Biological roles of enhancer RNA m6A modification and its implications in cancer

Abstract

Enhancers, as distal cis-regulatory elements in the genome, have a pivotal influence on orchestrating precise gene expression. Enhancer RNAs (eRNAs), transcribed from active enhancer regions, are increasingly recognized as key regulators of transcription. N6-methyladenosine (m6A), the most plentiful internal modification in eukaryotic mRNAs, has garnered significant research interest in recent years. With advancements in high-throughput sequencing technologies, it has been established that m6A modifications are also present on eRNAs. An accumulative body of evidence demonstrates that aberrant enhancers, eRNAs, and m6A modifications are intimately connected with carcinoma onset, progression, invasion, metastasis, treatment response, drug resistance, and prognosis. However, the underlying molecular mechanisms governing m6A modification of eRNAs in cancer remain elusive. Here, we review and synthesize current understanding of the regulatory roles of enhancers, eRNAs, and m6A modifications in cancer. Furthermore, we investigate the possible roles of eRNAs m6A modification in tumorigenesis based on existing literature, offering novel perspectives and directions for future research on epigenetic regulatory mechanisms in cancer cells.

Introduction

The mechanisms regulating gene expression in higher eukaryotes are extremely complex, with different genes being activated in a cell-type-specific manner, resulting in cell types with specialised morphology and function [1]. Enhancers are DNA sequences consisting of 200–1500 base pairs that function in regulating gene expression, influencing cell identity, and regulating cell-type-specific gene expression [2]. The human genome contains a significantly higher number of enhancers compared to genes [3], with one enhancer capable of regulating multiple genes, while a single gene can be influenced by numerous enhancers [4]. However, aberrant enhancer regulation can lead to dysregulation of gene transcription, ultimately contributing to tumorigenesis and progression. Enhancer remodelling, altered enhancer activity, and even transcription factor (TF) dysregulation, which indirectly alters enhancer activity, can be key factors in tumourigenesis.

In the activated state, the enhancer attracts TFs, recruits RNA polymerase II (RNA Pol II), and transcribes to produce eRNAs. eRNAs are widely present in human cells and possess multiple biological functions. In recent years, increasing evidence has revealed that eRNAs can regulate the expression of proximal or distal target genes through a variety of mechanisms [5]. Furthermore, eRNAs can contribute to cancer development by altering gene transcription, as well as protein-RNA interactions to regulate oncogenes, tumour suppressor genes, and cancer signalling pathways [6]. In terms of potential clinical applications, eRNAs are of great value for further research as cancer-related biomarkers and therapeutic targets.

m6A modification is a methylation modification that occurs on RNA, accounting for about 60% of methylation modifications [7]. m6A methylation is primarily regulated by three types of proteins: writing proteins, which catalyse the formation of m6A methylation; erasing proteins, which demethylate RNA; and reading proteins, which recognise m6A methylation sites [8]. The m6A modification plays a crucial role in diverse cellular processes and is closely associated with cancer development [9]. For example, dysregulated m6A modifications significantly influence glycolysis and tumour progression in hepatocellular carcinoma (HCC) [10]. Moreover, studies have demonstrated that targeted modulation of m6A can enhance anti-tumour therapy efficacy in gastric cancer (GC) [11].

With the advancement of new technologies, m6A modification on eRNAs has been gradually uncovered. Several studies have investigated how m6A modification affects eRNAs, including their stability and transcriptional activation, as well as the possible involvement of m6A-related eRNAs in cancer. In this review, we describe the characterisation of enhancers and eRNAs, their mechanisms of action in transcriptional activation, and their association with cancer. Furthermore, we extensively explore the pathophysiological utility of m6A modification on eRNAs and comprehensively analyse its implications in cancer, offering new perspectives on the epigenetic regulatory mechanisms within cancer cells.

Enhancer and enhancer RNAs

Enhancer

Enhancers, which are non-coding sequences in the genome, have the ability to activate or increase the transcription of linked genes in specific cell types. With the advancement of genomic research, our understanding of enhancers has become more comprehensive. Enhancers are typically characterised by the following features: (1) They can function independently of the direction, distance, and location of the target gene. Generally, enhancers are located in intergenic and intronic regions, with some found in exonic regions. The distance from the target gene can vary from a few hundred bases to several thousand bases. For instance, the distance between the mouse SHH gene and its enhancer ZRS sequence exceeds 1 million base pairs [12, 13]. (2) Enhancers contain specific DNA sequences that allow TF binding. These sequences primarily consist of dense clusters of transcription factor binding sites (TFBs) that bind to cell type-specific TFs, co-regulatory factors, structural proteins, chromatin modifiers, and some enzymes [1]. (3) They are typically tissue-specific or cell type-specific, as demonstrated by the first mammalian cell enhancer found in the intron region of the immunoglobulin (Ig) heavy chain gene [14]. (4) Enhancers are characterised by a lack of nucleosomes and heightened sensitivity to nucleases, such as DNase I [15]. (5) Enhancers are evolutionarily conserved in both sequence and function [16]. Based on these characteristics, approximately 400,000 potential enhancers have been identified in the mammalian genome [17].

Genes can be regulated by either a single enhancer or multiple enhancer elements, which may cluster together to form super-enhancers. Super-enhancers cover larger genomic regions and contain higher concentrations of TFs, mediator coactivators, and histone markers linked to transcriptional activity compared to regular enhancers [18]. In addition to super-enhancers, two specific types of enhancers exist: shadow enhancers and stretch enhancers. Shadow enhancers are essentially a set of enhancers that influence a shared target gene, resulting in overlapping expression patterns in both space and time [19]. Stretch enhancers, identified through computational analysis of type 2 diabetes data, are broad regions of enhancers larger than 3 kb. These enhancers also increase gene expression and function in a cell type-specific manner [20] (Table 1).

Table 1.

Comparison of super-enhancers, shadow enhancers, and stretch enhancers

Feature	Super-enhancers	Shadow enhancers	Stretch enhancers	References
Genomic organization	Large enhancer clusters formed by multiple neighboring ordinary enhancers	A set of enhancers on the genome that regulate the same gene	Large genomic regions characterized by enhancers	[20–23]
Typical size	8-20 kb	—	Larger than 3 kb	[20–22]
Regulatory function	Powerful activation of cellular identity genes	Gene expression stability	Directing cell-specific gene expression	[20–22, 24, 25]
Relevance to cancer	Driving aberrant activation of oncogenes to maintain tumor properties	Redundant regulation ensures high expression of oncogenes	Activation of distal oncogenes	[19, 20, 26–29]

The enhancer can be activated by specific histone modifications. During transcriptional activation, the enhancer recruits two key complexes: the Mll3/Mll4/COMPASS complex and the p300/CBP complex. The former has methyltransferase activity, which mediates the monomethylation of histone H3 lysine 4 (H3K4me1), while the latter has acetyltransferase activity and mediates acetylation of histone H3 lysine 27 (H3K27ac). Upon labelling with H3K4me1 or H3K27ac, the enhancer becomes active. Therefore, H3K4me1 and H3K27ac can serve as markers for enhancers in chromatin immunoprecipitation sequencing analyses and can be used for the identification of enhancers in the genome [30, 31]. Enhancers exist in multiple states beyond the activated state, including primed, poised, and inactive states. Activated enhancers feature histone modifications and open chromatin, whereas enhancers in the primed or poised states lack H3K27ac modifications, and silent enhancers exhibit few epigenetic signatures [32]. As cis-acting elements, enhancers can also be modified by methylation, such as the addition of a methyl group at position 5 of the cytosine ring (5-methylcytosine, 5mC). Studies have shown that active enhancers have lower levels of 5mC compared to silent enhancers, though the precise role of 5mC in regulating enhancer activity remains unclear [33].

Before an enhancer can activate a promoter for transcription, it must first establish a link with the promoter to facilitate the exchange of information or material. Several models have been proposed to explain how enhancers communicate with promoters, such as the sliding model, the linking model, the contact model, the action-at-a-distance model, and the kissing model. However, the exact mechanism of enhancer-promoter interaction is still not fully understood [13, 34–37]. Regardless of the interaction mechanism, once enhancers and promoters interact, transcription is activated. Mechanistic models of enhancer-activated transcription can be categorised into four types: (1) enhancer-recruited protein complexes are transferred directly from the enhancer to the promoter; (2) enhancer-mediated protein complexes are recruited to the promoter; (3) enhancers accumulate a high concentration of TFs, coactivators, and chromatin remodelling agents near the promoter, increasing the likelihood of transcriptional activation at the promoter; (4) enhancer-mediated post-translational modification of the promoter [38]. The relationship between enhancers and promoters, and the mechanics of enhancer-activated transcription, remains a challenging area of research that requires further investigation for a deeper understanding.

Enhancer RNAs

In 2010, two research teams independently observed differences in the transcription of enhancers in neurons and macrophages, leading to the discovery of eRNAs. This discovery expanded the non-coding RNA (ncRNA) family and prompted the scientific community to explore eRNAs and their role in transcriptional regulation. To date, it is estimated that humans possess between 40,000 and 65,000 eRNAs [39]. Most eRNAs contain a 5’cap structure but lack a 3’polyadenylation (polyA) modification, typically being no longer than 2 kb, and are bi-directionally transcribed from active enhancers. A small proportion of eRNAs are transcribed unidirectionally from active enhancers; these are relatively long, have both 3’polyA modifications and 5’cap structures, and are generally more stable [40–42]. However, not all enhancers are capable of producing eRNAs. Enhancers that generate eRNAs possess the following characteristics: (1) histone modifications (such as H3K4me1 and H3K27ac) or the presence of certain histone variants (e.g., H2AZ); (2) they are located in open chromatin regions; (3) they are bound by transcriptional coactivators; (4) they exhibit low levels of DNA methylation [43]. Additionally, super-enhancers, which were mentioned earlier, can also be transcribed, and transcription from these super-enhancers can generate larger amounts of eRNAs [44].

eRNAs, along with mRNAs and long non-coding RNAs (lncRNAs), belong to the ncRNA category, sharing similarities and notable differences. mRNAs and lncRNAs are longer, more stable, and typically undergo splicing and 3′ polyA modification, unlike most eRNAs. However, some unidirectional eRNAs transcripts do undergo splicing and 3′ polyA modifications. Studies have shown that mRNAs and lncRNAs are 90–100 times more stable than eRNAs [45]. eRNAs are rapidly degraded by nuclear RNA exosome complexes after synthesis, and this short lifespan complicates their sequencing. To detect eRNAs, techniques such as global run-on sequencing (GRO-seq) and its derivatives, such as precise nuclear run-on sequencing (PRO-seq), have been employed. Researchers have refined methods to detect eRNAs based on their 5’-cap structure, including PRO-cap and 5’GRO-seq. Additionally, nanopore direct RNA sequencing has been applied in the analysis of cellular mRNAs, non-coding RNAs (ncRNAs), and the detection of various RNA viruses [46], and single-molecule imaging has made breakthroughs in revealing the translation and decay of mRNAs [47], potentially applicable to eRNAs analysis through further technical advancements.

eRNAs are generated in enhancer regions marked by high levels of H3K4me1 and low levels of H3K4me3 histone modifications, in contrast to lncRNAs, which are produced in promoter regions with high H3K4me3 modifications. The active enhancer regions producing eRNAs are associated with high levels of serine-5-phosphorylated RNA Pol II, whereas mRNA-producing promoter regions are linked to serine-2-phosphorylated RNA Pol II [48]. Interestingly, it was found that eRNAs transcribed from super-enhancers can function similarly to lncRNAs. Thus, eRNAs and lncRNAs exhibit functional similarities, leading to the hypothesis that lncRNAs may have evolved from eRNAs, stabilised and acquired functionality over evolutionary time. However, this hypothesis requires further investigation to verify. Additionally, eRNAs expression is positively correlated with the active enhancer marker H3K27ac, which is commonly used to mark active enhancers. Notably, eRNAs themselves serve as a more precise indicator of active enhancers, aiding the identification of new enhancers [49].

Typically, eRNAs production occurs prior to the transcription of the corresponding gene and involves three key steps. In the first step, TFs and coactivators are recruited. Upon specific stimuli, TFs and coactivators are recruited to interact with the enhancer region, promoting nucleosome remodelling and recruiting additional TFs, cofactors, and complexes (e.g., P300/CBP) [50]. In the second step, histone modification occurs at the enhancer site, with H3K27 acetylation mediated by P300/CBP, which promotes chromatin accessibility, further opening the enhancer region and recruiting RNA Pol II and bromodomain-containing protein 4 (BRD4). In the third step, eRNAs are transcribed, elongated, and processed. BRD4 regulates the recruitment of transcription initiation factors and the formation of transcription mediator complexes, promoting the elongation of RNA Pol II at highly acetylated enhancers [51]. Estrogen receptor-α (ERα), a central TF, associates with estrogen-responsive enhancers in MCF-7 breast cancer cells and selectively recruits a prominent complex of other DNA-binding TFs, ultimately forming a large complex known as MegaTrans, which is essential for eRNAs transcription [52]. However, it remains unclear whether this complexity exists in other cell types. Furthermore, it has been found that eRNAs transcriptional pausing and elongation are mediated by Spt5 and the positive transcription elongation factor b (P-TEFb) complex [53]. Additionally, Integrator, a large 12-subunit complex with nucleic acid endonuclease activity, mediates eRNAs transcription termination by interacting with the Pol II C-terminal domain (CTD) and cleaving the 3′ end of the primary eRNAs transcript [54]. Once released from chromatin, nascent eRNAs are degraded by nuclear RNA exosome complexes in a 3′-5′ manner. The enzymatic activity of these complexes is supported by two proteins, Rrp44 (Dis3) and Rrp6 (Exosc10). During eRNAs production, harmful DNA/RNA hybrids, such as R-loops, may form, where eRNAs are enzymatically modified to prevent the formation of base pairs, significantly impacting genomic stability. In knockout cells for the Dis3 and Exosc10 genes, R-loop structures in the transcribed regions of eRNAs were significantly increased, indicating that nuclear RNA exosome complexes may play a role in degrading detrimental secondary structures during eRNAs formation [55]. Methylation modifications, including m6A and m5C (5-methylcytosine), have also been detected on eRNAs, influencing their stability and metabolism. At the initial stages of eRNAs transcription, the cap-binding complex (CBC) associates with eRNAs to form the m7G (N7-methylguanosine) cap at the 5′ end (Fig. 1).

