Key Epigenetic Players in Etiology and Novel Combinatorial Therapies for Treatment of Hepatocellular Carcinoma

Abstract

Hepatocellular carcinoma (HCC) is one of the leading causes of death in which the molecular tumorigenesis and cellular heterogeneity are poorly understood. The genetic principle that specific driver mutations in oncogenes, DNA repair genes, and tumor-suppressor genes can independently drive cancer development has been widely explored. Additionally, a repertory of harmful epigenetic modifications in DNA and chromatin—impacting the expression of genes involved in cellular proliferation, differentiation, genome stability, cell-cycle control, and DNA repair—are now acknowledged across various biological contexts that contribute to cancer etiology. Notably, the dynamic hypermethylation and hypomethylation in enhancer and promoter regions that promote activation or silencing of the master regulatory genes of the epigenetic programs is often altered in tumor cells due to mutation. Genome instability is one of the cancer hallmarks that contribute to transdifferentiation and intratumoral heterogeneity. Thus, it is broadly accepted that tumor tissue is dominated by genetically and epigenetically distinct sub-clones which display a set of genetic and epigenetic mutations. Here we summarize some functions of key genetic and epigenetic players and biochemical pathways leading to liver cell transformation. We discuss the role of the potential epigenetic marks in target genes thought to be involved in sequential events following liver lipid metabolism dysregulation, inflammation, fibrosis, cirrhosis, and finally hepatocellular carcinoma. We also briefly describe new findings showing how epigenetic drugs together with chemotherapy and immunotherapy can improve overall responses in patients with hepatic tumors.

1. Introduction

Hepatocellular carcinoma (HCC) is one of the leading causes of death in which the molecular tumorigenesis and cellular heterogeneity are poorly understood. Cellular and molecular mechanisms underlying liver diseases have been intensively investigated through basic and clinical studies [1,2,3]. Various endogenous and exogenous factors, including viral and bacterial agents, diabetes, drugs, alcohol abuse, autoimmune responses, and fat deposition can in due course lead to liver cell death, inflammation, fibrosis, cirrhosis, hepatocarcinogenesis, and late-on liver tumors [1,2,3]. Cell death is one of the major causes of liver injury, which is followed by sterile inflammation and long-last regeneration. Oxygen-free radicals (ROSs), lipid-derived metabolites; damage-associated molecular patterns (DAMPs); pathogen-associated molecular patterns (PAMPs); drugs; toxins; and cytokines, such as tumor necrosis factor-alpha (TNF-α), FasL/CD95-ligand, TNF-related apoptosis inducing ligand (TRAIL), and CD40-ligand, are the most frequent inducers of the cell-death processes named apoptosis, necroptosis, pyroptosis, and ferroptosis [4,5]. Diverse intrinsic and extrinsic factors that cause dysregulation of inflammation participate in liver tumorigenesis [6,7]. Genes involved in epigenetic transcriptional reprogramming with dominant roles during embryonic stem cell lineage commitment have been implicated in the malignant transformation of somatic cells into cancer cells [8]. There is also solid evidence that epigenetic modifications can either silence or activate oncogenes, tumor-suppressor genes, and genes involved in DNA repair and genomic maintenance as well as genes involved in cellular energy and metabolic signaling pathways [9]. Genome-wide promoter hypermethylation of CpG islands can contribute to oncogene activation by altering the expression of genes involved in cell growth, division, and survival. Deletion and mutation of epigenetic genes are also frequently found in many types of human cancers [9]. Understanding the mechanisms underlying these epigenetic changes is important for developing targeted therapies aimed at reversing aberrant DNA methylation patterns and restoring normal gene expression in cancer cells. In this narrative review, we update on recent findings in the literature supporting that epigenetics potentially contribute to the development of liver diseases and cancers. We briefly describe emerging innovative therapies to liver cancers that involve a combination of epigenetic drugs with chemotherapy and immunotherapy.

2. Epigenetic Pathways and Regulators

Epigenetics is broadly defined as the study of heritable changes in gene expression by specific biochemical modification in the DNA and histone complexes [10,11,12,13]. Contrary to mutations, DNA epigenetic marks are reversible. The structural changes brought by biochemical modifications regulate access to genes and control how genes are turned on or off in response to various physiological contexts. During germline development, DNA methylation patterns are extensively reprogrammed in order to reset the epigenome, allowing the development of germline cells into any cell type after fertilization. Epigenetic modifications in these cells ensure that genes essential for a cell’s specific function are expressed, while genes associated with other cell types are repressed. Epigenetic reprogramming is critical during transitions between cell states, such as the differentiation of stem cells into multiple somatic cell lines [10,11,12,13]. The family of DNA methyltransferases (DNMTs), including DNMT1, DNMT3A, and DNMT3B, are mediators of the covalent addition of the methyl group (-CH₃) from S-adenosylmethionine (SAM) on the fifth carbon of cytosine (5-mC) within the DNA sequence. The demethylase (KDM) family enzymes remove methyl (-CH₃) groups within the DNA sequence. The family of methylcytosine dioxygenases named ten-eleven translocation (TET), including TET1, 2, and 3, catalyzes hydroxylation of DNA 5-methylcytosine (5mC) resulting in 5-hydroxymethylcytosine (5hmC). The DNA repair machinery enzymes can replace 5hmC to cytosine (C) to maintain genome integrity. DNA methylation patterns serve as many proposals. They regulate the inactivation of the X-chromosome, silencing of the genes through genomic imprinting, reprogramming and stability of cellular differentiation as well as transposon inactivation [10,11,12,13]. CpG islands are stretches of DNA rich in cytosine (C) and guanine (G) nucleotides, often found near the promoters of genes, and their methylation status can influence gene expression. When CpG islands within gene promoters become hypermethylated, it can lead to the silencing or downregulation of gene expression. This is because DNA methylation at promoter regions typically inhibits the binding of transcription factors and other proteins necessary for gene transcription, effectively turning off the expression of those genes. The mammalian genome of embryonic stem cells (ESCs) is very highly methylated throughout the CpG dinucleotide sequence; thus, they lose their ability to properly differentiate while retaining their self-renewal capabilities. An epigenetic defect affecting the differentiation potential of stem cells rather than a gatekeeper mutation is much more frequently found in cancer cells. The CpG islands occurs within promoter regions near the transcription start sites and between enhancer regions in over 50% of most human genes [9]. When methylated, they correlate with gene-specific transcriptional repression [10]. While hypermethylation of CpG islands enhances the oncogene activation, hypomethylation of tumor-suppressor genes can promote proliferation and cell growth.

