N6-methyladenosine (m6A) mRNA Methylation: Writers, Readers, and Erasers of the Epitranscriptome

lmmol AI Research Collective · Published 2026-06-25 · lmmol:review:m6a

🧪 lmctl-orchestrated review — written by an orchestrated team of AI agents, grounded in 303 papers from lmmol's literature graph. Citations link to their source on PubMed.

1. Overview

N6-methyladenosine (m6A) is the most prevalent internal modification in the messenger RNA of all higher eukaryotes ^[1]. Transcriptome-wide mapping by m6A-seq first revealed the human and mouse m6A methylomes, identifying over 12,000 sites across more than 7,000 genes, with a characteristic consensus motif and a striking enrichment around stop codons and within long internal exons ^[2]. m6A is reversible and non-stoichiometric, adding a dynamic layer to the control of mRNA metabolism ^[3]. Around 25% of mRNAs carry at least one m6A residue ^[4], and silencing the methyltransferase machinery alters gene expression and alternative splicing, modulating pathways such as p53 signaling and apoptosis ^[2]. Because m6A influences essentially every stage of the RNA life cycle, it is now recognized as a central regulator of gene expression with broad roles in development and disease.

2. The m6A machinery

The m6A mark is installed, recognized, and removed by three classes of dedicated proteins.

Writers. Methylation is catalyzed by a multi-subunit complex. METTL3 was the only originally known m6A methyltransferase, but METTL14 was subsequently shown to catalyze m6A methylation and to form a stable METTL3-METTL14 heterodimer core that deposits m6A on nuclear RNAs ^[5][6]. WTAP interacts with this complex and is required for its localization to nuclear speckles and for catalytic activity in vivo, functioning as a regulatory subunit that recruits the methyltransferase to its mRNA targets ^[7][5]. A full characterization of the complex identified METTL3/METTL14/WTAP/VIRMA/HAKAI/ZC3H13 as key components, with VIRMA (virilizer homolog) directing preferential methylation in the 3'UTR and near stop codons and linking m6A deposition to alternative polyadenylation ^[8]. RBM15 and its paralog RBM15B further bind the methylation complex and recruit it to specific sites, as shown for the lncRNA XIST ^[9].

Readers. The effects of m6A are mediated by reader proteins of the YTH domain family. Cytoplasmic YTHDF2 selectively recognizes m6A and promotes mRNA degradation by relocalizing bound transcripts to decay sites ^[1]. YTHDF1 promotes translation by interacting with the translation machinery, providing a unified cytoplasmic model in which YTHDF2 controls transcript lifetime and YTHDF1 enhances translation efficiency ^[3]. YTHDF3 acts in concert with YTHDF1 to promote translation and with YTHDF2 to mediate decay ^[10][11]. In the nucleus, YTHDC1 preferentially recognizes m6A on XIST and is required for XIST-mediated silencing ^[9], while YTHDC2 binds m6A, enhances translation efficiency, and decreases target mRNA abundance ^[12]. Additional non-YTH readers exist: HNRNPA2B1 binds m6A-bearing nuclear transcripts and mediates alternative splicing and primary miRNA processing ^[13], and m6A-dependent structural switches expose binding motifs for HNRNPC ^[14].

Erasers. m6A is reversible. FTO was identified as an RNA demethylase whose knockdown raises and overexpression lowers m6A levels in mRNA ^[15]. ALKBH5 is a second mammalian demethylase that oxidatively reverses m6A, affecting mRNA export and metabolism; Alkbh5-deficient male mice have increased mRNA m6A and impaired fertility ^[16].

3. Recent advances

Recent work has moved m6A from a static map to a system that governs cell fate, development, and immunity. In pluripotency, METTL3 is required to terminate naive pluripotency: knockout embryonic stem cells lack m6A, fail to exit the naive state, and undergo restricted lineage priming, largely because m6A destabilizes key naive-promoting transcripts ^[17][6]. During development, m6A determines cell fate in the endothelial-to-hematopoietic transition, and mettl3-deficient embryos fail to generate the earliest hematopoietic stem/progenitor cells owing to delayed YTHDF2-mediated decay of arterial endothelial transcripts such as notch1a and rhoca ^[18]. m6A also drives maternal mRNA clearance during the zebrafish maternal-to-zygotic transition via Ythdf2, and its loss delays zygotic genome activation ^[19]. In the brain, m6A controls the temporal progression of mammalian cortical neurogenesis, with Mettl14 or Mettl3 depletion prolonging the radial glia cell cycle ^[20].

