Interplay between m⁶A epitranscriptome and epigenome in cancer: current knowledge and therapeutic perspectives

1 INTRODUCTION

The Greek prefix “epi” means on top of and is used to convey the idea that epi-modifications guide gene expression without altering underlying DNA and RNA sequences. Currently, the epi-signature comprises more than 170 RNA modifications discovered in eukaryotes,¹ 46 in vivo-verified DNA marks² and 114 histone posttranslational modifications as reported in the HISTome2 database.³ These chemical tags can be reversibly added or removed at either pretranscriptional, co-transcriptional or posttranscriptional level. The presence of enzymes that deposit (writers), remove (erasers) or decode (readers) these modifications is a common feature in epi-dynamics⁴ (ie, the mechanisms involving epigenetic factors that impact on RNA synthesis, processing and degradation influencing gene expression). Intriguingly, despite the huge number of marks known, only for a small minority has the complete enzymatic machinery been fully described. Very few DNA/histone modifications have been shown to play a role in transcription nal processes. Major efforts have therefore long been focused on studying such modifications and epigenetic mechanisms that impact gene expression, such as long noncoding RNAs (lncRNAs), microRNAs (miRNAs) and circular RNAs (circRNAs; reviewed in Allis and Jenuwein⁵). Conversely, the existence of epitranscriptomic marks has been known since 1960, when pseudouridine was first described.⁶ However, investigation into these marks lagged for decades, until the discovery of the m⁶A demethylase fat mass and obesity-associated gene (FTO).⁷ This finding demonstrated the reversibility of m⁶A, increasing interest in the field and leading to the breakthrough development of methylated RNA immunoprecipitation sequencing (MeRIP-seq).^{8, 9} m⁶A sequencing approaches are currently able to detect m⁶A profiles of RNAs.¹⁰ Emerging evidence suggests that m⁶A regulates many steps of mRNA biology, including stability, degradation, processing and translation (reviewed in Zaccara et al¹¹). Interestingly, aberrant distribution of m⁶A is reported in several cancer types, and many m⁶A effectors are described as playing an oncogenic or oncosuppressor role depending on the tumor-specific context (reviewed in Gu et al¹²). Thus, the involvement of m⁶A in oncogenesis and the reversibility of the m⁶A mark make this modification attractive for therapeutic investigations. In addition, increasing studies connect m⁶A and epigenetic mechanisms.¹³ Notably, m⁶A methylation was found on transcripts coding for chromatin modifiers, possibly altering their translation and, eventually, the cellular epi-signature.¹⁴ Similarly, factors associated with enzymes involved in the deposition of the m⁶A mark support the activity of epigenetic modifying enzymes at chromatin level. Epigenetic marks can also drive the methylation of specific RNAs by recruiting the m⁶A machinery complex.¹⁵ Importantly, both a direct interaction between m⁶A-related effectors and histone modifications and the presence of the m⁶A mark on various classes of ncRNAs have been reported.¹³ Here, we provide an overview of the current knowledge of m⁶A biology and discuss recent findings illustrating the role of m⁶A in cancer. We also describe the crosstalk between m⁶A and epigenetic regulatory pathways, looking at how this interplay affects the cellular transcriptome and its involvement in oncogenic processes. Lastly, we highlight the crucial role of the m⁶A modification in simultaneously linking multiple and diverse gene regulatory networks and its potential use in novel therapeutic applications.

2 THE m⁶A MODIFICATION

2.1 Insights into m⁶A biology

The m⁶A modification is dynamically regulated by a known set of proteins that act in synergy to deposit (writers), remove (erasers) or decode (readers) the mark. The deposition of the m⁶A modification in eukaryotic RNAs mainly relies on the activity of methyltransferase-like (METTL) proteins, of which more than 30 different types are encoded by the human genome (HGNC database). However, not all METTL proteins have been characterized or have enzymatic activity. To date, METTL3 is the most studied. METTL3 forms a complex with METTL14 called m6A methyltransferase complex (MAC),¹⁶ where METTL3 is the catalytic component, and which co-transcriptionally methylates the vast majority of RNA substrates (Figure 1).¹⁷ Conversely, METTL16 preferentially catalyzes the addition of m⁶A to small nuclear RNAs (snRNAs) and a few mRNAs.¹⁸ In addition, METTL5, in association with its allosteric activator TRMT112, and ZCCHC4 were found to methylate ribosomal RNAs (rRNAs).^{19, 20} Despite the variety of catalytically active METTL proteins, the silencing or genetic deletion of METTL3 and/or METTL14 abolishes up to 99% of m⁶A peaks in mRNA,²¹ suggesting that m⁶A deposition almost completely depends on the activity of METTL3. However, knockout or knockdown (KD) of METTL3-associated proteins reveals various degrees of m⁶A reduction in different contexts,²² pointing to the role of these proteins in driving deposition of m⁶A.