DNA is enfolded by specific classes of histone proteins, namely H2A, H2B, H3, and H4 [14]. This structure resembles beads on a string. Each bead in the extended chromosome is called a nucleosome, and the DNA–protein complex forms two types of chromatins: heterochromatin (condensed form) and euchromatin (extended form). Its access is only permitted when epigenetic chemical reactions change their highly condensed structures [14]. Histone methylation occurs on lysine (K) and arginine (R) residues. “K” is mono-, di-, and tri-methylated, while “R” is modified by mono- or di-methylation. Specific epigenetic modifications, such as acetylation (addition of CH3) of an ε-N-acetyl lysine amino acid on both histone and non-histone proteins are intermediated by histone acetyltransferases (HATs). Deacetylases (HDACs) promote the removal of CH3 groups on the chromatin, leading to the entrance and binding of transcription factors. A number of solid and hematologic cancers, especially those that appear in childhood, contain mutated histones, which are now entitled oncohistones [15]. Mutations, such as those that occur in the residues of a histone H3 tail (H3K27, H3G34 and H3K36) inhibit histone methyltransferase activity, causing dysregulation in the epigenetic programs. DNA transcription factors, repair proteins, and epigenetic regulators are also modified by ADP ribosylation, ubiquitylation (Ub), O-GlcNAcylation (O-Glc), SUMOylation (SUMO), and biotinylation (biotin). These chemical modifications also play pivotal roles in oncogenesis [10,11,12].

The methyltransferase (writer) and demethylase (eraser) enzymes together with specific effector proteins (readers) act on its chromatin target leading to the propagation cycles of an active or repressive epigenetic state. Two major chromatin remodeler complexes named PRC1 (polycomb repressive complex 1) and PRC2, formed by polycomb a (PcG) and trithorax (TrxG) core proteins, act together in epigenetic processes for transcriptional repression [15,16]. The catalytic subunit of PRC2 is the enhancer of zeste homolog 2 (EZH2), a histone methyltransferase. The PRC2 complex catalyzes the di- and tri-methylation of lysine 27 on histone H3 (H3K27me3), as well as other active histone marks, such as H3K4me3 and H3K36me3. The PRC1 complex formed by the chromobox (CBX), BMI1 (B lymphoma Mo-MLV insertion region 1 homolog), and RING1B, an E3 ubiquitin ligase, binds to H3K27me3 and catalyzes histone H2K119me ubiquitination, leading to chromatin compaction and gene silencing. Thus, the binding of PRC1/PRC2 complexes into specific sites, called polycomb response elements (PREs), and methylated histone tails temporally and dynamically control the transcription or silencing of the genes at physiological and pathological conditions [15,16]. Distinct H3K4me, H3K36me, and H3K79me methylation indicate active transcription, whereas H3K9me, H3K27me, and H4K20me methylation indicate transcriptional repression. Biochemical studies have found that H3K27M mutant histones inhibit the activity for the PRC2 methyltransferase complex that catalyzes H3K27 methylation, which is a critical step for oncogenesis [15]. The human switch/sucrose nonfermentable (SWI/SNF) complexes are ATP-dependent chromatin remodelers that exist in two main subtypes [15,16]. The SWI/SNF-A (also known as the BAF complex) contains the BRG/SMARCA4 ATPase subunit. The SWI/SNF-B (also known as the PBAF complex) contains the BRM/SMARCA2 ATPase subunit and other subunits, such as ARID1A (AT-rich interactive domain-containing protein 1A), ARID1B, and ARID2 [15,16]. The BRM and BRG subunits require ATP hydrolysis to disrupt the complex DNA and histones, thereby promoting nucleosome disassembly. These chromatin remodeling complexes when mutated and disrupted can independently promote oncogenic activities. The accessibility and activities of the transcriptional factors and RNA polymerases are also modulated by various families of non-coding RNAs (ncRNA) and microRNAs (miRNAs) [17,18]. NcRNAs act as scaffolds and mediate the recruitment and regulation of PRCs and their interactions with other chromatin-modifier complexes [17,18]. For example, the lncRNA Xist (X-inactive specific transcript) interacts with PRC2, and this is crucial for X-chromosome inactivation in females. The lncRNAs’ HOTAIR and miRNAs’ miR-101 and miR-214 that control polycomb complexes are frequently associated with cancer. Figure 1 shows a schematic outline of the most important enzyme groups and modifications on DNA molecules and histone proteins involved in the epigenetic programs for various cellular functions.

