N6-Methyladenosine (m6A)

N⁶-Methyladenosine (m⁶A) is a reversible post-transcriptional chemical modification in which a methyl group (−CH3) is covalently attached to the exocyclic N6 amino nitrogen of adenosine within ribonucleic acid (nucleotides). Accounting for more than 80% of all internal base methylations in eukaryote polyadenylated messenger RNA (mRNA), an average mammalian transcript contains three to five m⁶A residues—corresponding to ~0.1% to 0.4% of total adenosines across the cellular transcriptome. Rather than occurring randomly, m⁶A is co-transcriptionally deposited within the degenerate five-nucleotide consensus sequence DRACH (where D = A, G, or U; R = A or G; A = the methylated adenosine; C = cytidine; and H = A, C, or U), clustering preferentially near stop codons, within long internal exons, in 3′ untranslated regions (3′ UTRs), and along architectural super-enhancer RNAs (seRNAs).

Analogous to how DNA methylation and histone modifications establish reversible epigenetics states on chromatin, m⁶A forms the central pillar of the epitranscriptome. Its cellular lifecycle is governed by three classes of regulatory proteins: writers (the METTL3–METTL14–WTAP methyltransferase complex that transfers the methyl group from S-adenosyl-L-methionine), erasers (the Fe²⁺ and α-ketoglutarate-dependent dioxygenases FTO and ALKBH5 that oxidatively demethylate m⁶A back to unmodified adenosine), and readers (such as the YTH-domain family YTHDF1/2/3 and YTHDC1 that recognize methylated transcripts to control alternative splicing, nuclear export, translation efficiency, mRNA decay, and transcriptional condensate assembly).

The METTL3–METTL14–WTAP Writer Complex and Exon Junction Exclusion

In mammalian nuclei, co-transcriptional installation of m⁶A onto nascent pre-mRNA synthesized by RNA Polymerase II is catalyzed by a ~1,000 kDa multiprotein methyltransferase holocomplex (MAC + MACOM). The catalytic core is an obligate heterodimer of METTL3 (Methyltransferase-Like 3) and METTL14:

• METTL3 houses the functional Rossmann-fold catalytic pocket equipped with the conserved DPPW motif (Asp395–Pro396–Pro397–Trp398) that binds the methyl donor cofactor S-adenosyl-L-methionine (SAM) and positions the target adenosine N6 amine for nucleophilic attack.
• METTL14 possesses a degenerate, catalytically inactive SAM-binding pocket, but provides an essential positively charged concave groove (Together with METTL3 CCCH zinc-finger domains) that binds the RNA substrate and enforces sequence specificity for the DRACH motif (most frequently GGACU, AGACU, and GAACU).
• Accessory Subunits (MACOM): WTAP (Wilms' tumor 1-associating protein) anchors the METTL3–METTL14 heterodimer in nuclear speckles, while VIRMA (KIAA1429), ZC3H13, RBM15/RBM15B (which directs methylation of the X-inactive specific transcript XIST), and the E3 ubiquitin ligase HAKAI guide site-specific recruitment.

Although the 5-nt DRACH sequence occurs once every ~50–60 nucleotides across pre-mRNA, fewer than 5% of genomic DRACH motifs are methylated in vivo. High-resolution mapping demonstrated that this sharp topological restriction—specifically the enrichment of m⁶A in long internal exons (>400 nt) and near stop codons—is enforced by Exon Junction Complex (EJC) exclusion. During co-transcriptional spliceosome assembly, the EJC is deposited ~24 nt upstream of every exon–exon boundary, physically sterically blocking METTL3–METTL14 from accessing DRACH sites within ~100–150 nt of splice junctions across average-length (~140 nt) exons.

Oxidative Demethylation by FTO and ALKBH5 Erasers

The discovery by Chuan He and colleagues in 2011 and 2013 that FTO (Fat Mass and Obesity-Associated protein) and ALKBH5 (AlkB Homolog 5) remove m⁶A established that RNA methylation is dynamically reversible. Both enzymes belong to the non-heme Fe(II)/α-ketoglutarate (2-oxoglutarate)-dependent AlkB dioxygenase superfamily:

1. ALKBH5 localizes to nuclear speckles and catalyzes direct oxidative demethylation of internal m⁶A on mRNA and seRNAs, releasing formaldehyde and restoring unmodified adenosine; ALKBH5 activity is essential for proper mRNA export and mammalian spermatogenesis.
2. FTO oxidizes m⁶A through short-lived N⁶-hydroxymethyladenosine (hm6A) and N⁶-formyladenosine (f6A) intermediates (with half-lives of ~3 hours in physiological buffer), and exhibits even higher catalytic efficiency (k_cat/K_m) toward m6Am (N⁶,2′-O-dimethyladenosine) located at the first transcribed nucleotide adjacent to the 5′ m7G cap.

