The role of H3K27me3 methylation in cancer development

来源: 发布时间:2024-03-28 15:10:06 浏览次数: 【字体:

Review Article

Published: 04 January 2024

Longjiang Di & Wei-Guo Zhu, 

Genome Instability & Disease  Volume 5, pages 17–34, (2024)



Abstract

Although the importance of histone methylation in epigenetics was first suggested more than 50 years ago, research into histone modifications conducted in the past decade has led to an exponential increase in our understanding of histone H3 modifications. In particular, the involvement of H3 histone 27 lysine trimethylation in the development of various cancer phenotypes has been demonstrated. Unlike mutations in the DNA sequence, such epigenetic changes are reversible, suggesting that inhibitors of H3 histone 27 amino acid methylation enzymes could be used as anti-cancer agents. Here, we outline the regulatory functions of H3 histone 27 lysine trimethylation carried out by different enzymes, in carcinogenesis. We describe the role of H3 histone 27 lysine trimethylation as an important epigenetic regulatory mechanism in the development of various cancers via effects on inflammation, DNA damage repair, cell proliferation, cell metastasis, regulatory cell death, ferroptosis, and angiogenesis. Finally, we focus specifically on H3 histone 27 lysine trimethylation regulators and their future development as anti-cancer drugs.


Introduction

Cancer is a complex genomic disease with a high mortality rate (Nowell, 1976). Mutations in certain genes, including proto-oncogenes and tumor suppressor genes, play a key role in driving the development and progression of cancer (Vogelstein et al., 2013). Moreover, the increase in genetic research in recent years has highlighted the important contribution of epigenetic alterations to the mechanisms of cancer development (Chabanon et al., 2020). Changes in chromatin structure can lead to alterations in gene expression via proto-oncogene activation or tumor-suppressor silencing (Dawson & Kouzarides, 2012), for example, increased levels of H3K14me3 in cervical cancer cells lead to the activation of ATR, resulting in their resistance to replicative stress and affecting the prognosis of tumor patients (Zhu et al., 2021). SETD2 deficiency downregulates H3K36me3 enrichment and cholesterol efflux gene expression, leading to hepatocarcinogenesis (Li et al., 2021). Emerging evidence has highlighted a potential role for epigenetics in future therapeutic strategies for tumors. It has been found that more than 80% of patients with diffuse intrinsic pontine glioma (DIPG), a fatal malignant childhood tumor, have a heterozygous point mutation (H3K27M) at the histone H3K27 locus, which leads to a reduction in H3K27me3 levels. As surgical resection and conventional radiotherapy approaches are ineffective in treating DIPG, elevating intracellular H3K27me3 levels to inhibit tumor growth could represent a novel strategy for treating DIPG. Recent studies have identified a previously unknown mechanism via which α-ketoglutarate maintains the epigenetic status of low H3K27me3 levels. Based on this epigenetic mechanism, targeting α-ketoglutarate generation provide a new strategy for the treatment of DIPG (Chung et al., 2020).

Eukaryotic chromatin is made up of DNA and its major protein components, histones. This group of highly conserved proteins includes the core histones H2A, H2B, H3, and H4, as well as the splice histone H1. One pair of each core histone assembles with 146 base pairs of DNA to form nucleosomes, which comprise the basic units of chromatin (Kornberg, 1974; Margueron & Reinberg, 2010). When cells are stimulated by factors such as radiation, they initiate a rapid DNA damage response triggering various processes, including chromatin remodeling, cell cycle arrest, autophagy, apoptosis, and senescence (Harper & Elledge, 2007; Jackson & Bartek, 2009; Polo & Jackson, 2011; Soria et al., 2012). Dysregulation of these processes, resulting in cellular dysfunction and altered gene expression, is central to cancer development (Cacabelos et al., 2016; Sharma et al., 2010). By affecting chromatin structure, histone modification plays a key regulatory role in these crucial cellular processes. Important histone modifications include acetylation, ubiquitination, threonylation, glycosylation, phosphorylation, and methylation (Kouzarides, 2007). Methylation and acetylation, the most common types of modifications, were first observed to be involved in the regulation of RNA synthesis more than 50 years ago (Allfrey et al., 1964).

Over the years, our understanding of histone modifications has deepened. Methylation of histones mainly occurs at the positively charged amino acids lysine (K) and arginine (R), which are abundant within histone proteins. The most widely studied histone methylation sites include the histone H3 lysines 4, 9, 27, and 36 (H3K4, H3K9, H3K27, and H3K36). Methylation modifications of H3K4 and H3K36 mediate transcriptional activation, while methylation modifications of H3K9 and H3K27 mediate transcriptional repression. Different histone lysine residues have different—sometimes opposing—functions when methylated, and different degrees of methylation (mono-, di-, or tri-methylation) on the same residue may cause dramatic functional differences (Barski et al., 2007; Heintzman et al., 2009). In addition, histones can be methylated on arginine residues, a modification that can play a positive or negative role in transcription (Bedford & Clarke, 2009). To date, more than 20 methylation marks on lysine and arginine residues have been identified (Mosammaparast & Shi, 2010). Methylation of lysine residues in histone tails, in particular, has been a focus of epigenetics research in recent years (Fig. 1).

