UBE2T And CYP3A4: hub genes regulating the transformation of cirrhosis into hepatocellular carcinoma

Abstract

Cirrhosis is the primary driver of hepatocellular carcinoma (HCC). Blocking the deterioration of cirrhosis into HCC would be of benefit for long–term survival. We applied a series of bioinformatic online databases to select and analyze the hub genes concerning cirrhosis and HCC, and identified UBE2T and CYP3A4 as hub genes for cirrhosis-HCC transformation. An elevated UBE2T and a decreased CYP3A4 were expressed in HCC compared with cirrhosis, which have been confirmed by real-time polymerase chain reaction (RT–PCR), Western Blotting (WB) and immunohistochemistry (IHC). The specificity (89.9%) and sensitivity (74.1%) on the combination of UBE2T and CYP3A4 for predicting HCC development in cirrhosis patients were better than that of UBE2T or CYP3A4 alone, or alpha-fetoprotein. The effectiveness of the combination of UBE2T and CYP3A4 was further verified by 15 paired HCC and paracancerous cirrhosis samples using RT–PCR, with a specificity of 100% and a sensitivity of 80%. HCC patients with elevated UBE2T and decreased CYP3A4 expression demonstrated a poorer overall survival rate (P = 0.0016, 0.019) and disease-free survival rate (P = 0.0013, 0.041). In conclusion, UBE2T and CYP3A4 could be a new combination of biomarkers for the carcinogenesis and progression of cirrhosis into HCC.

KEYWORDS:

• Liver cancer
• hepatic fibrosis
• bioinformatic analysis
• seed gene

Introduction

As the fourth main cause of tumor-related deaths, liver tumors are a major threat to human health (Villanueva 2019). Hepatocellular carcinoma (HCC) encompasses 85–90% of all primary hepatic malignant tumors (Sartorius et al. 2015), the majority of which are caused by hepatitis-associated or alcohol-associated cirrhosis. It is therefore of great importance to understand the genetic processes by which cirrhosis develops into HCC.

Bioinformatics utilizes computers as tools to store, retrieve and analyze biological information; such ‘in silico’ systems are now frequently used to provide insight into the molecular mechanisms of disease. Most bioinformatic analyses of HCC have focused on comparing gene expression profiles between HCC biopsies and adjacent normal tissue samples (Li et al. 2017; Wang et al. 2017; Zhang et al. 2019), rather than comparisons between HCC and cirrhosis specimens, with few recent studies to further validate (Li et al. 2016; Xie et al. 2016; He et al. 2017; Chen et al. 2018; Fan and Ye 2018; Wu et al. 2019; Jiang et al. 2020).

For this study, we attempted to identify the gene expression discrepancies between HCC samples and cirrhosis samples; single out differentially expressed genes (DEGs); discover the hub genes; and build protein to protein interaction (PPI) networks of hub genes combined with their correlated genes by analyzing various bioinformatic online datasets to reveal the possible mechanism of cirrhosis developing into HCC, and further verify it by real-time polymerase chain reaction (RT–PCR), Western Blotting (WB) and immunohistochemistry (IHC).

Materials and methods

GEO datasets and GSE series

Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo) is an open format genomic database that contains array- and sequence-based data submissions (Edgar et al. 2002). GEO (https://www.ncbi.nlm.nih.gov/gds) contains multiple curated gene expression sets, as well as platform records and original series, in the GEO repository. We searched these datasets using the keywords ‘cirrhosis’ and ‘hepatocellular carcinoma’ and selected the study type as ‘Expression profiling by array’ and setting the organisms to ‘Homo sapiens’. This strategy identified eleven valid GSE Series: GSE63898 (Villanueva et al. 2015), GSE128726, GSE98383/107170 (Giacomo et al. 2018), GSE45050 (Darpolor et al. 2014), GSE89377, GSE17548 (Yildiz et al. 2013), GSE102451 (Torrecilla et al. 2017), GSE14323 (Mas et al. 2009), GSE3632, GSE10459 (Sakai et al. 2008) and GSE17967 (Archer et al. 2009) after eliminating duplicated data. The gene symbols were converted by the probes, as per the platform annotation information.

