Hepatocellular carcinoma (HCC) is the main type of primary liver cancer, it is clinically, molecularly and biologically heterogeneous, and highly resistant to treatment. Recently, the survival of patients with hepatocellular carcinoma (HCC) has been prolonged with improvements in various diagnostic tools and medical treatment modalities. However, there is a wide range of factors affecting the prognosis of patients, such as molecular changes, oncogenic mutations, tumor recurrence and metastasis, which are closely related to the prognosis of patients. Therefore, finding suitable targets to predict patients' prognosis and provide the potential therapeutic target in cancer treatment is of great practical significance to improve the outcome of hepatocellular carcinoma.
POLE is the core of the catalytic subunit of DNA polymerase epsilon (Polε), also known as POLE1, FILS and IMAGEI, etc. POLE contains both the catalytic active site and the proofreading exonuclease domain. Accordingly, the POLE gene confers to Polε both replicative and 3 to 5 repair capabilities for the new strand. Previous studies have suggested that different Polε gene members are associated with different stages of several different cancers, for example, Ma et al. [13] pointed out in their study that mutations in POLE are associated with the development of various tumors; Zhu et al. [14] found that people with POLE mutations are more prone to gastrointestinal tumors. However, the study of Raffone A et al.[15]pointed out that patients with POLE mutations have increased numbers of tumor-infiltrating lymphocytes in endometrial cancer, suggesting a better prognosis, this seems to contradict the previous studies to some extent. On the other hand Galati et al.[16] attempted to incorporate CRISPR/Cas9-mediated knock-in in a POLE mutant mouse model, and the results showed that the model exhibited a significantly lower survival rate, providing insights into human and mouse genotypic phenotypes and revealing POLE-associated mutational features. POLE2 is the second largest of Polε with a relative molecular mass of 59,000, it has also been mentioned that upregulation of POLE2 suggests deterioration and poor prognosis in patients with esophageal squamous carcinoma[17], Ye et al.[18] similarly showed in their study that high expression of POLE2 predicts poor prognosis in lung adenocarcinoma, and it also found that the expression level of POLE2 in NSCLC cells was significantly higher than that in human bronchial epithelial-like cells and overexpression of POLE2 promoted the proliferation and invasive migration of A549 cells and upregulated the expression of MMP9, which in turn promoted the invasive migration of NSCLC cells[19]. POLE3 and POLE4 are the other two smaller subunits with relative molecular masses of 17,000 and 12,000, respectively. POLE 4 was observed to be associated with tumor proliferation and invasion in neuroblastoma[20]. MiR-7-5p inhibits tumor cell proliferation, migration and invasion by targeting different genes[21]. It has been observed that miR-7-5p inhibits proliferation and migration by regulating the PI3K/Akt signaling pathway in HCC and glioblastoma migration[22], and it also targeted the binding of POLE4 to inhibit the progression of non-small cell carcinoma[23], and the high expression of POLE4 also correlates with melanoma progression and overall survival[24]. As for POLE3, it is mainly involved in the mitotic G1 phase, G1/S transition and DNA repair pathway. It is also associated with macular degeneration, aging, etc. Up to now, there is no mention of its association with tumor processes in the literature. And POLE1/2/3 are involved in synthetic regulation and cofactor binding[9].
Our analysis showed that the Polε gene members are significantly upregulated in HCC, and associated with lower OS and DSS in HCC patients. And also showed significant correlations between different Polε gene members and different clinical features, such as the expression level of POLE/POLE3 correlated with clinicopathological stage, and the expression of POLE2/POLE3 and POLE4 correlated significantly with the histologic grade of patients, respectively. the expression level of POLE1/POLE2/POLE3 correlated with the AFP expression level of patients, etc. It is suggested that the Polε gene members are closely associated with the proliferation and migration of tumor cells. By further exploration we found that POLE2 and POLE3 showed better predictive efficacy in both diagnostic ROC curves and time-dependent ROC curves for disease-specific survival, suggesting the better disease diagnostic efficacy of the two genes. All of the above suggested that POLE1/2/3/4 are prognostic markers of HCC and that POLE2/3 are potential diagnostic targets for HCC.
Next, unsupervised cluster analysis and typing of HCC using the expression of the four target genes showed that the samples were clearly classified into two subtypes and there were significant differences in survival prognosis between the two subtypes, and similar results were obtained by validation of their principal component analysis, suggesting that the four genes POLE1/2/3/4 are important characteristic genes in HCC. Further studies on them may bring new research hotspots for the diagnosis and treatment of HCC.
Functional and pathway studies of genes/molecules are important prerequisites for understanding their biological functions, and detailed functional/pathway studies largely help to further solve clinical problems. Our results show that the co-expressed genes of the four genes are mainly enriched in cell cycle, chromosome-related functions, nuclear division, DNA replication, and DNA enzyme catalysis, etc. They play crucial roles in these components and functional levels, and through these functions and pathways, they directly or indirectly influence cell cycle processes and processes such as value-added differentiation. Any biological process that disrupts this homeostasis will play a direct or indirect role in the pathogenesis and development of cancer.