Fig. 1figure 1

Methylation sites of the major histone lysine residues

In general, methylation is considered to be more complex than other forms of histone modification. Methyl groups are turned over more slowly than other modifications—indeed, histone methylation was initially believed to be irreversible (Byvoet et al., 1972); the later discovery of demethylases such as lysine-specific demethylase 1A (KDM1A) revealed that histone methylation is, in fact, reversible (Shi et al., 2004). A large number of specific histone methylation-modifying enzymes are now known to mediate the addition or removal of methyl groups from various lysine residues on histones, catalyzing post-translational events. This specificity reflects the fact that enzymes that catalyze histone methylation tend to produce well-defined modifying functions (Xiao et al., 2003). The discovery of enzymes capable of producing a diversity of stable and altered methylation states has helped to establish the increasingly important regulatory role of H3 histone methylation in oncogenesis. In particular, aberrant epigenetic regulation of histone H3K27 modification seems to be involved in the activation of multiple oncogenic and tumor suppressor signaling pathways and the progression of many types of cancer (Ezponda & Licht, 2014; Li et al., 2017). Below, we delineate the mechanisms that can lead to the altered methylation of histone H3K27.

Generation and elimination of the H3K27 methyl marker

During the past decade, the enzymes associated with histone H3K27 methylation have been studied to uncover the mechanisms contributing to genetic epistasis regulation. The enzymes responsible for catalyzing the methylation/demethylation of histone H3K27 (histone methyltransferases and demethylases) are considered to be potential therapeutic targets for the clinical treatment of cancer (Shi & Whetstine, 2007; Yoshimi & Kurokawa, 2011). The main H3K27 methyltransferases include the PRC2 complex (which consists of the SET structural domain-containing histone methyltransferases EZH2 or EZH1 and core accessory proteins), as well as the proteins ATXR5, ATXR6, TXR1, EZL2, EHMT2, and EHMT1. The major H3K27 demethylases are UTX/KMD6A, JMJD3/KDM6B, and KIAA1718/KDM7A.

The enzymes mediating H3K27 methylation differ between species. In Arabidopsis, the methyltransferases ATXR5, ATXR6, MEDEA, CLF, and SWINGER play important roles in H3K27 methylation (Makarevich et al., 2006). Inactivation of ATXR5 and ATXR6 leads to a significant reduction in H3K27 methylation (Jacob et al., 2009). MEDEA, CLF, and SWINGER reduce gene expression by increasing H3K27 methylation (Gu et al., 2014; Schubert et al., 2006). In the unicellular eukaryote Tetrahymena thermophila, H3K27 methylation was significantly reduced by deletion of TXR1 (Zhang et al., 2013); however, H3K27 methylation status was not completely lost when TXR1 was inhibited, demonstrating the existence of other methyltransferases or pathways regulating H3K27 methylation. Gao et al. and others also identified EZL1, EZL2, and EZL3 (homologs of the Drosophila EZ proteins) as methyltransferases in Tetrahymena (Gao et al., 2013; Liu et al., 2007; Zhang et al., 2014). EZL2 was expressed at significant levels, indicating its potential importance as a regulator of H3K27 methylation (Gao et al., 2013).

In mammalian cells, EZH1 was the first enzyme reported to methylate H3K27, and EZH1 and EZH2 were subsequently shown to form part of distinct PRC2 complexes. The PRC2 complex consists of four core protein subunits, including EZH2 (or EZH1), zeste 12 inhibitor (SUZ12), embryonic ectodermal development (EED), and retinoblastoma protein-associated protein 46/48 (RbAp46/48). Shen et al. observed significant reductions in H3K27me2 and H3K27me3, as well as partial reductions in H3K27me1, in Ezh2−/− embryonic stem cells (ESCs) (Shen et al., 2008), while knockdown of Ezh1 in Ezh2−/− ESCs completely suppressed H3K27me1. Genome-wide analysis showed that cancer-associated overexpression of EZH2 resulted in increased H3K27me3 levels and repression of tumor suppressor genes (Deb et al., 2014; Kim & Yu, 2012). In addition, Pasini et al. demonstrated the specific loss of H3K27me2 and H3K27me3 in Suz12-deficient embryos (Pasini et al., 2004). Similarly, Montgomery et al. reported that EED knockdown reduced H3K27me2 and H3K27me3 levels (Montgomery et al., 2005). G9a/EHMT2 and GLP/EHMT1 have also been identified as H3K27 methyltransferases in vitro and in vivo (Mozzetta et al., 20142015; Tachibana et al., 2005; Wu et al., 2011). Removal of the H3K27 methyl moiety is performed by the histone demethylases UTX/KMD6A and JMJD3/KDM6B (Lan et al., 2007; Lee et al., 2007; Tsukada et al., 2006), as demonstrated in ESC differentiation experiments conducted by Agger et al. (2007). De Santa et al. also identified KDM6B as an inducible enzyme that catalyzed H3K27me3 removal and altered macrophage transdifferentiation (De Santa et al., 2007). Notably, although KDM6A and KDM6B contributed significantly to H3K27me2 and H3K27me3 demethylation, these enzymes did not show significant H3K27me1 demethylation activity. Another enzyme, KDM7A, was later identified as a histone demethylase capable of acting on both H3K27me1 and H3K27me2 (Huang et al., 2010; Lin et al., 2010; Yang et al., 2010). The overall production and elimination of H3K27 methyl marks reflect the balance between methyltransferases and demethylases. Moreover, the responses of these enzymes to various signaling cascades affect the activation/repression of oncogenes at different stages of a tumor’s development.

