Comprehensive Analysis of the Clinical Signicance and Prognostic Value of the CBX Family in Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma (CESC)

Background: Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC) is the second most common type of cancer among gynecologic malignancies worldwide. Chromobox (CBX) family proteins are associated with the regulation of tumorigenesis, metastasis, and evolution of various cancers. Methods: The clinical features, expression levels, and prognostic value of CBXs in CESC were analyzed through several databases, including ONCOMINE, GEPIA, HPA, UALCAN, cBioPortal, Kaplan-Meier plotter, and Metascape. Results: We concluded that the expression level of CBX2/4/8 was upregulated, while the expression level of CBX6/7 was downregulated in CESC specimens. Immune inltration analysis revealed that CBX1/2/3/4/5/6/8 proteins were downregulated in normal cervical tissues, and upregulated in CESC specimens. In contrast, CBX7 protein expression was signicantly higher in normal adjacent cervical tissues and was not detected in CESC tissues. CBX1/3/6 mRNA expression was signicantly correlated with the pathological stage of CESC. Prognostic analysis showed that patients with high CBX7 levels of CESC had a favorable prognosis. Conclusions: Our study indicated that CBX7 could be an attractive biomarker for the prognosis of CESC.


Introduction
Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC) is the second most common type of cancer among gynecologic malignancies worldwide [1,2]. The estimated annual agestandardized incidence rate is 15/100 000 women globally, and CESC predominantly occurs in women aged 20-39 years. In recent years, due to advances in diagnosis and treatments, CESC survival rates have improved [3,4].
Although it has been discovered that CESC is caused by HPV (human papillomavirus) infection, there is a lack of effective biomarkers to detect CESC patients early [5,6]. Patients with CESC have a poor prognosis once they have metastatic tumors or are in advanced stages. On account of tumor heterogeneity, there is a critical need for valid biomarkers to improve individual treatment and the prediction of outcomes.
Additionally, abnormal epigenetic regulation plays a prominent role in tumorigenesis [7]. Chromobox  [8][9][10]. In previous studies, members of the CBXs family were found to have prognostic value. For example, CBX1 protein expression is increased in human hepatocellular carcinoma tissues, CBX2 is signi cantly elevated in human prostate cancer specimens [11,12].
Nevertheless, the roles of CBXs family proteins in the carcinogenesis and progression of CESC have not been completely investigated. Therefore, in our research, we extended the research eld to CESC based on various online databases to determine the prognostic values of distinct CBX family members.

ONCOMINE database
The ONCOMINE database (www.oncomine.org) is a web-based DNA and RNA sequence analysis data mining platform and cancer gene chip database that facilitates genome-wide expression analysis [13]. In our study, the transcriptional levels of CBXs between cancer tissues and their corresponding adjacent normal samples were analyzed using the ONCOMINE database. Student's t-tests were used to compare differences in transcription levels (p-value: 0.01, fold change: 1.5, gene rank: top 10%, data type: mRNA). UALCAN UALCAN (http://ualcan.path.uab.edu) is an interactive online tool. It can be used to analyze relative gene expression in matched samples and estimate associations between transcriptional expression and clinicopathological parameters of patient survival time [14]. In this study, UALCAN was used to analyze the relationship between the mRNA expression level of eight CBX family members and the grade of CESC cancer. The student's t-test was used to compare the differences, and a p-value <0.01 was considered statistically signi cant.

Human Protein Atlas (HPA)
The Human Protein Atlas (https://www.proteinatlas.org) is a web portal that contains nearly 20 types of cancer and transcriptome analysis of 8000 patient data and is based on immunohistochemical expression [15]. Users can identify the protein expression patterns of speci c genes in speci c types of tumors. In this study, we obtained immunohistochemical images to directly compare the expression of different CBX family members in normal tissues with CESC specimens.

GEPIA database
The GEPIA (Gene Expression Pro ling Interactive Analysis) database (http://gepia.cancer-pku.cn/) is an interactive online tool containing RNA expression data and provides fast and customizable capabilities based on Cancer Genome Atlas (TCGA) and Genotypic Tissue Expression (GTEX) data [16]. Using this database, we successively analyzed the differences in expression between CESC and normal tissues in eight members of the CBX family. Expression fold changes ≥1.5 and p-value<0.05 were considered statistically signi cant.

Kaplan-Meier plotter
Kaplan-Meier plotter (http://kmplot.com/analysis/) is an online tool to evaluate the in uence of 54,000 genes on 21 types of cancer survival and draw survival plots with gene expression data and survival information [17,18]. The results are shown by plotting a survival curve and simultaneously analyzing hazard ratios (HRs) with 95% con dence intervals (CIs) and log-rank p-values. When a p-value < 0.05, a statistically signi cant difference was considered. cBioPortal cBioPortal (www.cbioportal.org) is an online open-access portal that provides a way to visualize, explore, and analyze multidimensional cancer genomics data [19]. In our study, we compared the frequency of mutations in the CBXS gene in CESC and summarized all associated genomic changes.

