Identification of most representative hub-genes for diagnosis, prognosis, and therapies of hepatocellular carcinoma

Introduction

Hepatocellular carcinoma (HCC), a primary liver cancer type, currently stands as the third leading cause of cancer-related fatalities worldwide (1). According to the latest estimates, it ranks as the third leading cause of cancer-related mortality worldwide, with over 900,000 new cases and 830,000 deaths reported annually (1). HCC exhibits significant geographical variation, with the highest incidence rates observed in regions endemic for chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infections, such as sub-Saharan Africa and East Asia (2). However, its incidence is also increasing in Western countries due to the rising prevalence of non-alcoholic fatty liver disease (NAFLD) and alcoholic liver disease (ALD) (3). Globally, HCC disproportionately affects individuals with underlying liver cirrhosis and chronic liver disease, which serves as a significant risk factor for its development.

The diagnosis of HCC relies on clinical, radiological, and histopathological criteria. Imaging modalities such as ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI) are pivotal in detecting liver lesions and evaluating tumor characteristics. Serum biomarkers such as alpha-fetoprotein (AFP), AFP-L3, and des-gamma-carboxy prothrombin (DCP) are commonly used for HCC surveillance and diagnosis, albeit with limitations in sensitivity and specificity. Histopathological examination of liver tissue obtained via biopsy remains the gold standard for confirming HCC diagnosis and assessing tumor grade, stage, and molecular characteristics. However, the majority of HCC diagnoses occur at advanced stages, where effective treatment options are lacking, resulting in a bleak prognosis (4). For this reason, despite significant progress in the treatment approaches for HCC, including liver transplantation, extensive surgical resection, and interventional therapy, the extended-term survival rates for individuals with HCC remain dismal globally (5-7). Additionally, the lack of potential biomarkers has led to population-based treatments for HCC that frequently fall short of achieving the desired therapeutic outcomes (8,9). So, the limitations of existing approaches necessitate the urgent development of novel and more effective diagnostic, prognostic, and therapeutic strategies for HCC patient’s survival.

Nonetheless, the conventional process of developing new drugs is laborious, costly, and challenging. An alternative strategy entails repurposing drugs based on potential therapeutic targets, wherein the efficacy of already Food and Drug Administration (FDA)-approved drugs for different diseases is investigated instead of developing entirely new drugs from scratch (10). This strategy can significantly reduce failure rates by approximately 45% and save an average of 5–7 years compared to de novo drug development (11,12). However, it is crucial to identify potential drug targets from many alternatives in both drug discovery approaches.

The discovery of potential biomarkers to interrupt the pathophysiological processes of the disease could serve as a driving force for the advancement of both diagnostic and more efficacious therapeutic strategies in the future. Over the past decade, the integration of bioinformatics and high-throughput sequencing technologies has paved the way for researchers to pinpoint disease-related genes implicated in the pathogenesis of HCC. The utilization of hub differentially expressed genes (HubGs) in distinguishing between disease and control samples has been playing a crucial role in disease diagnosis and treatment strategies (13-19). There are some studies in the literature that reported HCC-associated HubGs for diagnosis and prognosis (20-24). Only a few studies identified HubGs-guided candidate drugs for HCC treatment (16,25).

While several studies have explored HCC-associated hub genes for diagnosis and prognosis, a critical limitation remains in identifying more effective and consistent therapeutic targets. Existing literature exhibits inconsistencies in the proposed sets of hub genes across various publications. Furthermore, no prior study has explored more representative and effective hub genes/proteins as therapeutic targets by integrating all hub gene sets. This study considered that issue. We detected more representative and effective hub genes/proteins as therapeutic targets by combining all hub-genes sets published in 59 articles using protein-protein interaction (PPI) and survival probability analysis, which may enhance the representativeness of our findings and reduce potential biases compared to the previous individual hub-genes set. Additionally, we incorporate a novel step: evaluating the performance of our proposed candidate drugs against the top-ranked target proteins mediated by hub genes identified across various studies. This cross-validation approach strengthens our findings’ generalizability and potential effectiveness in identifying more reliable and potentially effective candidate drugs for HCC treatment.

In this study, we conducted a systematic literature review on 59 individual studies that collectively identified 202 unique HCC-causing HubGs. PPI network analysis and survival analysis were constructed on 202 HubGs to identify top-ranked HubGs (tHubGs) associated with HCC. Then, Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway, and stage-specific expression analysis were performed to investigate the progression and prognosis of HCC. Moreover, the gene regulatory network analysis highlighted key transcription factors (TFs) and microRNAs (miRNA) that modulate these tHubGs at transcriptional and post-transcriptional. Finally, we identified three tHubGs-guided candidate drugs with our proposed and published protein receptors by molecular docking (MD).

Methods

In this study, we collected essential meta-drug targets and agents from various online sources and published articles. Our objective was to explore the most representative drug targets and agents for the treatment of HCC using an integrated bioinformatics approach. The research pipeline is depicted in Figure 1.

Figure 1 Overall pipeline of the study. HubGs, hub differentially expressed genes; tHubGs, top-ranked HubGs; PPI, protein-protein interaction; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; 3D, three-dimensional.

Collection of HCC-causing HubGs as the meta-receptors

We conducted a review of 59 articles that suggested HubGs associated with HCC. This comprehensive analysis, as presented in Table S1, allowed us to collect HubGs sets from these sources, resulting in a total of 202 unique HubGs. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Collection of meta-drug agents

To explore promising drug candidates for HCC treatment through MD analysis with our proposed receptors, we collected a total of 157 anti-HCC-related drugs from published articles and online databases, which we refer to as our meta-drug agents (Table S2).

Selecting tHubGs by ranking HubGs

In order to select tHubGs associated with HCC, as reported by various independent studies, we took into account both the PPI network and the prognostic significance of these HubGs. We conducted PPI network analysis using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (https://string-db.org/), configuring it with default parameters to elucidate the interactions and functional roles of the HubGs (26). We utilized the Cytoscape software to provide a more intuitive visualization of the constructed PPI network (27). Within this network, genes/proteins are represented by nodes, and interactions are depicted by edges. To rank the HubGs, we utilized the CytoHubba App (28) within Cytoscape. Three topological measures, including Degree (29), Maximal Clique Centrality (MCC) (28), and Maximum Neighborhood Component (MNC) (30) of the PPI network have been used to rank the HubGs separately. Additionally, the Gene Expression Profiling Interactive Analysis 2 (GEPIA2) web tool (http://gepia2.cancer-pku.cn/) (31) was used to perform survival analysis and calculate the log-rank P value of HubGs based on The Cancer Genome Atlas (TCGA) Program database (https://portal.gdc.cancer.gov/). This analysis allowed us to evaluate the significant differences in patient overall survival between low and high expressions of HubGs. Subsequently, we ordered the HubGs based on the average rank of Degree, MCC, MNC, and log-rank P value to select tHubGs (appendix available at https://cdn.amegroups.cn/static/public/CCO-23-151-1.xlsx).