3 THE m⁶A MODIFICATION IN CANCER

Several independent studies have investigated the role of m⁶A and its biological determinants as a widespread regulatory mechanism that controls abnormal gene expression via different pathological pathways, leading to cancer development. In this section, we describe the role of the m⁶A mark and its effectors in tumorigenesis.

3.1 Writers

MAC plays a dual role, either promoting or suppressing cancer progression depending on the cellular context, type and origin of the tumor. The catalytic component of MAC, METTL3, is overexpressed in cancer, including hepatocellular carcinoma (HCC),⁴⁵ prostate adenocarcinoma (PRAD),⁴⁶ gastric carcinoma (GC),⁴⁷ colorectal cancer (CRC),⁴⁸ pancreatic adenocarcinoma (PAAD),⁴⁹ ovarian cancer (OC),⁵⁰ breast cancer (BC),⁵¹ acute myeloid leukemia (AML),⁵² bladder cancer (BLC),⁵³ osteosarcoma (OS)⁵⁴ and others. The enzymatic activity of METTL3 is described as driving different hallmarks of cancer.⁵⁵ Mechanistically, METTL3-mediated methylation can be directed to oncogenic or oncosuppressor mRNAs, either stabilizing or destabilizing target transcripts and affecting expression of onco- and oncosuppressor genes. In most cancers, METTL3 acts as an oncogene (reviewed in Zeng et al⁵⁶). However, a tumor suppressor role for METTL3 is described in renal cell carcinoma (RCC)⁵⁷ and endometrial cancer (ENC).⁵⁸ Surprisingly, METTL14, the co-partner of METTL3, is reported to have an oncosuppressor function in cancers, including HCC, CRC and GC (reviewed in Liu et al⁵⁹). Therefore, the opposite role of METTL3 and METTL14 in the same cancer type suggests that these proteins may regulate oncogenesis not only via m⁶A, but possibly also in an m⁶A-independent manner. For example, has been recently shown that in senescent cells METTL3 and METTL14 individually localize at distinct genomic regions to regulate the expression of senescence-associated secretory phenotype (SASP) genes and drive tumor progression, independently of m⁶A.⁶⁰ However, the role of METTL3/METTL14 in supporting tumor progression besides m⁶A remains still unclear and may possibly be cancer context-dependent. Further investigations are needed to shed light on these new mechanisms. The METTL3's other co-partner, WTAP, also plays a major role as an oncogene, mediating tumor progression (Figure 2). Ancillary MAC subunits (VIRMA, ZC3H13, RBM15/B, HAKAI) are also involved in promoting or arresting cancer growth by acting as either tumor driver or tumor suppressor genes, depending on the context. VIRMA is reported with an oncogenic role in PRAD,⁶¹ HCC,⁶² BC,⁶³ GC⁶⁴ and lung cancer (LC),⁶⁵ whereas ZC3H13 exerts oncosuppressor functions in PAAD,⁶⁶ BC⁶⁷ and CRC.⁶⁸

4 INTERPLAY BETWEEN m⁶A FACTORS AND EPIGENETIC PROCESSES

The complex landscape of gene expression involves an ever-growing number of cellular players and mechanisms. Genetic mutation is the major cause of transcriptional alterations and plays a pivotal role in physio-pathological contexts. Epigenetic mechanisms are now widely accepted as a further layer controlling transcriptional regulation, and the importance of crosstalk between epigenetic and genomic pathways in driving disease has been extensively documented, especially in cancer. An additional level of gene regulation involving RNA modifications has recently become an exciting area of study, and the newly discovered relationship between genome, epigenome and epitranscriptome is proving to be crucial in depicting a comprehensive scenario of gene regulation mechanisms that drive disease. In the following subsections, we describe the latest advances in the interplay between epigenetic mechanisms and m⁶A effectors, focusing on its implication in cancer.