DEG screening

By using GEO2R (http://www.ncbi.nlm.nih.gov/geo/geo2r), the DEGs between cirrhosis and HCC were identified. These DEGs were chosen based on their p-values being significant after adjusting for multiple testing (adj. P-value) ∣logFC (fold change)∣ > 1 and adj. P-value < 0.05, and their having a log2-fold difference in expression between the two clinical conditions. DEGs with less than 10 samples in each group were excluded. Using GEO2R we selected the GSE groups which ranked in the top five in terms of unique element numbers as the final research objects.

DEGs’ GO and KEGG enrichment pathway research

We used the Database for Annotation, Visualization and Integrated Discovery (DAVID) bioinformatics resources version 6.8 (http://david.ncifcrf.gov) to analyze the DEGs’ biological features and functions (Huang et al. 2009a, 2009b). We used Gene Ontology (GO) for annotating genes and analyzing biological process (BP), cell component (CC), molecular function (MF) of these genes (Ashburner et al. 2000). We used the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/kegg/) for pathway analyses (Kanehisa et al. 2016). A P-value < 0.05 was considered statistically significant.

DEGs’ PPI network diagram and modularization research

We used the Search Tool for the Retrieval of Interacting Genes (STRING) (http://string-db.org) (Szklarczyk et al. 2019) version 11.0, to predict the DEGs’ PPI network. The biological interactions among proteins were analyzed to predict disease pathogenesis. A communication with a combined score > 0.4 was considered statistically significant. Subsequently, the PPI network diagrams were visualized via Cytoscape (version 3.6.0) (Shannon et al. 2003) and the key aspects were analyzed with MCODE (Molecular Complex Detection) (version 1.5.1) (Bader and Hogue 2003). We found clusters throughout the network based on the following filter criteria in MCODE: the cut-off of degree = 2, the cut-off of node score = 0.2, max depth = 100 and k-core = 2.

Hub genes’ screening and analysis

Through cBioPortal (http://www.cbioportal.org) (Cerami et al. 2012; Gao et al. 2013), we set the seed genes identified from MCODE as hub genes, and analyzed their expressions in HCC. We used Biological Networks Gene Oncology (BiNGO) (version 3.0.3) (Maere et al. 2005) to examine the hub genes’ BP, CC and MF annotations. We used the cancer genomics browser from UCSC (University of California, Santa Cruz) (http://genome-cancer.ucsc.edu) (Goldman et al. 2018), for multi-omic and clinical/phenotype data to build a hub genes-ordered cluster.

The expressions of hub genes related to hepatitis C virus (HCV) infection status, cancer type, HCC grade, vascular invasion and satellites, were examined with Oncomine (http://www.oncomine.com). Heat maps associated with the hub genes’ expressions in normal tissues versus those in HCC were also analyzed using Oncomine. Subsequently, the disease-free survival (DFS) and overall survival (OS) estimates of hub gene contributions were carried out through GEPIA (Gene Expression Profiling Interactive Analysis) (http://gepia.cancer-pku.cn/detail.php) (Tang et al. 2017).

Histological samples

The samples of HCC, paracancerous cirrhosis and distant cirrhosis tissues from 15 patients, and 15 normal liver tissues from liver transplant donors were retrieved from the biobank of Changzheng Hospital, Navy Medical University. All clinical samples used have obtained the informed consent of patients and donors. This study was approved by the Ethics Committee of Navy Medical University, and in line with the guidelines outlined in the Declaration of Helsinki.

Real-time polymerase chain reaction (RT–PCR) analysis

RT–PCR analysis was further used to perform validation of hub genes. Total cellular RNA was extracted from fresh tissue using TRIzol reagent (Invitrogen, 15596018). A spectrophotometer (NanoDrop 2000) was used to measure the quality and quantity of the RNA. For cDNA synthesis, 1 µg of total RNA was used as a template in a 20 µl reaction system using a reverse transcription kit (Takara, RR047A), and the cDNA was subjected to quantitative RT–PCR carried out with the TB Green qPCR kit (Takara, RR420A) via Real-Time System (Roche, LightCycler^® 480II). Primer sequences were as follows (Table 1).