Assessment of tHubGs as early diagnostic and prognostic biomarkers

UALCAN (http://ualcan.path.uab.edu/analysis.html) serves as an interactive web portal designed for comprehensive analyses of TCGA gene expression data (32). Within the Liver hepatocellular carcinoma (LIHC) dataset, there were 50 samples from normal tissues and 340 samples from HCC patients, distributed across stages 1, 2, 3, and 4, with 168, 84, 82, and 6 samples, respectively. We utilized the UALCAN online database to investigate the correlation between tHubGs and the different stages of HCC progression by plotting their independent RNA sequencing (RNA-seq) profiles.

Functional and pathway enrichment analysis of tHubGs

To gain insights into the pathogenic processes and pathways associated with HubGs, we conducted enrichment analyses using the GO (http://geneontology.org) functions and the KEGG (https://www.kegg.jp) pathways. For this purpose, we utilized the g:Profiler (https://biit.cs.ut.ee/gprofiler/gost), an updated database that facilitates GO functional and KEGG pathway enrichment analyses on input gene lists (33). Our functional and pathway enrichment analysis, focusing on HubGs in HCC, was carried out using the g:GOSt tool integrated into the g:Profiler web server. g:Profiler employs the widely used hypergeometric distribution to assess the significance and relies on the Ensembl database (34). We aimed to identify statistically significant GO terms related to biological processes (BPs), molecular functions (MFs), cellular components (CCs), and KEGG pathways. In this analysis, a significance threshold of P value <0.05 was employed, with results further subjected to the Benjamini and Hochberg false discovery rate (FDR) multiple testing correction procedure (35).

Gene regulatory network analysis of tHubGs

To explore key TFs and miRNAs of tHubGs, we explored the tHubGs-TFs and tHubGs-miRNAs interaction networks using the NetworkAnalyst (https://www.networkanalyst.ca) web platform (36). NetworkAnalyst provides support for constructing networks using two widely utilized topological measures: degree and betweenness centrality. The construction of the tHubGs-TFs and tHubGs-miRNAs interaction networks was facilitated through the NetworkAnalyst web server, incorporating data from the JASPAR and TarBase databases. To enhance the quality of these networks, we employed the Cytoscape software. Subsequently, the key regulators were identified based on their degree and betweenness within the networks.

Drug repurposing by MD analysis

In our study, we conducted MD analysis, aiming to identify effective candidate drugs for HCC treatment based on tHubGs and their corresponding regulatory TF proteins, which served as potential drug targets (p=11). Additionally, we considered 157 meta-drug agents (q=157) as promising candidates. To carry out the MD analysis, we required three-dimensional (3D) structures for both the target proteins and the meta-drug agents. The 3D structures of protein receptors were retrieved from the Protein Data Bank (PDB) (https://www.rcsb.org/) (37) and AlphaFold Protein Structure Database (https://alphafold.ebi.ac.uk/) (38), while the 3D structures of the meta-drug agents were obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) (39). We utilized Discovery Studio Visualizer 2019v (40) to visualize receptor protein’s 3D structures and deleted any irrelevant target chains in the process. The receptor proteins underwent processing using AutoDock Tools (41), involving the removal of water molecules and ligand heteroatoms, as well as the incorporation of polar hydrogens. The ligands were also subjected to preparatory steps for the MD study, including setting the torsion tree and specifying rotatable and non-rotatable bonds, which were accomplished using AutoDock Tools. The evaluation of binding affinities (BAs) between the target proteins and the drug agents was performed using AutoDock Vina (42), with an exhaustiveness parameter set to 10. Subsequently, the docked complexes underwent a comprehensive analysis to examine surface complexes, non-covalent bond types, and interatomic distances. This analysis was conducted using PyMol (43) and Discovery Studio Visualizer 2019v. For the purpose of selecting the top-ranked potential drugs, we ranked the target proteins and agents based on the descending order of the row sums ∑j=1qMij,i=1,2,…,p, and column sums ∑i=1pMij,j=1,2,…,q, of the BA M_ij between ith target protein (i=1, 2, ..., p) and jth drug agent (j=1, 2, ..., q) respectively.

Assessment of drug-likeness of candidate drugs

A significant challenge in drug discovery involves identifying candidate molecules with desired pharmacokinetic properties. These properties determine how a drug moves through the body, impacting its absorption, distribution, metabolism, and excretion (ADME). We utilized RDkit, a free and open-source Python library, to aid in this evaluation. RDkit leverages extensive experimental datasets to analyze the potential pharmacokinetic behavior of small molecules (44). Lipinski’s “Rule of Five” is a rule to assess the pharmacokinetic properties of small molecules in the context of drug development. This rule suggests that most successful orally administered drugs tend to have molecular weight (MW) ≤500 dalltons, number of hydrogen bond acceptors (HBAs) ≤10, number of hydrogen bond donors (HBDs) ≤5 (45), log P (a measure of lipophilicity) ≤5.

Results

Ranking HubGs to identify tHubGs

In this study, we retrieved metadata derived from previous studies focused on HCC by searching online databases, including PubMed and Google Scholar, spanning the years 2002 to 2023. We included studies based on the following criteria: (I) differentially expressed genes between HCC and normal samples; (II) hub genes, core genes, or key genes identified or proposed for HCC; and (III) biomarkers identified by bioinformatics or statistical analysis. Finally, we compiled the hub genes proposed in each of the chosen studies and determined the 202 unique meta HubGs for HCC. The PPI network analysis of collected 202 HubGs from the selected published papers was performed to rank HubGs. The construction of the PPI network of the HubGs’ was achieved using the STRING database, resulting in a network featuring 229 nodes and 4,488 edges (Figure 2A). We applied the Cytoscape plugin CytoHubba to apply three measures (Degree, MCC, and MNC) to rank the HubGs separately. Additionally, the GEPIA2 database was used to calculate the prognostic value by fitting the Cox proportional hazards regression model between low- and high-risk group patients in the TCGA-LIHC cohort based on the expression of HubGs. HubGs were then ranked according to their P values of the log-rank test. Finally, we calculated the average rank based on the ranks with the PPI network (Degree, MCC, and MNC) and survival analysis (ranks based on P values of the log-rank test) and selected top-ranked eight genes (CDK1, AURKA, CDC20, CCNB2, TOP2A, PLK1, BUB1B, and BIRC5) as tHubGs. We observed that the higher expression of tHubGs CDK1 (log-rank P=0.00017), AURKA (log-rank P=0.00022), CDC20 (log-rank P=0.00038), CCNB2 (log-rank P=0.00015), TOP2A (log-rank P=0.00047), PLK1 (log-rank P=0.00082), BUB1B (log-rank P=0.00066), and BIRC5 (log-rank P=0.000041) has a significant negative association with the overall survival of HCC patients (Figure 2B). Therefore, this indicates that tHubGs have strong prognostic power for HCC.

Figure 2 The PPI network and survival analysis. (A) PPI of HubGs where the pink-colored rectangular nodes represent tHubGs. (B) Kaplan-Meier survival plots show survival differences between low and high expressions of tHubGs. HR, hazard ratio; PPI, protein-protein interaction; HubGs, hub differentially expressed genes; tHubGs, top-ranked HubGs.