Different H3K27 methylation-related enzymes differ in the degree of methylation they can achieve. Monomethylation of H3K27 (addition of me1) is mainly carried out by ATXR5, ATXR6, TXR1, EZH1, EZH2, EHMT1, and EHMT2, while the methyltransferases capable of converting H3K27me1 to H3K27me2 are MEA, CLF, SWN, EZL2, EZH2, Suz12, and EED (Fig. 2). Demethylation of H3K27me1 and H3K27me2 is mainly performed by KDM6A, KDM6B, and KDM7A. In addition, MEA, CLF, SWN, EZL2, EZH2, Suz12, and EED are the major methyltransferases that convert H3K27me2 to H3K27me3, while demethylation of H3K27me3 is mostly carried out by KDM6A and KDM6B (Fig. 2).

Fig. 2figure 2

Enzymes involved in the methylation modification of histone H3K27

Although there have been many studies investigating H3K27me3 and the repression of gene expression, the functions of H3K27me1 and H3K27me2 in gene regulation remain poorly understood. Ferrari et al. demonstrated that H3K27me1 accumulates within transcribed genes, promotes transcription, and is regulated by Set2-dependent deposition of H3K36me3 (Ferrari et al., 2014). H3K27me2 is present on approximately 70% of total histone H3 and is distributed within large chromatin domains that exert a protective function by preventing the initiation of non-cell-type-specific enhancers. Considering the key role played by H3K27me3 in the regulation of gene expression and signaling pathways, the rest of this review will focus on new findings regarding H3K27me3 in cancer-related signatures, including cellular inflammation, DNA damage and repair, cell metastasis, proliferation, invasion, and apoptosis/ferroptosis (Table 1).

Signature

Description

No. of genes

Angiogenesis

Angiogenesis ensures that cancer cells receive continuous supplies of oxygen and other nutrients

73

Apoptosis

Cancer cell resistance to apoptosis leads to the persistence of grossly abnormal cells in tissues

66

DNA damage

Unrepaired DNA damage accumulates in replicating cells, potentially contributing to cancer progression

110

DNA repair

Deficits in DNA repair disrupt genomic integrity and may lead to carcinogenesis

119

Inflammation

Chronic inflammation can cause between 15 and 25% of human cancers

112

Invasion

Invasion is a critical carcinogenic event whereby cancer cells escape from their primary sites and spread into the blood or lymphatic vessels

97

Metastasis

Metastasis promotes the malignant transformation of cancer and causes most cancer deaths

166

Proliferation

Cell proliferation is responsible for tumor progression 

88

Table 1 Number of genes involved in key cancer signatures

H3K27me3 and inflammation

Inflammation is a self-induced response to tissue damage and often occurs in conjunction with the clinical manifestations of disease. Inflammation can be caused by a variety of types of trauma, including physical injury, ischemic injury, infection, and exposure to toxins. Acute inflammatory responses cause a series of physiological changes, giving rise to cellular changes at the damage site and immune stress in the body. If inflammation persists for physical reasons, acute inflammation becomes chronic, a process that leads to gradual cellular mutation; continued mutation is often a prerequisite for oncogenesis. Inflammation is caused by the activation of macrophages and the resulting production of inflammatory cytokines, including interleukin-1β (IL-1β), tumor necrosis factor-alpha (TNFα), and IL-10 (Arango Duque & Descoteaux, 2014). In human primary macrophages, lipopolysaccharide (LPS) was shown to stimulate KDM6B expression, promoting macrophage polarization toward a pro-inflammatory M1 phenotype via the NF-κB signaling pathway. By contrast, GSK-J4, an inhibitor of KDM6B, was shown to inhibit LPS-induced TNFα release in macrophages and attenuate the inflammatory response by promoting H3K27me3 accumulation (Kruidenier et al., 2012). In addition, during bacterial infection, deficiency in the m6A reader YTH-domain family 2 stabilizes KDM6B protein, promoting H3K27me3 demethylation and subsequently enhancing the transcription of the pro-inflammatory cytokines IL-6, IL-12, and ICAM (Wu et al., 2020).

The regulation of H3K27me3 status by different enzymes is closely associated with the activation of key signaling pathways that contribute to inflammation in various diseases. For example, a process of modulating KDM6B to regulate H3K27me3 methylation status and thus promote the persistence of inflammation has been demonstrated in necrotizing small bowel colitis (Ma et al., 2022), as well as during the normal physiological process of muscle repair in humans (Nakka et al., 2022). In addition, EZH-catalyzed H3K27me3 is involved in inflammatory processes associated with many diseases. For example, in bronchopulmonary dysplasia in preterm infants, EZH2-catalyzed H3K27me3 resulted in RUNX3 downregulation and caused intrauterine inflammation and infection (Zhu et al., 2015). A similar process has been identified in periodontal membrane cells in periodontal disease (Schuldt et al., 2022).

