Liver cancer (LC) is the fourth leading cause of cancer-related deaths worldwide (Llovet et al., 2021; Siegel et al., 2021). Hepatocellular carcinoma (HCC) is the second highest cause of cancer-related deaths globally, accounting for 85–90% of all primary liver malignancies (Bosch et al., 2005). Although alpha fetoprotein (AFP) has low specificity as a serologic diagnostic marker for hepatocellular carcinoma, its sensitivity remains relatively high (around 60%) (Trevisani et al., 2001). Previous studies have reported use of tumor stage in HCC prognosis, with curative therapies for early-stage HCC providing a 5-year survival rate of more than 70% (Llovet et al., 2016; Villanueva, 2019). However, most HCC patients are diagnosed during the late stage of disease development. Notably, prognosis of HCC patients has markedly improved due to advancements in surgical procedures coupled with the use of targeted medications. On the other hand, the median survival rate for asymptomatic advanced cases undergoing systemic therapy is 1–1.5 years (Llovet et al., 2008; Kudo et al., 2018). Therefore, early detection, coupled with use of effective therapy, and prognostic assessment of HCC cases are critical to effective management of the disease.

To date, up to eight members of the CBX family have been discovered in the human genome. CBX family proteins can be classified into two categories based on their molecular structure. Each of these can be further stratified into two subgroups, namely the heterochromatin protein 1 (HP1) (which includes CBX1, CBX3, and CBX5) and the polycomb (Pc) (which comprises CBX2, CBX4, CBX6, CBX7, and CBX8) groups. Specifically, the chromobox (CBX) family proteins represents typical members of the Pc group complex that regulate heterochromatin control, gene expression, and developmental programs (Desai and Pethe, 2020). Previous studies have shown that the HP1 group comprises both N-terminal and a C-terminal chromosome structural domains, whereas the PC group only contains a conserved N-terminal chromosome structural domain (Wotton and Merrill, 2007).

Numerous studies have described the role of members of the CBX family in malignancies of the brain and central nervous cancer (Duan et al., 2018), breast cancer (Iqbal et al., 2021), colorectal cancer (Wang et al., 2016), head and neck cancer (Wang et al., 2017) and sarcoma (Zhou et al., 2021). However, information on their function in HCC is dearth. The Most of the studies have focused on single nucleotide polymorphisms of CBX4 and CBX7 and their association with reduced risk of hepatocellular carcinoma (Tan et al., 2019). Previous research evidence demonstrated that CBX 1/2/3/6/8 is an independent predictor of survival HCC patients (Ning et al., 2018), whereas CBX4 was found to stimulate angiogenesis in HCC via its SUMO E3 ligase activity (Li et al., 2014). To date, however, the role of various members of the CBX family in development and progression of HCC remains unknown. Here, we analyzed expression and mutations across various members of the CBX family members in HCC patients, as well as their relationships to clinical indicators, with a view of elucidating their roles in development and progression of the disease.

Controlling gene expression precisely is critical for cell function. Although transcription is ultimately guided by DNA-binding transcription factors, it is also necessary for chromatin and histone post-translational modification. Eight CBX proteins have been found in the human genome, all of which have a similar chemical structure and contain a single N-terminal chromodomain. All of the CBX proteins are involved in heterochromatin control, gene expression, and developmental programs. CBX2, CBX4, CBX6, CBX7, and CBX8 are conventional PRC1 components that cooperate with PRC2 to regulate gene expression. PRC1 CBX protein components identify and bind to H3K27me3 mediated by PRC2, hence recruiting PRC1 to transcriptionally regulated target genes. The remaining three CBX proteins (CBX1, 3, and 5) are referred to as heterochromatin protein 1 (HP1), HP1 (HP1), and HP1 (HP1). They (CBX1, 3, and 5) are heterochromatin protein 1 (HP1) complex-associated methyl readers that decode H3K9me3 marks generated by H3K9 methyltransferases (van Wijnen et al., 2021). Using nonhomologous protein sequences, unique polycomb group CBX proteins were associated with diverse areas of chromatin, implying that each polycomb group protein has a distinctive target. Notably, members of the CBX family have been implicated in the development of a range of malignancies, including HCC (Supplementary Figure S1) (Kulis and Esteller, 2010). Despite the fact that numerous members of the CBX family have been implicated in the development of HCC, the precise functions of others remain uncertain. The primary goal of this work was to examine the expression patterns, prognostic value, and immune infiltration of several CBX family members in HCC and to develop a nomogram model to improve patient prognosis.