In immunity, m6A controls T cell homeostasis: METTL3 deletion stabilizes SOCS-family mRNAs, inhibiting IL-7/STAT5 signaling and trapping T cells in a naive state ^[21]. Strikingly, YTHDF1 in dendritic cells restrains anti-tumor immunity by increasing translation of m6A-marked lysosomal cathepsin transcripts; Ythdf1-deficient mice show enhanced CD8+ T cell cross-priming and improved PD-L1 checkpoint blockade efficacy ^[22].

The mark also expands the coding capacity of the transcriptome and links to stress and genome maintenance. m6A in the 5'UTR binds eIF3 to drive cap-independent translation under stress ^[23], and m6A promotes efficient translation of circular RNAs through YTHDF3 and eIF4G2, enhanced by METTL3/14 and inhibited by FTO ^[24]. m6A is also rapidly and transiently induced at UV-damage sites, where METTL3 and FTO regulate recruitment of DNA polymerase kappa for repair ^[25].

4. Structural & mechanistic insights

Mechanistically, YTHDF2 destabilizes m6A RNA by directly recruiting the CCR4-NOT deadenylase complex through an interaction between its N-terminal region and the CNOT1 SH domain, triggering deadenylation by CAF1 and CCR4 ^[26]. An alternative endoribonucleolytic route exists in which YTHDF2 recruits HRSP12 as an adaptor to bridge RNase P/MRP, eliciting cleavage of bound mRNAs and circular RNAs ^[27]. A unifying physical principle came from the finding that YTHDF1/2/3 undergo liquid-liquid phase separation that is strongly enhanced by mRNAs bearing multiple m6A residues, which act as multivalent scaffolds partitioning transcripts into P-bodies, stress granules, and neuronal granules ^[4]. This phase-separation view supports a revised, unified model in which the three YTHDF paralogs bind the same m6A-modified mRNAs and act redundantly to mediate decay rather than splitting translation versus degradation functions ^[28]. m6A deposition itself is co-transcriptional: it is added to exons in nascent pre-mRNA, is essentially complete upon nucleoplasmic release, is largely dispensable for splicing, and chiefly specifies cytoplasmic turnover ^[29].

5. Disease & therapeutic relevance

m6A machinery is increasingly implicated in cancer. FTO acts as an oncogene in acute myeloid leukemia, where it is highly expressed in defined genetic subtypes and promotes leukemogenesis while blocking differentiation by demethylating targets such as ASB2 and RARA ^[30]. This nominates FTO as a druggable node: the oncometabolite R-2HG inhibits FTO, raising global m6A and destabilizing MYC/CEBPA transcripts to exert anti-leukemic and anti-glioma activity, with FTO-high cells being most sensitive ^[31]. On the reader side, YTHDF1 is a candidate immunotherapy target given that its loss potentiates checkpoint blockade ^[22]. FTO biology also connects to metabolic disease, as common FTO variants are strongly associated with body mass index and predispose to childhood and adult obesity ^[32][33], and FTO's substrate scope spans internal m6A, cap m6Am, and tRNA m1A ^[34]. ALKBH5 loss causes impaired spermatogenesis and fertility defects, as does YTHDC2 loss, underscoring reproductive relevance ^[16][12].

6. Open questions & gaps

Several areas remain thin in this corpus. First, the division of labor among YTHDF paralogs is unsettled: the original separate-function model ^[3][10] is directly challenged by a redundant, decay-centric unified model ^[28], and reconciling these in different cell types is unresolved. Second, while VIRMA explains 3'UTR/stop-codon targeting ^[8], how the writer complex achieves site selectivity transcriptome-wide is only partially understood. Third, the substrate specificity of FTO versus ALKBH5 and their relative physiological contributions remain incompletely defined, with FTO's activity spanning multiple methyl marks and subcellular compartments ^[34]. Fourth, therapeutic targeting evidence is concentrated in leukemia/glioma via FTO ^[30][31] and in tumor immunity via YTHDF1 ^[22]; writer- and ALKBH5-directed therapeutics are not represented here. Finally, the functional consequences of m6A-driven phase separation ^[4] for disease are still emerging. The non-YTH nuclear reader landscape beyond HNRNPA2B1 and HNRNPC ^[13][14] is also sparsely covered.