An increasing number of studies have shown that inflammation-associated DNA methylation is common in tumors, with 60% of aberrant tumor-associated methylation also present in inflamed tissues (Hahn et al., 2008). This suggests that methylation may have important regulatory effects on cancer development. To investigate this association more closely, we obtained a total of 112 genes associated with cancer inflammation from the CancerSEA bioinformatics database (http://biocc.hrbmu.edu.cn/CancerSEA) (Supplementary Table 1) and performed protein–protein interaction (PPI) network analysis for the human H3K27me3-related enzymes KDM6A, KDM6B, EZH2, SUZ12, and EED (Fig. 3). Our results showed that these proteins are closely associated with inflammogenesis genes, suggesting that regulation of H3K27me3 by these enzymes may be involved in the development of cancer-related inflammation.

Fig. 3figure 3

Protein–protein interaction network of related enzymes regulating H3K27me3 and cancer inflammation-related genes


H3K27me3 and DNA damage repair

DNA damage, which involves alteration of the chemical structure of DNA, frequently accumulates in cells during cancer development. DNA repair contributes to the maintenance of genomic integrity, while defective DNA repair often leads to cancer progression. In clinical practice, conventional chemotherapeutic agents usually act by targeting tumor cells to induce DNA damage and cell death (Cheung-Ong et al., 2013; Espinosa et al., 2003; Nitiss, 2009); however, H3K27me3-labeled regions may not be accessible to such agents because of their compact chromatin structure (Pang et al., 2015). Decompaction of chromatin may therefore enhance the therapeutic effects of cytotoxic drugs. Consistent with this notion, inhibition of EZH2 in elderly patients with acute myeloid leukemia (AML) resulted in the decompaction of H3K27me3-labeled chromatin, thereby enhancing chromatin accessibility and promoting chemotherapy-induced DNA damage/leukemia suppression (Porazzi et al., 2022). This finding indicates that lowering the dose of chemotherapeutic agents in AML may be possible, which could reduce treatment-related side effects (Porazzi et al., 2022). Other studies have shown that the H3K27M mutation in a subset of pediatric high-grade gliomas reduced overall histone H3K27 trimethylation by affecting the enzymatic activity of EZH2, leading to enhanced DNA damage repair in tumor cells, thereby promoting tumorigenicity of cancer cells in vivo and resistance to tumor radiotherapy (Bender et al., 2013; Lewis et al., 2013). In glioblastoma, mutations in the telomerase reverse transcriptase promoter were shown to stimulate EZH2 expression and pharmacologically inhibit lipid accumulation and carcinogenesis by regulating H3K27 trimethylation (Ahmad et al., 2017). In addition, in various cancers such as hepatocellular carcinoma (HCC), triple-negative breast cancer, small cell lung cancer, and bone marrow tumors, H3K27me3 is influenced by protein modifications/gene mutations, with concomitant effects on DNA damage/repair-related pathways and tumorigenesis (Gao et al., 2017; Gardner et al., 2017; Imgruet et al., 2021; Yamamoto et al., 2019). For example, in small cell lung cancers that rapidly develop chemoresistance to cisplatin and etoposide, significant inhibition of SLFN11 (a factor associated with defective DNA damage repair) expression was shown, while silencing of SLFN11 expression in vivo was associated with significant deposition of H3K27me3; thus, combining an H3K27me3 inhibitor with standard cytotoxic therapies could improve small cell lung cancer chemotherapeutic efficacy (Gardner et al., 2017).

PPI network validation for genes related to cancer DNA damage and repair (obtained from the CancerSEA bioinformatics database, as before; Supplementary Table 1) revealed close associations with the human H3K27me3-related enzymes KDM6A, KDM6B, EZH2, SUZ12, and EED (Fig. 4). These findings suggest that regulation of H3K27me3 by these enzymes is involved in DNA damage and repair in cancer cells. These results are highly relevant to the study of the pharmacological modulation of H3K27 trimethylation in clinical treatment for cancer. Indeed, many anti-cancer drugs are known to induce DNA breakage by modulating H3K27me3 levels. For example, Pang et al. discovered that the anthracycline aclarubicin expelled histones from H3K27me3-labeled heterochromatin, exerting an important regulatory effect on diffuse large B-cell lymphoma cells with elevated H3K27me3 levels (Pang et al., 2015). Tang et al. showed that clinical treatment with metformin significantly reduced histone H3K27me3 levels and thus played an anti-tumor role in ovarian cancer (Tang et al., 2018). Similarly-acting drugs include tanshinone I, which is effective against AML (Huang et al., 2021), and sorafenib, which is effective against HCC (Kusakabe et al., 2021). Thus, interestingly, although anti-cancer drugs can produce differing genomic effects in different classes of tumors, regulating the DNA damage and repair effects caused by H3K27me3 seems to represent a common pathway for many agents.