DNA methylation is an epigenetic modification that is involved in several cellular processes including senescence, tumor progression, proliferation, apoptosis, autophagy and immunity by regulating gene transcription and chromatin structure [25–27], and our results showed that POLE in HCC had significantly elevated methylation levels in all CpG islands and their downstream regions. And by further exploration, the high methylation levels of POLE were associated with lower survival rates, suggesting that its hypermethylation process plays a currently unknown promoting function in the development and progression of HCC, thus reducing patient survival. Our study also found that POLE2/POLE3/POLE4 also showed higher methylation levels in some microarray data and that POLE2/POLE4 hypermethylation levels were shown to be associated with poorer patient survival in some regions, and POLE3 hypermethylation was associated with a better prognosis for patients. However, overall, the methylation levels of POLE2/POLE3/POLE4 were not uniform in most CPG islands and their downstream regions, and we could not draw clear conclusions from them.
In recent years, immune infiltration has increasingly become a hot topic and many studies have shown that it plays an important role at different levels in various cancers[28–29]. It has been proposed that some immune cells have both tumor-suppressive and tumor-promoting effects and are key determinants of immunotherapy efficacy and patient clinical prognosis[30–31]. We explored the phase relationship between the expression levels of four Polε gene members and the infiltration of eight common immune cells, and the four target genes in the study all showed a high correlation with the abundance of immune cell infiltration, especially with the infiltration of CD4 + T cells, macrophages, and dendritic cells. The high correlations of them suggest that the immune process of HCC primarily involves CD4 + T cells, macrophages and dendritic cells and play an immune role that is closely related to the four target genes and affects different stages of cancer. It is therefore easy to speculate that the different Polε gene members reveal to some extent the immune status of HCC patients and the prognosis of the patients.
We also analyzed the correlation between the four target genes and classical immune targets and the results showed that POLE with CD276/TNFSF4, POLE2 with ADORA2A/TNFSF4, POLE3 with TNFSF15/LAIR1/NRP1/CD80/CD86/CD276/
CD200R1/TNFSF4/CD274(PDL1)/HAVCR2/ICOS, POLE4 with CD70/LAIR1/
TNFSF4/TNFSF18/CD86/TNFSF9/LGALS9/LAG3/TNFSF9/CTLA4 all showed significant positive correlation (r ≥ 0.25), it is evident that there is a significant correlation between target genes and common immune targets, while targeted drugs based on immune targets such as PDCD1 (PDL1) and CTLA4 have been successfully marketed and applied, and further exploration of these immune checkpoints may lead to the discovery of new immune-targeted drugs. And the PD-1 therapeutic sensitivity between POLE1/2/3/4 high and low expression groups in multiple cancers, including liver cancer, was further analyzed comprehensively, and the results showed that the POLE high expression group showed better therapeutic sensitivity. This agrees with that although the mutation rate of POLE is not high, patients with POLE mutations have a younger incidence and lower disease recurrence and mortality rates as found by Churchet al.[32], and Husseinet al.[33] also pointed out that patients with POLE mutations showed a better prognosis compared to wild type in endometrial cancer cases. It has been hypothesized that in POLE mutated tumors, correction of loss of function causes enhanced genetic instability, and tumor point mutations increased, leading to loss of normal function and the accumulation of a large number of mismatched bases resulting in a large number of incidental mutations that can induce an early immune response, thereby inhibiting the proliferation and differentiation of tumor cells and even killing them[34]. The macroscopic manifestation is that high mutations and methylation levels in POLE are associated with a better prognosis for patients, which cautions that high expression of POLE may serve as a potential predictor of the effect of targeted therapies.
Liver cancer is generally insensitive to radiotherapy. Hepatocellular carcinoma develops insidiously and most patients miss the best treatment period when diagnosed, and the recurrence rate of liver cancer is high, and the sensitivity to chemotherapy drugs is further reduced after recurrence, therefore, finding new chemotherapeutic drugs and sensitizers has become an urgent problem for hepatocellular carcinoma treatment. At the end of the study, the prediction of drug IC50 based on the pRRophetic algorithm and TCGA gene expression profile was explored, and the results suggested that the high expression group of the four genes showed higher chemosensitivity to a variety of first- and second-line oncologic chemotherapeutic agents, especially the significantly upregulated group of POLE3 showed higher sensitivity to the classical drugs used in clinical chemotherapy for hepatocellular carcinoma, including sorafenib, suggesting that POLE3 might be used as a new biological factor to predict chemotherapy sensitivity in hepatocellular carcinoma. Moreover, the upregulated group of the four genes also showed higher sensitivity to dozens of chemotherapeutic drugs not currently used in hepatocellular carcinoma, suggesting that we can study the synergistic effects of the four genes on the sensitivity of these drugs and their mechanisms in a deeper way based on the present study, and further possibly explore new clinical drugs.
Finally, there are still some limitations in our study. The research samples in this study were taken from publicly available online data databases, and the results of this study are influenced by the results of the database data, and there is a certain amount of subjectivity in the interpretation of the results, etc.