Scientific Frontline: Extended "At a Glance" Summary: METTL3-Mediated DNMT1 Regulation in Ovarian Cancer
The Core Concept: METTL3-mediated regulation of DNMT1 is a newly identified molecular mechanism in which chemical modifications on messenger RNA (mRNA) control a DNA methylation enzyme to silence tumor-suppressor genes in ovarian cancer cells. By linking RNA modification to DNA methylation, this pathway enables aggressive ovarian cancer cells to suppress protective genes and accelerate tumor progression without altering the underlying DNA sequence.
Key Distinction/Mechanism: Rather than increasing the total amount of DNMT1 messenger RNA, the RNA-modifying enzyme METTL3 adds m6A chemical marks to DNMT1 mRNA to enhance how efficiently that RNA is translated into DNMT1 protein. Elevated DNMT1 protein levels subsequently drive DNA methylation across the genome, keeping tumor-suppressor genes switched off; whereas RNA modification and DNA methylation were previously studied as separate gene-regulatory systems, this mechanism demonstrates direct functional crosstalk between them.
Major Frameworks/Components:
- METTL3-Mediated m6A RNA Modification: METTL3 is an enzyme that deposits m6A chemical marks onto the messenger RNA encoding DNMT1, boosting DNMT1 protein production by increasing translation efficiency rather than mRNA abundance.
- DNMT1-Driven DNA Methylation: DNMT1 is an enzyme that methylates DNA without changing the genetic sequence itself, maintaining the transcriptional silencing of tumor-suppressor genes; inhibiting DNMT1 lowers genome-wide DNA methylation and increases tumor-suppressor gene activity.
- Epitranscriptomic-Epigenetic Crosstalk in HGSOC: Functional coupling between METTL3 and DNMT1 allows high-grade serous ovarian carcinoma (HGSOC) cells to exploit normal cellular gene-regulation machinery for uncontrolled proliferation and spread, as demonstrated in both cell-based colony-formation assays and mouse models where reducing METTL3 delayed tumor development and curtailed tumor growth.
Branches of Science: Epitranscriptomics, Cancer Epigenetics, Molecular Oncology, RNA Biology, and Gynecologic Oncology.
Future Application: Mapping how METTL3 controls DNMT1 establishes two interconnected molecular targets for future therapeutic development in high-grade serous ovarian carcinoma. In the long term, pharmacological strategies designed to inhibit METTL3 or DNMT1 could help suppress tumor growth, reduce aberrant DNA methylation, and restore the expression of silenced tumor-suppressor genes.
Why It Matters: High-grade serous ovarian carcinoma is the most aggressive and lethal subtype of ovarian cancer, accounting for the majority of deaths from the disease. Because symptoms are often vague and no effective screening method exists, most patients are diagnosed only after the cancer has already spread, underscoring the urgent need to uncover fundamental gene-regulatory mechanisms that can guide new treatment strategies.
An international research team led by scientists at Umeå University has mapped a molecular pathway in ovarian cancer cell lines, patient tissue specimens, and mouse xenograft models through which chemical modifications on messenger RNA (mRNA) boost the production of a DNA methylation enzyme to silence specific tumor suppressor genes. Published November 4, 2026, in Science Advances, the peer-reviewed experimental study establishes a direct post-transcriptional link between epitranscriptomic and epigenetic gene regulation in high-grade serous ovarian carcinoma (HGSOC), the most prevalent and aggressive subtype of ovarian cancer.
Led by Kanchan Kumari, a postdoctoral researcher in the Department of Molecular Biology and the Wallenberg Centre for Molecular Medicine at Umeå University and the study’s first and co-corresponding author, and Francesca Aguilo, associate professor in the same department and co-corresponding author, the team investigated the oncogenic role of the core \(N^6\)-methyladenosine (\(\text{m}^6\text{A}\)) methyltransferase complex—comprising the catalytic subunit methyltransferase 3, \(N^6\)-adenosine-methyltransferase complex catalytic subunit (METTL3; commonly known as methyltransferase-like 3) and its binding partner METTL14. Their findings demonstrate that METTL3-mediated \(\text{m}^6\text{A}\) deposition enhances the translation of DNA methyltransferase 1 (DNMT1), the primary maintenance enzyme responsible for preserving 5-methylcytosine (5mC) DNA methylation patterns and repressing tumor suppressor genes.