Table 1. The primer used in the real-time polymerase chain reaction (RT–PCR) analysis.

Gene	Primer	Sequence
GAPDH	Forward Primer	5′-ACAACTTTGGTATCGTGGAAGG-3′
	Reverse Primer	5′-GCCATCACGCCACAGTTTC-3′
UBE2T	Forward Primer	5′-TTAGGTGGAGCCAACACACC-3′
	Reverse Primer	5′-TGCGATGTTGAGGGATGGTC-3′
CYP3A4	Forward Primer	5′-ATTTCCACCACCCCCAGTTAG-3′
	Reverse Primer	5′-CTCCTTAATGTGCAGGAAAGC-3′

Western blotting (WB) analysis

WB combined with RT–PCR will assist in validating hub genes more effectively. The total protein from each fresh sample was extracted with RIPA lysis buffer (Siwuwu Bio, SD0006) and centrifuged at 12,000 rpm for 10 min to remove debris, and protein concentration was measured using a BCA protein assay kit (Siwuwu Bio, SD0012). Equal amounts of each protein extract were electrophoretically separated on 10% SDS polyacrylamide gels. Following transfer to polyvinylidene fluoride (PVDF) membranes (Siwuwu Bio, SD0043), membranes were incubated with 5% skimmed milk powder in Tris-buffered saline containing 0.1% Tween-20 (TBST) for 1 h at room temperature to block non-specific binding, then incubated with the primary antibody against the hub genes (dilution 1:1000, abcam154022; abcam124921) overnight at 4°C. Anti-rabbit (dilution 1:20000, SD0039) secondary antibody was used. Finally, protein expression was determined using an enhanced chemiluminescence detection system (ECL Hypersensitive Chemiluminescence Kit, SD0032) in strict accordance with the manufacturer's protocols. Anti-β-Actin antibody (dilution, SD0034) served as a loading control. Image quantitation of band intensity was performed using the Image J software.

Immunohistochemistry (IHC) analysis

We selected oncogenes in hub genes for IHC to further verify the difference in the expression of these genes between liver cirrhosis and HCC.

Hub genes’ sensitivity and specificity analysis

We extracted the expression of UBE2T and CYP3A4 from patients with liver cirrhosis and liver cancer in GSE datasets, and calculated the cut-off value, sensitivity and specificity of UBE2T and CYP3A4 by the receiver operating characteristics (ROC) curve, and compared with that of AFP. According to the cut-off value, UBE2T and CYP3A4 were divided into UBE2T positive group and UBE2T negative group, CYP3A4 positive group and CYP3A4 negative group. In order to improve the sensitivity, the combined diagnosis of UBE2T and CYP3A4 was defined as satisfying either UBE2T positive or CYP3A4 positive. Then, we used the RT–PCR data of 15 pairs of samples to further verify the sensitivity and specificity of UBE2T, CYP3A4 and the combination of UBE2T and CYP3A4.

Statistical analysis

The box diagrams for comparison between groups were drawn by GraphPad Prism 7 (GraphPad Software Inc., La Jolla, CA, USA). SPSS 24.0 software (IBM, Armonk, NY, USA) was applied for statistical analyses. The Kruskal–Wallis one-way ANOVA and paired t test were used for comparison between groups. The ROC curve was used to analyze the sensitivity and specificity of hub gene. A P-value < 0.05 indicated a statistically significant difference.

Results

DEGs’ identification in HCC

After screening, five datasets were identified for further analyses: GSE98383, GSE63898, GSE17548, GSE89377 and GSE102451. The GSE98383 dataset contained 11 HCC tissue samples, 29 cirrhosis tissues samples and 1907 DEGs. The GSE63898 contained 228 HCC samples, 168 cirrhosis samples and 1107 DEGs. The GSE17548 contained 17 HCC samples, 20 cirrhosis samples and 462 DEGs. The GSE89377 contained 35 HCC samples, 12 cirrhosis samples and 284 DEGs. GSE102451 contained 27 HCC samples, 30 cirrhosis samples and 240 DEGs. The overlap among these five datasets identified 52 genes that are presented as a Venn diagram (Figure 1(A)) in which it can be seen that there are 13 up-regulated and 39 down-regulated genes, between cirrhosis and HCC tissues.