Advanced epigenomic assays for DNA accessibility (DNase-seq, ATAC-seq), the genome-wide methylation state of CpG (WGBS), histone mark enrichment (ChIP-seq), and transcription factor binding (TF ChIP-seq) are providing new insights into the epigenomic mechanisms and deeper understanding of their biological implications in physiological and pathological states [19]. The epigenomic maps provided by the Encyclopedia of DNA Elements (ENCODE) project (https://www.encodeproject.org/) (accessed on 9 November 2024) and The Cancer Genome Atlas (https://www.cancer.gov/tcga) (accessed on 12 November 2024) have proven to be a powerful tool for investigating potential targets for drug development [19]. A growing number of chemical probes and epidrugs to various structural and biochemical targets have been under investigation [20,21]. The full description of the epigenome map of each single embryonic stem cell and its variation throughout differentiated cell types will help us to predict complex traits and diseases and design therapeutic interventions in future healthcare [22].

3. Liver Disease Transition to Hepatocellular Carcinomas: Role of Epigenetics

The contribution of epigenetics in liver diseases has been examined in great detail elsewhere [23,24,25,26]. Overall the studies have demonstrated that the aberrant DNA methylation patterns influence the development of various types of liver diseases, including metabolic-associated steatotic liver disease (MASLD), previously known as non-alcoholic fatty liver disease (NAFLD), and metabolic-associated steatohepatitis (NASH), viral hepatitis, and HCC. MASLD encompasses a spectrum of conditions, ranging from simple steatosis to steatohepatitis, and later on advanced fibrosis, and eventually to cirrhosis and cancer. In this spectrum of disorders, dysregulated gene expression plays a significant role, with DNA methylation changes being a major contributing factor [23,24,25,26]. Liver fat accumulation is a common trigger for systemic insulin resistance and NASH, and eventually contributes to the development of cirrhosis and hepatocellular carcinoma. The multistep events leading to the development of liver injury, viral hepatitis, liver fibrosis, and MASLD and the participation of lifestyle factors have been partially understood [23,24,25,26,27,28]. The expression of microRNAs and long non-coding RNAs have emerged as biomarkers and therapeutic targets for liver disease [29,30,31]. MiRNAs act on posttranscriptional events promoting degradation and inhibition of the translated target mRNAs. Recent studies have shown that various ncRNAs can modulate key pathways involved in the development and progression of NAFLD and its associated insulin resistance [29,30,31]. Environmental factors, such as nutrients, bioactive compounds, alcohol, and toxins can influence epigenetic programs that facilitate the specific changes in chromatin and gene expression states in the liver. For example, folic acid, vitamin B12, and choline are critical sources of one-carbon metabolism, which generates the methyl group used in DNA methylation [32]. Increased alcohol intake is the major cause of a spectrum of liver diseases, including fatty liver, alcoholic hepatitis, and chronic hepatitis, in the world [33]. Chronic alcohol consumption promotes hypomethylation by reducing the levels of SAM, a metabolite co-factor used by DNMTs and HMTs for introducing chemical modification in DNA and histone molecules [32,33]. Our comprehension of environmental factors and genetic and epigenetic mechanisms involved in liver diseases continues to emerge [34].

HCC is a complex disease arising from hepatocytes through the sequential accumulation of mutations and dysregulation of multiple genetic, epigenetic, metabolic, and biochemical signaling pathways [1,2,3]. The majority of HCC develops in patients suffering from chronic liver diseases, including B and C viral hepatitis (HBV and HCV), MASH, MASLD, cirrhosis, obesity, and alcohol-related fatty liver diseases [1,2,3]. Although both viral hepatitis and MASLD can lead to HCC, the pathways to tumorigenesis differ substantially. For example, in MASLD, inflammation is often linked to metabolic dysfunction and lipotoxicity associated with oxidative stress rather than infection, leading to different epigenetic alterations, such as lipid-induced histone modifications and miRNA changes. HBV’s viral protein HBx can interact with host epigenetic machinery, affecting histone acetyltransferases and deacetylases, which changes the transcriptional landscape and can promote oncogenesis. HBV and HCV can also affect the expression of various non-coding RNAs, particularly microRNAs (miRNAs) and long non-coding RNAs (lncRNAs). Some miRNAs, such as miR-122 (a liver-specific miRNA), are dysregulated in viral hepatitis-associated HCC and play roles in cell proliferation and apoptosis [17].

Diverse classes of genes are captured to subvert the liver cellular processes, including telomere maintenance, Wnt/β-catenin activation, P53/cell-cycle regulation, oxidative stress excess, MET (hepatocyte receptor tyrosine kinase), and Ras-phosphatidylinositol 3-kinase (PI3K)-Akt-mammalian target of rapamycin (mTOR) kinase signaling pathways [1,2,3]. Figure 2 shows a schematic view of the key events, environmental factors, and genes involved during the histological and molecular transition of acute to chronic liver disease to hepatocellular carcinoma.