Fig. 1.

A diagram illustrates eRNAs transcription process, including modifications and degradation. Following the preliminary attachment of general transcription factors (GTFs) and other TFs in response to specific signals, RNA Pol II is brought to the enhancer to start bidirectional transcription. The enhancer region is rich in H3K4me1 and H3K27ac modifications. The C-terminal domain (CTD) of Pol II in the enhancer region is primarily phosphorylated at the Tyr1 and Ser5 sites, which regulate the pausing and elongation of eRNAs transcription. Integrators can mediate the transcription termination of eRNAs. WDR82 (WD repeat domain 82) and PP1 (phosphatase 1) are engaged in the early ending of eRNAs transcription. Once eRNAs are released from chromatin, they are degraded by nuclear RNA exosome complexes in a 3’-5’ manner. Various RNA modifications, maintaining m5C, m6A, and m7G, are frequently present in eRNAs and affect their stability

The function of eRNAs was initially controversial. Some researchers argued that, due to their low abundance and short half-life, eRNAs were merely “noise” in the chromatin environment opened by enhancers and by-products of gene transcription. However, more studies have refuted this claim. eRNAs are not only markers of active enhancers, but also crucial regulators of transcription. eRNAs play the following roles in transcriptional regulation: (1) regulation of chromatin remodelling. eRNAs recruit Cohesin and bind to its components, SMC3 and RAD21, to regulate chromatin accessibility at the target promoter. Conversely, depletion of eRNAs decreases chromatin accessibility at enhancers and promoters [56]. eRNAs also stabilise enhancer-promoter loops by directly interacting with chromatin loop factors, such as MED12 [57]. (2) Regulation of histone modifications. eRNAs interact with CBP, stimulating its catalytic activity and increasing histone acetylation at enhancers [58]. eRNAs can also activate p300 and induce H3K27 acetylation at enhancers [59]. Similarly, eRNAs interacting with the EZH2 subunit antagonise its nucleosome-binding activity and inhibit the methyltransferase activity of PRC2, thus reducing the deposition of inhibitory H3K27me3 [60]. (3) Recruitment of TFs and co-regulators, thereby modulating their activity. eRNAs can capture TF YY1 and increase its local concentration at transcription sites [61]. eRNAs also interact directly with BRD4 to amplify the interaction between H3K27ac and BRD4, thus maintaining enhancer activity [62]. Recent studies have found that eRNAs capture dissociated TFs through weak RNA-mediated interactions, promoting their recombination with regulatory elements [63]. (4) Direct participation in or interference with transcription, enhancing transcription elongation. eRNAs can interact with the positive elongation factor (P-TEFb) to enhance transcription. In neurons, eRNAs act as a decoy for the negative elongation factor (NELF) after inducing the immediate early gene (IEG), luring it away from the IEG and converting the paused RNA Pol II into productive elongation [64]. Furthermore, eRNAs regulate the proper loading of target sites by CAAA bundles, which control the intermediate heterogeneous cytosolic nuclear ribonucleoprotein L (hnRNPL) [65].

Enhancer and enhancer RNAs in cancer

Enhancer in cancer

Enhancers function as cis-regulatory elements that elevate gene expression by cooperating with promoters to orchestrate transcriptional programmes. With growing research into enhancer biology, it is now well-established that enhancer dysregulation—manifesting as loss of native enhancers, emergence of de novo enhancers, or aberrant enhancer activation/inhibition—plays a pivotal role in tumorigenesis. The disruption of enhancer function in cancer can be broadly attributed to three primary categories: genetic alterations, epigenetic reprogramming, and TF dysregulation.

Firstly, genetic alterations can directly influence enhancer activity. These include copy number variations, chromosomal translocations, amplifications, insertions, deletions, and both intra- and inter-chromosomal inversions. Single nucleotide polymorphisms (SNPs) and structural variants in enhancer regions are now recognised as key contributors to cancer risk. Genome-wide association studies (GWAS) have revealed that a significant proportion of cancer-associated variants—including those linked to lung and breast cancers—are enriched within enhancer regions [66]. For instance, a point mutation within the enhancer region of the *TERT* gene enhances TF binding, leading to elevated telomerase activity and uncontrolled proliferation of tumour cells [67].

In a genome-wide CRISPR interference (CRISPRi) screen, researchers identified 66 enhancers within the melanoma genome associated with highly recurrent regions (HRRs), which appear to function as tumour-suppressive elements. These were stratified into four categories using a latent block model: (1) enhancers frequently deleted in patients, (2) enhancers containing recurrent mutations at functional loci, (3) enhancers with sparse mutations, and (4) enhancers with high-level duplications in proximity to oncogenes. Among these, 21 enhancers exhibited copy number gains and 17 showed heterozygous deletions. Functionally, these disrupted enhancers were implicated in the EGFR signalling pathway, central carbon metabolism, and the AMPK signalling cascade, collectively contributing to tumour progression [68]. Similarly, chromatin structure can be altered by enhancer-associated structural variants; in osteosarcoma, abnormal chromatin looping between the *MYC* enhancer and promoter augments oncogenic transcription by increasing spatial proximity [69]. Hence, enhancer amplification, hijacking, or mutation is critical to cancer initiation and progression.

Secondly, alterations in epigenetic modifications can also affect enhancer function, such as aberrant DNA methylation and modifications like histone methylation and acetylation. Aberrant hypermethylation of enhancers associated with tumour suppressor genes, such as CDKN2A, can result in gene silencing, whereas hypomethylation of enhancers linked to oncogenes, such as ERα in breast cancer (BC), facilitates the binding of TFs [70, 71]. TP63 has been associated with poor prognosis in prostate cancer (PCa) due to its elevated expression. The methylation level of enhancers in the epithelium of prostate basal cells decreases when TP63 is expressed, but it may be responsible for the enhanced hydroxymethylation of cytosine (hmC) at certain enhancers by TP63. However, not all CpG sites related to TP63 undergo demethylation when TP63 is expressed [72]. In castration-resistant prostate cancer (CRPC), inhibition of lysine-specific demethylase 1 (LSD1/KDM1A) blocks the demethylation of the AR pioneer factor FOXA1, which then recruits BRD4 to chromatin, disrupting FOXA1’s global binding and reducing enhancer accessibility. Genetic and epigenetic alterations occur simultaneously in most cancer cells [73].

In addition, abnormalities in TFs, such as mutations, aberrant expression, chimerism, altered function, altered stability, or even crosstalk between TFs, can also cause changes in enhancer activity and thus affect cancer. The overexpression of FOXA1 in BC remodels ER signalling pathway enhancers by recruiting the MLL3 complex, promoting cell proliferation [74]. Consistent with this, ΔNp63 recruits p300 to restructure the enhancer landscape in lung adenocarcinoma (LUAD), driving squamous differentiation phenotypes [75]. Knockdown of forkhead box protein D2 (FOXD2) in colonoid organs downregulates the expression of more than 200 genes, leading to the loss of colorectal identity and causing colorectal carcinogenesis. Mechanistically, FOXD2 opens chromatin and recruits K3K4 monomethyltransferase (MLL4) to introduce histone modifications, and enhancers and promoters further form differential chromatin loops to regulate chromatin structure, re-edit enhancers, and induce apoptosis [76]. In acute myeloid leukaemia, an in-frame mutation in the C-terminal region of the enhancer binding protein alpha (C/EBPα) alters its basic leucine zipper structural domain, producing a C/EBPα C-terminal mutant (C/EBPα-Cm). The resulting mutant has a reduced capability to bind with DNA, causing decreased enhancer activity of the ULBP2/5/6 genes and reduced gene expression [77]. This ultimately allows tumour cells to evade natural killer cell-mediated cell lysis.

Abnormal enhancer activity is intricately linked to tumour drug resistance, sensitivity to targeted therapy, and response to chemotherapy. Aberrant enhancer activation can stabilise the continuous expression of oncogenes and maintain survival signalling pathways in tumour cells, thereby reducing the cytotoxic effects of conventional chemotherapy or targeted therapy. For instance, the nuclear matrix binding protein SATB1 mediates the formation of an enhancer-promoter loop of the BCL2 gene, upregulating the expression of the anti-apoptotic protein BCL2 and leading to tumour cell resistance to apoptosis-inducing drugs such as cisplatin [78]. Therefore, individual classification based on enhancer genetic variations can provide new biomarkers for predicting treatment responses and guide drug selection. Thus, the enhancer regulatory network plays a “double-edged sword” role in cancer treatment responses: its abnormal activation can drive drug resistance, while targeted intervention may restore treatment sensitivity.

Enhancer-targeted therapeutic strategies can be categorised as follows: (1) development of small-molecule inhibitors targeting enhancers: the BET inhibitor JQ1 suppresses cancer stemness by disrupting enhancer architecture and downregulating BRD4-dependent stemness-related genes. Clinically, combining JQ1 with cisplatin has shown promise in enhancing chemosensitivity and improving patient survival outcomes [79]. Beyond JQ1, other enhancer-targeting agents, including the CDK7 inhibitor THZ1 and the CDK8/CDK19 dual inhibitor cortistatin A, have demonstrated efficacy in inhibiting tumour progression [80, 81]. Furthermore, clinical trials of BET inhibitors in haematologic malignancies have validated the feasibility of enhancer-targeted therapies. (2) Antisense Oligonucleotides (ASOs): ASOs are chemically synthesised nucleic acid analogues [82]. Modified ASOs can precisely downregulate enhancer-derived noncoding RNAs by disrupting enhancer-promoter looping [83]. Therapeutic ASOs exhibit high specificity and the ability to modulate enhancers that are traditionally considered “undruggable”. However, efficient delivery of ASOs to target tissues remains a major challenge for clinical translation [84]. (3) CRISPR/Cas9 genome editing systems: the CRISPR/Cas9 deletion system enables functional studies of enhancers by directly removing specific enhancer regions [85]. Advanced epigenome-editing tools, such as enCRISPRa (enhancer-targeted CRISPR activation) and enCRISPRi(enhancer-targeted CRISPR interference), can remodel epigenetic modifications at enhancers to either activate or suppress their transcriptional activity and downstream target genes [86]. Collectively, small-molecule compounds and ASOs targeting enhancer elements offer potential to restore enhancer homeostasis, modulate eRNAs levels, and thus interfere with tumour progression.

Tumour-specific hyperactive enhancers are frequently enriched at oncogenic loci (e.g., MYC, BCL2), and distinct enhancer signatures across tumour subtypes provide a foundation for precision targeting. However, several practical challenges must be addressed: (1) the vast complexity and functional redundancy of enhancer networks, coupled with high interindividual heterogeneity, complicate therapeutic targeting; (2) the blood-brain barrier penetration capacity of enhancer-targeted inhibitors requires optimisation for central nervous system malignancies.

In terms of potential risks, enhancer-targeted therapies carry inherent risks due to the involvement of enhancers in normal physiological processes. For instance, small-molecule inhibitors may exhibit off-target effects by simultaneously modulating multiple enhancers. BET inhibitors, while suppressing oncogenic enhancers, may inadvertently activate immune-related pathways, triggering excessive inflammatory responses [87]. Additionally, genes regulated by multiple enhancers (e.g., redundant enhancers) may resist single-enhancer inhibition, whereas over-editing could disrupt cellular homeostasis. Enhancer-targeted interventions might also perturb chromatin architecture, affecting non-target genes. For example, targeting the MYC enhancer could dysregulate adjacent noncoding RNAs like PVT1, leading to genomic instability [88]. Despite their transformative potential, the risks of enhancer-targeted strategies stem from the complexity of enhancer regulatory networks and insufficient technological specificity. Future efforts should integrate multi-omics data and machine learning models to predict critical enhancers, coupled with AI-driven design of next-generation inhibitors. Balancing innovation with risk mitigation will be essential to advance precision medicine sustainably.

Enhancer RNA in cancer

eRNAs play an indispensable role in enhancer activation as well as transcriptional regulation. Genetic alterations and epigenetic modifications in the enhancer region can significantly influence eRNAs expression. Growing evidence indicates that abnormal eRNAs expression can drive tumour progression.

eRNAs are crucial in the regulation of gene expression through various mechanisms. In cancer, they can either enhance the transcription of oncogenes or suppress the activity of tumour suppressor genes. For example, insulin receptor substrate 2 (IRS2), an oncogene located on chromosome 13, is essential for driving cell proliferation, clonogenic survival, and anti-apoptotic activity in oral squamous cell carcinoma (OSCC). The IRS2-linked enhancer RNA (IRS2e), located adjacent to the IRS2 locus, exhibits elevated expression in OSCC and is associated with unfavourable clinical outcomes [89]. Additionally, research has shown that CCAT1-L, an enhancer RNA embedded in a super-enhancer near the MYC locus, is expressed from a genomic region 515 kb upstream of MYC in human colorectal tumours [90]. Functionally, this long non-coding RNA (lncRNA) facilitates the formation of long-range chromatin loops, thereby modulating MYC transcription. Interestingly, the architecture of these loops undergoes remodelling during cancer progression, leading to dysregulated gene expression. In renal cancer cells, the level of eRNA p53BER2 is lower compared to normal tissues. Silencing p53BER2 mitigates the cytotoxic effects of nutlin-3 in TP53-WT cell lines, as well as the G1-phase blockade and senescence in these cells [91]. Furthermore, additional experiments have indicated that p53BER2 can induce cell-cycle arrest and DNA repair by interacting with BRCA2. Overall, eRNAs regulate oncogenes and tumour suppressors through various transcriptional, epigenetic, and structural mechanisms. Comprehensive studies combining eRNA profiling with functional assays are necessary to better understand their roles in cancer development and treatment.

In addition to regulating tumour-related genes, eRNAs are involved in aberrant signalling cascade response pathways in tumours. CCAT1 is a cancer-associated eRNA targeted downstream of TP63 and SOX2 in squamous cell carcinoma (SCC). TP63 and SOX2 co-occupy the promoter and super-enhancer of CCAT1 to activate its transcription in SCC cells, and CCAT1 subsequently forms complexes with TP63 and SOX2. Silencing CCAT1 in SCC significantly impairs cancer cell viability, clonogenic ability, and reduces tumour volume. Further studies found that CCAT1 interacts with SOX2 and TP63 to mediate EGFR expression by binding with its enhancer, consequently activating the MEK/ERK1/2 and PI3K/AKT signalling pathways [92].