Fig. 4figure 4

Protein–protein interaction network of related enzymes regulating H3K27me3 and cancer DNA damage or repair-related genes


H3K27me3 and cell proliferation, metastasis, and invasion

Aberrant proliferation is a hallmark of carcinogenesis. As an early step in tumor development, the strict regulatory mechanisms that control survival and proliferation in normal cells are lost in tumor cells, allowing cancer cells to survive beyond their normal lifespan and proliferate aberrantly. These changes involve the altered expression of cell cycle-related proteins and changes in many signal transduction pathways resulting from the dysregulation of epigenetic mechanisms (Feitelson et al., 2015). Following tumor development, cancer cells not only remain in a state of abnormal proliferation but also undergo metastasis and invasion. In most cancers, Metastasis involves the malignant transformation of cancer cells and their spread to other sites, increasing the risk of mortality. Invasive cancer cells spread from their primary site into the blood or lymphatic vessels and participate in tumorigenesis steps—including epithelial-to-mesenchymal transition, angiogenesis, and the triggering of autophagy—by producing excessive levels of hormones, inflammatory factors, and other cytokines (Jedinak et al., 2010; Qian & Pollard, 2010). Dysregulation of epigenetic mechanisms may act as an important driver of certain tumorigenesis mechanisms (Timp & Feinberg, 2013). For example, epigenetic changes driven by alterations in H3K27me3 can result in altered signaling pathways that allow cells to resist normal apoptotic mechanisms and promote their proliferation (Hanahan & Weinberg, 2011). Thus, H3K27me3 represents a meaningful target for the development of new therapeutic agents that act prior to or at the earliest stages of malignant transformation.

In recent years, research into the involvement of H3K27me3 in the proliferation, metastasis, and invasion of tumor cells has progressed rapidly. For example, in 2017, Wu et al. demonstrated that the long non-coding RNA (lncRNA) CASC9 negatively regulated PDCD4 expression in esophageal squamous cell carcinoma by recruiting EZH2 and subsequently altering the levels of H3K27me3, thereby promoting cancer cell proliferation (Wu et al., 2017). In 2019, Ma et al. found that in human glioma specimens, β-catenin/USP1/EZH2 promoted cell proliferation in vitro and tumor formation in vivo; moreover, USP1-mediated stabilization of EZH2 and promoted its recruitment to the CDKN1B, RUNX3, and HOXA5 promoters, leading to H3K27me3 enrichment and repression of target gene expression (Ma et al., 2019). In the same year, Yang et al. found that HOXD-AS1 inhibited the proliferation and migration of colorectal cancer (CRC) cells in vitro, as well as CRC tumorigenesis and metastasis in vivo, by suppressing HOXD3-mediated activation of the MAPK/AKT signaling pathway; specifically, HOXD-AS1 inhibited HOXD3 transcription by inducing the accumulation of H3K27me3 at the HOXD3 promoter (Yang et al., 2019). In 2020, Ricci et al. revealed that the proto-oncogene Bmi1 mediated tumor proliferation, metastasis, and invasion in a mouse model of glioblastoma-initiating cells via the EfnA5 gene, regulated by H3K27me3 (Ricci et al., 2020). In 2021, Wang et al. discovered that the lncRNA DNM3OS was significantly upregulated in HCC compared with its expression in adjacent normal liver tissue; tumor-associated mesenchymal stem cells accelerated HCC growth and metastasis via the DNM3OS/KDM6B/TIAM1 axis, whereby the binding of DNM3OS to KDM6B induced its expression by reducing H3K27me3 at the TIAM1 promoter (Wang et al., 2021). Yang et al. demonstrated that a combined intervention using inhibitors of PRMT5 (GSK591) and EZH2 (GSK126) synergistically inhibited the growth of CRC cells/xenografts in vitro and in vivo by reducing EZH2 binding and H3K27me3 deposition (Yang et al., 2021). In addition, in 2022, Ying et al. found that the silencing of zinc finger protein 280C in CRC cells inhibited cell proliferation, clone formation, cell migration, xenograft growth, and liver metastasis, for which regulation by H3K27me3 was essential (Ying et al., 2022). Finally, Peralta-Arrieta et al. showed that MEOX2/GLI1 increased tumor cell proliferation and resistance to anti-cancer chemotherapeutic agents in human lung cancer; this effect was mediated via aberrant epigenetic regulation of the EGFR/AKT/ERK signaling pathway, including reduced binding of histone H3K27me3 to EGFR regulatory regions (Peralta-Arrieta et al., 2022). These studies, combined with bioinformatics analyses (Fig. 5), have identified H3K27me3 regulation as a key factor in tumor cell proliferation, metastasis, and invasion, which may have implications for cancer treatment. Despite the improved efficacy of anti-cancer treatment in recent years, the prognosis for many patients with cancer remains poor. Cancer therapy often relies on cytotoxic drugs that kill highly proliferative tumor cells but also affect rapidly proliferating normal cells in the gastrointestinal tract, hair, and epithelium. Moreover, most targets of oncology drugs are multidirectional and can regulate different signaling pathways, thus these drugs can have potentially leading to irreversible side effects. By contrast, regulation of H3K27me3 may help to delay cancer onset and/or reverse cancer cell proliferation and reduce the toxicity associated with current anti-cancer drugs.