Figure 1. Venn diagram, PPI network and the significant modules of the differentially expressed genes (DEGs). (A) DEGs were selected with ∣logFC (fold change)∣ > 1 and adj. P-value < 0.05 among the mRNA expression profiling sets GSE98383, GSE63898, GSE17548, GSE89377 and GSE102451. The 5 datasets showed an overlap of 52 genes. (B) The PPI network of DEGs was constructed using Cytoscape. Up-regulated and down-regulated genes were marked in rectangle and oval, respectively. The hub genes UBE2T and CYP3A4 were marked with underline.

DEGs’ GO and KEGG enrichment pathway research

To analyze the DEGs’ biological classifications, GO and KEGG enrichment pathways research was performed using DAVID. The GO enrichment showed that the DEGs’ BP were associated with mitotic chromosome condensation, cell surface receptor signaling pathways, xenobiotic metabolic processes, exogenous drug catabolic processes, mitotic nuclear division and hyaluronan catabolic process. The DEGs’ CC were associated with the insulin-like growth factor ternary complex, the extracellular region, collagen trimers extracellular exosomes and spindle. The DEGs’ MF were associated with enzyme binding and oxidoreductase activity. KEGG enrichment pathway analyses identified the DEGs as being largely involved in chemical carcinogenesis, drug metabolism via cytochrome P450 (CYP450) and other enzymes, caffeine metabolism, retinol metabolism, and metabolism of xenobiotics by CYP450 (Table 2).

Table 2. GO and KEGG pathway enrichment analysis of DEGs in HCC samples.

Term	Description	Count	P-value
GOBP
GO:0007076	Mitotic chromosome condensation	3	9.26E-04
GO:0007166	Cell surface receptor signaling pathway	5	0.00950184
GO:0006805	Xenobiotic metabolic process	3	0.02344906
GO:0042738	Exogenous drug catabolic process	2	0.03585479
GO:0007067	Mitotic nuclear division	4	0.03943968
GO:0030214	Hyaluronan catabolic process	2	0.04461969
GOCC
GO:0042567	Insulin-like growth factor ternary complex	2	0.01093
GO:0005576	Extracellular region	11	0.010968
GO:0005581	Collagen trimer	3	0.026397
GO:0070062	Extracellular exosome	14	0.036567
GO:0005819	Spindle	3	0.04355
GOMF
GO:0019899	Enzyme binding	5	0.015696
GO:0016712	Oxidoreductase activity	2	0.042684
KEGGPATHWAY
hsa05204	Chemical carcinogenesis	4	0.003247709
hsa00983	Drug metabolism – other enzymes	3	0.012839084
hsa00232	Caffeine metabolism	2	0.018761194
hsa00830	Retinol metabolism	3	0.023991456
hsa00982	Drug metabolism – cytochrome P450	3	0.026861018
hsa00980	Metabolism of xenobiotics by cytochrome P450	3	0.031412565

DEGs: differentially expressed genes; GO: gene ontology; HCC: hepatocellular carcinoma; KEGG: Kyoto Encyclopedia of Genes and Genomes.

DEGs’ PPI network diagram and modularization research

By using Cytoscape, the DEGs’ PPI network diagram was constructed. Two significant modules in the PPI networks were identified using MCODE. UBE2T had the highest MCODE score in one significant module with 11 nodes and 55 edges, while CYP3A4 had the highest MCODE score in the other significant module with 3 nodes and 3 edges (Figure 1(B)).