4.1 Interplay m⁶A writers and epigenetic processes

Emerging evidence suggests that the deposition of the m⁶A mark on RNAs and the activity of MAC influence numerous epigenetic mechanisms. For example, METTL3 and its associated proteins can directly recognize and bind chromatin remodeling factors and histone modifications, while the presence of m⁶A on transcripts coding for epigenetic modifiers can affect mRNA stability, (directly) influencing its expression and (indirectly) governing the deposition of histone marks that in turn shape the chromatin state. Intriguingly, disruption of m⁶A installation by METTL3 or METTL14 KD reduced methylation of CBP/p300 and EZH2 mRNAs, increasing levels of H3K27ac (marker of transcriptional activation) and decreasing levels of H3K27me3 (marker of transcriptional repression), respectively.^{14, 109} METTL3 silencing also reduced levels of H3K4me3, resulting in increased chromatin accessibility and production of chromatin-associated RNAs (carRNAs).¹¹⁰ Additionally, depletion of METTL3 significantly increased H3K9me2 levels, whereas its overexpression rescued the aberrant upregulation of H3K9me2.¹¹¹ This finding suggests that the broad catalytic activity of METTL3 regulates the deposition of numerous histone marks, possibly playing a further role in determining chromatin conformation. Interestingly, METTL14 was shown to directly bind the H3K36me3 modification at histone level.¹⁵ Such binding stimulates the recruitment of MAC/MACOM complex to open chromatin regions and the co-transcriptional addition of the m⁶A mark, demonstrating that histone modifications also influence m⁶A dynamics (Figure 3).¹⁵ The deposition of epigenetic marks on the promoter of m⁶A genes and the addition of m⁶A to ncRNAs are further examples of epigenetic/epitranscriptomic interplay (Figure 3). Specifically, insertion of the H3K27ac mark in the promoter of METTL3 enhanced its expression,¹¹² while removal of H3K4me3 from the promoter of METTL14 caused the opposite effect.¹¹³ These findings highlight the importance of histone marks in the expression of m⁶A factors. Notably, hypomethylation at the METTL3 promoter not only decreases METTL3 transcription but also, in turn, reduces m⁶A levels on pri-miR-25, impairing its maturation.¹¹⁴ Lastly, ncRNAs play a role in regulating m⁶A. For example, lncRNAs can exploit m⁶A machinery to acquire the m⁶A modification. Interpretation of the m⁶A mark on modified lncRNAs by reader proteins leads to an increase in their stability,¹¹⁵ while miR-186 reduces the expression of METTL3 and disrupts the deposition of m⁶A.¹¹⁶

5 INTERPLAY BETWEEN CANCER m⁶A EPITRANSCRIPTOME AND EPIGENOME

Onset of tumorigenesis is due to the accumulation over time of genetic mutations and epigenetic changes that interplay altering correct gene expression patterns. The RNA modification m⁶A has a role in tumorigenesis and is involved in a third level of gene regulation. Studying the intersection of genetic, epigenetic and epitranscriptomic pathways in tumor settings is therefore crucial, and a growing body of compelling evidence points to an unexpected complexity. Examples of crosstalk between all classes of m⁶A effectors and epigenetic regulation in cancer are reported.¹³ In CRC, downregulation of METTL14 via removal of the H3K4me3 mark from its promoter reduced the deposition of m⁶A on SOX4 mRNA, rescuing it from YTHDF2-mediated degradation.¹¹³ This resulted in enhanced migration and SOX4-mediated epithelial-to-mesenchymal transition of CRC cells, both in vitro and in vivo.¹¹³ Similarly, overexpression of METTL3, though hypomethylation of its promoter, increased m⁶A deposition on miR-25-3p and stimulated its maturation, eventually driving the proliferation of PAAD cells via the AKT-p70S6K pathway.¹¹⁴ Further, MeRIP-seq analysis of GC cells depleted for METTL14 showed decreased m⁶A levels on the circRNA circORC5.¹²⁸ Absence of the m⁶A mark on circORC5 increased its expression by directing its activity toward the oncosuppressor miR-30c-2-3p, leading to GC progression.¹²⁸ These results indicate that METTL3/14 influences epigenetic pathways and could be a promising axis to target in cancer. The IGFBP protein family of m⁶A readers has also been shown to dialog with epigenetic pathways in cancer settings. Specifically, stabilization of the lncRNA LINRIS by the m⁶A-reading activity of IGF2BP1 or the interaction of IGF2BP2 with an oncopeptide produced by the lncRNA LINC00266-1 supported the growth of CRC cells by increasing expression of MYC and cellular glycolysis.^{119, 120} In this context, the use of inhibitors against IGF2BP1/2 could be a novel strategy to target CRC. Downregulation of YTHDF2 by miR-145 also suppressed the proliferation of HCC cells, although the exact mechanism of action has not yet been elucidated.¹²² m⁶A erasers also have a broad effect in crosstalk with epigenetic regulation in cancer. In esophageal squamous cell carcinoma (ESCC), FTO KD increased levels of the m⁶A-modified lncRNA LINC00022, altering its stability and impacting the proliferation of ESCC cells.¹²³ Similarly, overexpression of ALKBH5 via hypomethylation of CpG islands at its promoter stimulated the removal of m⁶A from the lncRNA LINC00278 and decreased both its stability and the production of the anticancer Yin Yang 1 (YY1)-binding micropeptide (YY1BM), eventually promoting the expression of eukaryotic elongation factor 2 kinase (eEF2K) and the proliferation of ESCC cells.¹²⁴ In addition, ALKBH5-mediated m⁶A demethylation of the lncRNA PVT1 supported OS progression,¹²⁹ suggesting a role for chromatin modifiers in regulating the expression of m⁶A genes. Lastly, overexpression of ALKBH5, caused by an increase in the deposition of H3K4me3 at its promoter, mediated by WDR5 in hepatitis B virus-infected cells, promoted the development of HCC.¹³⁰ Mechanistically, the removal of m⁶A from the viral mRNA of HBx protein increases its stability and stimulates the recruitment of methylation complexes at the ALKBH5 locus, creating a positive feedback loop.¹³⁰