Assessment of tHubGs as early diagnostic and prognostic biomarkers

Genes exhibiting stage-specific expression patterns may serve as potential therapeutic biomarkers. We conducted a boxplot analysis to examine the stage-specific expressions of tHubGs in the TCGA database. The aim of this analysis was to uncover differences in the expression levels of tHubGs among normal individuals and those with stage 1, 2, 3, and 4 HCC. From Figure 3, we observed consistently high expression levels of all tHubGs in stage 1 through stage 4 samples compared to normal. However, the high expression of all tHubGs in stages 1, 2, and 3 gradually increases, signifying HCC progression from less aggressive to more aggressive stages, in comparison to the expression observed at stage 4. The consistent expression patterns across stages 1, 2, and 3 suggest that these genes may be promising candidates for stage 1 HCC biomarkers. This finding suggests the potential utility of our proposed tHubGs as robust early diagnostic and prognostic biomarkers.

Figure 3 Boxplots depicting relative expressions of tHubGs in normal, stages 1, 2, 3, and 4 of HCC patients. tHubGs, top-ranked hub differentially expressed genes; HCC, hepatocellular carcinoma.

Functional and pathway enrichment analysis of tHubGs

The GO annotation and KEGG pathway enrichment analyses of 202 HubGs revealed significant biological roles of tHubGs in HCC (Table 1 and Table S3). The GO MFs showed significant enrichment of tHubGs in activities related to carbohydrate derivative binding, small molecule binding, protein binding, anion binding, and enzyme binding. The top 5 significantly enriched BPs terms encompassed mitotic cell cycle processes, mitotic cell cycles, cell cycle processes, cell cycles, and the regulation of cell cycles. In the case of CCs, tHubGs significantly enriched in chromosome structures, chromosomal regions, spindle components, the lumina of intracellular organelles, and organelle lumina. Furthermore, our investigation revealed the five most enriched pathways for tHubGs, namely the cell cycle, chemical carcinogenesis with receptor activation, pathways in cancer, cellular senescence, and progesterone-mediated oocyte maturation.

Gene regulatory network analysis of tHubGs

At both the transcriptional and post-transcriptional levels, gene expression is profoundly regulated by two paramount factors: TFs and miRNAs. An interaction network between tHubGs and TFs, using data from the JASPAR database in NetworkAnalyst, was constructed to identify key transcriptional regulatory factors for tHubGs (Figure 4A). Within this network, tHubGs were represented by round-rectangle nodes in pink, while TFs were denoted by diamond nodes in orange color. Utilizing a topological measure with a degree of ≥5, we identified FOXC1, GATA2, and NFIC as the top three key TFs regulating tHubGs. To delve into key post-transcriptional factors associated with tHubGs, we constructed the tHubGs-miRNAs interaction network from the TarBase database in NetworkAnalyst (Figure 4B). In this network, tHubGs were represented by pink round-rectangle nodes, while miRNAs were highlighted in green. We selected the top three miRNAs—namely, hsa-miR-107, hsa-miR-34a-5p, and hsa-miR-16-5p—based on topological measure degree ≥8 as the key post-transcriptional factors regulating tHubGs.

Figure 4 Analysis of the Gene Regulatory Network. (A) Interaction network between TFs and tHubGs. The circle-shaped nodes, pink rectangular nodes, and orange diamond-shaped nodes depict TFs, tHubGs, and key TFs, respectively. (B) Interaction network between miRNA and tHubGs. The rectangular green color and pink color nodes represent miRNAs and tHubGs, respectively. The larger green color nodes represent key miRNA. TFs, transcription factors; tHubGs, top-ranked hub differentially expressed genes; miRNA, microRNA.

Drug repurposing by MD analysis

We considered eleven key proteins for our investigation, including eight tHubGs and three TFs (FOXC1, GATA2, and NFIC) as our drug target proteins. We downloaded 3D structures for 11 target proteins. Specifically, for CDK1, AURKA, CDC20, TOP2A, PLK1, BUB1B, BIRC5, FOXC1, and NFIC, we obtained structures with the following source codes: 5lqf, 2j4z, 4ggc, 6zy5, 3p34, 3si5, 2qfa, 1d5v, and 7qqd, respectively, from the PDB (37). Additionally, we downloaded structures for CCNB2 and GATA2 with the identifiers AF_095067 and AF-P23769, respectively, from the AlphaFold Protein Structure Database (38). Moreover, the 3D structures for the 157 meta-drugs were retrieved from the PubChem database (39). Subsequently, we performed MD analysis between our selected (p=11) target proteins and (q=157) meta-drug agents. This process calculated BA for each pair of target proteins and drug agents. We ranked the target proteins based on their BA with drug agents, considering the row sums of the BA score matrix, denoted as M = M_ij. In contrast, we ranked the drug agents by considering the column sums of the BA score matrix to select potential drug candidates. To illustrate this ranking, we presented the BA matrix in Figure 5A, where the Y-axis represents proposed target proteins and the X-axis represents the top 50 ranked drug agents out of 157. Our criteria for identifying promising compounds against the 11 receptors involved considering candidates with a BA of −7.5 or less. Notably, we identified the top three drug agents—CD437, avrainvillamide, and LRRK2-IN-1—all producing BA scores below −8.3 kcal/mol with all of our proposed 11 target proteins. Moving further down the list, the subsequent seven top lead compounds, namely PF-573228, KU-60019, NVP-BSK805, PF-184, TW-37, NVP-231, sotrastaurin, and SNX-2112, displayed BA less than −7.5 kcal/mol with 10 out of the 11 proposed receptor proteins. Therefore, we considered the top 3 lead compounds, including CD437, avrainvillamide, and LRRK2-IN-1 as most probable candidate drugs for inhibiting HCC.

Figure 5 Results of MD for proposed and published proteins with drug agents. Red represents strong binding between proteins and drug agents, while green indicates weak binding. (A) The Y-axis represents target proteins (proposed), and the X-axis represents the top 50 ranked drug agents out of 157. (B) The Y-axis represents published proteins, and the X-axis represents the top 50 ranked drug agents out of 157. MD, molecular docking.

To validate our top three candidate drugs, we conducted MD with receptor proteins associated with HCC from independently published research. In our review of 59 published articles, we identified eight hub genes (CCNB1, CDK1, CDC20, TOP2A, CCNB2, BUB1B, CCNA2, and PRC1) consistently mentioned in at least nine of these articles (Table S1). From these eight hub genes, we found that five genes (CDK1, CDC20, TOP2A, CCNB2, and BUB1B) were common with our selected tHubGs. So, we downloaded the 3D structures of the remaining three hub genes (CCNB1, CCNA2, and PRC1) from the PDB with source codes 5hq0, 1h1s, and 4l6y, respectively. Subsequently, we performed MD analysis between these eight published receptor proteins and 157 meta-drugs. The BA matrix, denoted as M* = M_ij*, was visualized in Figure 5B, with the Y-axis representing published proteins and the X-axis representing the top 50 ranked drug agents out of 157. We observed that eight out of the top 10 drugs overlap between our proposed and published protein receptors. Notably, CD437, avrainvillamide, and LRRK2-IN-1 produced higher BA scores for both proposed and published receptors. Therefore, we selected these top-ranked three drug agents, CD437, avrainvillamide, and LRRK2-IN-1, as candidate drugs due to their higher BA scores for all receptors. Furthermore, we presented the structural interaction profiles of the top-ranked three docked complexes (CDK1-CD437, AURKA-avrainvillamide, and CDC20-LRRK2-IN-1) in Table 2.