Because eRNAs are associated with immune escape and treatment resistance in cancer cells, they can be clinically used as markers of tumour treatment responsiveness and resistance. The eRNA LINC02257 is notably linked to poor prognosis in colorectal adenocarcinoma patients. Abnormal LINC02257 expression alters the tumour immune microenvironment across different tumour categories [93]. Additionally, the relocation of proto-oncogenes to transcriptionally active regions is a common chromosomal rearrangement in cancer. For example, the transfer of the proto-oncogene MYC to the immunoglobulin (IGH) heavy chain motif occurs in 85% of Burkitt lymphoma cases, leading to increased MYC proto-oncogene expression. The IGH/MYC translocation results in overexpression of the eRNA AL928768.3. Overexpression of AL928768.3 increases lymphoma cell growth and resistance to chemotherapy, whereas no such resistance is observed in B-cell lines without this translocation. Targeting AL928768.3 to reduce resistance in Burkitt’s lymphoma may become a common therapeutic approach in the clinic [94].

Furthermore, eRNAs can serve as therapeutic targets and potential biomarkers for diagnosis and prognosis. The proto-oncogene MYC is a key driver in leukemogenesis and AML (Acute myeloid leukemia) progression, while the orphan nuclear receptor NR4A1, a tumour suppressor frequently silenced in AML patients, directly binds to MYC super-enhancers, eliminating the necessary coactivators to disrupt AML-selective MYC super-enhancer activity, thus inhibiting tumour cell growth both in vivo and in vitro [95]. To better predict the prognosis of cervical adenocarcinoma (CESC) patients, researchers identified 7936 eRNAs that were distinctly expressed in CESC patients through bioinformatic analysis and further identified a prognostic model composed of eight eRNAs. The 8-eRNA model demonstrates higher accuracy and stability and is an independent prognostic indicator compared to clinicopathological classification alone [96]. The construction of the 8-eRNA model provides important insights for exploring novel therapeutic targets for cervical cancer patients. As of now, many eRNAs in cancer cells remain untapped, and we anticipate further research to uncover the relationship between eRNAs and cancer.

In conclusion, eRNAs emerge as both dynamic biomarkers and functional executors of enhancer activity, playing pivotal roles in oncogenesis, metastatic progression, and therapeutic resistance. Mechanistically, eRNAs directly regulate cell cycle-associated genes to drive tumour cell proliferation and survival. Notably, Li et al. demonstrated that among 626 cancer-associated eRNAs, a significant subset (including CCND1 and MYC targets) exhibit potent pro-proliferative effects, with experimental eRNAs suppression leading to marked downregulation of these oncogenes and subsequent suppression of PCa growth [97]. Importantly, eRNA quantitative trait loci (eRNA-QTLs) show spatial colocalisation with cancer risk variants, establishing their regulatory significance in tumourigenesis. Beyond primary tumour growth, eRNAs orchestrate metastatic programmes through microenvironmental remodelling. This is exemplified in BC models where eRNA-mediated activation of epithelial-mesenchymal transition (EMT) genes promotes invasive phenotypes [98]. Furthermore, eRNAs drive therapeutic resistance via multimodal regulation of drug metabolism enzymes (e.g., cytochrome P450 family) and survival signalling cascades. Illustrating this paradigm, Shi et al. have characterised the dual functionality of super-enhancer-derived LINC02454 in modulating temozolomide (TMZ) sensitivity through epigenetic and post-transcriptional mechanisms in glioma [99]. These collective findings position eRNA-targeted therapies as a transformative frontier in oncology. Strategic modulation of eRNAs networks may overcome current limitations in precision oncology approaches, particularly for treatment-refractory malignancies.

RNA m6A modification

RNA m6A modification

Post-transcriptional modifications, a crucial domain of epigenetics, have come out as key regulators of numerous physiological and pathological processes. According to MODOMICS, over 170 distinct chemical modifications of RNA had been established in living organisms by the end of 2023 [100]. Among these, N6-methyladenosine (m6A) stands out as the most prevailing, abundant, and evolutionarily conserved internal co-transcriptional modification in eukaryotic RNAs. As a dynamically reversible process, RNA m6A modification is catalyzed by writer proteins, removed by erasers, and recognized by reader proteins, predominantly accumulating near termination codons, the 3’ untranslated region (3’UTR), and internal exons. At the molecular level, m6A influences almost every dimension of RNA metabolism, including mRNA translation, degradation, splicing, folding, nuclear export, stabilization, and decay. Moreover, m6A modifications play integral roles in diverse biological processes, such as transcriptional regulation, intracellular signaling, and the DNA damage response [101–103]. In the last several years, accumulating researches have demonstrated that RNA m6A modification further modulates tumor development by regulating tumor cell metabolism, underscoring its significance in cancer biology.

Regulators of m6A (writer, eraser, reader)

Writer

The most extensively studied m6A-modifying complex is the m6A methyltransferase complex, which comprises methyltransferase-like protein 3 (METTL3), methyltransferase-like protein 14 (METTL14), and the cofactor Wilms Tumor 1 Associated Protein (WTAP). Each of these components plays a crucial role in the enzymatic process. In recent years, additional accessory proteins participating in m6A modification have been discovered, maintaining methyltransferase-like protein 16 (METTL16), methyltransferase-like protein 5 (METTL5), human methyltransferase (ZCCHC4), RNA-binding motif protein 15 (RBM15), and Vir-like m6A methyltransferase-associated protein (VIRMA/KIAA1429). These proteins contribute to m6A deposition on various structured RNAs, such as U6 snRNA and 28 S rRNA, and, under certain conditions, within mRNA introns [104–108].

Eraser

m6A erasers function as demethylases, removing m6A modifications via oxidative reactions that require ferrous iron (Fe²⁺) as a cofactor and α-ketoglutarate as a co-substrate. The finding of the first m6A eraser, fat mass and obesity-associated protein (FTO), by Professor He Chuan in 2011 provided compelling evidence that m6A modifications are dynamically reversible, leading to a surge in m6A research. The second major m6A eraser, alkylation repair homolog 5 (ALKBH5), was discovered in 2013. Like FTO, ALKBH5 belongs to the α-ketoglutarate-dependent dioxygenase family. Aside from FTO and ALKBH5, other erasers such as ALKBH1 and ALKBH3 have been identified. ALKBH1 is involved in tRNA demethylation [109], whereas ALKBH3 functions as an m1A demethylase (N1-methyladenosine), acting on both m1A residues in tRNA and m6A modifications in RNA [110, 111].

Reader

m6A-modified readers encompass a varied range of proteins, containing the YTH domain-containing family (YTHDF1, YTHDF2, YTHDF3, YTHDC1, and YTHDC2), the insulin-like growth factor 2 mRNA-binding protein family (IGF2BP1, IGF2BP2, IGF2BP3), the HNRNP family (HNRNPA2B1, HNRNPC, HNRNPG), eukaryotic translation initiation factors (eIF3, eIF4E, eIF4G), and embryonic lethal abnormal visual-like protein 1 (ELAVL1). Recently, a novel m6A reader, proline-rich coiled-coil 2 A (Prrc2a), has also been identified. These readers regulate downstream RNA metabolism either by indirectly binding RNA through m6A-induced structural changes or by directly recognizing and interacting with m6A-modified sites. Their functions influence various cellular processes, maintaining RNA splicing, nuclear export, stability, and translation, thereby playing critical roles in disease development (Table 2).

Table 2.

Functions of m6A regulators in RNA metabolism

Type	m6A regulators	Function	References
Writers	METTL3	Catalyzes m6A modification	[112]
	METTL14	A core subunit of the m6A methyltransferase forms a heterodimer with METTL3 to facilitate m6A modification	[112]
	WTAP	The regulatory subunit of the m6A methyltransferase and recruits METTL3 and METTL14 into the nuclear speckles	[113]
	METTL16	Accountable for m6A modification of U6 snRNA, snRNA and other lncRNAs	[104]
	METTL5	Accountable for m6A modification of 18 S rRNA	[105]
	ZCCHC4	Accountable for m6A modification of 28 S rRNA	[106]
	RBM15	Binds the m6A complex and recruits it to a special RNA site	[114]
	VIRMA/KIAA1429	Directs the methyltransferase components to specific RNA region	[115]
Erasers	FTO	Demethylates m6A, also has activity towards m6Am and m1A	[116, 117]
	ALKBH5	Principally demethylates m6A	[118]
	ALKBH1	Acting as a tRNA demethylase by removing N(1)-methyladenine, which regulates translation initiation and elongation	[109]
	ALKBH3	tRNAs demethylase	[111]
Readers	YTHDF1	Enhances mRNA translation	[119]
	YTHDF2	Enhances mRNA degradation	[119]
	YTHDF3	Prompts translation and degradation by interacting with YTHDF1 and YTHDF2	[120]
	YTHDC1	Leads to RNA splicing and export	[121]
	YTHDC2	Improves the translation efficiency of target mRNA	[121]
	IGF2BP1/2/3	Enhances the stability and translation of mRNA	[122]
	HNRNPA2B1	Spurs primary microRNA processing	[123]
	eIF3	Promotes mRNA translation	[124]
	ELAVL1	RNA-binding protein	[125]
	Prrc2a	Recognizes m6A modification and associates with myelin formation and meiosis	[126, 127]

Cross talk between m6A modification and other posttranscriptional modifications

Post-transcriptional modification represents an essential regulatory mechanism in gene expression, involving a complex interplay of various processes. The interaction between m6A modification and other RNA modifications in both physiological and pathological contexts has been increasingly recognized; however, research in this field remains in its early stages. As early as 2018, genome-wide analyses revealed that m6A modification negatively regulates adenosine-to-inosine (A-to-I) RNA editing. Adenosine deaminases acting on RNA (ADARs) catalyze the conversion of adenosine to inosine in double-stranded RNA substrates. In mouse 3T3 cells, knockdown of METTL3 contributed to an enhanced A-to-I editing, whereas depletion of FTO resulted in its downregulation. Despite these observations, the precise mechanism by which m6A influences A-to-I editing within the same transcript remains unclear. It is hypothesized that m6A-modified RNAs may exhibit reduced binding affinity for ADAR, or alternatively, that the presence of m6A-bound reader proteins may sterically hinder subsequent A-to-I editing [128]. N-acetyltransferase 10 (NAT10) is the only enzyme identified to catalyze N4-acetylcytosine (ac4C) modification [129]. In osteosarcoma, NAT10 knockdown impairs the translation of the m6A reader YTHDC1. Through disrupting this pathway, NAT10 depletion reduces osteosarcoma cell growth, migration, and invasion while increasing global m6A levels [130]. Beyond acetylation and A-to-I editing, m6A also exhibits crosstalk with various methylation modifications. Under normal physiological conditions, decapping enzyme 2 (DCP2) regulates mRNA degradation when the 5’ cap is modified by m7G [131]. If the first nucleotide following the m7G cap is 2’-O-dimethyladenosine (Am), this Am residue can be further methylated by phosphorylated CTD-interacting factor 1 (PCIF1), forming the m6Am modification. FTO is capable of removing this m6Am mark, thereby maintaining its dynamic reversibility. Notably, the presence of an m7G cap adjacent to m6Am protects the transcript from DCP2-mediated decapping, underscoring the functional coordination between these modifications [132]. m6Am modification has also been detected within the interior regions of U2 small nuclear RNA (snRNA) alongside m7G-modified caps. The cooperative action of m6Am and m7G modifications has an essential impact on regulating RNA splicing [133]. In bladder cancer (BCa), trophoblast cell surface antigen 2 (TROP2) is highly upregulated. METTL3-mediated m6A modification at the 3’UTR of TROP2 mRNA enhances its translation, while METTL1-mediated m7G modification in tRNA improves translation efficiency. These modifications act synergistically to increase TROP2 expression, thereby promoting BCa progression [134]. Interestingly, IGF2BP family proteins recognize both m6A and m7G modifications within mRNAs in cancer cells. Notably, the spatial distribution of m6A and m7G is similar when bound by IGF2BP, but their regulatory effects on mRNA stability differ. IGF2BP-mediated recognition of m7G promotes mRNA degradation, whereas recognition of m6A enhances mRNA stability. This suggests a potential functional interplay between these modifications, as more than half of m7G-modified mRNAs were found to be simultaneously modified by m6A [135]. 5mC methylation is a well-characterized epigenetic modification in mammalian DNA. Remarkably, single-molecule quantum sequencing of miR-200c-5p—a microRNA derived from colorectal cancer (DLD-1) cells—revealed simultaneous m6A and 5mC modifications. Further studies demonstrated that RNA 5mC methylation is catalyzed by NOP2/Sun RNA methyltransferase 2 (NSUN2), an enzyme possessing two distinct structural domains: a YTH-like domain responsible for recognizing m6A, and an RsmB domain that catalyzes cytosine methylation. As a result, m6A modifications at positions #7 and #11 in miR-200c-5p are consistently associated with 5mC modification at position #13, highlighting the coordinated regulatory interplay between these two modifications [136] (Fig. 2).

Fig. 2.