Our findings indicated that CBX1 and CBX8 mRNAs were significantly upregulated in HCC tumor, relative to adjacent normal, tissues. Apart from CBX2, whose data was missing, all other CBX family proteins were significantly differentially expressed between tumor and normal tissues in HCC patients. Moreover, CBX1/3/4/5/8 was significantly upregulated in tumor, relative to normal tissues, which was consistent with the trend of expression observed in the TGCA database. The expression pattern for CBX1/3/4/5/8 proteins was also consistent with findings from previous studies (Jiao et al., 2015; Yang et al., 2018; Tang et al., 2019; Zhong et al., 2019). Although expression of CBX6/7 proteins were significantly upregulated in tumor tissues, its mRNA levels were almost similar between tumor and normal tissues based on the TCGA database. There was no consistency between mRNA and protein expression for this factor. Our results on CBX6 protein expression were in contrast to previous findings (Wang et al., 2020a), thus datasets from larger clinical trials are required to validate this phenomenon. Expression of CBX1/2/3/5 mRNA was significantly associated with each patient’s cancer stage, suggesting that it be playing a crucial role in HCC progression. Based on this, we hypothesize that the HP1 subtype plays a direct role in HCC progression. Additionally, CBX4 and CBX8 of the Pc subtype may exert their effects on HCC via post-transcriptional changes such as phosphorylation and sumoylation (Zhan et al., 2018; Wang et al., 2020b). The particular mechanism via which the Pc subtype affects HCC requires experimental confirmation. To summarize, the CBX family was expressed significantly differentially in HCC, and the trend of expression of the eight CBX families were inconsistent. The mechanism by which the CBX family is involved in HCC requires further investigation.

Our results further indicated that expression of members of the CBX family was strongly associated with prognosis of HCC patients. Moreover, expression of CBX1/2/3/6/8 was significantly associated with prognosis of HCC patients, with its upregulation correlated with poor OS rates. Considering the high expression of members of the CBXs family in HCC, we postulate that CBX1/2/3/8 and CBX6 are oncogenes. Results from LASSO regression analysis indicate that CBX2 and CBX3 were more essential for survival of HCC patients than the other five genes. Additionally, CBX2 and CBX3 acted as oncogenes in HCC, as illustrated in Figure 3B. This result was consistent with our preceding conclusion.

There was a significant association between the CBXs family and the prognosis of HCC. Using LASSO regression, we determined that CBX2 and CBX3 were more essential for survival prognosis than the other five genes. Additionally, as illustrated in Figure 3B, CBX2 and CBX3 operate as oncogenes in HCC. This finding is completely consistent with the prior conclusion. The Nomogram supports our hypothesis that CBX3 expression and T-stage together have an effect on the OS prognosis of HCC, that CBX3 has a significantly bigger effect on HCC than T-stage, and that CBX3 will function as an independent prognostic biomarker for HCC. Simultaneously, we validated our results against a variety of other databases (Supplementary Figure S5). However, we discovered no statistically significant relationship between the N and M phases and the prognosis of HCC, which spreads primarily through hematologic metastasis and extremely rarely via lymph node metastasis. Although lymph nodes are the second site of extrahepatic metastasis in hepatocellular carcinoma, their rarity in HCC may explain why HCC is not associated with N stage (Liang et al., 2022). Intrahepatic metastasis was the most frequent kind of metastasis in HCC, while extrahepatic metastasis was less frequent. There were four patients with M1 and nine individuals with Mx in TGCA-HCC (Table 1). Inadequate data may skew the outcomes. This study contains several flaws. TGCA-HCC did not include liver function or vascular invasion (Bruix et al., 2016). These two indicators have a strong predictive value for the prognosis of HCC. Investigators should incorporate liver function and vascular invasion in subsequent big clinical trials. This has a stronger bearing on clinical work.