YTH-Domain Readers and the Trp Aromatic Cage

Downstream biological outcomes of m⁶A are executed by reader proteins, foremost the five human YTH (YT521-B homology) domain proteins: nuclear YTHDC1, nucleocytoplasmic YTHDC2, and cytoplasmic YTHDF1, YTHDF2, and YTHDF3. X-ray crystal and NMR structures of the ~150-amino-acid YTH domain reveal a conserved globular α/β fold containing a hydrophobic aromatic cage built from three tryptophan residues (or two tryptophans and one tyrosine/leucine, such as Trp377, Trp428, and Leu439 in YTHDC1; Trp411, Trp465, and Trp470 in YTHDF1).

When an m⁶A-modified DRACH motif docks into the YTH pocket, the N6-methyl group inserts directly into the tryptophan cage, forming favorable CH–π dispersion contacts and van der Waals interactions that elevate binding affinity 20-fold to 50-fold (K_d ≈ 0.1–1.5 μM for methylated RNA versus >20–50 μM for unmethylated RNA). Outside the globular YTH domain, each reader carries low-complexity intrinsically disordered regions (IDRs):

• YTHDC1 remains in the nucleus, where it recruits the splicing factor SRSF3 (while displacing SRSF10) to promote exon inclusion, facilitates nuclear export via SRSF3–NXF1, and undergoes phase separation on methylated seRNAs to stabilize transcriptional condensate droplets.
• YTHDF1, YTHDF2, and YTHDF3 operate in the cytoplasm, where YTHDF2 (and redundantly YTHDF1/YTHDF3 in unified dosage models) recruits the CCR4–NOT deadenylase complex (via direct binding to the CNOT1 SH domain) and the HRSP12–RNase P/MRP endonuclease pathway to accelerate mRNA deadenylation and turnover in P-bodies, while YTHDF1 interacts with eukaryotic initiation factor eIF3 to stimulate cap-dependent translation in specific cellular contexts.

Thermodynamic Destabilization of RNA Duplexes (The m6A-Switch)

Although N6-methylation does not alter the canonical Watson–Crick hydrogen-bonding stoichiometry of an A•U base pair (one bond between N1 and H3–N3; one bond between N6–H and O4), optical melting and NMR relaxation dispersion experiments by Eric Kool, Tao Pan, and colleagues revealed a fundamental thermodynamics effect: m⁶A destabilizes RNA duplexes while stabilizing single-stranded stacked regions.

In single-stranded RNA, the N6-methyl group of m⁶A adopts the energetically preferred syn rotamer (with the methyl group pointing toward the Watson–Crick N1 edge and away from the N7 imidazole ring), which enhances hydrophobic stacking against adjacent bases (ΔΔG° ≈ −0.4 to −0.7 kcal mol⁻¹ at 37 °C). However, to form a Watson–Crick m6A•U base pair inside an A-form double-stranded stem, the N6-methyl group must rotate 180° around the C6–N6 bond into the sterically hindered anti rotamer, bringing the methyl protons into a high-energy steric clash with the N7 nitrogen. This conformational penalty destabilizes every m6A•U base pair by +0.5 to +1.7 kcal mol⁻¹ (lowering hairpin melting temperatures by 2–6 °C per modification). Cells exploit this m6A-switch mechanism: methylating a DRACH site inside an RNA stem-loop locally melts the duplex to expose buried single-stranded poly(U) tracts, enabling heterogeneous nuclear ribonucleoproteins (hnRNPC, hnRNPG, and IGF2BP1/2/3) to bind the newly unfolded transcript.