Linking \(\text{m}^6\text{A}\) RNA Modification to DNMT1 Translation
Nearly all somatic cells share an identical genomic sequence, yet distinct cell types express different subsets of genes. Alongside genetic alterations, gene expression is regulated by chemical modifications to DNA (epigenetic marks such as 5mC) and RNA (epitranscriptomic marks such as \(\text{m}^6\text{A}\), the most abundant internal modification on mammalian mRNA) that do not alter the underlying nucleotide sequence. Whereas DNA methylation typically sustains repressive chromatin states and transcriptional silencing, \(\text{m}^6\text{A}\) regulates mRNA stability, splicing, and translation.
Using methylated RNA immunoprecipitation sequencing (MeRIP-seq) in the HGSOC cell line OVCAR8, the researchers mapped 6,178 \(\text{m}^6\text{A}\) peaks enriched near stop codons and within coding sequences, identifying DNMT1 as a modified transcript carrying an \(\text{m}^6\text{A}\) site in exon 8 of its coding sequence (position 3013). Targeted immunoprecipitation showed an approximately 2.5-fold enrichment of \(\text{m}^6\text{A}\)-modified DNMT1 mRNA, and nucleotide-resolution SELECT assays confirmed that genetic knockdown of METTL3 or METTL14—or treatment with the selective small-molecule METTL3 inhibitor STM2457 (\(10\,\mu\text{M}\))—significantly reduced \(\text{m}^6\text{A}\) levels at that site.
Crucially, suppressing METTL3 or METTL14 markedly reduced DNMT1 protein abundance without altering DNMT1 mRNA levels or transcript stability over 48 hours of actinomycin D treatment. Re-expressing wild-type METTL3 restored DNMT1 protein levels in METTL3-depleted cells, whereas a catalytically inactive METTL3 mutant failed to rescue expression, confirming that METTL3’s methyltransferase activity is required. Dual-luciferase reporter assays carrying DNMT1 exon 8 showed that STM2457 treatment reduced translational output and that mutating the \(\text{m}^6\text{A}\) site (an A-to-C substitution) lowered baseline translation and blunted STM2457 sensitivity. Polysome profiling further confirmed reduced DNMT1 mRNA association with actively translating polysomes upon METTL3 depletion, alongside broader defects in ribosomal RNA processing and ribosome biogenesis. The team also detected physical co-immunoprecipitation between endogenous METTL3 and DNMT1 and reduced DNMT1 chromatin occupancy upon METTL3 knockdown, noting that chromatin-level recruitment may act alongside translational control.
Clinical tissue analyses supported this post-transcriptional mechanism. In public datasets (Gene Expression Omnibus GSE37582, covering 74 ovarian carcinomas and 47 normal ovarian tissues), METTL3 mRNA was significantly elevated in ovarian tumors, whereas METTL14 and DNMT1 mRNA levels showed no significant difference. In The Cancer Genome Atlas (TCGA) HGSOC cohort (\(n = 487\)), high METTL3 expression correlated with significantly shorter overall survival (log-rank \(P = 0.0168\)), whereas METTL14 did not (\(P = 0.164\)). At the protein level, however, immunohistochemical analysis of tissue microarrays comprising 180 ovarian carcinoma specimens revealed significantly higher DNMT1 protein expression in HGSOC than in low-grade serous carcinoma. Similarly, immunoblotting of 11 HGSOC patient specimens and 6 normal ovarian tissues demonstrated a statistically significant positive correlation between METTL3 and DNMT1 protein levels in tumors (\(r = 0.67\), \(P = 0.04\)), but not in normal tissue (\(r = -0.02\), \(P = 0.97\)).