After identifying CBX3 as an independent prognostic factor for HCC, we analyzed its role in patient prognosis. Firstly, we eliminated the idea that CBX3 was mutating and impacting HCC development. We detected CBX3 mutations in 1.6% of the 366 HCC patients, but found no statistically significant differences in outcome between persons with and without CBX3 mutations (Supplementary Figure S2A). Further analysis indicated that CBX3 mutations did not appear to play a role in HCC progression (Supplementary Figure S2B, C). Next, we employed the HPA database to analyze the relationship between CBX3 expression and immune cell infiltration and found that CBX3 was significantly upregulated in immune cells relative to other members of the CBX family (Supplementary Figure S3). Moreover, CBX5 was dramatically elevated in T cells, but remained significantly downregulated relative to CBX3. Secondly, we found that CBX3 was significantly upregulated in immune cells relative to other cell subpopulations, with B cells and T cells being the most abundant. Results from the TIMER database indicated that CBX3 expression was strongly correlated with B- and DC-cell infiltration. Notably, CBX3 most likely affected the microenvironment of HCC tumors via B cells, CD4⁺ T cells, and DC cells. Recent research targeting HCC and TME have demonstrated that both CD4⁺ and CD8⁺ T cells possess anticancer potential. In fact, researchers have proved that CD8⁺ T cells and the widely used PD-1 therapy significantly improve prognosis of HCC patients (Iwai et al., 2002), whereas CD4⁺ T cells are also components of activated CD8⁺ T cells (Borst et al., 2018). Collectively, these findings demonstrated the crucial role played by T cells in the immunological treatment of HCC, which has subsequently expanded the use of DC cells in liver immunotherapy. A combination of DC immunization and PD-L1 inhibitors has shown promise as a novel treatment approach for HCC (Teng et al., 2020). In fact, a dose-dependent concentration of this regimen resulted in significantly longer overall survival, decreased tumor volume, and increased tumor cell apoptosis in mice relative to either treatment alone (Teng et al., 2020). This was achieved by inducing a stronger anti-tumor cytotoxic T cell response. Results of the present study indicated that CBX3 expression was strongly correlated with infiltration of DC cells, suggesting that CBX3 may be a key target for future development of DC cells in hepatocellular carcinoma. However, there was a substantial link between CBX3 expression and B cells, although there are diverse opinions regarding the role of B cells in HCC. While B cells have been shown to promote CTL activity through secretion of cytokines, cytokines released by B cells reportedly promote angiogenesis and tumor formation (de Visser et al., 2005; Tsou et al., 2016). Immunotherapy is expected to drastically improve prognosis of HCC patients in the near future.

Since proteins represent the primary mode of execution for CBX family functions, we analyzed expression of proteins known to interact with CBX3. Results indicated that CBX3 has a strong affinity for the CBX and HIST families. Notably, CBX3 was expressed in the nucleus, and is a member of the HP1 subfamily of the CBX family (Supplementary Figure S4A). Previous studies have shown that HP1 proteins directly interact with the methylated H3K9 promoter region, thereby regulating chromatin, gene expression, and growth and development (Lomberk et al., 2006; Bot et al., 2017). Additional research evidences have associated aberrant expression of members of the HP1 group with progression of several human diseases and malignancies (Thorne et al., 2012; Slezak et al., 2013). In the present study, GO terms revealed that CBX3 was not only involved in nuclear chromatin regulation and Histone methyltransferase activity, but also played a role in regulation of the cell cycle (Supplementary Figure S4B). Moreover, CBX3 was not only highly expressed in the G1 phase but also acted as a barrier to cell progression from the G0 to the G1 phase (Supplementary Figure S4C, D). Therefore, CBX3’s unique and central biological functions suggests its potential as a target for development of treatment therapies for HCC patients.

Currently, HCC is mostly treated with three types of drugs. Chemotherapy is predominantly FOLFOX4, targeted treatment is predominantly sorafenib, and immunotherapy is either atenizumab plus bevacizumab or sindilizumab plus bevacizumab. The website GDSC estimated the sensitivity of CBX3^high HCC patients to standard chemotherapy and targeted therapies (Supplementary Figure S6). As can be shown in Supplementary Fig. 6D, CBX3^high HCC patients were responsive to sorafenib but insensitive to platinum chemotherapeutic treatments (Supplementary Figure S6A). As a result, platinum chemotherapy agents should be considered when treating CBX3^high HCC patients with the usual FOLFOX4 regimen. Sorafenib was indicated for CBX3^high HCC patients.

In conclusion, CBX3 is an independent prognostic factor for HCC. Functionally, it inhibits development of HCC tumor via immune cell infiltration. Overall, these findings indicate that CBX3 is a novel biomarker for hepatocellular carcinoma, and a potential target for future development of therapies for HCC treatment.