Hub genes’ screening and analysis

UBE2T and CYP3A4 were recognized as hub genes. The full name, alternative name, location, chromosome and function of hub genes are shown in Figure 2. By applying the cBioPortal platform, a system of the hub genes combined with their correlated genes was examined (Figure 3(A)). The hub genes’ GOBP and GOMF analyses are presented in Figure 3(B). Ordered cluster analysis showed that the expression of the hub genes distinguished between HCC samples and cirrhosis samples (Figure 3(C)). In the Wurmbach Liver dataset, an elevated expression of UBE2T and a decreased expression of CYP3A4 were observed in HCC compared to cirrhosis. These same expression patterns were also found to be related to HCV infection status, tumor grade, vascular invasion and satellites compared to controls (Figure 4(A)). Oncomine analysis of normal tissue versus cancer showed that UBE2T was significantly over-expressed in diverse datasets in HCC (Chen et al. 2002; Wurmbach et al. 2007; Guichard et al. 2012), but the relative expression levels of the CYP3A4 gene varied among the datasets (Figure 4(B)) (Chen et al. 2002; Wurmbach et al. 2007; Mas et al. 2009; Roessler et al. 2010; Guichard et al. 2012). Finally, the OS and DFS attributable to the hub genes were calculated through GEPIA. HCC patients with elevated expression of UBE2T had worse OS (P = 0.0016) and DFS (P = 0.0013), and HCC patients with decreased CYP3A4 expression also had worse OS (P = 0.019) and DFS (P = 0.041) (Figure 4(C)).

Figure 2. Functional characteristics of hub genes in homo sapiens.

Figure 3. Interaction network, biological process and molecular function analysis of the hub genes. (A) Hub genes and their co-expression genes were analyzed using cBioPortal. Nodes with bold and thin outline respectively represent hub genes and the co-expression genes. (B) The biological process and molecular function analysis of hub genes were constructed using BiNGO. The filled nodes refers to the corrected P-value of ontologies. The size of nodes refers to the numbers of genes that are involved in the ontologies. A P-value < 0.05 was considered statistically significant. (C) Hierarchical clustering of hub genes was constructed using UCSC. Solid tissue normal samples, recurrent hepatocellular carcinoma (HCC) samples and primary HCC samples were involved. Upregulation of genes is marked in dark; downregulation of genes is marked in light.

Figure 4. The gene expression, overall survival and disease/progress-free survival analysis of UBE2T and CYP3A4 in hepatocellular carcinoma (HCC). (A) The expression of UBE2T and CYP3A4 associated with cancer type, HCC grade, hepatitis virus infection status, satellites and vascular invasion in the Wurmbach Liver dataset. (B) Heat maps of UBE2T and CYP3A4 gene expression in HCC vs. normal tissues were analyzed using Oncomine. (C) Overall survival and disease-free survival analysis of UBE2T and CYP3A4 were performed using GEPIA. A P-value < 0.05 was considered statistically significant.

Expression of UBE2T and CYP3A4 in human liver tissue

In order to further confirm the research results mentioned above, RT–PCR, WB and IHC analysis were used to examine levels of UBE2T and CYP3A4 expression in human tissues. We obtained 15 paired HCC tissue samples and their corresponding paracancerous cirrhosis tissues, cirrhosis tissues, and normal liver tissues to confirm UBE2T and CYP3A4 expression by RT–PCR and WB analysis. The results in RT–PCR analysis showed that levels of UBE2T mRNA expression in HCC tissues were significantly higher than that in paracancerous cirrhosis tissues and cirrhosis tissues (P = 0.001, <0.001) (Figure 5(A)), and CYP3A4 were significantly lower in HCC tissues (P = 0.020, < 0.001) (Figure 5(B)). However, there was no significant difference between cirrhosis tissues and paracancerous cirrhosis tissues in the expression of UBE2T and CYP3A4 (P > 0.05). The WB analysis results indicated that protein expression levels of UBE2T and CYP3A4 are increased and decreased during the carcinogenesis and progression of cirrhosis into HCC respectively (Figure 5(C)). The IHC results showed that the expression of UBE2T in HCC tissues was higher than that in cirrhotic tissues, while the result of CYP3A4 was just the opposite (Figure 5(D)).