Cancer hallmarks are the distinct acquired capabilities of cancer cells to subvert critical processes, such as cellular replication, proliferation, differentiation, genomic stability, metabolism, immune response, and cell-death mechanisms [35,36]. Cancer cells undergo significant changes in their transcriptional program through dysregulation of HATs, HDACs, and chromatin remodeling complexes [37,38]. Global hypoacetylation of histones, particularly histone H4, following general transcriptional repression of tumor-suppressor genes, is frequently observed in cancers, including HCC [37,38]. It is now widely acknowledged that single cancer-cell evolution to multiple cancer clones is driven by the integration of diverse genetic and non-genetic modifications, which encompass changes in genomic integrity driven by DNA methylation, chromatin remodeling, and regulation of RNA, miRNA, and lncRNA expression [39,40]. It is worth noting that epigenetic genes are disrupted by mutation in close to 50% of hepatocellular carcinomas, bladder cancer, and medulloblastoma [37,38,39,40]. Therefore, a deeper understanding of genetic and epigenetic mechanisms will offer new clues for subsets of targets for new treatment of cancers.

Several types of cancers have been linked with microbial infection, including bacteria, viruses, and fungi, which produce toxins with pro-inflammatory and pro-carcinogenic compounds activated metabolically by the cytochrome P450 (CYP) system [41,42,43]. The gastrointestinal and hepatobiliary tracts are colonized by a great diversity of commensal bacterial species [44,45]. Loss of bacterial diversity, a major feature of dysbiosis, can lead to changes in the gut microbiota’s metabolism and progressive decline of the production of bioactive compounds, such as short-chain fatty acids (SCFAs) and secondary bile acids, which are essential to maintaining liver health [44,45]. Dysregulation of the gut–liver axis, referred to as the bidirectional connection between the gut and liver and their microbiotas, is frequently caused by the rupture of the intestinal mucosal barrier and bacterial translocation [44,45,46]. This complex interplay is particularly significant in patients with inflammatory bowel disease (IBD), as chronic gut inflammation and dysbiosis lead to increased translocation of bacterial components to the liver. The liver, in turn, responds to these microbial signals by activating hepatic inflammation, immune cells, and fibrogenic pathways. Studies in animal models and patients have confirmed that gut vascular barrier disruption contributes to hepatic fibrosis, NASH, and cirrhosis as well as HCC [47,48,49]. High dietary cholesterol consumption can promote liver cancer by modulating the production of toxic metabolites, as demonstrated in mouse models [47,48,49]. These studies showed that bacterial species, including Mucispirillum, Desulfovibrio, Anaerotruncus, and Desulfovibrionaceae increased substantially, while probiotic bacteria species, including Bifidobacterium and Bacteroides, are depleted in the mice gut microbiome [47,48,49]. The intestinal microbiota is colonized with Lactobacillus, Bifidobacterium, and bacteriodes rich in unmethylated CpG motifs which are poor at stimulating the TLR-9 signaling pathway for an innate immune response. Studies in mouse models of chronic liver disease showed that unknown bacterial species of intestinal microbiota can promote HCC by activation of the TLR4 signaling pathway [50]. The possible bacterial origin involved in the etiology of liver tumors remains to be confirmed [51]. Various molecular studies have detected live pathogenic microbiomes within solid tumors, for example, Fusobacterium nucleatum, E. Coli, and Helicobacter Pylori. It is now proven that the intratumoral and tumor microenvironment microbiome influence the patient’s response to immunotherapy with immune checkpoint inhibitors [41,42,43]. Many more studies are required to characterize anti- and pro-tumorigenic bacterial species and their role in inflammation, immune surveillance, and liver carcinogenesis.

The telomerase enzymatic complex promotes telomere shortening at the terminal of linear chromosomes. This cell-cycle checkpoint, and pro-senescence mechanism, act as a strong tumor-suppressor mechanism for human somatic cells. Studies in pre-lesions of hepatocellular tumors have uncovered a series of mutations in the telomerase (TERT) reverse transcriptase gene promoter region [52]. Increased expression of telomerase transcripts causes genomic amplification and instability, activation of DNA damage response (DDR), and mitochondrial dysfunction and increases intracellular oxidative stress. Notably, cirrhosis lesions have an accumulation of nodules consisting of senescent hepatocytes with shortened telomeres [52]. Histone phosphorylation at H2AX (forming γ-H2AX) is a well-known marker of DDR and is critical for recruiting DDR proteins to damaged sites. However, in diabetic and obese patients, chronic inflammation can lead to aberrant histone phosphorylation patterns, impairing DDR efficiency and contributing to genomic instability. Due to the combination of impaired DDR and abnormal histone modifications, HCC in diabetic and obese patients tends to be more aggressive [24,26]. TERT is a transcriptional co-activator in Wnt/β-catenin signaling involved in cellular growth, division, and differentiation. Between 10 and 15% of hepatocellular adenomas display mutations in exon 3, 7, and 8 of the beta-catenin gene (CTNNB1). This normally occurs together with mutations in the hepatocyte nuclear transcription factor (HNF1A). The brain-expressed X-linked protein 1 (BEX1) is a member of the protein family that controls the liver stem cell population through the activation of Wnt/β-catenin signaling [53]. Accordingly, inhibition of DNMT1-mediated methylation of the BEX1 promoter inhibited the activation of Wnt/β-catenin signaling, stemness, and tumorigenesis of liver cells [53].