Table 2

Overview of the top 3 protein-drug complexes (CDK1-CD437, AURKA-avrainvillamide, and CDC20-LRRK2-IN-1) from MD analysis

Potential targets	2D structure of compounds	3D structure of drug target-agent complexes	BA (kCal/mol)	Hydrogen bonds	Hydrophobic interactions	Electrostatic
CDK1	CD437	CDK1-CD437	−11.6	LYS89, SER84	–	–
AURKA	Avrainvillamide	AURKA-avrainvillamide	−10.4	ASP146	ILE10, LEU135, AL18, ALA31, LYS33, VAL64, PHE80, VAL18, ALA145, YS89	ASP86
CDC20	LRRK2-IN-1	CDC20-LRRK2-IN-1	−10.2	ASP256, ASP274, ARG255, GLY276	–	–

MD, molecular docking; 2D, two-dimensional; 3D, three-dimensional; BA, binding affinity.

Assessment of drug-likeness of candidate drugs

We assessed the potential drug-likeness of three top-ranked drug candidates (avrainvillamide, CD437, and LRRK2-IN-1) using Lipinski’s “Rule of Five” criteria (Figure 6). Two molecules, avrainvillamide and CD437, had a MW below 500 g/mol, while LRRK2-IN-1 was slightly larger. All three candidates had nHBA and nHBD values less than 10 and 5, respectively. The calculated log P value for CD437 (6.3784) was higher than the suggested limit of 5, whereas the calculated log P values for avrainvillamide and LRRK2-IN-1 were less than 5. Overall, they have potential as drug candidates according to Lipinski’s “Rule of Five” criteria.

Figure 6 Basic pharmacodynamics properties of the top-ranked three compounds, such as MWs, the number of HBAs, the number of HBDs, and CLog P. MW, molecular weight; HBA, hydrogen bond acceptor; HBD, hydrogen bond donor; CLog P, calculated log P.

Discussion

HCC currently is the third leading cause of cancer-related deaths worldwide. To improve the survival rates and reduce mortality among HCC patients, further studies are needed to validate effective biomarkers and candidate drug agents. Consequently, target-based therapy for HCC has gained significant attention in the medical field. While multiple biomarkers for HCC have been identified in recent years (Table S1), the more representative biomarkers associated with HCC also need to be explored and investigated for early diagnosis, prognosis, and therapies. In this study, we reviewed 59 individual studies that have suggested 202 HCC-causing HubGs in total. Among them, we selected the top-ranked eight HubGs (CDK1, AURKA, CDC20, CCNB2, TOP2A, PLK1, BUB1B, and BIRC5) by the PPI network and prognostic power analyses highlighting their GO functions and pathways, key regulatory components (TFs and miRNAs), and therapeutic candidates for the treatment of HCC. We observed that these tHubGs significantly influence the overall survival of HCC patients (Figure 2B). The boxplots generated from the stage-specific expression of tHubGs using the ULACAN web server demonstrated that tHubGs significantly differed among the HCC progression stages as compared to the normal group (Figure 3). These findings suggest that the proposed tHubGs may have a crucial role in the diagnosis of HCC at an earlier stage.

Protein kinase CDK1, a vital member of the serine/threonine protein kinase family, plays a crucial role in progression into the mitotic phase and is frequently seen to be over-expressed in human cancers (18,46). Previous research indicated a clear association between CDK1 overexpression and negative clinicopathological characteristics and a lower survival rate among patients with HCC (47), which is consistent with our present study. AURKA, a mitotic serine/threonine kinase with regulatory influence over mitosis, cell division, and cell cycle progression, has exhibited conspicuous overexpression in both cell lines and tissue specimens of HCC (48). Such overexpression in AURKA levels has been linked to aggressive tumor characteristics, adverse prognoses, and resistance to chemotherapy in HCC cases (19,49-52). The CDC20 serves as a regulatory factor within the cell cycle. Numerous studies have consistently associated heightened CDC20 expression with the development and progression of HCC (53-55), which aligns with our results. CCNB2, overexpressed in various cancer types such as lung, bladder, gastric, and breast cancers, has previously been investigated for its overexpression in HCC and its correlation with poor prognostic indicators (56-62). Our current findings align with this previous research. TOP2A is intricately involved in essential BPs, encompassing cell proliferation, DNA replication, transcription, recombination, and chromosome segregation (63). Among various topoisomerases, TOP2A predominantly influences cell growth, the development of aggressive diseases, and resistance to chemotherapy. Its overexpression has been observed across a spectrum of cancers, including breast, cervical, bladder, and ovarian cancer (14,64). For HCC, TOP2A expression and amplification have displayed diagnostic, prognostic, and therapeutic implications, affecting disease progression, survival, and treatment response (16,20). PLK1, a key cell cycle regulator, is increasingly recognized as a promising therapeutic target in HCC treatment (65-67). Previous studies have offered strong support for our findings, underscoring PLK1’s potential as a therapeutic target for HCC (68,69). The oncogenic effect of BUB1B on HCC cell proliferation, migration, and invasion is partially mediated through its impact on mitochondrial function (70). Previous studies suggested that BUB1B is a more potential therapeutic target for HCC (21,71,72). Overexpression of BIRC5 has been related to a negative prognosis for patients suffering from HCC, lung cancer, pancreatic cancer, renal clear cell carcinoma, renal papillary cell carcinoma, sarcoma, and endometrial carcinoma (73). The high expression of survivin during the progression and development of HCC stimulates cancer cell growth, hinders cancer cell death, triggers angiogenesis in the tumor stroma, decreases the efficacy of radiation and chemotherapy on cancer cells, and substantially affects the survival rates of HCC patients (74). Therefore, BIRC5 emerges as a pivotal target for HCC treatment. CDK1, AURKA, CDC20, CCNB2, PLK1, and BUB1B are crucial regulators of the cell cycle, a fundamental process in cell division. The disruption of this cycle is a common characteristic of cancer, making these molecules attractive targets for the prevention and treatment of HCC. To counteract the uncontrolled proliferation of HCC cells, several therapeutic strategies are being explored. Developing or repurposing small molecule inhibitors to specifically target these proteins can directly impede their activity in cancer cells, potentially halting tumor growth. Immunotherapy offers an alternative approach by boosting the immune system’s response to cells that overexpress these targets, utilizing methods such as creating antibodies or chimeric antigen receptor (CAR)-T cells designed to attack these proteins. Furthermore, combination therapy, which includes the use of cell cycle regulator inhibitors alongside chemotherapy, targeted therapy, or immunotherapy, could amplify treatment effectiveness and circumvent mechanisms of resistance. These strategies represent a concerted effort to leverage our understanding of the cell cycle in cancer to improve outcomes for HCC patients.