Figure 5. The expression of UBE2T and CYP3A4 in samples of normal liver (n = 15), cirrhosis (n = 15), paracancerous cirrhosis (n = 15) and HCC (n = 15). (A–B) The different expressions of UBE2T and CYP3A4 were verified by RT–PCR. (C–D) The different expressions of UBE2T and CYP3A4 were verified by WB. (E) Representative grayscales of UBE2T and CYP3A4 expressions (n = 3 in each group). (F) The results of immunohistochemistry showed that the expression of UBE2T was higher in HCC tissues than that in cirrhotic tissues, and the expression of CYP3A4 was lower in HCC tissues than that in cirrhotic tissues. Notes: * P < 0.05, ** P < 0.01, *** P < 0.001.

The sensitivity and specificity of hub genes

ROC curves were used to analyze the sensitivity and specificity of hub genes. The sample data were collected from 259 patients with liver cirrhosis and 288 patients with liver cancer in the GSE98383, GSE63898, GSE17548, GSE89377 and GSE102451 datasets, and the expressions of UBE2T and CYP3A4 were determined by whole genome shotgun sequencing. The area under curve (AUC) of UBE2T was 0.834 (95%CI: 0.799∼0.869), with a specificity of 83.0% and a sensitivity of 76.1% at the cut-off value of 4.975. The AUC of CYP3A4 was 0.704 (95%CI: 0.657∼0.750), with a sensitivity of 54.2% and a specificity of 92.3% at the cut-off value of 11.204. The AUC of AFP was 0.529 (95%CI: 0.480∼0.577), with a sensitivity of 35.8% and a specificity of 82.2% at its cut-off value (Figure 6(A)). Then according to the cut-off values, we divided them into UBE2T positive group (UBE2T ≥ 4.975), UBE2T negative group (UBE2T < 4.975), CYP3A4 positive group (CYP3A4 < 11.204) and CYP3A4 negative group (CYP3A4 ≥ 11.204). The sensitivity and specificity on the combination of UBE2T and CYP3A4 was 89.9% and 74.1%, respectively. In order to further verify the prediction values of hub genes, 15 paired HCC tissues and cirrhosis tissues were used as validation cohort. The expressions of UBE2T and CYP3A4 in validation cohort were determined by RT–PCR. The AUC of UBE2T in validation cohort was 0.920 (95%CI: 0.822∼1.000) with a specificity of 100% and a sensitivity of 80% at the cut-off value of 4.580. The AUC of CYP3A4 was 0.862 (95%CI: 0.725∼0.999), with a sensitivity of 86.7% and a specificity of 80% at the cut-off value of 0.533. The AUC of AFP was 0.516 (95%CI: 0.282∼0.749), with a sensitivity of 40% and a specificity of 100% at its cut-off value (Figure 6(B)). The sensitivity and specificity on the combination of UBE2T and CYP3A4 were 100.0% and 80.0%, respectively.

Figure 6. ROC curves of UBE2T, CYP3A4 and AFP for predicting the development trend of hepatocellular carcinoma in patients with cirrhosis. (A) Through the datasets collected from GSE98383, GSE63898, GSE17548, GSE89377 and GSE102451, 259 samples with liver cirrhosis and 288 samples with liver cancer were involved. The areas under ROC curve of UBE2T, CYP3A4 and AFP were 0.834, 0.704, and 0.529, respectively. (B) For the validation cohort of 15 paired HCC tissues and cirrhosis tissues, the areas under ROC curve of UBE2T, CYP3A4 and AFP were 0.920, 0.862, and 0.516, respectively.

Discussion

Cirrhosis, can lead to a series of fatal complications, which include HCC, infection, upper gastrointestinal hemorrhage, hepatic encephalopathy (HE), hepatorenal syndrome (HRS), hepatopulmonary syndrome (HPS), electrolyte disorder and portal vein thrombosis (PVT) (Schuppan and Afdhal 2008). Cirrhosis and HCC place huge physical, psychological and financial burdens on patients and their families (Bajaj et al. 2011; Tapper et al. 2015; Patel et al. 2017). In the United States, the annual cirrhosis death rate increased by 65% to 34,174 between 1999 and 2016 and is expected to triple again by 2030. Moreover, the annual HCC death rate doubled to 11,073 between 1999 and 2016 and is expected to double again by 2030 (Estes et al. 2018; Tapper and Parikh 2018). Thus, the pathogenesis of cirrhosis progressing to HCC is of great interest.