Antibody-Independent Mapping and Direct RNA Nanopore Sequencing

Early transcriptome-wide m⁶A maps (MeRIP-seq / m6A-seq) relied on anti-m⁶A immunoprecipitation of ~100-nt RNA fragments, which lacked single-nucleotide resolution and cross-reacted with cap-adjacent m6Am. Second-generation biochemical methods achieved single-base precision:

• miCLIP (UV cross-linking of anti-m⁶A antibodies to induce C-to-T transitions or truncations at the +1 position during reverse transcription).
• GLORI (glyoxal and nitrite-mediated deamination of unmodified adenosines to inosine while m⁶A remains protected, reading out absolute stoichiometric m⁶A percentages via standard sequencing).
• eTAM-seq (enzymatic deamination of unmodified adenosines by an evolved TadA8.20 deaminase).
• Direct RNA Nanopore Sequencing (ONT + Modkit / Dorado): Native RNA molecules translocated through engineered biological nanopores (CsgG/Remora models) generate distinct ionic current shifts (ΔpA) and dwell-time signatures when an m⁶A-modified DRACH k-mer passes through the pore constriction, enabling single-molecule, isoform-resolved calling of m⁶A stoichiometry alongside poly(A) tail length and co-occurring RNA modifications without reverse transcription or PCR amplification.

Why does m6A weaken double-stranded RNA stems if it still forms two hydrogen bonds with uracil?

In an unpaired single strand, the N6-methyl group of m⁶A sits in the low-energy syn conformation pointing toward the Watson–Crick N1 atom. To form a base pair with uracil (m6A•U), the methyl group must rotate into the high-energy anti conformation so the remaining N6–H proton can hydrogen-bond to uracil O4. In the anti orientation, the bulky methyl group clashes sterically with the purine N7 nitrogen, imposing a thermodynamic penalty of +0.5 to +1.7 kcal mol⁻¹ per base pair that locally unfolds RNA hairpins.

Why are only a small fraction of DRACH motifs methylated in mammalian mRNAs?

Although the five-nucleotide DRACH consensus occurs roughly once every 50 bases across pre-mRNA, the spliceosome deposits an Exon Junction Complex (EJC) ~24 nucleotides upstream of every spliced exon–exon boundary. The bulky EJC physically shields ~100–150 nucleotides on either side of the splice junction from the METTL3–METTL14 writer complex. As a result, average-length internal exons (~140 nt) are protected from methylation, restricting m⁶A deposition to long internal exons (>400 nt), terminal exons near the stop codon, and unspliced non-coding RNAs.

How does METTL3 inhibition act as a therapeutic strategy in acute myeloid leukemia (AML)?

In acute myeloid leukemia (AML), METTL3 is recruited by the transcription factor CEBPZ to the promoters of oncogenic transcription factors (MYC, BCL2, SP1, MYB), depositing m⁶A on their nascent transcripts to sustain high translation and block myeloid differentiation. Orally bioavailable small-molecule inhibitors of the METTL3 SAM-binding pocket (such as STM2457 and clinical candidate STC-15) deplete m⁶A on these leukemogenic transcripts, inducing terminal myeloid differentiation and apoptosis of leukemia stem cells while sparing normal hematopoietic progenitors.

1. Dominissini, D., Moshitch-Moshkovitz, S., Schwartz, S., Salmon-Divon, M., Ungar, L., Osenberg, S., Cesarkas, K., Jacob-Hirsch, J., Amariglio, N., Kupiec, M., Sorek, R., & Rechavi, G. (2012). "Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq." Nature, 485(7397), 201–206. https://doi.org/10.1038/nature11112
2. Meyer, K. D., Saletore, Y., Zumbo, P., Elemento, O., Mason, C. E., & Jaffrey, S. R. (2012). "Comprehensive analysis of mRNA methylation reveals enrichment in 3' UTRs and near stop codons." Cell, 149(7), 1635–1646. https://doi.org/10.1016/j.cell.2012.05.003
3. Roost, C., Lynch, S. R., Batista, P. J., Qu, K., Chang, H. Y., & Kool, E. T. (2015). "Structure and thermodynamics of N6-methyladenosine in RNA: a spring-loaded base modification." Journal of the American Chemical Society, 137(5), 2107–2115. https://doi.org/10.1021/ja513080v
4. Uzonyi, A., Dierks, D., Nir, R., Kwon, O. S., Toth, U., Barbosa, I., Burel, C., Brandis, A., Rossmanith, W., Le Hir, H., Slobodin, B., & Schwartz, S. (2023). "Exclusion of m6A from splice-site proximal regions by the exon junction complex dictates m6A topologies and mRNA stability." Molecular Cell, 83(2), 237–251. https://doi.org/10.1016/j.molcel.2022.12.026