The analysis of the interaction network between TFs and tHubGs revealed three key TFs: FOXC1, GATA2, and NFIC (Figure 4A). FOXC1 emerges as a prospective molecular drug target for mitigating tumor metastasis in HCC patients (75-77). It has recently garnered attention as a potential regulator in various cancers, including basal-like breast cancer (78), gastric cancer (GC) (79), nasopharyngeal carcinoma (NPC) (80), and ovarian cancer (81). GATA2 exhibits a strong association with poor prognoses in HCC (82). Its links to tumor aggressiveness and unfavorable survival extend to other malignancies such as GC (83), prostate cancer (84), clear cell renal cell carcinoma (85), and KRAS mutant lung cancer (86). NFIC is a member of the NFI family that has the potential to influence developmental processes through the regulation of transcription in various cell types. It may play a crucial role in the development of HCC (87). The above discussion indicates that the proposed three TFs play crucial roles in HCC and other cancers. Therefore, we utilized these TFs as the drug target proteins along tHubGs. We also explored three post-transcriptional regulatory factors of tHubGs (hsa-miR-107, hsa-miR-34a-5p, and hsa-miR-16-5p) (Figure 4B). MiR-107 has been demonstrated to enhance the proliferation of HCC cells through its impact on Axin2, making it a promising therapeutic intervention target in HCC (88). Studies have consistently shown that hsa-miR-34a-5p is often downregulated in various cancer types, including liver cancer, and its restoration has been proven to inhibit tumor cell growth and induce apoptosis (89). MiR-16-5p, on the other hand, hinders the invasion and migration of HCC cells by directly targeting and inhibiting IGF1R protein expression (90). This discussion underscores the significance of the identified regulatory TFs and miRNAs in relation to HCC and their potential as valuable therapeutic targets.

In this study, we selected the top 5 GO functions associated with MFs, BPs, and CCs to investigate the pathogenic processes related to HCC based on tHubGs. The top 5 MFs selected were carbohydrate derivative binding, small molecule binding, protein binding, anion binding, and enzyme binding. The association of these MFs with HCC has been previously supported by several studies (91-94). Furthermore, the top 5 BPs, including mitotic cell cycle process, mitotic cell cycle, cell cycle process, cell cycle, and regulation of cell cycle that were found to exhibit significant association with HCC have also been supported by several independent studies (61,62,95). Similarly, the top 5 CCs (chromosome, chromosomal region, spindle, intracellular organelle lumen, and organelle lumen) were revealed to have a significant association with HCC, as supported by existing literature (96-99). The analysis of the proposed tHubGs using the KEGG pathway method revealed some significant pathway enrichments. The top 5 significant KEGG pathways included cell cycle, chemical carcinogenesis-receptor activation, pathways in cancer, cellular senescence, and progesterone-mediated oocyte maturation. Notably, these pathways have previously been reported in the context of HCC in earlier studies (62,100-103).

We conducted an MD analysis of 157 meta-drug agents with both our proposed protein receptors and top-ranked published receptors to explore potential candidate therapeutics for HCC. Then, we selected three drugs (CD437, avrainvillamide, and LRRK2-IN-1) as candidate drugs for HCC treatment as they showed strong binding with all of our proposed and published protein receptors. Several other independent studies have also suggested the use of proposed drugs, including CD437 (104,105), avrainvillamide (106,107), and LRRK2-IN-1 (108-110) for the treatment of HCC and other types of cancer. Additionally, we evaluated the drug-likeness properties of these candidate drugs and showed potentiality to be bioavailable drugs. However, the present study emphasizes the need for additional laboratory experiments to validate both the proposed target proteins and candidate drugs. Therefore, the findings of this study may play an effective role in early diagnosis, prognosis, and therapies for HCC.

Conclusions

HubGs play crucial roles in HCC, contributing to the understanding of its molecular mechanisms. HubGs may serve as potential diagnostic and prognostic biomarkers, also offering therapeutic targets for precise therapies. In this study, we selected the top eight tHubGs using the PPI network and prognostic power analyses. We observed that these eight tHubGs are statistically significant in the overall survival of HCC patients. The enrichment analysis of the tHubGs set using GO annotations and KEGG pathways revealed significant biological functions and signaling pathways associated with HCC. We identified the top three TF proteins and three miRNAs as the transcriptional and post-transcriptional factors of HCC through gene regulatory network analysis. Additionally, three potential drug agents (CD437, avrainvillamide, and LRRK2-IN-1) were identified through MD analysis, and their validity was further confirmed by the state-of-the-art alternatives already published for the top-ranked 8 HubGs related to HCC. Therefore, the results of our study could serve as valuable diagnostic, prognostic, and therapeutic resources for HCC.

Acknowledgments

We would like to thank all the other lab members for their excellent cooperation during this work.

Funding: This work was supported by the National Science Foundation of China (No. 32370790 to C.H.) and the Initiative Scientific Research Program, Institute of Zoology, Chinese Academy of Sciences (No. 2023IOZ0102).