The comprehensive genomic, proteomic and phosphoproteomic analyses of HCC using bioinformatic tools and online datasets have given us a deeper comprehension concerning the molecular mechanisms driving HCC (Cancer Genome Atlas Research Network 2017; Gao et al. 2019). If it was possible to identify and modify the expression of the hub genes which cause the deterioration of cirrhosis into HCC, the prevalence of HCC could be significantly reduced. In this study, we identified five GSE series using preset conditions, and found 52 DEGs using a Venn graphing program that included 13 up-regulated genes and 39 down-regulated genes. We then applied a module analysis and PPI network construction in order to obtain the hub genes UBE2T and CYP3A4.

UBE2T receives ubiquitin from an E1 complex and catalyzes its covalent binding to other proteins; it also functions as an explicit E2 ubiquitin binding enzyme by connecting with the E3 ubiquitin-protein ligase FA Complementation Group L (FANCL) of Fanconi anemia complex and catalyzing the monoubiquitination of FA Complementation Group D2 (FANCD2), which plays a pivotal role in the DNA damage pathway (Machida et al. 2006; Alpi et al. 2007, 2009; Longerich et al. 2009; Morreale et al. 2017). Additionally, UBE2T is involved in mitomycin-C (MMC)-induced DNA repair, catalyzing monoubiquitination, mediating monoubiquitination of FA Complementation Group I (FANCI) and FANCL (Machida et al. 2006; Alpi et al. 2007, 2009; Longerich et al. 2009), possibly contributing to BRCA1 degradation and ubiquitination (Ueki et al. 2009), and the promotion of polyubiquitination in vitro, particularly ‘Lys-63’, ‘Lys-48’, ‘Lys-27’ and ‘Lys-11’, and -linked polyubiquitination (David et al. 2010).

As an up-regulated gene in the pathophysiological process of HCC, UBE2T promotes the deprivation of the P53 protein by improving its ubiquitination and activating the carcinogenic process. A higher UBE2T expression level is related to a poor prognosis, tumor vascular invasion, late TNM stages, and higher pathological grades in HCC patients, which is consistent with the results from other studies. A multivariate analysis showed that UBE2T is an independent risk factor for OS in HCC patients (Liu et al. 2017). It has also been shown that over-expressed UBE2T accelerates tumorigenesis, migration, proliferation, colony formation, and HCC invasion, while the knockout of UBE2T suppresses HCC growth with a G2/M phase arrest in the cell cycle (Liu et al. 2019). In one study, the ectopic expression of UBE2T led to decreased expressions of P53, P21 and NOXA. The newly discovered miR-543/UBE2T/P53 axis could be a likely new target for the treatment of HCC (Liu et al. 2017). Furthermore, through the activation of the AKT/glycogen synthase kinase-3β (GSK3β)/β-catenin signaling pathway and cell proliferation, UBE2T promotes invasion, and metastasis (Hu et al. 2016). MicroRNA-1305 could act as an effective target for UBE2T to inhibit the AKT/GSK3β/β-catenin signaling pathway, thereby inhibiting the propagation and tumorogenicity of liver cancer stem cells (LCSCs) (Wei et al. 2019).