Fig. 5figure 5

Protein–protein interaction network of related enzymes regulating H3K27me3 and cancer cell proliferation-, metastasis-, and invasion-related genes


H3K27me3 and apoptosis/ferroptosis

Dysregulation of apoptosis and ferroptosis persists in cancerous tissues. Apoptosis is a natural, programmed form of cell death induced by a variety of cellular signals. The process of apoptosis includes cell breakage and shrinkage, chromatin condensation, and apoptosome production (Saraste & Pulkki, 2000), as well as nuclear fragmentation. The apoptotic process depends on extrinsic and intrinsic pathways. The intrinsic pathway is triggered when cells are exposed to certain harmful stimuli that cause DNA damage, such as radiation, toxins, or hypoxia, resulting in the formation of pores in the mitochondrial membrane via Bcl-2-associated X protein (BAX) and Bcl-2 antagonist killer 1 (BAK). The resulting massive release of cytochrome c activates effector caspases that break up the nucleus and degrade cytoskeletal elements, leading to the eventual formation of apoptotic bodies (Kalkavan et al., 2018; Kharbanda et al., 1997; Zhang et al., 2017). By contrast, the exogenous pathway depends on ligand-receptor interactions, such as TNFα binding to the TNFα receptor and Fas ligand (FasL) binding to the Fas receptor, which induce apoptosis by activating caspases (Lavrik et al., 2005; Strasser et al., 2009). Pro- and anti-apoptotic proteins, including BAX, BAK, Bcl-2, Bcl-xL, p53, and TNF-related apoptosis-inducing ligands (TRAIL), have been shown to regulate the apoptotic process by influencing mitochondrial membrane impermeability, cytochrome c release, and intracellular Fas signaling (el-Deiry, 1998; Kale et al., 2018; Nakano & Vousden, 2001; Rathore et al., 2017; Yuan et al., 2018).

The H3K27me3 marker has been found to play a key role in apoptosis. A systematic examination of the relationship between 1923 apoptosis-related genes and the associated histone modifications during neuronal development, by Jiang et al., showed that downregulation of apoptotic genes could be attributed to reduced H3K4me3 signals at promoters, while upregulation of apoptotic genes resulted from the loss of H3K27me3 marks (Jiang et al., 2021). As expected, many other studies have shown the regulation of apoptosis by H3K27me3. For example, miR-143 activates the BAX-dependent caspase 3 signaling pathway to promote apoptosis, and H3K27me3 can inhibit miR-143 transcription by affecting the recruitment of transcription factors and binding proteins, thereby affecting the process of apoptosis (Zhong et al., 2020); subsequently, increased H3K27me3 labeling combined with diminished apoptosis caused unstable cell growth, leading to rapid proliferation and promoting the initiation of the malignant process. This finding demonstrates the key role played by H3K27me3 in the regulatory processes of tumor cells. Another study reported that nuclear receptor-binding SET structural domain protein 1 (NSD1) increased the viability of breast cancer cells and inhibited apoptosis via the NSD1/H3K27me3/Wnt10b/β-catenin signaling pathway, thereby causing resistance to drugs such as paclitaxel (Chen et al., 2022). Evasion of apoptosis is one of the characteristics of cancer cell formation. Indeed, many treatment-resistant cancers have shown an association between H3K27me3 and the ability of tumor cells to evade apoptosis, including glioblastoma, oral cancer, AML, cervical cancer, B-cell lymphoma, CRC, breast cancer, and ovarian cancer (Akpa et al., 2019; Benard et al., 2014; Chu et al., 2020; Dev et al., 2021; Huang & Ling, 2017; Li et al., 2012; Lin et al., 2019; Zhao et al., 2021). Bioinformatics analysis conducted using the CancerSEA database (http://biocc.hrbmu.edu.cn/CancerSEA), as before (Supplementary Table 1), also showed associations between KDM6A, KDM6B, EZH2, SUZ12, and EED and multiple cancer-related apoptosis genes (Fig. 6). These findings confirm that H3K27me3, which is regulated by these enzymes, modulates the resistance of cancer cells to apoptosis.

Fig. 6figure 6

Protein–protein interaction network of related enzymes regulating H3K27me3 and cancer apoptosis-, ferroptosis-related genes

Ferroptosis is a novel, iron-regulated form of programmed cell death—distinct from apoptosis, cell necrosis, and autophagy—that was initially defined as iron-dependent cell death induced by erastin and RAS-selective lethal factor 3 (Dixon et al., 2012). In the presence of divalent iron or ester oxygenase, the levels of unsaturated fatty acids in cell membranes increase and lipid peroxides and reactive oxygen species (ROS) gradually accumulate, resulting in cell death. Morphological features of ferroptosis include mitochondrial shrinkage, mitochondrial membrane condensation, and reduction or elimination of mitochondrial cristae (Agmon et al., 2018); these features are caused by excessive ROS production and lipid peroxide accumulation. Ferroptosis is also induced by the inhibition of GPX4, a core antioxidant enzyme, because of the resulting accumulation of cytotoxic lipid peroxides.

Some studies have shown that H3K27me3 may play a role in diseases related to ferroptosis in various organs, including the liver, kidney, and heart. For example, hepatitis B virus (HBV) X protein promotes ferroptosis in acute liver failure via EZH2/H3K27me3-mediated inhibition of SLC7A11 (Liu et al., 2021), while palmitic acid stimulation of cardiomyocytes decreases H3K27me3 levels and promotes ferroptosis. Moreover, GSK-J4, a potent dual inhibitor of the H3K27me3/me2 demethylases JMJD3/KDM6B and UTX/KDM6A, was shown to reduce ferroptosis in cardiomyocytes by preserving H3K27me3 levels (Xu et al., 2022). H3K27me3 is also involved in the pathophysiological processes of tumors via the regulation of ferroptosis. For example, in melanoma, UNC 1999, an inhibitor of EZH1 and EZH2, downregulates the levels of H3K27me3 and inhibits cancer cell growth via the ferroptosis pathway (Hou et al., 2022). Ferroptosis plays different roles in different diseases, and H3K27me3 can influence disease progression by inhibiting or promoting ferroptosis, providing a promising target for the treatment of many diseases. As before, using ferroptosis-related regulators and markers from a ferroptosis database (http://www.zhounan.org/ferrdb) (Supplementary Table 2) to conduct PPI network analysis, we confirmed that the H3K27me3-associated enzymes KDM6A, KDM6B, EZH2, SUZ12, and EED are closely associated with ferroptosis-related genes in humans (Fig. 6). These findings indicate that H3K27me3 may play an important role in the regulation of ferroptosis in cancer cells.

