3.1 The Expression Levels of the CBXs in ESCA Patients
We firstly used Oncomine to explore the expression levels of the CBXs in ESCA. The results are depicted in Figure 1 and (Table 1). Obviously, the transcriptional levels of CBX1, CBX3 and CBX6 were raised, while the level of CBX7 was significantly reduced in ESCA tissues compared with the normal group (Figure 1). In detail, the expression level of CBX1 was upregulated in ESCA tissues vs. normal tissues (fold change=1.662), while that of CBX3 was significantly overexpressed in ESCA vs. normal tissues. The results based on the data acquired from the dataset showed a fold change of 2.188 (p = 3.75E-29) (Table 1).
Table 1 The mRNA levels of CBXs in different types of ESCA and normal tissues.
Name
|
ESCA VS. Nomal
|
Fold Change
|
P Value
|
t-test
|
Source(Oncomine)
|
CBX1
|
ESCC
|
1.662
|
5.56E-13
|
8.101
|
Su Esophagus
|
|
EAC
|
3.301
|
2.78E-04
|
3.72
|
Hao Esophagus
|
|
ESCC
|
1.77
|
2.50E-04
|
4.158
|
Hu Esophagus
|
CBX2
|
ESCC
|
2.308
|
5.78E-04
|
6.846
|
Su Esophagus
|
CBX3
|
ESCC
|
2.188
|
3.75E-29
|
15.587
|
Su Esophagus
|
|
ESCC
|
2.389
|
3.21E-10
|
10.596
|
Hu Esophagus
|
|
EAC
|
2.559
|
1.20E-04
|
4.887
|
Kimchi Esophagus
|
|
Barrett's Esophagus
|
1.629
|
1.06E-05
|
5.127
|
Kim Esophagus
|
CBX6
|
ESCC
|
2.101
|
1.43E-06
|
6.336
|
Hu Esophagus
|
Afterwards, the GEPIA dataset was used to compare the mRNA levels of CBXs between ESCA and the normal groups. The results show that the transcriptional levels of CBX3 and CBX5 were higher in ESCA patients compared with normal tissues, and the expression levels of CBX7 were highly lower in the former compared with the control group (Figure 2). We later evaluated the relation between the expression levels of CBXs and the pathological stages of ESCA patients. Clearly, the CBX1 group was significantly varied, while the CBX2, CBX3, CBX4, CBX5, CBX6, CBX7, and CBX8 groups were not obviously different (Figure 3), demonstrating that that CBX1 is significantly associated with the progression of ESCA. Consequently, CBX1 might be considered as a pathological stage marker.
3.2 Prognostic Value of the CBXs in ESCA Patiens
In order to assess the prognostic value of the CBX members in the progression of ESCA, GEPIA was used to analyze the correlations between different CBXs and the clinical outcomes. The disease-free survival (DFS) and overall survival (OS) curves are shown in Figure 4. The transcription level of CBX1 (p = 0.044) was significantly correlated with short DFS. Except for CBX1, other CBX protein members have little effect on OS or DFS.
Then, the Kaplan-Meier plotter was used to analyze the prognostic values of CBXs in ESCC patients and EAC patients, which is a subgroup analysis (Figures 5 and 6). It is obviously observed that high CBX4 (HR = 2.93, p = 0.008) mRNA expression was significantly associated with short OS in patients with ESCC (Figure 5). Consistently, CBX3 (HR = 3.12, p = 0.00028) and CBX8 (HR = 2.27, p = 0.035) mRNA expressions were highly associated with short OS in patients with EAC (Figure 6). All these show that CBX3/4/8 have a significant effect on the prognosis of ESCA, and may provide some ideas for the research and treatment of ESCA.
In addition, in order to further explore the prognostic value of CBXs genes with prognostic value, we performed immunohistochemistry (IHC) to detect the protein expression of CBX1, CBX3 and CBX5 in normal tissues and tumor tissues. We found that the expressions of CBX1, CBX3 and CBX5 proteins in ESCA tissues are higher than that in normal tissues(Figure 7). Therefore, we can reasonably infer that these genes play key roles in protein expression and may affect the occurrence and development of ESCA.
3.3 Genetic Alteration, Expression, and Cancer Pathway Activity of CBXs in ESCA Patients
To analyze the genetic alteration of CBXs in the patients with ESCA, we detected two or more alterations of CBXs in the two subtypes of ESCC and EAC in ESCA. The results show that the alterations frequency of CBXs were high in ESCC samples (Figure 8A). In addition, CBXs were altered in 138 samples of 927 ESCA patients, accounting for 15% (Figure 8B). Among them, CBX3 and CBX8 are the two most frequently mutated genes.
Then, we used GSCALite to evaluate the role of the CBXs in ESCA pathway activity. And we found that most members of CBXs are involved in the activation of DNA damage response, EMT and hormone AR pathways. Also, we found that CBX6/7 are mainly related to the inhibition of cell cycle and apoptosis pathway (Figure 8C).
3.4 Immune Infiltration of CBXs in ESCA Patients
In addition, we also used the TIMER to explore the relation between the CBXs and immune infiltration. The results depict that CBX1 is significantly correlated with the abundance of B cells (cor = 0.071, P = 3.48e-01) and macrophages (cor = 0.171, P = 2.20e-02) (Figure 9). Also, we found that the CBX2, CBX5, CBX6 and CBX8 expression levels have positive correlations with B cells, CD4+ cells, and macrophages. Besides, CBX6 is also significantly related to the abundance of dendritic cell (cor = 0.03, P = 6.92e-01) (Figure 9). Interestingly, we found that CBX7 is positively correlated with these 6 immune cells. Meanwhile, the abundance of other immune cells (CD4+ cells and dendritic cells) is negatively correlated with CBX3 and CBX4. All these results reveal that CBXs may have a certain effect on immunotherapy of ESCA.
3.5 Enrichment and PPI Analysis of CBXs in ESCA
Further, to explore the biological functions of CBXs in ESCA, the GO and KEGG analysis were also carried out. We first used GEPIA to investigate the top 100 genes associated with each CBX member, and screened out the 50 most frequently alterd adjacent genes. Then, we submitted the CBX family and related genes to DAVID for enrichment analysis. The results reveal that the CBXs participated in various biological processes, including the negative regulation of RNA polymerase II transcription through promoters, DNA replication, negative regulation of transcription, and mRNA splicing through spliceosome (Figure 10A). With respect to the cellular component, the CBXs are involved in nuclear chromatin, nucleus, nuclear ribonucleoprotein complex, and nucleoplasm (Figure 10A). In addition, molecular function analysis disclosed that the CBX-related genes were enriched in chromatin binding, protein binding, ATP binding, DNA binding, and enzyme binding (Figure 10A). Besides, the results of KEGG show that the CBXs and related genes are enriched in mismatch repair, DNA replication, cancer pathways, and spliceosomes (Figure 10B). Obviously, these signaling pathways are involved in the tumorigenesis and pathogenesis of ESCA.
Additionally, in order to show the direction and degree of association between different nodes of CBXs and related genes, and to indicate the type and intensity of interaction between CBXs and related genes, we constructed a PPI network. The results show that the CBXs is related to nuclear chromatin, histone blinding,PcG protein complex, chromatin, transcriptional coactivator activity, and nuclear ubiquitin ligase complex (Figure 11). All these pathways are closely related to gene coding and transcription. Consequently, we can speculate that CBXs play certain roles in the generation and development of ESCA.