2.1 Differential expression and co-expression analysis of CBX between tumor and normal samples
First, the differences in CBXs mRNA expression between 22 tumor samples and normal tissues were compared. It was then determined whether they were upregulated or downregulated in tumor tissues. Specifically, CBX2/CBX4/CBX8 were highly expressed in most types of tumors. CBX7 was the only gene commonly downregulated in CBX family. However, CBX6 was inconsistently upregulated or downregulated in different cancers. The expression of CBXs caused a proportional predominance of up-regulation or down-regulation in pan-cancer, suggesting that they may play a role in tumor development. cBioPortal online database is available to understand CBXs genetic alterations,The spectrum of genetic alterations in CBX2/CBX4/CBX8 suggests that its amplification was one of the most important single factors for alteration in SKCM,BRCA,OV,LIHC,UCS,MESO,etc.In addition,CBX6/CBX7 have less mutation and amplification.Results are shown in Figure 1.
2.2 Correlation analysis of CBXs expression and survival
The distinct differential expression pattern of CBXs in pan-cancer prompted us to explore their prognostic value.We further tested whether the expression of CBX could predict the survival of patients. For survival analysis, all cancer types were analyzed by Kaplan-Meier. According to the median value of gene expression, patients were divided into two groups: high and low expression groups. The differential expression was expressed by P value, and the P < 0.05 meant significant difference. The results of P value were shown in Table Ⅰ. According to Figure 2, The high expression of CBX2 in ACC, KIRP, MESO and CBX4 in ACC, KIRC and KIRP predicted poor prognosis.In contrast, CBX6 and CBX8 show inconsistent prognostic outcomes in different cancers. CBX7 was highly expressed in pan cancer and showed better prognosis. In the analysis of clinical stage correlation (Figure 3), it was found that compared to stage Ⅰ, CBX2 and CBX4 were expressed more in stage Ⅱ and Ⅲ patients (P < 0.05). It indicates that the increase in its possible expression may lead to poor survival results. The expression of CBX7 decreased with the increase of clinical stage. Cox proportional hazard regression model was used to measure the prognostic value of CBXs in 33 types of cancers. CBX with HR > 1 was considered as a risk factor for this type of cancer, which was unfavorable to the prognosis. The forest map (Figure 4) showed that CBX4 had pan cancer significance in ACC, DLBC, ESCA, KICH, KIRC and KIRP, HR > 1. CBX2 in DLBC and THCA, CBX6 in PCPG, CBX7 in PCPG and PRAD, and CBX8 in TGCT and THYM were considered to be favorable to the survival, HR < 1. In this part, CBX6 does not have uniform consistency in pan-cancer, considering that it may not have the potential of tumor markers.In some cancer types especially exhibited recurrent negative associations between high CBX2 mRNA levels and lower survival,and CBX7 predicted a better prognosis. Therefore, our study indicated that the majority of its high expression and poor prognosis suggest CBX2, possibly as an adverse factor , and CBX7 may be a protective factor in patients .
Table 1
survival P value result
Gene
|
CancerType
|
P Value
|
CBX2
|
ACC
|
2.69E-07
|
CBX2
|
BRCA
|
0.04319
|
CBX2
|
KIRP
|
0.001561
|
CBX2
|
LAML
|
0.024321
|
CBX2
|
LGG
|
0.03992
|
CBX2
|
LIHC
|
0.001288
|
CBX2
|
MESO
|
0.000215
|
CBX2
|
OV
|
0.042056
|
CBX2
|
PCPG
|
0.035438
|
CBX2
|
THYM
|
0.031716
|
CBX4
|
ACC
|
0.012857
|
CBX4
|
KIRC
|
4.99E-08
|
CBX4
|
KIRP
|
0.001216
|
CBX6
|
KIRC
|
0.046466
|
CBX6
|
KIRP
|
0.029178
|
CBX6
|
LGG
|
2.68E-06
|
CBX6
|
PAAD
|
5.99E-05
|
CBX6
|
PCPG
|
0.005805
|
CBX6
|
READ
|
0.019745
|
CBX6
|
THYM
|
0.018514
|
CBX6
|
UVM
|
0.04051
|
CBX7
|
ACC
|
0.021448
|
CBX7
|
CESC
|
0.002526
|
CBX7
|
KIRC
|
0.001729
|
CBX7
|
KIRP
|
0.008273
|
CBX7
|
LGG
|
0.002094
|
CBX7
|
LUAD
|
0.024615
|
CBX7
|
MESO
|
0.000256
|
CBX7
|
OV
|
0.012516
|
CBX7
|
PAAD
|
0.04281
|
CBX7
|
SARC
|
0.006135
|
CBX7
|
SKCM
|
0.005267
|
CBX7
|
UVM
|
0.048189
|
CBX8
|
ACC
|
4.21E-05
|
CBX8
|
KIRC
|
2.38E-08
|
CBX8
|
MESO
|
0.040008
|
CBX8
|
PAAD
|
0.000995
|
CBX8
|
STAD
|
0.036372
|
CBX8
|
THYM
|
0.02429
|
2.3 Correlation analysis of CBX gene and immune response