H3K27me3 and angiogenesis

The survival and proliferation of cancer cells require the establishment of a blood supply to meet their demand for oxygen and nutrients and perform other metabolic functions (Potente et al., 2011). The formation of this vascular network occurs via angiogenesis. Vascular homeostasis is regulated by a large number of pro-angiogenic and anti-angiogenic factors. Tumors will remain dormant if these factors are in balance. Conversely, when homeostasis is imbalanced and pro-angiogenic signaling dominates, tumors are released from dormancy and new blood vessels are generated, leading to the rapid growth of the associated cancer cells. In 2003, anti-angiogenic therapies were successfully used to treat cancer in clinical trials and were subsequently approved by the US Food and Drug Administration (Hurwitz et al., 2004). This clinical process has accelerated research into the mechanisms of tumor angiogenesis and the associated signaling pathways. Studies have shown that in the immediate early stages of cancer, the basal lamina isolates the tumor from normal tissue containing blood vessels (Bluff et al., 2009; Bossi et al., 1995). As the tumor progresses from a benign to a malignant stage, the cancer cells become more aggressive, inducing a powerful angiogenic response that triggers the development of an actively growing and infiltrating vascular network (Bergers & Benjamin, 2003).

The extent of tumor vascularization can vary considerably as a result of regulation by tumor-inducing drivers, tumor-associated stromal cells, bioactive products, and other signaling molecules (Blouw et al., 2003; Bluff et al., 2009; Fukumura et al., 1997; Jubb et al., 2011; Monsky et al., 2002; Morrissey et al., 2008). Notably, the key drivers of angiogenesis, i.e., pro-angiogenic factors, can be triggered by epigenetic alterations in tumor cells. Among these, H3K27me3 has an important regulatory role. Recently, the combination of EZH and H3K27me3 was found to regulate TPM1 to promote cancer cell proliferation and angiogenesis in CRC (Liang et al., 2021). EZH2 and H3K27me3 were also found to be involved in the proliferation, angiogenesis, and metastasis of HCC cells via the downregulation of miR-138-5p (Bai et al., 2022). Several studies have shown that the involvement of H3K27me3 in angiogenesis also synergistically affects the proliferation, migration, and invasion of various tumor cell types, including prostate, breast, and ovarian cancer cells (Jain et al., 2020; Li & Zhang, 2013; Lyu et al., 2013; Ning & Ma, 2019; Xu et al., 2017). Analysis of cancer angiogenesis-related genes in the CancerSEA database (http://biocc.hrbmu.edu.cn/CancerSEA) (Supplementary Table 1) indicated their association with H3K27me3-related enzymes (Fig. 7), demonstrating the involvement of H3K27me3 in tumor angiogenesis.

Fig. 7figure 7

Protein–protein interaction network of related enzymes regulating H3K27me3 and cancer angiogenesis-related genes


H3K27me3 and autophagy

Autophagy is a physiological process involving the degradation of cytoplasmic macromolecules, aggregated proteins, damaged organelles, and pathogens by lysosomes (Cuervo, 2004; Klionsky, 2007; Levine, 2007; Levine & Klionsky, 2004; Levine & Kroemer, 2008; Mizushima et al., 2011; Rogov et al., 2014). There are three autophagic pathways: macroautophagy (also referred to simply as autophagy), microautophagy, and molecular chaperone-mediated autophagy (Glick et al., 2010; Xie & Klionsky, 2007). Macroautophagy occurs as a physiological stress response and consists of several steps, including induction, assembly, and formation of autophagosomes; docking and fusion of autophagosomes with lysosomal membranes; and degradation of autophagosomal contents (Hansen & Johansen, 2011; Mizushima, 2007) (Fig. 8). Damaged membrane-bound organelles such as mitochondria and other entities such as invading pathogens are processed in this way (Galluzzi et al., 2017; Levine & Kroemer, 2019). In microautophagy, cellular material is sequestered via invagination of late endosomal or lysosomal membranes (Mejlvang et al., 2018; Sahu et al., 2011; Uytterhoeven et al., 2015). By contrast, chaperone-mediated autophagy involves the recognition of KFERQ-containing proteins by the heat shock cognate 71 kDa protein HSC70/HSPA8 and binding to lysosomal-associated membrane protein 2A, followed by entry into the lysosome and degradation (Kaushik & Cuervo, 2018).