Footnote

Peer Review File: Available at https://cco.amegroups.com/article/view/10.21037/cco-23-151/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://cco.amegroups.com/article/view/10.21037/cco-23-151/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394-424. [Crossref] [PubMed]
2. Jafri W, Kamran M. Hepatocellular Carcinoma in Asia: A Challenging Situation. Euroasian J Hepatogastroenterol 2019;9:27-33. [Crossref] [PubMed]
3. Paschos P, Paletas K. Non alcoholic fatty liver disease and metabolic syndrome. Hippokratia 2009;13:9-19. [PubMed]
4. Liu CY, Chen KF, Chen PJ. Treatment of Liver Cancer. Cold Spring Harb Perspect Med 2015;5:a021535. [Crossref] [PubMed]
5. Reig M, da Fonseca LG, Faivre S. New trials and results in systemic treatment of HCC. J Hepatol 2018;69:525-33. [Crossref] [PubMed]
6. Cauchy F, Zalinski S, Dokmak S, et al. Surgical treatment of hepatocellular carcinoma associated with the metabolic syndrome. Br J Surg 2013;100:113-21. [Crossref] [PubMed]
7. Luo JJ, Zhang ZH, Liu QX, et al. Endovascular brachytherapy combined with stent placement and TACE for treatment of HCC with main portal vein tumor thrombus. Hepatol Int 2016;10:185-95. [Crossref] [PubMed]
8. Faivre S, Rimassa L, Finn RS. Molecular therapies for HCC: Looking outside the box. J Hepatol 2020;72:342-52. [Crossref] [PubMed]
9. Ko KL, Mak LY, Cheung KS, et al. Hepatocellular carcinoma: recent advances and emerging medical therapies. F1000Res 2020;9:F1000 Faculty Rev-620.
10. Rudrapal M, Khairnar SJ, Jadhav AG. Drug repurposing (DR): an emerging approach in drug discovery. In: Badria FA. editor. Drug Repurposing: Hypothesis, Molecular Aspects and Therapeutic Applications. London: IntechOpen; 2020.
11. Chong CR, Sullivan DJ Jr. New uses for old drugs. Nature 2007;448:645-6. [Crossref] [PubMed]
12. Xue H, Li J, Xie H, et al. Review of Drug Repositioning Approaches and Resources. Int J Biol Sci 2018;14:1232-44. [Crossref] [PubMed]
13. Mosharaf MP, Reza MS, Kibria MK, et al. Computational identification of host genomic biomarkers highlighting their functions, pathways and regulators that influence SARS-CoV-2 infections and drug repurposing. Sci Rep 2022;12:4279. [Crossref] [PubMed]
14. Reza MS, Harun-Or-Roshid M, Islam MA, et al. Bioinformatics Screening of Potential Biomarkers from mRNA Expression Profiles to Discover Drug Targets and Agents for Cervical Cancer. Int J Mol Sci 2022;23:3968. [Crossref] [PubMed]
15. Ahmed FF, Reza MS, Sarker MS, et al. Identification of host transcriptome-guided repurposable drugs for SARS-CoV-1 infections and their validation with SARS-CoV-2 infections by using the integrated bioinformatics approaches. PLoS One 2022;17:e0266124. [Crossref] [PubMed]
16. Chen X, Xia Z, Wan Y, et al. Identification of hub genes and candidate drugs in hepatocellular carcinoma by integrated bioinformatics analysis. Medicine (Baltimore) 2021;100:e27117. [Crossref] [PubMed]
17. Alam MS, Rahaman MM, Sultana A, et al. Statistics and network-based approaches to identify molecular mechanisms that drive the progression of breast cancer. Comput Biol Med 2022;145:105508. [Crossref] [PubMed]
18. Reza MS, Hossen MA, Harun-Or-Roshid M, et al. Metadata analysis to explore hub of the hub-genes highlighting their functions, pathways and regulators for cervical cancer diagnosis and therapies. Discov Oncol 2022;13:79. [Crossref] [PubMed]
19. Hossen MA, Reza MS, Harun-Or-Roshid M, et al. Identification of Drug Targets and Agents Associated with Hepatocellular Carcinoma through Integrated Bioinformatics Analysis. Curr Cancer Drug Targets 2023;23:547-63. [Crossref] [PubMed]
20. Wang M, Wang L, Wu S, et al. Identification of Key Genes and Prognostic Value Analysis in Hepatocellular Carcinoma by Integrated Bioinformatics Analysis. Int J Genomics 2019;2019:3518378. [Crossref] [PubMed]
21. Li Z, Lin Y, Cheng B, et al. Identification and Analysis of Potential Key Genes Associated With Hepatocellular Carcinoma Based on Integrated Bioinformatics Methods. Front Genet 2021;12:571231. [Crossref] [PubMed]
22. Song X, Du R, Gui H, et al. Identification of potential hub genes related to the progression and prognosis of hepatocellular carcinoma through integrated bioinformatics analysis. Oncol Rep 2020;43:133-46. [PubMed]
23. Zhang H, Liu R, Sun L, et al. Comprehensive Analysis of Gene Expression Changes and Validation in Hepatocellular Carcinoma. Onco Targets Ther 2021;14:1021-31. [Crossref] [PubMed]
24. Zhang L, Yuan L, Li D, et al. Identification of potential prognostic biomarkers for hepatocellular carcinoma. J Gastrointest Oncol 2022;13:812-21. [Crossref] [PubMed]
25. Whelan JS, Stebbings W, Owen RA, et al. Successful treatment of a primary endodermal sinus tumor of the liver. Cancer 1992;70:2260-2. [Crossref] [PubMed]
26. Szklarczyk D, Franceschini A, Kuhn M, et al. The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res 2011;39:D561-8. [Crossref] [PubMed]
27. Shannon P, Markiel A, Ozier O, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003;13:2498-504. [Crossref] [PubMed]
28. Chin CH, Chen SH, Wu HH, et al. cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 2014;8:S11. [Crossref] [PubMed]
29. Jeong H, Mason SP, Barabási AL, et al. Lethality and centrality in protein networks. Nature 2001;411:41-2. [Crossref] [PubMed]
30. Lin CY, Chin CH, Wu HH, et al. Hubba: hub objects analyzer--a framework of interactome hubs identification for network biology. Nucleic Acids Res 2008;36:W438-43. [Crossref] [PubMed]
31. Tang Z, Kang B, Li C, et al. GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res 2019;47:W556-60. [Crossref] [PubMed]
32. Chandrashekar DS, Karthikeyan SK, Korla PK, et al. UALCAN: An update to the integrated cancer data analysis platform. Neoplasia 2022;25:18-27. [Crossref] [PubMed]
33. Raudvere U, Kolberg L, Kuzmin I, et al. g:Profiler: a web server for functional enrichment analysis and conversions of gene lists (2019 update). Nucleic Acids Res 2019;47:W191-8. [Crossref] [PubMed]
34. Martin FJ, Amode MR, Aneja A, et al. Ensembl 2023. Nucleic Acids Res 2023;51:D933-41. [Crossref] [PubMed]
35. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol 1995;57:289-300. [Crossref]
36. Zhou G, Soufan O, Ewald J, et al. NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis. Nucleic Acids Res 2019;47:W234-41. [Crossref] [PubMed]
37. Berman HM, Battistuz T, Bhat TN, et al. The Protein Data Bank. Acta Crystallogr D Biol Crystallogr 2002;58:899-907. [Crossref] [PubMed]
38. Jumper J, Evans R, Pritzel A, et al. Highly accurate protein structure prediction with AlphaFold. Nature 2021;596:583-9. [Crossref] [PubMed]