Mutations in the components of SWI/SNF-A and -B chromatin-remodeling complexes lead to uncontrolled cell proliferation, genomic instability, and impaired DNA repair, contributing to cancer development [54]. The subunits ARID1A, ARID2, SMARCA4 (BRG1), and SMARCA2 (BRM) are recurrently mutated in cancers, including HCC [55,56]. Mutations in ARID2 are particularly detected in HCC associated with HBV and HCV infections. The target of rapamycin complex 1 (TORC1) is another important regulator of various signaling pathways that controls cell growth and metabolism. Mutation is one important step for driven hepatocellular carcinogenesis since it enables TORC1 to enter the nucleus, promoting the ubiquitination and degradation of ARID1A via the proteasomes [57]. The degradation of ARID1A causes the activation of the Hippo–YAP signaling pathway and acceleration of the transcription of genes for cell growth and liver cell transformation [57].

BMI-1 is a core protein component of the PRC1 complex, regulating stem cell lineage decisions during development and tissue regeneration. The PRC1/BMI-1 multifunction complex plays a significant role in many malignancies, including hepatocellular carcinomas [58]. BMI1 functions in the transcriptional regulation of many genes, including the p16^INK4a and p14^ARF genes, whose products are the CDK4 (cyclin-dependent kinase) and CDK6 proteins that function as inhibitor cell cycles [58]. In addition to its critical role in cell division, the PRC1/BMI1 complex inhibits TGFβ2 (transforming growth factor) expression by binding to its promoter, and this promotes hepatocarcinogenesis by maintaining the pluripotency of liver cancer stem cells [59]. Accordingly, BMI1 is co-expressed with liver stem cell markers CD133+ (Prominin-1), CD90 (Thy-1), and EpCAM (epithelial cell adhesion molecule), therefore confirming the critical role of stem cells in the origin of human cancer cells [59].

The SET1/mixed-lineage leukemia (MLL) family of histone methyltransferases, the SET1/MLL family, consists of several members, including SETD1A and SETD1B (part of the SET1A/B complexes), MLL1 (KMT2A), MLL2 (KMT2B), MLL3 (KMT2C), and MLL4 (KMT2D). Mutations in MLL family members may disturb the balance between gene activation (H₃K₄ methylation by MLL) and gene repression (H₃K₂₇ methylation by PRC2). Using the genome-wide CRISPR/Cas9 knockout library (GeCKO v2), Shen and colleagues identified the SETD1A gene as a critical diver for HCC stemness and progression [60]. SETD1A knockdown promoted upregulation of various tumor-suppressor genes and reduced the expression of oncogenes associated proliferation, differentiation, and collagen biosynthesis. Moreover, SETD1A inhibits EZH2 expression, thereby histone H₃K₂₇ metilation, a key event for transcriptional repression [60]. Moreover, the blockade of SETD1A methyltransferase activity inhibited the growth and expansion of CD133+ liver cancer cell clones and HCC progression [60].

Hepatocarcinogenesis increases exponentially in patients with chronic HCV infection even after long treatment with antivirals [61,62]. Recent studies have identified persistent epigenetic signatures associated with HCC development in HCV-infected patients post antiviral treatment [63]. It was also demonstrated that HCV disrupts the HCC circadian transcriptome and epigenome in humanized mouse models, and this correlated with disease persistence [64]. Thus, methylome analysis may provide a powerful tool to effectively differentiate those patients who may respond to epigenetic therapy. Various studies have demonstrated the importance of non-coding RNAs, microRNAs, small nucleolar RNAs, and long non-coding RNAs in HCC tumorigenesis. For example, miR-181, miR-194, miR-206, and miR-365 contribute to HCC progression by regulating liver stem cell expansion [65,66,67]. LncRNAs and microRNAs can be released by tumor cells through exosomes. A positive correlation between circulating and tissue levels of miR-519d, miR-494, miR-21, miR-365, HULC (highly upregulated in liver cancer), HOTAIR (HOX transcript antisense RNA), MALAT1 (metastatic-associated lung adenocarcinoma transcript 1), and H19 were found in HCC patients at diverse stages of disease [64,65,66,67]. However, despite the suspected roles of ncRNAs in disease progression, the precise mechanism remains limited.

Epidemiological clinical data, combined with the results from experimental animal models and cell-culture systems, continue to supply further fundamental information supporting the significant roles of epigenetics in the transition of liver diseases to liver cancers [34,47]. Accordingly, recent studies have endeavored to validate where manipulating epigenetic modifications could improve the outcomes of liver cancer patients undergoing different modalities of treatment. Next, we will discuss some findings in this direction.