As a member of the CYP450 family, CYP3A4 is located in the endoplasmic reticulum and participates in drug metabolism and lipid synthesis. Its expression can be enhanced by steroid hormones and certain medicines (El-Merhibi et al. 2012). This enzyme helps in the metabolism of > 60% of clinical drugs, including erythromycin, diazepam, cyclosporin A, codeine, and acetaminophen (Fujimura et al. 2012; Zanger and Schwab 2013). It also acts on some carcinogens and steroids. It was previously believed to be down-regulated after birth, and has been found to be highly expressed in the fetal liver (Fanni et al. 2015). The CYP3A3 gene, which was previously believed to be an independent gene, has been shown to be a transcriptional variant of CYP3A4. CYP3A4 expression is reduced in hepatocellular adenomas (HCA) and HCC, and has been correlated with type 2 diabetes mellitus (T2DM) (Haas et al. 2009). CYP3A4 expression is stimulated by the hepatitis B virus X protein (HBX) through co-activating a pregnane X receptor (PXR), which affects hepatic carcinogen aflatoxin B1 (AFB1) bioactivation (Ayed-Boussema et al. 2012). The expression of CYP3A4 and PXR was high in stage I of HCC, but reduced in stages II and III (Niu et al. 2013). This indicates that down-regulated expressions of PXR and CYP3A4 are potential biomarkers for HBV-related HCC, particularly in advanced HCC. This is consistent with the image which can be seen in Figure 4(A).

Sorafenib, as a first-line drug for HCC; it is metabolized by CYP450, in particular by CYP3A4 isoforms, creating two main metabolites: N-hydroxymethyl metabolite and N-oxide. The overexpression of CYP3A4 could lead to enhanced drug degradation and thus resulting in clinical ineffectiveness. Therefore, CYP3A4-up-regualted HCC samples could indicate low responses to sorafenib (Fanni et al. 2016; Guo et al. 2018). It was discovered that rutin significantly reduced the protein and mRNA expression of the CYP3A4 gene in HepG2 cells by 75.3% and 73.2%, respectively. Such reduced levels of CY3A4 following rutin treatment could improve the action of sorafenib (Karakurt 2016). C-1311 is an operative regulator of CYP3A4 enzymatic activity, and is an inhibitor in Hep3A4 cells of its isoenzyme. C-1311 inhibits the overexpression of CYP3A4 in HepG2 cells, and HepG2 cells stalled in the G2 phase of the cell cycle (Augustin et al. 2013).

At present, there are few bioinformatics articles concerning hub genes for the transformation of cirrhosis into HCC (Li et al. 2016; Xie et al. 2016; He et al. 2017; Chen et al. 2018; Fan and Ye 2018; Wu et al. 2019; Jiang et al. 2020). Most of them are limited to the screening process and have not been further verified and analyzed. This is the first article that hub genes transformed from cirrhosis to HCC has been verified by RT–PCR, WB and IHC after bioinformatics analysis, which is a major advantage of this paper. In addition, combined diagnosis of UBE2T and CYP3A4 is superior to AFP in predicting the development of HCC in patients with cirrhosis, especially in patients with AFP-negative HCC, which is another advantage of this study.

This study had certain limitations. The origins of cirrhosis were not classified by this study with a finite number of online datasets. In addition, most databases were only used for the comparison of HCC with normal liver tissue, but there was a lack of comparison between HCC and cirrhosis. Although the difference between the expression of UBE2T and CYP3A4 in HCC group and cirrhosis group has been confirmed by bioinformatics analysis and a variety of molecular and histiocytic studies, it remains to be further verified by animal models. Furthermore, we will research their upstream and downstream genes and related signaling pathways to find a way to block the development of cirrhosis into HCC by inhibiting targets.

In conclusion, UBE2T and CYP3A4 were screened as tumor-promoting and tumor-suppressing hub genes, for cirrhosis-HCC transformation by a series of bioinformatics analysis and were verified by RT–PCR, WB and IHC. These genes may be applied as combined biological markers to diagnose and predict the HCC tendency in cirrhosis patients, which is beneficial to the early detection, diagnosis and treatment of HCC.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Due to the nature of this research, participants of this study did not agree for their data to be shared publicly, so supporting data is not available.

Additional information

Funding

This project was supported by the National Natural Science Foundation of China [grant number 81702923, 81971503], the Science and Technology Commission of Shanghai Municipality [grant number 18ZR1439300], the Precision Medicine Project of Navy Medical University, China [grant number 2017JZ50].