias due to confounding factors.
Although the carcinogenic pathways of different HNSCS are different in molecular level, this study is a preliminary exploration of the carcinogenic molecules of HNSCS in general and lays a foundation for in-depth research on the specific anatomical location and clinical stage. Moreover, the researchers themselves intend to further refine MRGBP to one or more subtypes of HNSC.
In addition, it is a comparative study of the same individual to use paracancer tissue as the control. Due to the different gene expression levels among different individuals, paracancer tissue as the control can be more convenient to explain the experimental results.

Conclusions
In conclusion, our study suggests that in HNSC, overexpression of MRGBP is associated with poor prognosis and is considered an independent factor in HNSC patients. In addition, GPCR ligand binding, interleukin receptor binding, and neutrophil degranulation may be regulated by MRGBP, and high MRGBP is associated with T cells and mast cells. In head and neck squamous cell carcinoma, increased expression of MRGBP may be an independent prognostic biomarker and is related to the immune invasion.

Data Availability
The ( Figures 5 and 6) data used to support the findings of this study are available from the corresponding author upon request.

Ethical Approval
TCGA and GEO belong to public databases. The patients involved in the database have obtained ethical approval. Users can download relevant data for free for research and publish relevant articles. Our study is based on open source data, so there are no ethical issues and other conflicts of interest.

Conflicts of Interest
There is no conflict of interest between authors.