Fig. 8figure 8

The process of autophagy and the protein–protein interaction network of related enzymes regulating H3K27me3 and cancer autophagy-related genes

Autophagy is essential for cellular homeostasis in vivo, acting to maintain the metabolism and development of cells under endogenous or exogenous stress. In addition, the elimination of damaged proteins and organelles by autophagy maintains protein quality and quantity (Mizushima & Komatsu, 2011; Rabinowitz & White, 2010). In recent years, the various regulatory roles of autophagy have been studied in depth. In healthy physiological states, autophagy has been demonstrated to be directly involved in the regulation of multiple processes, including developmental mechanisms (Allen & Baehrecke, 2020; Mizushima & Levine, 2010), maintenance of stem cell self-renewal activity (Chen et al., 2018; Dong et al., 2021), and cell differentiation and survival (Boya et al., 2018; Clarke & Simon, 2019). Autophagic mechanisms are also involved in intercellular communication, mediating the secretion of non-classical proteins (Ponpuak et al., 2015; Zahoor & Farhan, 2018), regulating immune cell function (Deretic, 2021), and maintaining the integrity of tissue barriers (Galluzzi & Green, 2019; Levine & Kroemer, 2019).

Autophagy regulation plays a complex role during the development of cancer. For example, dysfunctional regulation of autophagic cell death can contribute to the development of some human cancers (Amaravadi et al., 2011). Interestingly, the potential ability of autophagy to regulate cell death makes it a therapeutic target for cancer (Mathew et al., 2007; White & DiPaola, 2009). In liver cancer, elevated levels of hyperhomocysteine (Hcy) can cause persistent liver injury, which increases the incidence of hepatocellular carcinoma. Studies have reported that Hcy inhibits cystic fibrosis transmembrane conductance regulator (CFTR) expression through the interaction between H3K27me3 and DNA methylation, thereby activating autophagy and leading to persistent liver injury (Yang et al., 2018). Used in myeloma therapy, chidamide is a novel inhibitor of the histone deacetylase benzamide with potent antimyeloma activity. The autophagy-inhibitory effect of chidamide in myeloma cells is mediated primarily by significant downregulation of the histone deacetylase SIRT1 and upregulation of the histone methyltransferase EZH2, which lead to the acetylation of histone H4 lysine 16 (H4K16ac) and elevated overall levels of histone H3 lysine 27 trimethylation (H3K27me3) in the promoter region of the autophagy-associated gene LC3B, which enhances the therapeutic efficacy of antimyeloma therapy (Xu et al., 2020). In the past decade, many research groups have identified a multilevel role for H3K27me3 in regulating autophagy during tumor processes in various cancers, including head and neck cancers, myeloma, and glioblastoma (Lee et al., 2021; Liu et al., 2018). In addition, we downloaded 231 genes associated with autophagy (Supplementary Table 1) from the autophagy database (http://www.autophagy.lu/clustering/index.html). Our PPI network analysis showed that autophagy genes were closely related to H3K27me3 methylation-related enzymes (Fig. 8), confirming the role of H3K27me3 in autophagic processes.

Future prospects for H3K27me3

Histone methylation is a relatively stable epigenetic mark that is often involved in tumorigenesis, with alterations in H3K27 trimethylation being particularly critical for cancer development. The discovery of H3K27 histone methyltransferases and demethylases provides a theoretical basis for the use of H3K27 trimethylation as a therapeutic target for cancers, and the future may see the discovery of additional methylation-modifying enzymes that can be used as new therapeutic targets for anti-tumor drugs. Currently, drugs targeting H3K27me3 modification are mainly focused on anti-tumor therapy for DIPG. The small molecule ONC201, a dopamine receptor D2 (DRD2) antagonist, has been shown to be effective in treating DIPG by altering H3K27me3 activity, in both phase 2 clinical and phase 3 placebo-controlled trials for the treatment of H3K27M-DIPG (Venneti et al., 2023). Drugs targeting H3K27me3 can help overcome the limitations of single-target therapy for patients with cancer and increase patients’ chances of survival. However, while overcoming the problem of a single target, multi-target histone epitope inhibition may increase drug toxicity, so it will be important to find ways to increase the specificity of such drugs if they are to be successfully used in cancer therapy.

In this review, we have detailed the current understanding of the role H3K27 trimethylation plays in carcinogenesis. However, several questions regarding the dynamics of H3K27 methylation remain to be explored. For example, how do the complex regulatory mechanisms of H3K27me3 play a role in cancer progression and what impact do non-coding RNAs have on the relationship between H3K27me3 and cancer development? The major signaling pathways involving H3K27me3 during tumorigenesis also need to be elucidated. A clearer understanding of these unresolved areas will help us to better understand the epigenetic regulatory role of histone modifications in cancer.

Data availability

All the raw data generated in this study are available from corresponding authors on request.

Change history



12 March 2024

A Correction to this paper has been published: https://doi.org/10.1007/s42764-024-00127-7

About this journal

Electronic ISSN
2524-7662
Abstracted and indexed in
  1. CNKI
  2. Dimensions
  3. EBSCO Discovery Service
  4. Google Scholar
  5. Institute of Scientific and Technical Information of China
  6. Meta
  7. Naver
  8. OCLC WorldCat Discovery Service
  9. ProQuest-ExLibris Primo
  10. ProQuest-ExLibris Summon
  11. TD Net Discovery Service
Copyright information

Rights and permissions

Springer policies

© Shenzhen University School of Medicine; Fondazione Istituto FIRC di Oncologia Molecolare

×

用户登录