39. Kim S, Chen J, Cheng T, et al. PubChem 2019 update: improved access to chemical data. Nucleic Acids Res 2019;47:D1102-9. [Crossref] [PubMed]
40. Visualizer DS. v4. 0.100. 13345. Accelrys Software Inc. 2005.
41. Morris GM, Huey R, Lindstrom W, et al. AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility. J Comput Chem 2009;30:2785-91. [Crossref] [PubMed]
42. Trott O, Olson AJ. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem 2010;31:455-61. [Crossref] [PubMed]
43. DeLano WL, Bromberg S. PyMOL user's guide. Palo Alto: DeLano Scientific LLC.; 2004.
44. RDKit. 2024. Available online: https://www.rdkit.org/docs/index.html
45. Lipinski CA, Lombardo F, Dominy BW, et al. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev 2001;46:3-26. [Crossref] [PubMed]
46. Massagué J. G1 cell-cycle control and cancer. Nature 2004;432:298-306. [Crossref] [PubMed]
47. Agarwal R, Narayan J, Bhattacharyya A, et al. Gene expression profiling, pathway analysis and subtype classification reveal molecular heterogeneity in hepatocellular carcinoma and suggest subtype specific therapeutic targets. Cancer Genet 2017;216-217:37-51. [Crossref] [PubMed]
48. Vader G, Lens SM. The Aurora kinase family in cell division and cancer. Biochim Biophys Acta 2008;1786:60-72. [PubMed]
49. Li X, Xu W, Kang W, et al. Genomic analysis of liver cancer unveils novel driver genes and distinct prognostic features. Theranostics 2018;8:1740-51. [Crossref] [PubMed]
50. Simon EP, Freije CA, Farber BA, et al. Transcriptomic characterization of fibrolamellar hepatocellular carcinoma. Proc Natl Acad Sci U S A 2015;112:E5916-25. [Crossref] [PubMed]
51. Chen C, Song G, Xiang J, et al. AURKA promotes cancer metastasis by regulating epithelial-mesenchymal transition and cancer stem cell properties in hepatocellular carcinoma. Biochem Biophys Res Commun 2017;486:514-20. [Crossref] [PubMed]
52. Zhang K, Chen J, Chen D, et al. Aurora-A promotes chemoresistance in hepatocelluar carcinoma by targeting NF-kappaB/microRNA-21/PTEN signaling pathway. Oncotarget 2014;5:12916-35. [Crossref] [PubMed]
53. Li J, Gao JZ, Du JL, et al. Increased CDC20 expression is associated with development and progression of hepatocellular carcinoma. Int J Oncol 2014;45:1547-55. [Crossref] [PubMed]
54. Zhang X, Zhang X, Li X, et al. Connection Between CDC20 Expression and Hepatocellular Carcinoma Prognosis. Med Sci Monit 2021;27:e926760. [Crossref] [PubMed]
55. Yang G, Wang G, Xiong Y, et al. CDC20 promotes the progression of hepatocellular carcinoma by regulating epithelial‑mesenchymal transition. Mol Med Rep 2021;24:483. [Crossref] [PubMed]
56. Qian X, Song X, He Y, et al. CCNB2 overexpression is a poor prognostic biomarker in Chinese NSCLC patients. Biomed Pharmacother 2015;74:222-7. [Crossref] [PubMed]
57. Lei CY, Wang W, Zhu YT, et al. The decrease of cyclin B2 expression inhibits invasion and metastasis of bladder cancer. Urol Oncol 2016;34:237.e1-10. [Crossref] [PubMed]
58. Shubbar E, Kovács A, Hajizadeh S, et al. Elevated cyclin B2 expression in invasive breast carcinoma is associated with unfavorable clinical outcome. BMC Cancer 2013;13:1. [Crossref] [PubMed]
59. Shi Q, Wang W, Jia Z, et al. ISL1, a novel regulator of CCNB1, CCNB2 and c-MYC genes, promotes gastric cancer cell proliferation and tumor growth. Oncotarget 2016;7:36489-500. [Crossref] [PubMed]
60. Li R, Jiang X, Zhang Y, et al. Cyclin B2 Overexpression in Human Hepatocellular Carcinoma is Associated with Poor Prognosis. Arch Med Res 2019;50:10-7. [Crossref] [PubMed]
61. Zhou L, Du Y, Kong L, et al. Identification of molecular target genes and key pathways in hepatocellular carcinoma by bioinformatics analysis. Onco Targets Ther 2018;11:1861-9. [Crossref] [PubMed]
62. Wen DY, Lin P, Pang YY, et al. Expression of the Long Intergenic Non-Protein Coding RNA 665 (LINC00665) Gene and the Cell Cycle in Hepatocellular Carcinoma Using The Cancer Genome Atlas, the Gene Expression Omnibus, and Quantitative Real-Time Polymerase Chain Reaction. Med Sci Monit 2018;24:2786-808. [Crossref] [PubMed]
63. Porter AC, Farr CJ. Topoisomerase II: untangling its contribution at the centromere. Chromosome Res 2004;12:569-83. [Crossref] [PubMed]
64. Ali Y, Abd Hamid S. Human topoisomerase II alpha as a prognostic biomarker in cancer chemotherapy. Tumour Biol 2016;37:47-55. [Crossref] [PubMed]
65. Mok WC, Wasser S, Tan T, et al. Polo-like kinase 1, a new therapeutic target in hepatocellular carcinoma. World J Gastroenterol 2012;18:3527-36. [Crossref] [PubMed]
66. Wang D, Chang R, Wang G, et al. Polo-like Kinase 1-targeting Chitosan Nanoparticles Suppress the Progression of Hepatocellular Carcinoma. Anticancer Agents Med Chem 2017;17:948-54. [Crossref] [PubMed]
67. Xun R, Lu H, Wang X. PLK1 Is Implicated in the Poor Prognosis of Hepatocellular Carcinoma. Yangtze Medicine 2020;4:193. [Crossref]
68. Xu L, Zhu Y, Shao J, et al. Dasatinib synergises with irinotecan to suppress hepatocellular carcinoma via inhibiting the protein synthesis of PLK1. Br J Cancer 2017;116:1027-36. [Crossref] [PubMed]
69. Zheng DW, Xue YQ, Li Y, et al. Volasertib suppresses the growth of human hepatocellular carcinoma in vitro and in vivo. Am J Cancer Res 2016;6:2476-88. [PubMed]
70. Fu J, Zhang X, Yan L, et al. Identification of the hub gene BUB1B in hepatocellular carcinoma via bioinformatic analysis and in vitro experiments. PeerJ 2021;9:e10943. [Crossref] [PubMed]
71. Yang WX, Pan YY, You CG. CDK1, CCNB1, CDC20, BUB1, MAD2L1, MCM3, BUB1B, MCM2, and RFC4 May Be Potential Therapeutic Targets for Hepatocellular Carcinoma Using Integrated Bioinformatic Analysis. Biomed Res Int 2019;2019:1245072. [Crossref] [PubMed]
72. Mi N, Cao J, Zhang J, et al. Identification of hub genes involved in the occurrence and development of hepatocellular carcinoma via bioinformatics analysis. Oncol Lett 2020;20:1695-708. [Crossref] [PubMed]
73. Xu L, Yu W, Xiao H, et al. BIRC5 is a prognostic biomarker associated with tumor immune cell infiltration. Sci Rep 2021;11:390. [Crossref] [PubMed]
74. Su C. Survivin in survival of hepatocellular carcinoma. Cancer Lett 2016;379:184-90. [Crossref] [PubMed]
75. Xu ZY, Ding SM, Zhou L, et al. FOXC1 contributes to microvascular invasion in primary hepatocellular carcinoma via regulating epithelial-mesenchymal transition. Int J Biol Sci 2012;8:1130-41. [Crossref] [PubMed]
76. Xia L, Huang W, Tian D, et al. Overexpression of forkhead box C1 promotes tumor metastasis and indicates poor prognosis in hepatocellular carcinoma. Hepatology 2013;57:610-24. [Crossref] [PubMed]