Preclinical Effects in Cell Lines and Mouse Models
In OVCAR8 cells, short hairpin RNA (shRNA) knockdown of either METTL3 or METTL14 significantly impaired cell proliferation, colony formation, and migration, increased apoptosis, and induced \(\text{G}_0\)–\(\text{G}_1\) cell cycle arrest. The two enzymes also mutually stabilized one another at the protein level, with METTL3 depletion triggering proteasomal degradation of METTL14 and lowering levels of other writer complex subunits, including WTAP and VIRMA. Pharmacological METTL3 inhibition with STM2457 produced comparable though milder reductions in proliferation, colony formation, and migration, shifted cells toward S-phase accumulation, and yielded a statistically nonsignificant increase in apoptosis—consistent with METTL3 exerting both \(\text{m}^6\text{A}\)-dependent and \(\text{m}^6\text{A}\)-independent effects.
To test tumor progression in vivo, the researchers subcutaneously injected \(1 \times 10^6\) luciferase-labeled OVCAR8 cells into 6- to 7-week-old female Crl:NU(NCr)-Foxn1nu nude mice (10 tumors per experimental group). Mice receiving control cells developed tumors rapidly, with a median tumor-free survival of 12 days. By contrast, depleting METTL3 or METTL14 extended median tumor-free survival to 21 days and 28 days, respectively. Bioluminescence imaging through day 48 and tumor excisions at 8 weeks confirmed markedly slower tumor growth, lower final tumor volume and weight, and reduced DNMT1 protein levels in both knockdown groups.
Downstream Silencing of TNFAIP3 and FBXO32 and Functional Divergence
To dissect DNMT1’s specific contribution downstream of METTL3, the researchers treated OVCAR8 cells with GSK3484862 (\(10\,\mu\text{M}\)), a selective inhibitor that promotes proteasome-dependent degradation of DNMT1 without altering its mRNA. GSK3484862 induced widespread genome-wide DNA hypomethylation—altering 11,141 hypomethylated CpG sites and zero hypermethylated sites on methylation microarrays—and suppressed proliferation and migration while inducing apoptosis and S-phase arrest. Integrating RNA sequencing (222 differentially expressed genes: 196 upregulated and 26 downregulated) with methylation profiling identified 51 transcripts (26%) that were both hypomethylated and transcriptionally reactivated.
Cross-referencing those candidates with METTL3-knockdown transcriptomes highlighted two well-characterized tumor suppressor genes that are not themselves \(\text{m}^6\text{A}\)-modified: TNFAIP3 (tumor necrosis factor alpha-induced protein 3) and FBXO32 (F-box protein 32). Both genes were downregulated in ovarian cancer datasets, and either METTL3 knockdown, STM2457 treatment, or DNMT1 depletion elevated TNFAIP3 and FBXO32 protein levels. Conversely, re-expressing DNMT1 in METTL3-depleted cells restored repression of both tumor suppressors. The METTL3–DNMT1–TNFAIP3/FBXO32 regulatory axis was conserved across six additional human cell lines representing ovarian (OVCAR4), breast (MDA-MB-231), pancreatic (SUIT2, MiaPaCa2), osteosarcoma (U2OS), and embryonic kidney (HEK293T) lineages.
Orthogonal genetic and rescue experiments revealed a critical mechanistic distinction between migration and proliferation. Unlike acute pharmacological degradation with GSK3484862, shRNA knockdown of DNMT1 suppressed cell migration without reducing proliferation or inducing apoptosis. Likewise, re-expressing DNMT1 in METTL3-depleted or STM2457-treated cells significantly rescued cell migration but failed to restore proliferation or reverse apoptosis. Co-silencing TNFAIP3 and FBXO32 via small interfering RNAs rescued the antiproliferative effect of STM2457 but did not overcome the broader growth arrest triggered by GSK3484862. Together, these experiments show that DNMT1 specifically mediates METTL3-dependent cell migration and TNFAIP3/FBXO32 repression, whereas METTL3’s wider effects on proliferation and survival rely on additional downstream \(\text{m}^6\text{A}\) targets.