Metascape
Metascape (http://metascape.org) was used for functional enrichment and pathway annotation of genes [20]. We researched the genes in the resulting CBX family coexpression network in STRING employing the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways.

Results
The expression levels of eight CBX family members in patients with CESC ONCOMINE, GEPIA, and HPA were used to analyze the expression levels of the CBX family in CESC. In the ONCOMINE database, the mRNA expression of eight CBX family members in various tumors was analyzed and compared with normal tissues ( Fig. 1 and Table 1). Concerning CBX1, we found overexpression levels in the Dohun cervical statistics dataset (FC = 3.762, p = 1.72E-08). Regarding CBX2, low expression levels were found in the Luigi cervical statistics dataset (FC=-1.794, p = 0.000145). Notably, we found that CBX3/5 mRNA is highly expressed in CESC tissues in multiple datasets. In the Yali dataset, for CBX3, compared to normal tissue, cervical squamous cell carcinoma (CSCC) tissues and high-grade cervical squamous intraepithelial neoplasias (HGCEIN) had fold changes of 1.911 (p = 1.26E-08) and 1.576 (p = 0.003), respectively. While Dohun found that CBX3 mRNA expression had a 2.  After inspecting the mRNA expression patterns of CBXs in CESC, we subsequently attempted to analyze the expression pattern of CBXs family proteins in CESC by HPA (Human Protein Atlas). As shown in Fig. 3, the normal cervical tissues presented medium protein expression of CBX1/2/3/4/5, while medium and high protein expression was discovered in CESC tissues ( Fig. 3A-3E). Analogously, low protein expression of CBX6/8 was observed in normal cervical tissues, while medium and high protein expression of CBX6/8 was observed in CESC samples (Fig. 3F, 3H). In contrast, we found that the CBX7 protein was medially expressed in healthy cervical tissues and was not detected in CESC tissues (Fig. 3G).
The clinicopathological parameters associated with CESC patients were investigated using models of mRNA expression of eight members of the CBX family Afterward, we analyzed the correlation between CBXs family mRNA expression levels and clinicopathological parameters in patients with CESC by GEPIA and UALCAN, containing patients' cancer stages and tumor grades. In the Violin plot (Fig. 4), we found that the expression of CBX2/6/8 mRNA was correlated with the pathological stage of CESC patients. Figure 5 shows the relationship of the mRNA expression level of distinct CBX family members with the tumor grades of CESC patients. The mRNA expression levels of CBX1/3/6 were related to tumor grade. When the tumor grade advanced, CBX1/3 tended to have a higher mRNA expression level, and CBX6 had the opposite trend. In tumor grade 4, we found the highest mRNA expression of CBX1/2/3 (Fig. 5A, 5B,   5C). The highest mRNA expression of CBX4/5/8 was found in tumor grade 2 (Fig. 5D, 5E, 5H), while the highest mRNA expression of CBX6/7 was found in normal tissues (Fig. 5F, 5G).
In summary, the results showed that the mRNA expression of several CBX family members was signi cantly correlated with clinicopathological parameters in patients with CESC.
The relationship between CBXs mRNA expression levels and prognosis in patients with CESC Next, we analyzed the relationship between the CBXs mRNA expression levels and the prognosis of patients with CESC (Fig. 6). Distinctly, CBX1/2/3/4/5/6/8 mRNA expression showed no correlation with prognosis in CESC patients (Fig. 6A-6G), while higher mRNA expression of CBX7 (HR = 0.43, 95% CI: 0.26-0.69, p = 0.00042) was signi cantly correlated with favorable OS (overall survival) in CESC patients (Fig. 6H). The results showed that the mRNA expression level of CBX7 could be used as an important biomarker to predict the OS of patients with CESC.

Cbxs Family Gene Mutations In The Patients With Cesc
CBXs family gene mutations in the patients with CESC We analyzed genetic changes in the patients using the cBioPortal tool and found mutations in the CBXS gene in 64 of 188 patients, with a mutation rate of 34%. In addition, the mutation rates of the CBX4, CBX1, CBX2, and CBX3 genes were 11%, 9%, 7%, and 7%, respectively (Fig. 7).