The crosstalk between m6A modification and other posttranscriptional modifications. (a) There is a negative regulatory effect of m6A modification on A-to-I editing. Knockdown of METTL3 in mouse 3T3 cells upregulates A-to-I while knockdown of FTO downregulates A-to-I. (b) Knockout of NAT10, a major enzyme for ac4C acetylation, reduces the stability of YTHDC1, rendering the m6A sites on PFKM and LDHA mRNA unrecognized and ultimately positively regulating glycolysis in osteosarcoma cells. (c) METTL14 facilitates the m6A modification of NAT10 mRNA, while IGF2BP1 interacts with the m6A site on NAT10 mRNA to enhance its translation, resulting in NAT10 production. NAT10 mediates the ac4C modification of ACOT7 mRNA, promoting its translation to form ACOT7. (d) Catalyzes m6Am modification near the m7G cap, which increases resistance to DCP2-mediated decapping. (e) METTL14 installs m6Am modification at an internal site in U2 snRNA, which culminates in influencing global alternative RNA splicing. (f) METTL3 regulates m6A modification at the 3’-UTR in TROP2 mRNA to enhance translation of TROP2 mRNA, and METTL1 mediates m7G modification in tRNA to improve mRNA translation efficiency, which synergistically increase TROP2 expression to promote bladder cancer cell development. (g) Both m6A and m5C modifications are present in GPX4 mRNA, and both modifications work together to maintain redox homeostasis in COAD. (h) METTL3 mediates m6A modification of miRNAs, and NSUN2 recognizes m6A through the YTH-like structure domain and mediates 5mC methylation modification through the RsmB structural domain

The relationship between enhancer and m6A modification

Disruptions in epigenetic regulation play a pivotal role in numerous physiological and pathological conditions, with particularly profound implications for cancer development. Two key components in the field of epigenetics are m6A RNA methylation and enhancers. m6A is widely recognised as a post-transcriptional modification, while enhancers primarily regulate gene transcription. However, the relationship between m6A and enhancers remains unclear. Some potential effects of m6A on enhancers have been documented. In the context of acute myeloid leukaemia, the establishment of active enhancer elements potently induces IGF2BP2 and IGF2BP3 expression. These RNA-binding proteins, in turn, stabilise m6A-modified DDX21 transcripts. Functionally, DDX21 orchestrates an oncogenic programme by recruiting TF YBX1 to activate ULK1 expression, ultimately promoting leukemogenesis through enhanced proliferation and apoptosis resistance [137]. In clear cell renal cell carcinoma (ccRCC), the m6A eraser enzyme FTO regulates bromodomain-containing protein 9 (BRD9) expression by demethylating its mRNA to enhance stability. Importantly, the TF SOX17 mediates BRD9 recruitment to de novo enhancers that control key oncogenic networks, facilitating ccRCC progression [138]. Notably, we identified a METTL14-m6A-BPTF regulatory axis in renal cell carcinoma metastasis. While m6A modification destabilised BPTF mRNA, BPTF expression remained crucial for promoting lung metastasis. METTL14 deficiency led to BPTF accumulation, which subsequently altered the enhancer repertoire to reinforce oncogenic pathways and form active enhancers that reprogrammed cellular metabolism towards glycolysis. This work uncovers previously unrecognised functional links between RNA epigenetics and chromatin remodelling in cancer progression [139]. In conclusion, these results provide greater insight into the connection between enhancers and m6A. Nevertheless, only a limited number of studies have verified that m6A plays a role in enhancer regulation. It is still uncertain whether enhancers influence m6A in a mechanistic way, and further research is required to shed light on this issue.

RNA m6A modification in cancer

Numerous studies have demonstrated that m6A modification and its regulatory proteins exert significant influence over cancer initiation, progression, and prognosis. This section summarises the key roles of m6A modifications across various cancers (see Table 3).

Table 3.

The roles of m6A regulators in cancer

Cancer types	m6A regulators	Expression	Targets	Function	Role in cancer	References
AML	IGF2BP2	Upregulation	MYC/GPT2/SLC1A5	Prompting cell growth and proliferation Hindering cell differentiation and apoptosis	Oncogene	[144]
Glioblastoma	ALKBH5	Upregulation	VEGFA/CHK1/CD73	Prompting cell migration, invasion, and resistance to radiotherapy	Oncogene	[146–148]
	SRSF7	Upregulation	PBK	Enhancing cell proliferation and migration	Oncogene	[149]
LC	METTL3	Upregulation	circIGF2BP3/DCP2	Prompting cell immune escape and resistance to chemotherapy	Oncogene	[151, 153]
	IGF2BP3	Upregulation	COX6B2	Prompting cell resistance to chemotherapy	Oncogene	[154]
	FTO	Downregulation	ESR1	Inhibiting cell growth and metastasis	Anti-oncogene	[152]
	YTHDF1/IGF2BP3	Upregulation	ESR1	Prompting cell growth and metastasis	Oncogene	[152]
ESCA	METTL3	Upregulation	MYC	Promoting cell proliferation, invasion, metastasis, and immune escape	Oncogene	[157, 158]
GC	METTL3/IGF2BP2	Upregulation	PABPC1/miRNA-118/STAT5A	Prompting cell proliferation and migration	Oncogene	[159–161]
	YTHDF1	Upregulation	IFN-γ	Inhibiting cell growth and metastasis	Oncogene	[164]
	YTHDF2	Upregulation	CyclinD1	Prompting cell growth, invasion, and resistance to radiotherapy and chemotherapy	Oncogene	[165]
	YTHDF3	Upregulation	lncRNAs/miRNAs/mRNA	Prompting cell growth and invasion	Oncogene	[166]
HCC	METTL3/METTL14	Upregulation	ACLY, SCD1	Prompting cell growth and invasion	Oncogene	[167]
	WTAP/IGF2BP3	Upregulation	circCCAR1	Prompting cell growth, metastasis, resistance to chemotherapy, and immunosuppression	Oncogene	[170]
	YTHDF1	Upregulation	ATG2A	Prompting cell growth and metastasis	Oncogene	[234]
PC	WTAP/IGF2BP3	Upregulation	GBE1	Promoting cell proliferation and stemness	Oncogene	[175]
CRC	YTHDF2	Downregulation	STEAP3-AS1	Inhibiting cell growth, proliferation and metastasis	Anti-oncogene	[178]
	METTL3	Upregulation	POU6F2-AS1	Promoting cell growth and invasion	Oncogene	[179]
	METTL16/IGF2BP1	Upregulation	SOGA1	Promoting cell growth and proliferation	Oncogene	[180]
	METTL17	Upregulation	LRPPRC/MSSUs	Promoting cell growth, proliferation and invasion	Oncogene	[181]
	ALKBH5	Upregulation	AXIN2	Promoting cellular immunosuppression	Oncogene	[182]
RCC	HNRNPC	Upregulation	PPAP2B	Prompting cell proliferation and metastasis	Oncogene	[183]
	METTL14	Upregulation	TRAF1	Prompting cell resistance to chemotherapy	Oncogene	[184]
	ALKBH5	Upregulation	AURKB	Prompting cell growth, proliferation, invasion	Oncogene	[185]
BCa	RBM15/METTL3	Upregulation	ENO1	Promoting cell growth	Oncogene	[186]
	WTAP	Upregulation	circ0008399	Promoting cell resistance to chemotherapy	Oncogene	[187]
	YTHDF2	Upregulation	DHCR7	Promoting cell invasion and metastasis	Oncogene	[188]
	FTO	Upregulation	MALAT1	Promoting cell growth and proliferation	Oncogene	[189]
PCa	YTHDC1	Upregulation	SLC12A5	Prompting cell metastasis, and resistance to chemotherapy	Oncogene	[190]
	YTHDF3	Upregulation	SPOP/NXK3.1/TWIST1/SNAI2	Prompting cell growth and proliferation	Oncogene	[191]
	METTL3	Upregulation	circRBM33	Prompting cell growth, invasion, and resistance to chemotherapy	Oncogene	[192]
	RBM15	Upregulation	FTO-IT1	Prompting cell growth and proliferation	Oncogene	[193]
	FTO	Downregulation	DDIT4	Inhibiting cell growth and invasion	Anti-oncogene	[194]
BC	METTL3/YTHDF2	Upregulation	LATS1	Prompting cell growth and invasion	Oncogene	[195]
	ALKBH5	Upregulation	FOXO1	Prompting cell stemness, and resistance to chemotherapy	Oncogene	[196]
	IGF2BP2	Upregulation	CDK6	Prompting cell growth and proliferation	Oncogene	[197]
	hnRNPA2B1	Upregulation	ALYREF/NXF1	Prompting cell stemness and proliferation	Oncogene	[198]
OC	METTL3	Upregulation	circPLPP4	Prompting cell resistance to chemotherapy	Oncogene	[199]
	ALKBH5	Upregulation	ITGB1/FSH	Prompting cell proliferation and metastasis	Oncogene	[200, 201]
	YTHDC1	Downregulation	PIK3R1	Prompting cell growth and proliferation	Anti-oncogene	[202]
CC	ALKBH5	Downregulation	circCCDC134	Prompting cell proliferation and metastasis	Anti-oncogene	[203]
	HNRNPC	Upregulation	FOXM1	Prompting cell invasion and metastasis	Oncogene	[204]
	METTL3	Upregulation	HSPA9	Prompting malignant transformation of cells	Oncogene	[205]
EC	WTAP	Downregulation	EGR1	Hindering cell growth, and resistance to chemotherapy	Anti-oncogene	[207]
	METTL3	Downregulation	NLRC5	Hindering cell proliferation and metastasis	Anti-oncogene	[208]
OSCC	METTL14	Upregulation	CALD1	Promoting cell growth and metastasis	Oncogene	[209]
osteosarcoma	METTL14	Upregulation	MN1	Prompting cell growth and resistance to chemotherapy	Oncogene	[210]
melanoma	METTL14	Upregulation	RUNX2	Prompting cell growth and metastasis	Oncogene	[211]
pitNET	FTO	Upregulation	DSP	Prompting cell growth and resistance to chemotherapy	Oncogene	[212]
NPC	WTAP	Upregulation	DIAPH1-AS1	Prompting cell growth and metastasis	Oncogene	[213]
Laryngeal cancer	METTL3	Upregulation	HOXA10-AS	Prompting cell proliferation and oxidative resistance	Oncogene	[216]
HSCC	METTL3	Upregulation	circCUX1	Prompting cell resistance to radiotherapy	Oncogene	[217]
seminoma	METTL3	Upregulation	TFAP2C	Prompting cell resistance to chemotherapy	Oncogene	[218]

AML acute myeloid leukemia, LC lung cancer, ESCA esophageal cancer, GC gastric cancer, HCC hepatocellular carcinoma, PC pancreatic cancer, CRC colorectal cancer, RCC renal cell carcinoma, BCa bladder cancer, PCa prostate cancer, BC breast cancer, OC ovarian cancer, CC cervical cancer, EC endometrial cancer, OSCC oral squamous cell carcinoma, NPC nasopharyngeal carcinoma, pitNET pituitary neuroendocrine tumors, HSCC hypopharyngeal squamous cell carcinoma

Acute myeloid leukemia

AML, the most prevalent form of adult acute leukaemia, is an aggressive haematological malignancy characterised by the clonal expansion of immature myeloid cells, impaired differentiation of myeloid progenitors, and infiltration of malignant clones into the bone marrow, peripheral blood, and other tissues. These pathological processes lead to uncontrolled proliferation, blocked differentiation, and disrupted haematopoiesis. The development of AML has been linked to dysregulation of m6A regulators including METTL3, METTL14, ALKBH5, YTHDF1, YTHDF2, and IGF2BP2 [140–144]. Among these, IGF2BP2 expression is significantly elevated in AML, particularly in leukaemia stem cells (LSCs) and leukaemia-initiating cells (LICs), and is strongly associated with poor prognosis. IGF2BP2 enhances the expression of MYC, GPT2, and SLC1A5 in an m6A-dependent manner, promoting glutamine uptake and metabolism. This leads to improved mRNA stability and enhanced translation, fuelling the tricarboxylic acid (TCA) cycle and accelerating AML cell proliferation. These findings suggest that IGF2BP2 inhibitors may hold therapeutic promise in future AML treatment [144].

Glioblastoma

Glioblastoma is the most common and lethal primary brain tumour, known for its invasive growth, extensive vascularisation, and neurodestructive behaviour. Expression levels of YTHDF family members and METTL14 are elevated in glioblastoma tissues compared to normal brain tissue. In particular, YTHDF2 and YTHDF3 are markedly upregulated in high-grade gliomas [145]. ALKBH5 is also highly expressed in gliomas, and its knockdown has been shown to reduce angiogenesis both in vitro and in vivo [146]. Previous research has linked ALKBH5 with glioma cell invasion, as well as radioresistance and chemoresistance [147, 148]. Furthermore, ALKBH5 expression correlates with glioma malignancy and patient prognosis, positioning it as a potential clinical biomarker [146]. Additionally, SRSF7 (Serine/Arginine-Rich Splicing Factor 7), a novel m6A regulator, has been found to promote glioma cell proliferation and migration. It does so by recruiting the methyltransferase complex to selectively enhance m6A methylation near its RNA binding sites, which are often located in genes associated with cell growth and motility [149]. These findings provide important insights into the development of glioma-targeted therapies.

Lung cancer

Lung cancer (LC) is one of the most aggressive malignancies worldwide and accounts for approximately 18% of all cancer-related deaths. Compared to adjacent normal tissues, expression of METTL3, METTL16, YTHDF1, and IGF2BP3 is elevated in LC [150–152], while FTO expression is typically reduced [152]. A circular RNA derived from IGF2BP3, known as circIGF2BP3, is significantly overexpressed in non-small cell lung cancer (NSCLC). This circRNA (circular RNA) is generated via back-splicing of exons 4 and 13 and undergoes m6A modification mediated by METTL3, with YTHDC1 facilitating its circularisation. CircIGF2BP3 impairs CD8⁺ T cell responses and promotes tumour immune evasion [153]. In addition, METTL3 enhances m6A methylation of DCP2, leading to its degradation and the promotion of mitochondrial autophagy, which in turn contributes to chemoresistance in NSCLC [151]. Recent findings indicate that IGF2BP3 also promotes metabolic reprogramming, thereby contributing to acquired drug resistance [154]. FTO expression is significantly lower in NSCLC tumour tissues compared to adjacent non-cancerous tissues. One of FTO’s targets is ESR1 (oestrogen receptor alpha). YTHDF1 and IGF2BP3 each recognise distinct m6A sites on ESR1 mRNA, enhancing its stability and promoting tumour growth. Overexpression of FTO leads to reduced m6A methylation of ESR1 mRNA. These insights into the FTO–YTHDF1–IGF2BP3–ESR1 axis present new opportunities for NSCLC treatment [152]. However, studies on ALKBH5 in NSCLC have yielded conflicting results. Some research suggests ALKBH5 overexpression promotes NSCLC progression, while other studies indicate that elevated ALKBH5 levels inhibit tumour cell proliferation and promote apoptosis [155, 156]. These discrepancies highlight the need for further investigation into the context-dependent role of ALKBH5 in LC.

Esophageal cancer

Oesophageal cancer (ESCA) is a common and highly fatal malignancy of the digestive tract. Recent studies have revealed a positive correlation between Fusobacterium nucleatum (Fn) infection and METTL3 expression in ESCA tissues, with Fn significantly upregulating METTL3. c-Myc mRNA has been identified as a downstream target of METTL3, and the m6A reader YTHDF1 binds to the m6A site on c-Myc mRNA, enhancing its stability and thereby increasing c-Myc expression, which promotes ESCA cell invasion and metastasis [157]. Serine hydroxymethyltransferase 2 (SHMT2), a key mitochondrial enzyme involved in one-carbon unit metabolism, is significantly elevated in ESCA tissues and associated with poor prognosis. METTL3, ALKBH5, and FTO mediate the m6A modification of c-Myc mRNA, and SHMT2 further regulates this modification via the one-carbon metabolic pathway by increasing intracellular SAM (S-adenosylmethionine) levels, ultimately promoting c-Myc expression and facilitating immune escape [158]. Thus, targeting the one-carbon unit metabolic pathway may represent a promising therapeutic strategy for ESCA.