77. Lin Z, Huang W, He Q, et al. FOXC1 promotes HCC proliferation and metastasis by Upregulating DNMT3B to induce DNA Hypermethylation of CTH promoter. J Exp Clin Cancer Res 2021;40:50. [Crossref] [PubMed]
78. Ray PS, Wang J, Qu Y, et al. FOXC1 is a potential prognostic biomarker with functional significance in basal-like breast cancer. Cancer Res 2010;70:3870-6. [Crossref] [PubMed]
79. Xu Y, Shao QS, Yao HB, et al. Overexpression of FOXC1 correlates with poor prognosis in gastric cancer patients. Histopathology 2014;64:963-70. [Crossref] [PubMed]
80. Ou-Yang L, Xiao SJ, Liu P, et al. Forkhead box C1 induces epithelial‑mesenchymal transition and is a potential therapeutic target in nasopharyngeal carcinoma. Mol Med Rep 2015;12:8003-9. [Crossref] [PubMed]
81. Wang LY, Li LS, Yang Z. Correlation of FOXC1 protein with clinicopathological features in serous ovarian tumors. Oncol Lett 2016;11:933-8. [Crossref] [PubMed]
82. Li YW, Wang JX, Yin X, et al. Decreased expression of GATA2 promoted proliferation, migration and invasion of HepG2 in vitro and correlated with poor prognosis of hepatocellular carcinoma. PLoS One 2014;9:e87505. [Crossref] [PubMed]
83. Song SH, Jeon MS, Nam JW, et al. Aberrant GATA2 epigenetic dysregulation induces a GATA2/GATA6 switch in human gastric cancer. Oncogene 2018;37:993-1004. [Crossref] [PubMed]
84. Rodriguez-Bravo V, Carceles-Cordon M, Hoshida Y, et al. The role of GATA2 in lethal prostate cancer aggressiveness. Nat Rev Urol 2017;14:38-48. [Crossref] [PubMed]
85. Peters I, Dubrowinskaja N, Tezval H, et al. Decreased mRNA expression of GATA1 and GATA2 is associated with tumor aggressiveness and poor outcome in clear cell renal cell carcinoma. Target Oncol 2015;10:267-75. [Crossref] [PubMed]
86. Tessema M, Yingling CM, Snider AM, et al. GATA2 is epigenetically repressed in human and mouse lung tumors and is not requisite for survival of KRAS mutant lung cancer. J Thorac Oncol 2014;9:784-93. [Crossref] [PubMed]
87. Zhang L, Wang K, Deng Q, et al. Identification of Key Hydroxymethylated Genes and Transcription Factors Associated with Alpha-Fetoprotein-Negative Hepatocellular Carcinoma. DNA Cell Biol 2019;38:1346-56. [Crossref] [PubMed]
88. Zhang JJ, Wang CY, Hua L, et al. miR-107 promotes hepatocellular carcinoma cell proliferation by targeting Axin2. Int J Clin Exp Pathol 2015;8:5168-74. [PubMed]
89. Zhang HF, Wang YC, Han YD. MicroRNA‑34a inhibits liver cancer cell growth by reprogramming glucose metabolism. Mol Med Rep 2018;17:4483-9. [Crossref] [PubMed]
90. Cheng B, Ding F, Huang CY, et al. Role of miR-16-5p in the proliferation and metastasis of hepatocellular carcinoma. Eur Rev Med Pharmacol Sci 2019;23:137-45. [PubMed]
91. Grant TJ, Bishop JA, Christadore LM, et al. Antiproliferative small-molecule inhibitors of transcription factor LSF reveal oncogene addiction to LSF in hepatocellular carcinoma. Proc Natl Acad Sci U S A 2012;109:4503-8. [Crossref] [PubMed]
92. Wan Z, Zhang X, Luo Y, et al. Identification of Hepatocellular Carcinoma-Related Potential Genes and Pathways Through Bioinformatic-Based Analyses. Genet Test Mol Biomarkers 2019;23:766-77. [Crossref] [PubMed]
93. Li HT, Wei B, Li ZQ, et al. Diagnostic and prognostic value of MCM3 and its interacting proteins in hepatocellular carcinoma. Oncol Lett 2020;20:308. [Crossref] [PubMed]
94. Shen S, Kong J, Qiu Y, et al. Identification of core genes and outcomes in hepatocellular carcinoma by bioinformatics analysis. J Cell Biochem 2019;120:10069-81. [Crossref] [PubMed]
95. Yang L, Yin W, Liu X, et al. Identification of a five-gene signature in association with overall survival for hepatocellular carcinoma. PeerJ 2021;9:e11273. [Crossref] [PubMed]
96. Mou T, Zhu D, Wei X, et al. Identification and interaction analysis of key genes and microRNAs in hepatocellular carcinoma by bioinformatics analysis. World J Surg Oncol 2017;15:63. [Crossref] [PubMed]
97. Wu Q, Jiang S, Cheng T, et al. A Novel Pyroptosis-related Prognostic Model for Hepatocellular Carcinoma. Front Cell Dev Biol 2021;9:770301. [Crossref] [PubMed]
98. Fan G, Tu Y, Chen C, et al. DNA methylation biomarkers for hepatocellular carcinoma. Cancer Cell Int 2018;18:140. [Crossref] [PubMed]
99. Huang WT, Wang HL, Yang H, et al. Lower expressed miR-198 and its potential targets in hepatocellular carcinoma: a clinicopathological and in silico study. Onco Targets Ther 2016;9:5163-80. [Crossref] [PubMed]
100. Jin Z, Wang M, Meng Y, et al. Prognostic Implication of a Cuproptosis-Related miRNA Signature in Hepatocellular Carcinoma. J Healthc Eng 2022;2022:4694323. [Crossref] [PubMed]
101. Lou W, Liu J, Ding B, et al. Identification of potential miRNA-mRNA regulatory network contributing to pathogenesis of HBV-related HCC. J Transl Med 2019;17:7. [Crossref] [PubMed]
102. Mittermeier C, Konopa A, Muehlich S. Molecular Mechanisms to Target Cellular Senescence in Hepatocellular Carcinoma. Cells 2020;9:2540. [Crossref] [PubMed]
103. Chen QF, Xia JG, Li W, et al. Examining the key genes and pathways in hepatocellular carcinoma development from hepatitis B virus‑positive cirrhosis. Mol Med Rep 2018;18:4940-50. [Crossref] [PubMed]
104. Gonda K, Tsuchiya H, Sakabe T, et al. Synthetic retinoid CD437 induces mitochondria-mediated apoptosis in hepatocellular carcinoma cells. Biochem Biophys Res Commun 2008;370:629-33. [Crossref] [PubMed]
105. Han T, Goralski M, Capota E, et al. The antitumor toxin CD437 is a direct inhibitor of DNA polymerase α. Nat Chem Biol 2016;12:511-5. [Crossref] [PubMed]
106. Hu L, Zhang T, Liu D, et al. Notoamide-type alkaloid induced apoptosis and autophagy via a P38/JNK signaling pathway in hepatocellular carcinoma cells. RSC Adv 2019;9:19855-68. [Crossref] [PubMed]
107. Hoter A, Rizk S, Naim HY. Heat Shock Protein 60 in Hepatocellular Carcinoma: Insights and Perspectives. Front Mol Biosci 2020;7:60. [Crossref] [PubMed]
108. Weygant N, Qu D, Berry WL, et al. Small molecule kinase inhibitor LRRK2-IN-1 demonstrates potent activity against colorectal and pancreatic cancer through inhibition of doublecortin-like kinase 1. Mol Cancer 2014;13:103. [Crossref] [PubMed]
109. Nevi L, Di Matteo S, Carpino G, et al. DCLK1, a Putative Stem Cell Marker in Human Cholangiocarcinoma. Hepatology 2021;73:144-59. [Crossref] [PubMed]
110. Jung K, Choi JS, Koo BM, et al. TM4SF4 and LRRK2 Are Potential Therapeutic Targets in Lung and Breast Cancers through Outlier Analysis. Cancer Res Treat 2021;53:9-24. [Crossref] [PubMed]