4. Combinatorial Therapies for the Treatment of HCC

The current treatment options for HCC are mainly based on the tumor stage and a variety of grading criteria [1,2,3,68]. Surgical resection, TACE (transcatheter arterial chemoembolization), RFA (radiofrequency ablation), radiation, chemotherapy, liver transplantation, and their multiple combinations continue to be most effective in the early and intermediate stages of HCC [68]. Various novel therapeutic approaches combining immunotherapy with checkpoint inhibitors, namely PD-1, PD-L1, or CTLA-4 monoclonal antibodies (mAb) plus one tyrosine kinase inhibitor (TKI) or VEGF/VEGFR monoclonal antibodies, are being investigated to treat intermediate and advanced stages of HCC [68]. Figure 3 shows a short guide to possible combinatorial therapies for HCC. Clinical trials have shown that the additive and synergistic effects of certain combinations can improve the overall survival rate, with a reduced resistance and progression of the disease [68]. Next, we update some fundamental molecular mechanisms underlying the combinatorial regimes for HCC treatment. Cancer cells can overgrow by evading their clearance by cytotoxic T lymphocytes (CTLs) through upregulating costimulatory molecules that regulate T-cell activation at early or late immune response, named immune checkpoints [69]. The cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) has a high affinity for B7 ligands, then CD28, and it inhibits T-cell proliferation and IL-2 secretion and works as an early inhibitory signal for T-cell-receptor (TCR) activation and cell proliferation. Programmed death protein-1 (PD-1) and PD-2 along with its ligand 1 (PD-L1) and PD-L2 are immune checkpoints that prevent prolonged activation of T cells. Hepatitis A virus cellular receptor 2 (TIM3), the T-cell immunoglobulin and ITIM domain (TIGIT), lymphocyte-activating 3 (LAG3), and their ligands are secondary T-cell checkpoint regulators involved in stimulatory and immunosuppressive control of cellular immune responses. Tumor-infiltrating CD8+ T cells gradually become exhausted and hyporesponsive to tumor-associated antigens due to dysregulation in the mitochondrial metabolism and expression of immune checkpoint molecules [70]. Administration of monoclonal antibodies against the immune checkpoints inhibitors (ICI) or their ligands, by blocking the suppressive immune checkpoint proteins, can reactivate or recover the CD8+ T-cell immune response of cancer patients [70]. However, due to the low response rate to ICI therapy across cancer types, the associations with the classic treatment methods, such as standard chemotherapy and targeted therapy, have been explored [71].

Systemic antiangiogenic therapy based on monoclonal antibodies and small molecules targeting the vascular endothelial growth factor and receptor (VEGF/VEGFR) axis has notably increased both overall survival and progression-free survival of patients with various types of cancers [72]. Monoclonal antibodies primarily impede signal transduction by blocking ligand binding or interfering with receptor dimerization. On the other side, small-molecule tyrosine kinase inhibitors competitively bind to the receptor kinase domain to inhibit signal transduction, thereby impacting cellular growth, differentiation, survival, and migration. Sorafenib and lenvatinib were the first tyrosine kinase inhibitors approved for patients with advanced unresectable HCC. These molecules exhibit moderate to high affinity to a large array of targets, including VEGFR1-3, c-Kit (V-kit Hardy–Zuckerman 4 feline sarcoma viral oncogene homolog), PDGFR-β (platelet-derived growth factor), PDGFRa, FLT-3 (FMS-like tyrosine kinase-3), VEGFR1-3, FGFR1-4 (fibroblast growth factor receptor), RET (Ret proto-oncogene), and also some of their variants. These growth factors are released in an autocrine and paracrine fashion in the tumor microenvironment acting via angiogenic and non-angiogenic mechanisms to build up the formation of novel vessels [72]. Therapeutic benefits from the treatment using VEGF/VEGFR drugs have been attributed to transient normalization of tumor vessels, which favor oxygen flux, drug delivery, and migration of immune cells within the tumor microenvironment [72]. Various clinical trials have been conducted to explore the combination of immune checkpoint blockade and angiogenesis inhibitors. The combination of atezolizumab, a monoclonal antibody to PD-L1, and bevacizumab, a monoclonal antibody targeting VEGF-A (Atezo+Bev) was explored for the treatment of advanced HCC [68,73,74,75,76,77]. The combination of durvalumab, a monoclonal antibody to PD-L1, and tremelimumab, a monoclonal antibody to CTLA-4 immune inhibitory receptor (Durva+Treme) with sorafenib was also accessed in HIMALAYA/STRIDE clinical trials in patients with advanced HCC [68,73,74,75,76,77]. Overall, the combinatorial therapies significantly improved overall survival (OS) and progression-free survival of the patients. The ASCO Guideline Update published in May 2024 recommended that the Atezo + Bev and Durva + Treme treatments may be offered as first-line treatments for appropriate patients with advanced HCC. However, the clinical potential of these monoclonal antibodies is limited by the fact that the majority of patients develop immune-related adverse effects [78]. Drug-induced lesions are often seen in the gut, skin, endocrine glands, lung, and liver. In addition, the blockade of immune checkpoint receptors can lead to a distinct form of hepatotoxicity due to their role in inducing adaptive immune tolerance in the liver [79]. A critical step of liver immune tolerance is the expression of PD-L1 and PD-L2 ligands. The expression occurs mainly in non-parenchymal cells, including hepatic stellate cells, Kupffer cells, liver sinusoidal endothelial cells, and intrahepatic lymphocytes. The role of PD-L1 is dominant in Th1 and Th17 immune responses, while PD-L2 acts mainly in the Th2 immune response [80,81]. The interactions of these pathways can affect the activation and proliferation of CD4⁺ and CD8⁺ T-cell subsets. The methylation state at CpG islands within the promoter regions of PD-1, CTLA-4, and TIM-3 significantly influences their expression in both tumor cells and immune cells [82]. Hence, the potential application of ICI therapy with targeted therapy encounters obstacles, including acquired drug resistance and the occurrence of serious adverse events. Thus, various clinical trials combining ICI therapy with epigenetic regulatory enzyme inhibitors have been conducted to overcome these constraints.