Figure 5 showed the expression of CBX in different subtypes of immune infiltration. Among the 6 immune subtypes, C3 and C5 subtypes were associated with better survival rate, while C4 and C6 subtypes were on the contrary (15). Different CBX genes had different correlation patterns with different immune subtypes. The expression of CBX6 and CBX7 were increased in C5 and decreased in C1, C2 and C6, suggesting that higher expression of CBX6 and CBX7 might have better prognosis. The highly expressed genes were correlated with the favorable immune components, which indicated that these genes might play an anti-tumor role to some extent. On the contrary, CBX2 was highly expressed in C1 and C2, suggesting that CBX2 played a pro-tumor role. There was no significant difference in the expression of CBX4 among all 6 subtypes. The different expression patterns of CBX in different immune subtypes indicated that the function of each gene might be immune subtype dependent.Although the relationship of CBXs in them varies by tumor types,At least we observed that CBX2 and CBX7 have different roles in tumorigenesis, CBX2 may act as a covariate of patient survival and promote the aggressive progression of the disease, while CBX7 may be a tumor suppressor gene.CBX4,CBX6, and CBX8 cannot determine their homogeneous properties due to the incomprehensible laws in tumor types.Combined with the previous performance,In the following studies, we will focus on CBX2 and CBX7 to discuss the impact of their different expressions in tumors.
2.4 Stemness Indices and Microtumor Environment in Pan-Cancer
An increasing number of reports indicate that the TME plays a crucial role in tumor initiation and progression.Therefore, it is important to further explore the pan-cancer relationship between TME and CBXs expression.According to the estimation algorithm, a correlation matrix was drawn to show the relationship between the expression level of CBX family genes and the stromal scores of 33 different types of cancer. Spearman correlation analysis was generated. As shown in Figure 6A-C, the positive correlation between the CBX gene and stromal cells and immune cells in tumors is shown in red. It can be seen from the graph that the CBX family genes are negatively correlated with the three scores in most tumors. This means that, the higher the gene expression, the lower the content of stromal cells and immune cells. Furthermore, relatively speaking, tumor cells have increased.Our results demonstrate that CBX2 expression is significantly negatively correlated with immune scores in GBM, UCEC, CESC, LUAD, ESCA, SARC, STAD, HNSC, LUSC, SKCM, BLCA, THCA, TGCT, LAML and ACC, but except in PRAD, BRCA, THYM and KICH.According to the R and P values, The correlation between CBX2 or CBX7 expression in the top 4 tumors and the three scores is shown in Figure 7.In this study, we found that the expression level of CBX2 and CBX7 gene was significantly correlated with immune cell infiltration of T cells ,NK cells,Monocytic lineage,The latter was also significantly associated with Neutrophils, Endothelial cells infiltrating cells.As shown in Figure 8, CBX2 could at least reflect immune status of GBM, SARC, LUAD, LUSC,etc.CBX2 reduces the number of immune cells, promotes the increase in the number of tumor cells, and may also shape an inflamed TME for tumor progression, while CBX7 can increase the number of immune cells, contribute to immune surveillance, and reduce tumorigenesis.
2.5 Correlation analysis of CBX gene and drug sensitivity
Figure 9 showed the scatter diagram of the correlation between CBX gene expression and drug sensitivity in P value order. The abscissa was the gene expression, and the ordinate was the Z-score. P< 0.05.As shown in Figure 9, CBX2 was positively correlated with the sensitivity of Acrichine (cor =0.428, P<0.001), Curcumin (cor=0.418, P<0.001), Nelarabine (cor=0.412,P<0.001), Ixabepilone (cor=0.408, P<0.001), Ifosfamide (cor =0.386, P<0.001), tfdu (cor=0.366, P=0.004) and Tamoxifen (cor = 0.365, P= 0.004), but was positively correlated with the resistance of Dasatinib (cor = -0.437, P<0.001) and Staurosporine (cor = -0.673, P<0.001). The high expression of CBX7 was positively correlated with the drug resistance of cells to Aminoflavone (cor = -0.482, P < 0.001), Afp464 (cor = -0.399, P =0.002) and lificguat (cor = -0.389, P=0.002). We also noted that CBX2 and CBX4 were correlated with the drug resistance of cells to Dasatinib and Staurosporine.