Gastric cancer

GC is the fifth most prevalent malignancy worldwide. Both METTL3 and its downstream target STAT5A act as oncogenes in GC. Helicobacter pylori infection has been shown to further elevate the expression of METTL3 and STAT5A, which correlates with poor patient prognosis. The reader protein IGF2BP2 binds to the m6A-modified site on STAT5A mRNA, enhancing its stability and protein expression, thereby increasing the invasive potential of GC cells [159]. In addition, METTL3 can promote gastric carcinogenesis by facilitating EMT and by binding to non-m6A-modified mRNAs [160, 161]. Although ALKBH5 has been explored for its role in GC cell invasion and progression [162], its expression is significantly lower in GC tissues compared to normal tissues and is further reduced in patients with distant metastases [163]. This inconsistency has led to debate regarding the role of ALKBH5 in GC, necessitating further research. Moreover, engineered small extracellular vesicles have been developed to deliver siRNA targeting YTHDF1, effectively reducing its expression and thereby inhibiting GC progression and metastasis, offering a novel therapeutic avenue [164]. YTHDF2 has also been shown to promote GC cell proliferation both in vitro and in vivo, as well as to increase resistance to radiotherapy [165]. Similarly, YTHDF3 enhances GC cell invasiveness by activating the PI3K/AKT pathway and modulating the immune microenvironment. Its upregulation is often associated with poor clinical outcomes [166].

Hepatocellular carcinoma

HCC is the second leading cause of cancer-related death globally. In HCC cells, overexpression of METTL3 and METTL14 leads to m6A-mediated stabilisation of oncogenic transcripts, thereby accelerating tumour progression [167]. CircRNAs also play critical roles in HCC. For example, circGPR137B is significantly downregulated in HCC and forms a feedback loop with miR-4739 and FTO, collectively suppressing metastasis [168]. Paradoxically, FTO is upregulated in HCC and contributes to tumour proliferation. Targeting FTO with inhibitors has shown potential to suppress HCC growth and enhance chemosensitivity [169]. Additionally, EIF4A3 promotes the generation of circCCAR1, whose stability is enhanced by WTAP-mediated m6A modification via IGF2BP3 interaction. Elevated levels of circCCAR1 impair CD8⁺ T-cell function, facilitating HCC progression and metastasis [170]. Multi-omics analyses have also revealed that NOTCH1 is a direct downstream target of YTHDF1, which binds to the m6A-modified NOTCH1 mRNA, enhancing its stability and translation. Inhibiting YTHDF1 reduces cancer cell stemness and improves therapeutic responses, making it a promising target for HCC treatment [171].

Pancreatic cancer

Pancreatic cancer (PC) is the most lethal gastrointestinal malignancy, known for its insidious onset, rapid progression, and dismal prognosis. METTL3 has been found to promote PC cell proliferation, whereas METTL16 inhibits PC cell metastasis and FTO reduces drug resistance [172–174]. The m6A modification of GBE1 (Glucan branching enzyme 1) mRNA by WTAP and IGF2BP3 enhances its stability and expression, correlating with poor prognosis in PC [175]. Additionally, DDIT4-AS1, a downstream target of ALKBH5, is upregulated via ALKBH5-mediated m6A demethylation, increasing stemness and chemoresistance [176]. Under hypoxic conditions, ALKBH5 decreases global m6A levels in PC cells, thereby enhancing glycolysis. YTHDF2 recognises m6A-modified HDAC4 (histone deacetylase type 4) mRNA, stabilising its expression. HDAC4, in turn, regulates the acetylation of HIF1α, a direct target of ALKBH5. This establishes a positive feedback loop involving ALKBH5/HDAC4/HIF1α, which drives glycolytic metabolism and promotes PC cell migration under hypoxia [177]. This regulatory axis may offer a therapeutic opportunity for hypoxia-targeted PC treatment.

Colorectal cancer

Colorectal cancer (CRC) is one of the most common malignancies globally, ranking third in incidence and second in mortality. The relationship between m6A modification and CRC progression has been well established. The antisense lncRNA STEAP3-AS1, induced by hypoxia, promotes CRC cell growth and migration, correlating with poor prognosis. STEAP3-AS1 and YTHDF2 compete for binding, disrupting the interaction between YTHDF2 and STEAP3 mRNA. This competition protects STEAP3 mRNA from m6A-mediated degradation, thereby enhancing the expression of STEAP3 protein to support CRC progression [178]. Another long non-coding RNA, POU6F2-AS1, is also upregulated in CRC. METTL3-induced m6A modification of POU6F2-AS1 promotes CRC cell proliferation and adipogenesis [179]. METTL16 expression correlates with the inhibition of SOGA1 (suppressor of glucose by autophagy) and PDK4 (pyruvate dehydrogenase kinase 4). METTL16 enhances SOGA1 expression by binding to IGF2BP1, leading to the upregulation of PDK4, which subsequently promotes glycolytic reprogramming and accelerates CRC progression [180]. METTL17 expression is elevated in CRC patients, and its depletion increases sensitivity to ferroptosis while inhibiting CRC cell proliferation, migration, invasion, and tumour growth in xenograft models [181]. ALKBH5 interacts with and demethylates AXIN2, a Wnt repressor, leading to the degradation of IGF2BP1 and promoting immunosuppression, thus fostering CRC development. Targeting ALKBH5 may provide therapeutic benefits for CRC patients [182].

Renal cell carcinoma

Among urologic malignancies, renal cell carcinoma (RCC) has a comparatively high mortality rate, and its incidence continues to rise. Many studies have linked m6A regulators to the development of RCC. However, it has now been discovered that METTL3, the IGF2BP family, and FTO all appear to play conflicting roles in RCC, with even the function of the same m6A regulator differing depending on the RCC subtype. CircPPAP2B was found to be overexpressed in highly aggressive RCC cells and associated with poor prognosis. Furthermore, it was demonstrated that circPPAP2B interacts with HNRNPC in an m6A-dependent manner and promotes HNRNPC nuclear translocation. On-degradable ubiquitination of HNRNPC leads to subcellular re-localisation, which causes metastasis of RCC [183]. Tumour necrosis factor receptor-associated factor TRAF1 expression is up-regulated in sunitinib-resistant RCC cells. METTL14 regulates the mRNA stability of TRAF1 in an IGF2BP2-dependent manner to maintain sunitinib resistance [184]. ALKBH5, which is highly expressed in RCC, stabilises the AURKB mRNA (encoding aurora kinase B) in an m6A-dependent manner to promote RCC cell growth and invasion [185]. Therefore, targeting METTL14 and ALKBH5 signalling pathways could offer a promising therapeutic avenue for RCC.

Bladder cancer

BCa is the second most common genitourinary carcinoma. Recurrence, metastasis, and drug resistance are three key concerns of BCa that need to be addressed. METTL3, METTL14, and WTAP are elevated in BCa compared to adjacent normal tissues. The RBM15/METTL3 complex mediates the m6A modification of ENO1 mRNA, and the oncogene ENO1 stimulates the growth of BCa cells. YTHDF1 recognises this m6A modification and accelerates the effectiveness of ENO1 protein translation, thus promoting the progression of BCa [186]. Binding of WTAP to circ0008399 speeds up the establishment of the m6A methyltransferase complex, improving circ0008399’s mRNA stability and triggering the expression of TNFα-inducible protein 3 (TNFAIP3). TNFAIP3 overexpression decreases BCa’s chemosensitivity [187]. BCa invasion in vitro and metastasis in vivo were facilitated by the upregulation of 7-dehydrocholesterol reductase (DHCR7) in BCa tissues. Decreased mRNA degradation due to YTHDF2-mediated increased DHCR7 in BCa induced the cAMP/protein kinase A/FAK pathway, greatly enhancing the invasive and metastatic capability of BCa, leading to poor prognosis [188]. Notably, the m6A eraser FTO functions as an oncogenic factor in BCa, as well as m6A writers. It has been shown that FTO catalyses the demethylation of metastasis-associated lung adenocarcinoma transcript 1 (MALAT1), regulates microRNA miR-384 and MAL T-cell differentiation protein 2 (MAL2) expression, and ultimately promotes the progression of BCa [189]. Thus, whether some crosstalks between m6A erasers and writers occur in BCa, causing co-carcinogenesis, remains unclear.

Prostate cancer

PCa is the second most prevalent malignant tumour, which can become more challenging to treat after a few years if it develops into CRPC. The dysregulation of m6A modification is strongly linked to PCa progression. SLC12A5, a neuron-specific potassium chloride co-transporter, can form a complex with YTHDC1 and then upregulate the TF HOXB13 to promote PCa progression [190]. In CRPC, YTHDF3 expression is upregulated and is associated with poor survival and high-grade metastasis. YTHDF3 promotes the RNA degradation of SPOP and NXK3.1 but stabilises the RNA expression of m6A-dependent TWIST1 and SNAI2, ultimately promoting the proliferation of CRPC cells [191]. The m6A-modified circRBM33 could bind to FMR1 and form a binary complex. The knockdown of METTL3 disrupted the formation of the complex. The circRBM33 in the complex can stimulate PCa cells’ mitochondrial respiration, while FMR1 can regulate cellular respiration. When these two work together, they can accelerate the proliferation of PCa cells and promote invasion [192]. FTO-IT1 is a lncRNA that is negatively associated with patient survival and is elevated in anti-androgenic and chemotherapy-resistant PCa. FTO-IT1 can directly interact with RMB15 to hinder the m6A methyltransferase complex, which subsequently decreases the mRNA m6A levels and stability of a subset of p53 target genes, ultimately resulting in increased cancer cell growth [193]. Besides, FTO can regulate the EMT in PCa cells, which is linked to bone metastasis [194].

Breast cancer

BC is the most prevalent carcinoma and the leading cause of cancer-related mortality in women. m6A regulators are essential in the development and treatment of BC. LAST1 is identified as an oncogene, with its m6A modification mediated by METTL3. YTHDF2 recognises the m6A modification in LAST1 mRNA, resulting in reduced stability. By increasing LATS1 mRNA expression and activating YAP/TAZ in the Hippo pathway, the knockdown of METTL3 or overexpression of YTHDF2 can prevent the onset of BC [195]. ALKBH5 expression was increased in BC patients and positively correlated with poor prognosis and chemotherapy resistance. ALKBH5 promotes m6A demethylation of FOXO1 mRNA and enhances the stability of FOXO1 mRNA in DOX-resistant BC cells, increasing the resistance of BC cells to chemotherapy [196]. Cell cycle protein-dependent kinase 6 (CDK6) is a direct target of IGF2BP2 in BC cells. IGF2BP2 directly regulates CDK6 in an m6A-dependent manner to accelerate BC progression [197]. Knockdown of HnRNPA2B1 or inhibition of m6A-tagged mRNA export can cause maintenance of nuclear m6A-tagged mRNAs linked to stemness maintenance. This would inhibit BC cell self-renewal and effectively improve the efficacy of DOX treatment [198]. Consequently, inhibition of m6A-tagged mRNA and its nuclear export may be a prospective therapeutic approach.

Ovarian cancer

Ovarian cancer (OC) is one of the three most common gynaecologic malignancies, and patients often have poor prognosis due to the lack of effective screening strategies. m6A modification has an impact on OC’s development, migration, invasion, and drug resistance. m6A can modify circPLPP4 and increase RNA stability, resulting in elevated circPLPP4 in OC cells. CircPLPP4 can act as a microRNA sponge to sequester miR-136, which further competitively up-regulates the expression of phosphatidylinositol-3-kinase regulatory subunit 1 (PIK3R1), leading to an increase in drug resistance in OC cells. Targeting m6A-modified circPLPP4 may be a prospective therapeutic target for OC-resistant patients [199]. Integrin subunit β1 (ITGB1) is a downstream target of ALKBH5 regulation. High expression of ALKBH5 demethylates m6A modifications in ITGB1 mRNA, which in turn prevents YTHDF2 from mediating the m6A-dependent degradation of this mRNA. This results in increased ITGB1 expression and enhanced phosphorylation of focal adhesion kinase (FAK) and Src proto-oncogene, ultimately promoting lymph node metastasis of OC [200]. Moreover, follicle stimulating hormone (FSH) was also able to promote epithelial-mesenchymal transition in OC cells, leading to OC metastasis in an m6A-dependent manner [201]. YTHDC1 is an oncogene in OC progression, and overexpression of YTHDC1 hindered OC advancement both in vivo and in vitro, but YTHDC1 expression was down-regulated in OC patients [202]. More investigation is required to explore whether YTHDC1 could be a novel therapeutic avenue for OC.

Cervical cancer

Cervical cancer (CC) is the fourth most common gynaecological tumour worldwide and is primarily induced by persistent infection with high-risk human papillomavirus (HPV). Drug resistance, metastasis, and CC cell proliferation are all closely related to m6A modifications. It has been found that, except for ALKBH5, whose expression is down-regulated in CC and exerts an anti-tumour effect, all other m6A modifiers are up-regulated in expression and exert an oncogenic effect. CircCCDC134, a cyclic RNA with m6A modification, stimulates CC cell growth and invasion both in vivo and in vitro. ALKBH5 has the capacity to mediate the demethylation of circCCDC134 to decrease its stability, while YTHDF2 recognises the m6A on circCCDC134 in an m6A-dependent manner and improves its stability. As ALKBH5 expression was reduced in CC and YTHDF2 was increased in CC, they synergistically enhanced circCCDC134 levels and accelerated CC progression [203]. HNRNPC significantly decreased the patients’ overall survival by recognising the m6A site on FOXM1 RNA (oncogene) in an m6A-dependent manner and controlling the isoform conversion of FOXM1 to encourage lymph node metastasis in CC [204]. Expression of the CC-derived exosome Mortalin was increased in plasma exosomes isolated from CC patients. m6A methylation in the 3’UTR of METTL3-catalysed HSPA9 mRNA conferred higher mRNA stability and enhanced translational efficiency to Mortalin [205]. Overexpression of Mortalin, in turn, facilitated proliferation, migration, and invasion of CC cells.