6 CONCLUSIONS AND THERAPEUTIC PERSPECTIVES

Despite the increasing number of studies describing the crosstalk between the epigenome and epitranscriptome, how all the actors interact to finely regulate gene expression and shape the overall cellular transcriptional output remains largely unknown. DNA methylation, ncRNAs, histone marks and RNA modifications including m⁶A are only a small part of the transcriptional landscape.¹³¹ The interplay between different levels of regulation involved in transcription makes the scenario even more complex. However, since the ever-growing burden of cancer worldwide urgently requires the identification of novel therapeutic targets and prognostic biomarkers, elucidating the mechanisms of crosstalk among different networks of gene expression regulation is fundamental for achieving new milestones in drug discovery. The reversible m⁶A mark that contributes to several cancer hallmarks may offer new therapeutic opportunities, and rapid advancements are being made in this direction. In 2014, meclofenamic acid (MA) was identified as FTO inhibitor,¹³² and the anticancer activity of MA and its derivative^{133, 134} was successfully tested in preclinical studies.^133-137 Since MA is already used in the clinic, a drug repurposing approach may represent a promising strategy in cancer. Several molecules targeting m⁶A effectors, mainly writers and erasers, have been developed^{133, 138-151} (Table 1). However, most of these m⁶A-targeting drugs have low target selectivity and poor pharmacokinetic properties which hinder clinical applications. To date, the METTL3 inhibitor STC-15,¹⁵² developed by Storm Therapeutics, is the only epitranscriptomic compound that has entered clinical trials (NCT05584111). Interestingly, natural compounds¹⁵³ and AI-based approaches have also been used to speed up the lengthy process of drug discovery, but with mixed results.⁸⁷ Since m⁶A writer and eraser proteins possess a broad range of substrates, it would also be worthwhile to develop molecules specifically targeting m⁶A readers. A better understanding of the three layers of gene regulation and the critical involvement of m⁶A in cancer could provide a platform to design new drugs co-targeting multiple pathways and to identify new druggable targets. For example, competitive inhibitors of IGFBPs may affect the interaction between m⁶A readers and their lncRNA targets, exerting a therapeutic effect in CRC.¹¹⁹ In addition, the combinatory use of drugs targeting METTL3 or ALKBH5 and epigenetic regulators of m⁶A players might revert the oncogenic effects of m⁶A in PAAD and ESCC.^{114, 124} In the future, combination therapies using drugs simultaneously targeting the epitranscriptome and epigenome might open up new therapeutic avenues, especially in patients experiencing poor efficacy of epigenetic drugs or drug resistance. In sum, the complex world of transcription regulation has been dramatically changed by the discovery of the interplay between genetic, epigenetic and epitranscriptomic mechanisms. Dissecting this interplay will be fundamental for therapeutic intervention.

AUTHOR CONTRIBUTIONS

Nunzio Del Gaudio and Lucia Altucci: Supervision and Writing. Nunzio Del Gaudio and Guglielmo Bove: Conceptualization and Writing. Sajid Amin and Mehrad Babaei: Contributed to conceptualization. Angela Nebbioso and Rosaria Benedetti: Review and Editing. All authors read and approved the final manuscript. The work reported in the paper has been performed by the authors, unless clearly specified in the text.

CONFLICT OF INTEREST

The authors declare that they have no competing interests.