EZH2 is one of the core proteins of polycomb repressive complex 2 (PRC2), which catalyze histone protein methylation and DNA methylation within CpG islands, causing the transcriptional repression of many gene promoter regions [83]. EZH2 is a gene frequently found upregulated in various types of cancers. Various compounds capable of inhibiting EZH2 activity via different mechanisms have been tested clinically [83]. Treatment of hepatocellular carcinomas with sorafenib reduces the levels of EZH2, which is consistent with its degradation via the proteasomes [84]. It is noteworthy that EZH2 has the ability to repress the expression of immune checkpoint inhibitor PD-L1 in hepatocellular carcinoma cells [85]. EZH2 catalyzes H3K27me3 on the PD-L1 and IRF1 (interferon regulatory factor 1) promoters, leading to transcriptional repression of both genes [86]. On the contrary, the histone lysine methyltransferase 2 A (MLL1) increased the transcription of PD-L1 via H3K4me3 methylation of its promoter [87]. Histone deacetylase 8 (HDAC8) is a class I histone deacetylase that is overexpressed in many cases of NAFLD in obese and diabetic patients as well as cancer patients [88]. HDAC8 promotes cell proliferation via the Wnt/β-catenin pathway, a mechanism that is dependent on HDAC8 interaction with EZH2. A selective pharmacological inhibition of HDAC8 and, consequently, the release of its tumor-suppressive effects, increased the cytotoxic activity of tumor-infiltrating CD8+ T cells in immunodeficient and humanized HCC pre-clinical models [88]. It also demonstrated that pharmacological inhibition of EZH2 by reducing the NOTCH1 signaling pathway diminished the resistance of tumor cells to the tyrosine kinase inhibitor sorafenib [89]. Together, these studies have shown that EZH2 and HDAC8 would serve as molecular switches controlling tumor cells´ escape and induction of resistance during T-cell immunotherapy and receptor tyrosine kinase therapy of HCC patients.

DNA demethylating inhibitors (DNMTis) are chemical molecules that block DNA methylation by acting as a nucleoside analog of cytosine, thereby reducing cellular DNA methylation [90]. When incorporated into a DNA double-helix structure, the complex directly impairs the catalytic actions of DNMTs causing its degradation. The cytosine nucleoside analogs, including 5-azacytidine, decitabine, zebularine, and guadecitabine, are the most successful epigenetic drugs in clinical practice. However, the in vivo mechanism of action remains undefined. Numerous classes of mobile repetitive elements, mainly endogenous retroviruses (ERVs) from young and ancient RNA viruses, are dormant in all eukaryotic genomes as a result of massive H3K9me and H4K20me methylation [91]. Pioneer studies demonstrated that colorectal cancer cell lines treated with 5-aza-2′-deoxycytidine (decitabine or dacogen) and 5-azacytidine (Vidaza^®) at low doses induce a viral mimicry response. This immune response refers to the cellular defense mechanism for the detection and control of virus infection [92,93]. Specifically, the viral mimicry response in cancer cells treated with DNMTi causes the reactivation and expression of truncated derived long-terminal repeats (LTR)-ERVs, which are integrated into the host’s genome as a provirus. The anomalous transcripts derived from ERV families, and possibly Alu, SINE and LINE repetitive elements, are converted into double-stranded RNAs (dsRNAs) in treated cancer cells. These transcripts are sensed by the different classes of endogenous pattern recognition receptor (PRR) and toll-like receptors (TLR) families. Members of the PRR family include the RIG-I (retinoic acid-inducible gene I) receptor, MDA5 (melanoma differentiation-associated protein 5), and MAVS (mitochondrial antiviral-signaling protein). Oligomerization and activation of intracellular RNA sensors ultimately results in the activation of the interferon regulatory transcription factors IRF3 and IRF7, which in turn induce the synthesis of type I and III interferons (IFNs). IFNs bind to their receptors and trigger the phosphorylation of the Janus kinase (JAK) and signal transducer of activator of transcription (STAT) factors. This initiates the transcriptional induction of large numbers of genes, collectively named the IFN-stimulated genes (ISGs), whose products carry out various antiviral activities, among them, MHC class I alleles and antigen peptide transporter 1 (TAP1) proteins required for the processing and presentation of the antigen peptides derived from LTR-ERVs. This is followed up by the recognition and death of cancer cells by infiltrated cytotoxic T cells (CTLs) [94,95,96].

More than 30% of ERV promoters in the human genome contain p53 binding. Thus, p53 may regulate ERVs through epigenetic mechanisms. One study has shown that p53 reactivation by nutlin, an MDM2 (murine double minute 2) inhibitor, triggers viral mimicry in response to the accumulation of dsRNA [97]. MDM2 inhibitors block p53 degradation, thereby activating wild-type p53. ERVs are silenced by transcriptional repression by histone demethylase LSD1 and DNA methyltransferase DNMT1, which are two genes negatively regulated by p53. This study showed that during the time of increased p53 activation induced by the MDM2 inhibitor, p53 inhibits LSD1 and DNMT1 expression, and this promotes the anomalous expression of ERVs [97]. The data are online with previous studies showing that inhibition of LSD1 and DNMT1 induce expression of ERVs and viral mimicry response genes, including IFN-stimulated genes and MHC class I genes [94,95,96]. The reactivation of ERVs and other retroelements, such as retrotransposons, can elicit either an oncogenic process or an anti-tumor immune response [98,99]. For instance, onco-exaptation is an alternative adaptive genomic mechanism that occurs when the LTR promoter region drives the expression of fused oncogenic proteins that drive malignancy [98,99]. On the other hand, small peptides generated from retrotransposon exaptation mechanisms may work as vaccines inducing the formation of antibodies. Various vaccines based on LTR-derived neoantigens have been clinically investigated [100].