Endometrial cancer

Endometrial cancer (EC) is a highly prevalent gynaecologic carcinoma with increasing morbidity and mortality worldwide. m6A modifications can interact with circRNAs to influence the growth, invasion, and migration of EC cells. CircCHD7 connects with IGF2BP2 in an m6A-dependent manner by increasing mRNA stability and increases the expression of the downstream target gene, platelet-derived growth factor receptor β (PDGFRB), which ultimately enhances EC cell proliferation and hampers apoptosis [206]. Reduced levels of m6A modification and WTAP in EC patients correlate with the stemness of endometrial cancer stem cells (ECSC). The discovery of the WTAP/EGR1/PTEN axis offers a novel therapeutic strategy for EC [207]. Low METTL3 levels were related to poor prognosis in EC patients, while METTL3 expression was positively correlated with CD8 levels. Increased METTL3 in patients raises the ratio of CD8 + T cells and impedes EC progression, whereas METTL3 depletion has the opposite effect. The NLR family CARD structural domain containing 5 (NLRC5) is a target for METTL3-mediated m6A modification, which prevents its degradation through a YTHDF2-dependent mechanism. Elevated METTL3 levels enhance immunosurveillance and reduce EC immune evasion, suggesting that METTL3 and YTHDF2 could serve as diagnostic and prognostic factors for EC [208].

Other cancers

There is a connection between m6A modification and tumourigenesis in major systems of the body. Here, we present an overview of other cancers that have been comparatively less studied in relation to m6A. Research has revealed that m6A modifications on mRNAs and regulators of m6A are closely associated with the progression of various cancers. For example, in OSCC, METTL14 is involved in OSCC cell growth and metastasis by regulating the m6A level of CALD1 mRNA [209]. Additionally, METTL14 mediates m6A modification on MN1 mRNA to promote osteosarcoma progression and drug resistance [210], m6A modification on RUNX2 mRNA, mediated by METTL14, prompts migration and invasion of melanoma cells [211]. In pituitary neuroendocrine tumours, elevated levels of FTO cause a decline in the m6A level of DSP mRNA, which increases resistance to tumour therapy [212]. In nasopharyngeal carcinoma (NPC), WTAP mediates m6A modification of the lncRNA DIAPH1-AS1, enhancing its stability and promoting NPC metastasis [213]. In thyroid cancer and lymphoma, m6A-related lncRNAs (LINC02471, DOCK9-DT) can be useful biomarkers [214, 215]. Additionally, m6A modifications in various RNA types across different cancers can lead to distinct forms of tumour resistance. In laryngeal cancer, HOXA10-AS sequesters miR-29b-3p to function as a competitive endogenous RNA, enhancing tumour cell oxidative resistance and accelerating laryngeal cancer progression [216]. In hypopharyngeal squamous cell carcinoma, METTL3 mediates the m6A modification of circCUX1 to increase its expression and enhance cancer cell resistance to radiotherapy [217]. In seminoma, METTL3 acts as a TF-activated enhancer binding protein 2C mRNA (TFAP2C) m6A writer, increasing the stability of TFAP2C mRNA, which subsequently causes enhanced DNA repair-related gene expression, including WEE1 and BRCA1, ultimately resulting in cancer cells becoming resistant to cisplatin chemotherapy [218].

m6A modification exhibits dual regulatory roles in cancer phenotypes. First, m6A can drive pro-tumorigenic effects by activating key oncogenic pathways, promoting immune evasion, and reprogramming metabolic processes. For instance, m6A methylation of EphA2 and VEGFA enhances vasculogenic mimicry in CRC through activation of the PI3K/AKT and ERK1/2 signalling pathways [219]. Additionally, METTL14 modulates mitochondrial metabolism-related genes to enhance tumour cell energy acquisition [220]. Conversely, m6A modifications can exert anti-tumorigenic effects by suppressing malignant transformation and enhancing chemosensitivity. For example, m6A modification-mediated lncRNA TP53TG1 inhibits GC progression by regulating CIP2A stability [221]. The m6A modification of TGFB2 triggers a reconfiguration of lipid metabolism, which subsequently contributes to the development of gemcitabine resistance in pancreatic ductal adenocarcinoma (PDAC) [222]. Notably, the same regulatory factors may display opposing roles across cancer types. FTO promotes oncogene expression in AML but suppresses tumour progression in gliomas, a divergence potentially linked to tissue-specific signalling pathways or microenvironmental interactions [223, 224]. In chemotherapy resistance, m6A modifications may either promote drug resistance by regulating resistance-associated genes, such as METTL3 upregulation, which induces cisplatin resistance in OC [225]. In immunotherapy, m6A modifications regulate immune checkpoints by modulating PD-1/PD-L1 expression. YTHDF1 enhances PD-L1 mRNA translation to facilitate immune evasion in PCa [226]. Furthermore, ALKBH5 deficiency promotes M1-polarised macrophage infiltration, improving immunotherapy outcomes [155]. The dual roles of m6A modifications and their microenvironment-dependent effects underscore the need for personalised intervention strategies in cancer therapy.

Strategies targeting m6A mRNA can be categorised as follows: (1) targeting m6A regulatory factors. For instance, YTHDF2 has been shown to effectively regulate leukaemia cell proliferation [227], while targeting METTL3 holds potential efficacy in improving gemcitabine therapy for PC [228]. (2) Utilising RCas9 technology to modify m6A sites, enabling precise alteration of specific m6A loci [229]. (3) Combination therapies integrating conventional tumour treatments with novel epigenetic drugs. m6A mRNA regulators have been implicated in tumour radioresistance and demonstrated significant clinical utility when combined with kinase inhibitors (KIs), 5-fluorouracil (5-FU), gemcitabine, anthracyclines, and cisplatin [230]. Moreover, the integration of m6A mRNA regulation with immune checkpoint blockade has been recognised as a promising therapeutic approach for the treatment of malignancies [231].

Currently, no FDA-approved m6A-targeted anticancer drugs are available. The METTL3 inhibitor STC-15 is undergoing Phase I clinical trials (NCT05584111) for various malignancies, including lung, hepatic, bone, colorectal, cutaneous, breast cancers, and malignant melanoma [232]. CCS1 (Bisantrene), a specific FTO-targeting agent that inhibits m6A demethylase activity, has entered Phase II clinical trials (NCT04989335) in combination with fludarabine and clofarabine for relapsed or refractory AML [233]. These developments indicate that clinical trials of m6A-targeted therapies remain in their preliminary stages.

Several challenges and potential risks persist in developing m6A-targeted cancer therapeutics: firstly, current pharmacological strategies demonstrate insufficient target discrimination. Small-molecule inhibitors targeting m6A-associated enzymes might interact with other structurally similar enzymes or molecular pathways, leading to unintended consequences. Antisense oligonucleotide (ASO) approaches carry inherent risks of sequence-dependent off-target binding, which could inadvertently disrupt RNA processing mechanisms and compromise transcriptomic fidelity. Secondly, differential m6A modification patterns between tumour and normal tissues may result in unexpected toxicities. Additionally, dynamic changes in m6A modification during disease progression and tumour microenvironment evolution complicate therapeutic outcome prediction. Thirdly, potential side effects such as immunosuppression or hepatotoxicity require careful evaluation. Most current research remains at the preclinical stage, presenting significant challenges in clinical translation. Overall, while m6A targeting shows therapeutic promise, overcoming these obstacles necessitates further research to develop more specific, efficient, and safer therapeutic strategies.

m6A modification of enhancer RNAs

The revelation that a large number of eRNAs are produced within the human genome has transformed our perception of enhancers, indicating that they can act not just as DNA but also as regulatory RNAs [235]. Pathological samples show a significant production of eRNAs, and mutations linked to diseases often coincide with enhancers that display eRNAs transcription, underscoring the potential involvement of enhancer transcription or eRNAs transcripts in disease development [236].

Among the more than 170 types of cellular RNA modifications, N6-adenosine methylation (m6A) stands out as the most common and important post-transcriptional modification in eukaryotic cells [237]. Recent findings indicate that m6A modification is widespread in eRNAs and may affect their stability, localisation, and function [101, 238]. However, the molecular function of m6A has mainly been explored in more abundant mRNA and post-transcriptional regulatory processes, leaving the role of m6A-modified eRNAs in transcriptional regulation largely unexplored. Additional research is needed to investigate the impact of other post-transcriptional modifications on eRNAs.

m6A modification of enhancer RNAs detection

eRNAs are generated during the transcription process and are known to be unstable [239]. Most eRNAs have half-lives of under 10 min, making them difficult to capture in steady-state RNA samples [240]. Therefore, more sophisticated and sensitive sequencing technologies are necessary to accurately analyse m6A modifications on eRNAs. Recent advancements in m6A sequencing technology offer new methods to explore the post-transcriptional regulation of eRNAs. Detection techniques can be categorised into two types: antibody-dependent and antibody-independent (Table 4). Despite several improvements in this technology in recent years, including the availability of commercial m6A MeRIP kits, the localisation of m6A is less precise, typically spanning peak regions of about 50–200 base pairs that may include multiple DRAC motifs, and there is a risk of false-positive enrichment in areas of gene alternative splicing [241]. Furthermore, the need for several dozen to hundreds of micrograms (µg) of total RNA limits the application of m6A epitranscriptome profiling in clinical contexts, as most patient samples yield only a few micrograms of total RNA. Therefore, it is crucial to develop an m6A profiling method suitable for low-input RNA samples to better understand m6A dynamics in various diseases.

Table 4.

The advantages and limitations of m6A detection methods

m6A detection methods		Advantages	Limitations	References
Anti-body dependent	MeRIP-seq	1.Simple to perform 2.Good universality and wide applicability	1.False-positive 2.False-negative 3.Limited resolution	[248]
	m6A-seq2	1.Merge samples for IP 2.Reduce the technical variability	1.False-positive	[249]
	miCLIP-seq	1.Single-base resolution	1.Unable to quantify and locate	[250]
	MINT-Seq	1.Recognize m6A on nascent RNA	1.False-positive	[242]
Anti-body independent	m6A-SEAL-seq	1.High sensitivity and specificity	1.Tedious and time-consuming 2.Unable to achieve single-base resolution	[251]
	m6A-SAC-seq	1.Single-base resolution 2.Test of small samples	1.Based on an enzyme 2.Sequence perference	[252]
	Nanopore direct RNA sequencing	1.Single-base resolution	1.High cost and low accuracy 2.Require high RNA quality and quantity	[253]
	GLORI-seq	1.Absolute quantification of single-base resolution	1. High cost 2. The calculation complexity is high	[254]
	DART-seq	1.High sensitivity	1.Low accuracy	[255]

Recently, researchers have employed Methylation Inscribed Nascent Transcript Sequencing (MINT-Seq), a new sequencing technique that effectively enriches and quantifies m6A distribution on nascent RNA, leading to the identification of several novel m6A sites on introns [242]. Additionally, a recent study introduced a method for identifying m6A sites across the transcriptome in single cells, uncovering unexpected variability in the presence and abundance of m6A sites among individual cells and highlighting differentially methylated mRNAs throughout the cell cycle [243]. Furthermore, researchers have created CRISPR/dCas13 (Rx)-mediated m6A editors by combining them with RNA methyltransferase GhMTA (TME) or demethylase GhALKBH10 (TDE) to induce specific m6A site demethylation or methylation. This demonstrates that these m6A editors can be used to investigate the effects of specific m6A modifications [244]. Although high-throughput sequencing technologies have facilitated the swift accumulation of data regarding numerous m6A sites, these methodologies remain susceptible to generating false-positive or false-negative outcomes due to inconsistencies in the sequencing process or technical complications. Consequently, it is imperative to seek more reliable approaches for the detection of m6A modifications.

Most eRNAs degrade so rapidly that they are challenging to detect, and current sequencing techniques struggle to capture their entire length. However, recent studies utilising sequencing-based nascent RNA methods have introduced effective tools for identifying RNA synthesis and unstable transcripts, such as eRNAs. Techniques like NET-seq, GRO-seq, and TT-seq have emerged [240, 245–247]. These nascent RNA sequencing approaches have opened new possibilities for capturing eRNAs sequences. In the future, researchers may integrate these nascent RNA sequencing techniques with m6A detection methods to address the rapid degradation of eRNAs, thereby enabling more precise identification of m6A modifications on eRNAs.

m6A modification of enhancer RNAs in cancer

In 2019, researchers employed methylated RNA immunoprecipitation sequencing (MeRIP-Seq) to profile the m6A landscape across the entire transcriptome in 21 human fetal tissue samples, identifying 1,908 enhancer-derived lncRNAs containing m6A modifications [256]. Notably, the pattern of m6A deposition on eRNAs closely resembled that observed in mRNAs, both conforming to the conserved RRACH motif [257]. Since this discovery, successive researches have discovered the impact of m6A modifications on eRNAs stability. For instance, in mouse embryonic stem cells (mESCs), m6A methylation was investigated to promote eRNAs degradation. Knockdown of METTL3 or YTHDC1 contributed to reduced m6A levels on eRNAs as well as a subsequent increase in eRNAs abundance. Mechanistically, METTL3 functions as an m6A writer, catalyzing m6A modifications on eRNAs, while YTHDC1, an m6A reader, facilitates eRNAs decay through nuclear exosome-mediated degradation. Consequently, depletion of either METTL3 or YTHDC1 leads to eRNAs stabilization. Additionally, stabilized eRNAs actively recruit TFs, such as YY1 and CBP/EP300, to maintain an open chromatin state while simultaneously displacing repressive complexes like PRC2. This process ultimately enhances chromatin accessibility and activates the transcription of downstream genes in an m6A-dependent manner, inducing a positive feedback loop [258]. These findings illustrate that m6A modifications act as a regulatory switch controlling eRNAs abundance, stability, and chromatin dynamics, thereby influencing gene transcription. Conversely, an independent discovery in BC cell line MCF-7 reported the opposite effect—m6A modifications were found to protect eRNAs from degradation. In METTL3-knockdown MCF-7 cells, eRNAs production were significantly reduced, and premature termination of transcription was observed due to enhanced recruitment of the Integrator complex, which mediates RNA cleavage. Interestingly, loss of METTL3 caused a marked enhancement in the nucleic acid endonuclease subunit (INTS11) at enhancer regions. However, there is a lack of clarity about how Integrator is recruited to target genes and how its nucleic acid endonuclease activity is regulated. Mechanistically, METTL3-mediated m6A modifications appeared to impair Integrator recruitment by stabilizing RNA polymerase II (Pol II). Additionally, m6A modifications facilitated the recruitment of hnRNPG and YTHDC1, protecting eRNAs from endonucleolytic cleavage [259]. Thus, these findings suggest that m6A modifications, along with m6A readers, synergistically enhance eRNAs stability in certain cellular contexts. However, the strength of antagonism between METTL3 and Integrator may not fully explain the ability of METTL3 to positively regulate transcription. Thus, the m6A deposited by METTL3 on nascent eRNAs may also play other regulatory roles in transcription and/or post-transcription. In summary, the conflicting conclusions drawn from these two studies are likely attributed to differences in cellular environments—mESCs versus MCF-7 cells—highlighting the context-dependent nature of m6A regulation. This discrepancy emphasizes the demand for further research to elucidate the precise impact of m6A modifications on eRNAs stability across different cell types. Notably, it is well validated that m6A has a pivotal impact on promoting mRNA degradation in cytoplasmic lysates. However, its effects on various RNA species appear to vary relying on the stage of the RNA life cycle or cellular context. Furthermore, a recent study challenged the long-standing notion that m6A modifications primarily accelerate mRNA degradation in coding regions. Instead, the study demonstrated that m6A modifications in the 3’UTR stabilize mRNAs more efficiently than those in coding regions [260]. This suggests that m6A modifications exert differential regulatory effects depending on their position within the transcript. However, whether this principle extends to eRNAs remains to be determined and requires further investigation.