Predicted Functions And Pathways Of The Mutations In Cbxs
After analyzing CBXs family gene mutations in patients with CESC, we analyzed neighboring genes associated with mutations in the CBXs family of genes and constructed a complete network. As shown in

Discussion
CESC is one of the deadliest diseases originating from cervical cells. Squamous cell carcinoma and are the most common pathological subtypes accounting for approximately 70%, and adenocarcinoma accounting for approximately 25% of all CESC. Since the 3-year to 5-year survival rate of patients with CESC in many developing countries is < 50%, early detection, effective therapy, and precise prediction of the outcome are of great signi cance for CESC patients [1,5,21].
Although most cervical cancer cases are due to human papillomavirus (HPV) infection, additional genetic and epigenetic changes are required for cervical carcinogenesis [22]. Recently, advances in understanding the biology of CESC have shown that epigenetic alterations are common in the tumorigenesis of CESC and metastasis [23]. Therefore, novel biomarkers related to the prognosis and treatment of CESC are urgently needed.
In our study, the expression level, mutation status, and prognostic value of eight members of the CBX family in CESC were analyzed. We mainly focused on the relationship between the CBXs mRNA expression level and the prognosis of patients with CESC. We revealed that CESC patients with high CBX7 levels had a favorable prognosis, demonstrating CBX7 as a potential prognostic biomarker for CESC.
The CBX family is a canonical component of PcG and is implicated in the development of diversi ed cancers [24]. CBX1 is a member of the heterochromatin protein 1 (HP1) family, also known as HP1β. Yang et al. found that the expression level of CBX1 was signi cantly increased in HCC tissues and cell lines and was signi cantly correlated with tumor size, degree of differentiation, and tumor aggressiveness [11]. Analogously, Shiota et al. found that CBX1 expression was increased in prostate cancer samples. Additionally, its high expression could enhance the activity of the androgen receptor pathway and accelerate the growth of prostate cancer [25]. CBX4 has been thoroughly studied in a variety of cancers. Jiao et al. and Wang et al. respectively found that CBX4 was highly expressed in clinical specimens and multiple hepatocellular carcinoma cells. It was also associated with clinical parameters, including the α-fetoprotein level, tumor size, lymph node metastasis, distant metastasis, pathologic differentiation, shorter OS, and relapse-free survival (RFS) [31,32]. Additional studies carried out by Li et al. showed that CBX4 was overexpressed in HCC mice and promoted tumorigenesis by increasing VEGF production and angiogenesis, while knockdown of CBX4 suppressed the tumorigenicity of CBX4 [33]. Furthermore, Hu et al. indicated that CBX4 is overexpressed in lung cancer cells and promotes proliferation and metastasis by regulating the BMI-1 [34]. CBX5 also known as heterochromatic protein 1α (HP1α), and overexpression of CBX5 was found in many types of malignant tumors, including pancreatic cancers, breast cancers, and lung cancers [35].
In a previous report, Zheng et [39]. And Li et al. found that CBX7 inhibits tumor growth and metastatic potential by binding to the E-box to inhibit TWIST-1 function in ovarian cancer [40]. In our study, we found that CBX7 expression was decreased in CESC tissues compared to adjusted normal tissues, while higher mRNA expression of CBX7 was signi cantly associated with a favorable OS (overall survival) in CESC patients. In addition, The CBX7 protein expression was not detected in CESC tissues demonstrated by immunohistochemistry pictures. These results indicate that CBX7 is an effective marker for predicting the outcomes of patients.
Similarly, CBX8 plays a signi cant role in lymphoma, breast carcinoma, esophageal squamous cell carcinoma, esophageal cancer, and other human cancers [41]. Studies carried out by Yuan et al. showed that upregulated CBX8 was associated with the progression of the stages in muscle-invasive bladder cancer tissues by repressing the p53 pathway [42]. Coincidentally, CBX8 inhibits metastasis by suppressing the Snail pathway in esophageal squamous cell carcinoma while promoting cell proliferation [43].
In conclusion, we found that high expression of CBX7 mRNA is signi cantly related to the favorable prognosis and survival of CESC patients. Other members of the CBX family have no obvious relationship with prognosis and survival. Certainly, there are some limitations to our study, and all of the data researched in our study were acquired from online databases. Further research requires a larger sample size to test our ndings and to explore the clinical application and potential mechanisms of CBX members in the treatment of CESC.  Immunohistochemistry images of distinct CBX family members in CESC tissues and normal cervical tissues by Human Protein Atlas (HPA). CBX1/3/4/5/6/8 protein expression was higher in CESC tissues (A, C-F, H), and CBX2 was not differentially expressed between CESC tissues and healthy cervical tissues (B). CBX7 protein expression was not signi cantly expressed in CESC tissues (G).

Figure 4
Relevance between mRNA expression of distinct CBX family members and cancer stages of CESC patients presented by the violin plot. The mRNA expression levels of CBX2/6/8 was associated with the pathological stage in CESC with a p-value<0.05.

Figure 5
Relationship of the mRNA expression level of distinct CBX family members with tumor grades of CESC patients. The mRNA expression of CBX1/3 was related to tumor grade, and when the tumor grade increased, the mRNA expression level of CBX1/3 tended to be higher, while that of CBX6 was the opposite. *p<0.05, **p<0.01.  CBXs gene expression and mutation analysis in CESC (cBioPortal). A high mutation rate (34%) of CBXs was observed in CESC patients. CBX4, CBX1, CBX2, and CBX3 ranked as the four genes with the most genetic alterations, and their mutation rates were 11%, 9%, 7%, and 7%, respectively.