Many other chemical probes and inhibitors for the subunits of PRC1/2, including EZH2; DOT1L/DOT1-like histone lysine methyltransferase; G9a/euchromatic histone-lysine N-methyltransferase 2; SETDD7; protein arginine methyltransferases PRMT3, PRMT4, PRMT5; and lysine demethylase LSD1; as well as for the bromo- and extra-terminal (BET) proteins BRD2, BRD3, and BRD4 have been tested in vitro and in vivo [20,21]. BRD4 is overexpressed in HCC; thus, various studies have been conducted to explore the efficacy of the inhibitors of BRD4 in liver disease transition to HCC. One study showed that JQ1-induced BRD4 inhibition significantly reduced the tumorigenesis in a mouse model of HCV and NASH-induced HCC [101]. The PRMT family of arginine methyltransferases are RNA-binding proteins involved in mRNA splicing, and their inhibition triggers dsRNA accumulation from intron-retained RNAs [102]. It has been demonstrated that inhibitors of PRMTs can increase the therapeutic effects of ICTs by inducing the migration of CD45.1+leukocytes, NK cells, CD4+T cells, CD8+T cells, DC, and monocytes in the HCC tumor environment of mouse models [103]. Various epigenetic-related genes (ERGs) are differentially expressed by immune and non-immune cells in the tumor microenvironment The tumor stromal cells, fibroblasts, pericytes, endothelial cells, and bone marrow mesenchymal stromal cells perform important activities for tumor resistance [104,105]. These host stroma cells and tumor cells interact constantly with T cells, neutrophils, monocytes, and myeloid-derived suppressor cells (MDSCs), resulting in potent immunosuppressive activity [104,105]. Together with tumor-infiltrating lymphocytes, microsatellite instability (MSI), tumor mutation burden (TMB), and mismatch repair enzymes, a variety of immune-oncology biomarkers assessed by next-generation sequence (NGS) have been used to predict HCC patient response to ICI treatment [106]. A recent study investigated the association between ERGs and inflammatory response-related genes (IRRGs) in tumor tissue from patients with HCC [103]. Different bioinformatics approaches were used to determine the scores of the immune and stromal cells, which allowed the stratification and estimation of responsiveness of patients to CTLA-4 and PD-1 immunotherapy [107]. A set of ERG genes, including BCL6 corepressor-like 1 (BCORL1), a transcriptional corepressor, the polycomb group of proteins BMI-1, chromobox 2 (CBX2), chromobox 3 (CBX3), the cyclin-dependent kinase 1 (CDK1), and the cyclin-dependent kinase 5 (CDK5), were highly expressed in one-third of HCC patients. The group of patients with high ERG expression showed great responsiveness to ICI therapy as compared to patients with low ERG expression. Interestingly, the combination of PD-L1 therapy with epigenetic therapy greatly induced the expression of cytokines and chemokines involved in the attraction and maturation of dendritic cells into TME [106]. The antiangiogenic treatment targeting the VEGF/VEGFR axis normalizes abnormal tumor vasculature. This leads to reprogramming the immunosuppressive tumor microenvironment and making it more receptive to emerging immunotherapies that are under investigation. Overall, these studies suggest that the combination of ICI, VEGF inhibitors, and epigenetic modifier drugs could be a new strategy for an effective, novel therapy for advanced HCC. Figure 4 displays step-by-step the cellular signaling molecules and biochemical mechanisms that may increase the synergistic actions of combinatory therapies being explored to treat HCC patients.

5. Conclusions and Future Directions

The discovery of various key gene players in the genetic and epigenetic signaling pathways that initiate and drive chronic liver diseases to hepatocellular carcinomas has increased our ability to formulate biological and pharmacological approaches for prevention and treatment. Advanced biochemical studies have chemically characterized a myriad of epigenetic marks in DNA catalyzed by methyltransferases and demethylases, and in the histone complex catalyzed by acetyltransferases and deacetylases and their associated chromatin remodeling complexes. Epigenetic regulators are frequently mutated or deleted, causing epigenetic reprogramming in tumor cells that potentiate genetic instability and the metabolic capacity of cells to proliferative in distinct clonal subpopulations. A growing number of epidrugs with known and unknown molecular mechanisms have been explored for many types of cancers, including liver cancers. Emerging new combinatorial therapies adding together monoclonal antibodies targeting immune checkpoint inhibitors and monoclonal antibodies and small tyrosine kinase (TKI) targeting the VEGF/VEGFR axis, as well as DNMTi, HDACi, and MDMi targeting the epigenome have shown promising results. There are many problems and challenges related to untangling pharmacological approaches that combine multiple drugs. Searching for specific biomarkers for the stratification of patients is crucial. Randomized trials are necessary for dosing safety and toxicity and for finding the balance between drug-induced tumor death and harmful adverse effects to patients. Significantly more efforts are needed to incorporate genomic and epigenomic molecular profiles into clinic diagnostics for early detection and more personalized effective treatments for patients with liver cancers.