In addition to regulating eRNAs stability, m6A modifications have also been implicated in cancer development, progression, invasion, and metastasis. For instance, in breast cancer MCF-7 cells, m6A modifications are widely distributed on eRNAs, and eRNAs harboring m6A tags tend to be longer, more stable, and highly cell-type specific. These modifications serve as essential recognition sites for the m6A reader protein YTHDC1, which directly activates enhancers and induces eRNAs transcription. Furthermore, YTHDC1 binding to eRNAs facilitates chromatin condensation, engages in crosstalk with transcriptional coactivators (BRD4), and enhances enhancer activity, ultimately promoting oncogene expression and breast cancer progression. Thus, m6A modifications on eRNAs play a coordinated role with reader proteins in enhancer activation and gene transcriptional regulation, extending their influence beyond eRNAs function alone [101]. Interestingly, while previous findings in mESCs indicated that YTHDC1 depletion increased eRNAs stability, this study in MCF-7 cells reported that YTHDC1 depletion did not affect eRNAs stability but rather hampered enhancer function and gene transcription. This distinction may be attributed to differences in cell types or variations in experimental methodologies, though the precise underlying mechanism remains unclear [258]. A similar regulatory role of m6A-modified eRNAs has been observed in PDAC. Super-enhancer-associated RNAs (seRNAs) harboring m6A modifications were found to be longer and enriched with RRACH motifs compared to their non-m6A-modified counterparts. Notably, chromatin regions associated with highly m6A-modified seRNAs exhibited significantly increased H3K4me3 levels. Mechanistically, cofilin-1 (CFL1), a filament-severing protein, was established as a cofactor of METTL3 that interacts with its zinc finger domain, promoting METTL3 nuclear co-localization and facilitating localized chromatin modifications that drive oncogene expression. Additionally, YTHDC2, an m6A reader of seRNAs, was found to interact with the histone methyltransferase mixed-lineage leukemia protein 1 (MLL1/KMT2A) to cooperatively enhance localized H3K4me3 deposition and promote PDAC progression. Notably, silencing METTL3 or YTHDC2 significantly suppressed the malignant phenotype of PDAC. These findings suggest that m6A modifications of seRNAs, mediated by METTL3 and CFL1, are identified by YTHDC2, which subsequently employs MLL1 to enhance chromatin accessibility and drive oncogene transcription [261, 262]. Thus, the CFL1-YTHDC2-MLL1 axis plays an important role in the epigenetic regulation of chromatin remodeling during PDAC development. In uterine corpus endometrial cancer (UCEC), IGFBP7 antisense RNA1 (IGFBP7-AS1) functions as a critical eRNA regulating IGFBP7, a member of the IGFBP gene family. IGFBP7-AS1 and IGFBP7 were substantially lower in UCEC tissues than in normal tissues, and reduced IGFBP7-AS1 expression was linked to poor clinical outcomes. Further analysis disclosed a link between IGFBP7-AS1 expression and key m6A regulators. In addition to FTO, declined expression of METTL14, WTAP, RBM15, YTHDF1, YTHDF2, YTHDC1, YTHDC2, and HNRNPC led to an increase in IGFBP7-AS1 expression. Notably, patients with high IGFBP7-AS1 expression exhibited elevated basophil and immune cell infiltration, as well as improved responses to immunotherapy. Mechanistically, IGFBP7-AS1 may improve the tumor immune microenvironment and affect immune lymphocytes such as T cells, mast cells, macrophages M1 and M2, thereby inducing protective immunity in UCEC. These findings suggest that increased m6A modifications on IGFBP7-AS1 lead to its downregulation, contributing to poorer clinical outcomes in UCEC patients. However, the association between m6A modifications and eRNAs expression levels is unclear, perhaps by affecting the stability of eRNAs. In UCEC, m6A affects eRNAs stability possibly by recruiting m6A reading proteins to regulate degradation or stabilization, or by disrupting the local secondary structure of eRNAs to expose or hide nuclease cleavage sites, or by regulating the post-transcriptional processing of eRNAs, which no study has yet been able to demonstrate [263]. Surprisingly, a recent study revealed that enhancer-driven eRNAs can indirectly prevent m6A site recognition on mRNAs, thereby accelerating HCC progression. Super-enhancer (SE) profiling in HCC cell lines identified a novel lncRNA, LINC01089, which was markedly elevated in HCC tissues and in relation to poor patient prognosis. While LINC01089 overexpression had no influence on HCC cell growth, it considerably boosted EMT, migration, invasion, and metastasis. Mechanistically, LINC01089 was found to interact with the RRM1 domain (aa71-149) of hnRNPM, promoting exon 3 skipping in DIAPH3 mRNA during alternative splicing. Exon 3 of DIAPH3 harbors an m6A site recognized by IGF2BP3, which enhances DIAPH3 mRNA stability. Consequently, exon 3 skipping leads to decreased DIAPH3 mRNA stability, resulting in elevated levels of phosphorylated ERK (pERK), phosphorylated Elk1 (pElk1), and Snail, thereby driving HCC metastasis. The depletion of LINC01089 resulted in elevated levels of DIAPH3 protein, suppressed the ERK/Elk1/Snail signaling pathway, and ultimately impeded the process of epithelial-mesenchymal transition in HCC cells [264].

In addition to their role in cancer progression, m6A-modified eRNAs have also been implicated in cancer drug resistance and hold potential as clinical biomarkers. One notable example is SUCLG2-AS1, a super-enhancer (SE)-derived lncRNA that is highly expressed in NPC tissues. This lncRNA harbors 14 highly conserved m6A modification sites, and its elevated expression has been correlated with poorer prognosis and reduced survival rates in NPC patients. Mechanistically, the m6A reader IGF2BP3 identifies the METTL3-mediated m6A modifications on SUCLG2-AS1, preventing its degradation and thereby enhancing its stability. Additionally, knockdown of SUCLG2-AS1 significantly reduced SOX2 protein levels in NPC cells. Furthermore, CTCF (CCCTC-binding factor), a multifunctional TF, was found to be recruited to the SOX2 promoter and enhancer regions by SUCLG2-AS1. Through long-range chromatin looping, SUCLG2-AS1 facilitated SOX2 transcriptional activation, leading to SOX2 accumulation and subsequent NPC metastasis and radioresistance. These findings suggest that m6A-modified and stabilized SUCLG2-AS1 enhances SOX2 transcription by recruiting CTCF, thereby regulating the interaction between the super-enhancer and SOX2 promoter regions. Given its oncogenic role, targeting the m6A modification sites on SUCLG2-AS1 to reduce its stability may provide a novel therapeutic method for NPC [265]. In bone-metastatic prostate cancer (mPCa), MLXIPe has been identified as a tumor-specific eRNA modified by m6A. m6A levels of MLXIPe were considerably elevated in mPCa tissues compared to primary PCa and adjacent non-tumor tissues. Importantly, increased MLXIPe m6A levels were linked to poorer survival in mPCa patients, and m6A-modified MLXIPe was specifically enriched in bone-metastatic lesions. Further analysis revealed that PSMD9, a gene linked to mPCa progression, was a downstream target of MLXIPe. Notably, PSMD9 expression was correlated with lower recurrence-free survival following radiotherapy, and its 5’UTR contained m6Am modifications. Mechanistically, KHSRP, an RNA-binding protein that serves as a dual reader of m6A and m6Am, was found to regulate both MLXIPe and PSMD9. Specifically, KHSRP utilized its KH1/2 domain to recognize m6Am on PSMD9 mRNA and its KH3/4 domain to recognize m6A modifications on MLXIPe. This dual recognition mechanism stabilized both MLXIPe and PSMD9 transcripts. Knockdown of XRN2, a highly active 5’ > 3’ ribonuclease, led to increased PSMD9 mRNA stability. Additionally, MLXIPe was shown to prevent XRN2-mediated degradation of PSMD9 mRNA via m6A modification and the KHSRP complex, ultimately leading to increased PSMD9 expression and enhanced production of DNA repair proteins. As a result, m6A-modified MLXIPe contributed to radiotherapy resistance in mPCa cells. Silencing the m6A sites on MLXIPe markedly reduced radiotherapy resistance, thereby restoring the sensitivity of mPCa cells to radiotherapy. These findings underscore the potential of targeting m6A modifications on MLXIPe as a strategy for overcoming radiotherapy resistance in mPCa [238]. Another study identified five m6A-associated eRNAs that were significantly connected with prognosis in patients with head and neck squamous cell carcinoma (HNSCC). These eRNAs were utilized to establish a prognostic risk model according to their expression and m6A modification status. The model included the following five eRNAs: AC005562.1, COPDA1, LINC00271, MIAT, and MIR4435-2HG. Functional enrichment analysis suggested that the m6A-associated eRNAs risk model was involved in tumor progression and immune-related biological pathways. Notably, the expression levels of COPDA1 and MIAT were positively correlated with the infiltration rates of naïve B cells, CD8 + T cells, and activated CD4 + memory T cells, whereas their expression was inversely correlated with MIR4435-2HG levels. Compared to normal tissues, MIAT and MIR4435-2HG were significantly upregulated in HNSCC. Interestingly, while high MIAT expression was in association with enhanced tumor immune cell infiltration and improved anti-tumor immunity, resulting in a better prognosis, elevated MIR4435-2HG expression was associated with worsened disease progression and increased tumor aggressiveness. Further functional assays demonstrated that silencing MIR4435-2HG suppressed cancer cell migration and invasion, further supporting its oncogenic role in HNSCC. However, the roles of the remaining three eRNAs (AC005562.1, COPDA1, and LINC00271) in HNSCC remain unclear and require further investigation. Taken together, these findings indicate that the 5m6A-associated eRNAs risk model has potential as a predictive tool for patient survival and response to immunotherapy. By leveraging this risk model, clinicians may be able to identify high-risk HNSCC patients and develop personalized immunotherapy or targeted therapy strategies based on their m6A-associated eRNA profiles [266] (Fig. 3).

Fig. 3.

The role of eRNAs m6A in cancer. The regulators of eRNAs m6A are engaged in the progression of numerous cancers, including BC, PDAC, mPCa, NPC, HNSCC and UCEC

In clinical diagnostics, eRNAs m6A modifications demonstrate unique advantages over traditional biomarkers, including high sensitivity, robust specificity, and multi-omics correlations. The development of a “liquid biopsy 2.0” system—integrating Visium HD spatial transcriptomics, eRNA m6A profiling, cfDNA methylation, protein biomarkers (e.g., PD-L1), and nanopore sequencing—holds significant potential. Within this framework, deep neural networks (DNNs, e.g., Transformer models) can be trained to automatically decipher modification signatures in nanopore sequencing signals for AI-assisted interpretation, while rapid detection of cell-free RNA may be achieved through microfluidic chips and CRISPR-Cas13a systems. For future therapeutic strategies targeting eRNAs m6A modifications, several directions hold promise: firstly, small-molecule inhibitors could be designed by targeting the m6A writer complex or blocking reader protein functions, such as developing METTL3 allosteric inhibitors and YTHDF1/2-specific antagonistic peptides. Secondly, RNA-targeted therapies may include precision editing of m6A modification patterns to reduce off-target effects. Thirdly, combination therapeutic strategies, such as PARP inhibitors combined with METTL3 inhibitors or immune checkpoint inhibitors paired with YTHDF2 antagonists, warrant exploration. In summary, eRNAs m6A modifications serve as a critical nexus linking epigenetic regulation and transcriptional programmes, offering novel perspectives for cancer diagnosis and treatment. It is anticipated that through the integration of multi-omics approaches and cross-disciplinary applications of nanotechnology, targeting eRNAs m6A modifications will enable a comprehensive clinical workflow—from early screening to targeted intervention—ultimately improving patient survival outcomes and advancing the field of precision oncology.

Conclusions and future directions

Despite accounting for approximately 98% of the genome, the precise function of enhancers—one of the most significant classes of non-coding regulatory factors—remains largely elusive. Within the last couple of years, advances in high-throughput sequencing technologies have given deeper perception into the mechanisms by which enhancers drive cell-type-specific gene expression. However, these discoveries have simultaneously led to a more fluid and nuanced definition of enhancers. The identification of hundreds of thousands of enhancer-like regulatory elements that function through mechanisms distinct from classical enhancers has prompted a reconsideration of how enhancers should be defined. As a cutting-edge technological approach, crispr-based editing has been applied to identify novel enhancer elements, but challenges remain in applications across the genome and in multiple cell types. Perhaps in the future it will be possible to combine crispr-based editing with single-cell sequencing to reveal more in-depth functional mechanisms of enhancers or enhancer RNAs [267]. Moving forward, the refinement and integration of enhancer identification methodologies will be essential to distinguishing enhancer-like elements from classical enhancers. A more comprehensive understanding of their molecular mechanisms will not only contribute to redefining enhancers but also enhance our ability to interpret their regulatory roles. Furthermore, the mechanisms governing enhancer-promoter interactions and the precise processes by which enhancers activate transcription remain largely unresolved. In addition, enhancer dysregulation has emerged as a critical area of focus in cancer research. Structural variations, epigenetic modifications, and alterations in enhancer activity have been implicated in oncogene overexpression and tumor suppressor gene inactivation. However, the full repertoire of enhancers involved in tumorigenesis has yet to be elucidated. Key questions persist regarding the regulatory influence of known enhancers on their target genes, the extent of shared and distinct biological effects of enhancer dysregulation across different tumor types, and the feasibility of targeting enhancers for therapeutic intervention. Addressing these challenges will be essential for leveraging enhancer biology in cancer diagnosis and treatment.