Notably, unlike GSK3484862, STM2457 treatment did not produce detectable genome-wide 5mC hypomethylation on methylation arrays, indicating that residual DNMT1 following partial METTL3 inhibition remains above the threshold required to preserve bulk DNA methylation. The authors noted that partial DNMT1 reduction may derepress TNFAIP3 and FBXO32 by disrupting noncatalytic DNMT1 scaffolding with repressive chromatin complexes such as LSD1 or EZH2 prior to bulk 5mC loss, a hypothesis requiring direct experimental testing.
Clinical Context, Limitations, and Translational Outlook
Ovarian cancer is the most lethal gynecological malignancy and the fifth leading cause of cancer-related death among women, with HGSOC frequently diagnosed only after widespread metastasis has occurred. While DNA methyltransferase inhibitors such as 5-azacytidine and 5-aza-2′-deoxycytidine are approved by the U.S. Food and Drug Administration for hematologic malignancies, effective targeted epigenetic or epitranscriptomic therapies for HGSOC remain undeveloped.
Several caveats limit immediate clinical translation. First, the mechanistic and efficacy findings derive from established cancer cell lines and subcutaneous xenografts in immunodeficient nude mice rather than orthotopic, immunocompetent, or patient-derived clinical models. Second, although trends toward improved survival with higher TNFAIP3 and FBXO32 expression and negative protein correlations with DNMT1 and METTL3 appeared in patient cohorts, those associations did not reach statistical significance in the analyzed TCGA and 11-tumor clinical immunoblot sets. Third, the study did not establish which specific \(\text{m}^6\text{A}\) reader protein—if any—mediates enhanced DNMT1 exon-8 translation, and because both METTL3 and DNMT1 perform essential physiological roles in healthy somatic tissues, systemic inhibition carries toxicity risks that must be evaluated before clinical testing.
Additional information: In addition to Umeå University, the study involved collaborators from the Hospital del Mar Medical Research Institute, the Josep Carreras Leukaemia Research Institute, the University of Vic–Central University of Catalonia, the Sant Pau Research Institute, the Institució Catalana de Recerca i Estudis Avançats, and the University of Barcelona in Catalonia; the Centro de Investigación Biomédica en Red Cáncer in Madrid and the University of Extremadura in Badajoz, Spain; the University of Trento in Italy; and the University of Chicago in the United States.
Funding: Research funding was provided by the Kempe Foundation, Swedish Research Council Formas, Fundación Científica Asociación Española Contra el Cáncer, Cancerfonden, the Cancer Research Foundation in Northern Sweden, the Knut and Alice Wallenberg Foundation, Umeå University, and the Västerbotten County Council.
Disclosure: Co-author Toni Celià-Terrassa is a cofounder of VIOLET Pharmaceuticals, which develops LCOR-based cancer therapies; all other authors declared no competing interests. Next-generation sequencing and DNA methylation datasets are publicly archived in ArrayExpress (E-MTAB-17353) and GEO (GSE299247), respectively.
Title: METTL3-mediated m6A modification of DNMT1 enhances ovarian cancer progression
Authors: Kanchan Kumari, Iván Pérez-Núñez, Aleix Noguera Castells, Raquel Pluvinet, Carlos Peula, Annika Patthey, Ghazala Naz Khan, Kinga Pajdzik, Ángel-Carlos Román, Erik Dassi, Chuan He, Eva Lundin, Manel Esteller, Toni Celià-Terrassa, and Francesca Aguilo
Source/Credit: Umeå University
Edited by: Scientific Frontline
Reference Number: epi111926_01