An accumulating number of researches has underlined the pivotal impact of m6A RNA modification on cancer. Although m6A-related studies are still in their early stages, multiple studies strongly demonstrate that dysregulation of m6A regulators is closely linked to key oncogenic processes, maintaining cancer cell proliferation, invasion, metastasis, and drug resistance. Interestingly, some m6A regulators display dual roles, acting as both oncogenes and tumor suppressors in different cancer types. For instance, METTL3 promotes tumorigenesis in LC, ESCA, GC, HCC, CRC, PCa, BCa, BC, OC, CC, laryngeal cancer, hypopharyngeal squamous cell carcinoma, and seminoma, yet exerts tumor-suppressive effects in EC. Similarly, FTO has been investigated to enhance tumor advancement in BCa and pituitary neuroendocrine tumors while playing a tumor-suppressive role in LC and PCa. More strikingly, certain regulators display dual roles within the same cancer type. ALKBH5, for example, acts as both an oncogene and tumor suppressor in LC, GC, and PC, while FTO exhibits similar paradoxical effects in HCC. Beyond their roles in tumorigenesis, m6A modifications also mediate various forms of therapy resistance. For instance, METTL3 has been implicated in oxidative stress resistance in laryngeal cancer, radioresistance in hypopharyngeal squamous cell carcinoma, and chemoresistance in seminoma. These observations raise several fundamental questions: Why can both m6A writers and erasers be dysregulated within the same cancer type, contributing to comparable oncogenic or tumor-suppressive effects? Why do the same regulatory factors exert opposing roles across different cancers? How can a single regulator display dual functions within the same tumor type? What determines the ability of an m6A regulator to mediate distinct forms of therapy resistance across different malignancies? The underlying molecular mechanisms remain unclear, potentially involving variations in the tumor microenvironment (TME) or complex epigenetic crosstalk. Further investigation is required to elucidate these mechanisms. Additionally, m6A regulators have demonstrated potential as biomarkers for cancer diagnosis, prognosis, and targeted therapy. METTL3, FTO, ALKBH5, and YTHDF1/2/3 have been identified as prognostic biomarkers for multiple cancers, while METTL3 and FTO also serve as diagnostic markers. YTHDF1/2, METTL17, ALKBH5, and YTHDC1 have emerged as promising therapeutic targets, with interfering RNA against YTHDF1 successfully synthesized to inhibit GC progression. However, the translation of such therapeutic strategies into clinical applications remains in its infancy, with significant gaps in preclinical and clinical validation. Given the extensive regulatory influence of m6A on gene expression, its therapeutic potential must be carefully assessed to determine possible side effects and clinical viability. Furthermore, our findings highlight an intricate interplay between m6A RNA modifications and other epigenetic modifications, including DNA methylation, histone modifications, and RNA demethylation. This interplay has been implicated in pivotal biological processes such as RNA splicing and decay, significantly impacting the progression of osteosarcoma, OC, BCa, and CRC. However, it remains unclear whether these interactions are cancer-type specific or represent a more universal mechanism. Additionally, whether m6A modifications engage in broader crosstalk with other DNA and histone modifications in cancer progression remains an open question, lacking systematic study. Future research should address these pressing questions, refining our understanding of m6A modifications and their broader implications in oncogenesis. By integrating advanced methodologies and exploring novel regulatory networks, we can unlock new dimensions of cancer epigenetics and pave the way for more effective diagnostic and therapeutic strategies.

The number of eRNAs in the human transcriptome is extraordinarily large, and their functions are largely dependent on enhancers. However, due to their inherent instability and low abundance, there is an urgent need to develop more sensitive sequencing technologies beyond GRO-seq, PRO-seq, and ChIP-seq to enable more precise identification of eRNAs. Advancements in this area will offer deeper perspectives into the functional roles of eRNAs. While eRNAs are known to enhance enhancer activity by facilitating the formation of enhancer-promoter loops, the mechanisms by which they regulate the expression of distal genes remain unclear. eRNAs have a dual impact on cancer, contributing to both oncogene activation and tumor suppressor regulation while also participating in signaling cascades. However, these represent only a fraction of the functional mechanisms underlying cancer-associated eRNAs, and further investigation is needed to illuminate additional regulatory pathways. Moreover, in oncogenic signaling pathways, eRNAs are not only subject to regulation but also act as key regulatory elements themselves. The precise molecular mechanisms governing these interactions remain largely undefined. Additionally, eRNAs have been implicated in immune evasion and therapeutic resistance and have demonstrated potential as biomarkers for cancer diagnosis, prognosis, and even therapeutic targeting. However, several significant challenges hinder their clinical application. The complexity of obtaining human tissue samples, the difficulty of in vivo functional validation, and the potential toxicity and limitations of eRNA-based therapies necessitate extensive foundational research before transitioning to clinical trials. Our findings indicate that m6A-modified eRNAs tend to be longer in certain cancer cells, but whether differences exist in the transcription and function of long versus short eRNAs remains an open question. Furthermore, we strongly suspect that m6A-mediated regulation of eRNAs stability is highly context-dependent. In breast cancer MCF7 cells, m6A modifications did not significantly alter eRNAs stability, whereas in NPC, m6A modifications increased eRNAs stability, and in mESCs, m6A modifications led to eRNAs destabilization. However, a definitive conclusion regarding the influence of different cellular environments on m6A-mediated eRNAs stability has yet to be reached. Another critical yet unresolved question is whether m6A modifications exert differential effects on eRNAs stability depending on their specific modification sites. Existing evidence suggests that m6A modifications predominantly influence eRNAs stability, causing the activation of cancer-related genes, which subsequently drives cancer cell proliferation, invasion, metastasis, and drug resistance. Additionally, eRNAs can interfere with the recognition of m6A sites in mRNAs, thereby altering mRNA stability and subsequently promoting cancer metastasis. To explore the therapeutic potential of m6A-modified eRNAs, future investigation should concentrate on characterizing their advanced molecular structures to assist the exploration of more precise targeting strategies. Although the influence of m6A modifications on eRNA-mediated carcinoma advancement is of significant biological importance, current research in this area remains limited, highlighting the need for further comprehensive investigations. Beyond m6A, other RNA modifications, such as m1A, m5C, and 5-hydroxymethylcytosine (5hmC), have also been reported in eRNAs. However, the involvement of these modifications in cancer remains poorly understood and will require extensive long-term studies.

In conclusion, an in-depth investigation into the m6A modifications of eRNAs has the potential to reshape our understanding of cancer biology. The intricate interplay between enhancers, eRNAs, and m6A modifications reveals a complex yet promising frontier in epigenetics. Future research should adopt innovative, multidimensional, and interdisciplinary approaches to explore this field further. We anticipate that these efforts will lead to the development of precise epigenetic-targeted cancer therapies in the near future.

Abbreviations

eRNAs

Enhancer RNAs

m6A

N6-methyladenosine

TF

Transcription factor

TFBs

Transcription factor binding sites

Ig

Immunoglobulin

H3K4me1

Histone H3 lysine 4

H3K27ac

Histone H3 lysine 27

5mC

5-methylcytosine

ncRNAs

Non-coding RNAs

lncRNA

Long non-coding RNAs

miRNAs

microRNAs

polyA

3’polyadenylation

GRO-seq

Global run-on sequencing

PRO-seq

Precise nuclear run-on sequencing

BRD4

Bromodomain-containing protein 4

ERα

Estrogen receptor-α

P-TEFb

Positive transcription elongation factor b

CTD

C-terminal domain

WDR82

WD repeat domain 82

PP1

Phosphatase 1

CBC

Cap binding complex

NELF

Negative elongation factor

IEG

Immediate early gene

hnRNPL

Heterogeneous cytosolic nuclear ribonucleoprotein L

GTFs

General transcription factors

CRISPRi

CRISPR interference

HRRs

Highly recurrent regions

hmC

Hydroxymethylation of cytosine

LSD1/KDM1A

Lysine-specific demethylase 1

FOXD2

Forkhead box protein D2

C/EBPα

Enhancer binding protein alpha

ULBP2

UL16-binding protein 2

C/EBPα-Cm

C/EBPα C-terminal mutant

OSCC

Oral squamous cell carcinoma

HNSCC

Head and neck squamous cell carcinoma

IRS2e

IRS2-linked enhancer RNA

SCC

Squamous cell carcinoma

AKT/PI3K

Phosphatidylinositol 3-kinase

MEK1/2

Mitogen-activated protein kinase kinase 1/2

ERK1/2

Extracellular signal-regulated kinase 1/2

TME

Tumor microenvironment

IGH

Immunoglobulin heavy chain

CESC

Cervical adenocarcinoma

TNM

Tumor-lymph node-metastasis

METTL3

Methyltransferase-like protein 3

METTL14

Methyltransferase-like protein 14

WTAP

Wilms tumor 1 associated protein

SAM

S-adenosylmethionine

RRACH

Recognizing specific RNA sequences

METTL16

Methyltransferase-like protein 16

METTL5

Methyltransferase-like protein 5

ZCCHC4

Human methyltransferase

RBM15

RNA-binding motif protein 15

VIRMA/KIAA1429

Vir-like m6A methyltransferase-associated protein

3'UTR

3’ untranslated region

Fe²⁺

Ferrous iron

FTO

Fat mass and obesity-associated protein

ALKBH5

Alkylation repair homolog 5

HCC

Hepatocellular carcinoma

EMT

Epithelial-mesenchymal transition

HNRNP

Heterogeneous nuclear ribonucleoprotein

ELAVL1

Embryonic lethal abnormal visual-like protein 1

circRNA

Circular RNA

Prrc2a

Proline-rich coiled-coil 2 A

A-to-I

Adenosine-to-inosine

ADARs

Adenosine deaminases acting on RNA

NAT10

N-acetyltransferase 10

ac4C

N4-acetylcytosine

PFKM

Phosphofructokinase

LDHA

Lactate dehydrogenase A

ACOT7

Acyl-coenzyme A thioesterase 7

DCP2

Decapping enzyme 2

m7G

N7-methylguanosine

PCIF1

Phosphorylated CTD-interacting factor 1

TROP2

Trophoblast cell surface antigen 2

COAD

Colon adenocarcinoma

hnRNA

Heterogeneous nuclear RNA

GPX4

Glutathione peroxidase 4

m5C

5-methylcytosine

NSUN2

NOP2/Sun RNA methyltransferase 2

AML

Acute myeloid leukemia

LSCs

Leukemia stem cells

SRSF7

Serine/arginine-rich splicing factor 7

LC

Lung cancer

NSCLC

Non-small cell lung cancer

LUAD

Lung adenocarcinoma

ESR1

Estrogen receptorα

ESCA

Esophageal cancer

Fn

Fusobacterium nucleatum

GC

Gastric cancer

PC

Pancreatic cancer

GBE1

Glucan branching enzyme 1

HDAC4

Histone deacetylase type 4

CRC

Colorectal cancer

SOGA1

Suppressor of glucose by autophagy

PDK4

Pyruvate dehydrogenase kinase 4

RCC

Renal cell carcinoma

AURKB

Encoding aurora kinase B

BCa

Bladder cancer

TNFAIP3

TNFα-inducible protein3

DHCR7

7-dehydrocholesterol reductas

MALAT1

Metastasis-associated lung adenocarcinoma transcript 1

MAL2

Mal T-cell differentiation protein 2

PCa

Prostate cancer

CRPC

Castration-resistant prostate cancer

BC

Breast cancer

CDK6

Cell cycle protein-dependent kinase 6

ALYREF

Aly/REF export factor

NXF1

Nuclear RNA export factor 1

OC

Ovarian cancer

PIK3R1

Phosphatidylinositol-3-kinase regulatory subunit 1

ITGB1

Integrin subunit β1

FAK

Focal adhesion kinase

FSH

Follicle stimulating hormone

CC

Cervical cancer

HPV

Human papillomavirus

EC

Endometrial cancer

PDGFRB

Platelet-derived growth factor receptorβ

ECSC

Endometrial cancer stem cells

EGR1

Early growth response factor-1

NLRC5

NLR family CARD structural domain containing 5

NPC

nasopharyngeal carcinoma

TFAP2C

TF-activated enhancer binding protein 2 C

MeRIP-Seq

Methylated RNA immunoprecipitation sequencing

MINT-Seq

Methylation-inscribed nascent transcript sequencing

mESCs

Mouse embryonic stem cells

INTS11

Nucleic acid endonuclease subunit

Pol II

RNA polymerase II

PDAC

Pancreatic ductal adenocarcinoma

seRNAs

Super-enhancer-associated RNAs

CFL1

Cofilin-1

MLL1/KMT2A

Mixed-lineage leukemia protein 1

UCEC

Uterine corpus endometrial cancer

CTCF

CCCTC-binding factor

mPCa

Bone-metastatic prostate cancer

m1A

N1-methyladenosine

5hmC

5-hydroxymethylcytosine

enCRISPRi

Enhancer-targeted CRISPR interference

SNPs

Single nucleotide polymorphisms

GWAS

Genome-wide association studies

ASOs

Antisense Oligonucleotides

enCRISPRa

Enhancer-targeted CRISPR activation

IRS2

Insulin receptor substrate 2

eRNA-QTLs

eRNA quantitative trait loci

TMZ

Temozolomide

ccRCC

Clear cell renal cell carcinoma

BRD9

Bromodomain-containing protein 9

KIs

Kinase inhibitors

5-FU

5-fluorouracil

Author contributions

Conception and design of the review article: Y. H., L. M. and M. L. Interpretation of relevant literature and manuscript writing: J. S. and M. Y. The author(s) read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 32300465 and No. 82260493), Natural Science Foundation of Xinjiang Uygur Autonomous Region (Grant No. 2022D01C725) and Xinjiang Key Laboratory of Molecular Biology for Endemic Diseases (XJDFB2024G02).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.