Nuclear, Fe(II)- and 2-oxoglutarate-dependent oxygenase of the AlkB superfamily that oxidizes methyl groups on single-stranded nucleic acids. Its AlkB-like catalytic domain is paired with a C-terminal domain of novel fold, and an active-site loop excludes duplex substrates, so FTO acts only on single-stranded DNA and RNA. The best-supported cellular substrates are methylated adenosines in small RNAs and mRNA: N6,2'-O-dimethyladenosine (m6Am) at the transcription-start nucleotide of snRNAs and of capped mRNAs, N6-methyladenosine in U6 snRNA, and N1-methyladenosine in tRNA, the last of which represses translation. FTO can also oxidize internal N6-methyladenosine in mRNA and, in vitro, repair-type lesions such as 3-methylthymine in ssDNA and 3-methyluracil in ssRNA. Two aspects of its activity are actively disputed. Chemically, isolated FTO appears to stop at the hemiaminal N6-hydroxymethyladenosine rather than releasing formaldehyde itself, so demethylation may complete non-enzymatically, unlike the related ALKBH5 and AlkB enzymes. Biologically, whether internal mRNA m6A is a physiological substrate is contested: antibody-based transcriptome mapping supports it, while quantitative direct-RNA sequencing finds no change in mRNA m6A stoichiometry on FTO loss, with snRNA m6Am rising instead. FTO is mainly nuclear, concentrating in nuclear speckles, with a cell-type-dependent cytoplasmic pool. Complete loss of catalytic activity causes a lethal recessive polymalformation syndrome in humans, and the protein is widely studied as a candidate target in acute myeloid leukemia. Common variants in FTO intron 1 give the strongest known common-variant association with body mass index, but that association acts largely by altering enhancer-driven expression of the distant IRX3 and IRX5 genes rather than by altering FTO protein function.
Definition: Catalysis of the hydroxylation of the N6-methyl group of N6-methyladenosine within an RNA molecule, using Fe(II), 2-oxoglutarate and O2, to give N6-hydroxymethyladenosine, with concomitant formation of succinate and CO2. The hemiaminal product may subsequently fragment non-enzymatically to adenosine and formaldehyde.
Justification: GO currently offers no term for a nucleic-acid m6A hydroxylase whose enzymatic product is the hemiaminal rather than the demethylated base. Both terms available for FTO assert methyl removal by the enzyme: GO:0035515 oxidative RNA demethylase activity is defined as "Catalysis of the removal of a methyl group", and GO:1990931 mRNA N6-methyladenosine dioxygenase activity as releasing "oxidized methyl group on N6-methyladenosine as formaldehyde". A 2025 comparative study of FTO, ALKBH5, ALKBH2/3 and bacterial AlkB finds that isolated FTO differs from all of the others in stopping at N6-hydroxymethyladenosine. If that behaviour is confirmed in cells, FTO would need a term distinguishing it from bona fide demethylases, and the hemiaminal would become a curatable RNA modification in its own right. Proposed now so the distinction is recorded, not asserted for FTO here.
Parent term: 2-oxoglutarate-dependent dioxygenase activity
Supporting Evidence:
Definition: Catalysis of the oxidative removal of the N6-methyl group from N6,2'-O-dimethyladenosine (m6Am) at the transcription-start nucleotide of a small nuclear RNA, yielding 2'-O-methyladenosine, with 2-oxoglutarate and O2 as co-substrates.
Justification: GO has terms for m6A dioxygenase activity on mRNA (GO:1990931) and for tRNA demethylase activity (GO:1990984), but no term for the m6Am cap-adjacent substrate, in either snRNA or mRNA - a QuickGO search for m6Am returns nothing. This is a curation gap precisely where the FTO evidence is strongest and least disputed: snRNA m6Am is the substrate whose stoichiometry rises reproducibly and by large margins on FTO depletion, in the same experiments that find no change in mRNA m6A. Without such a term, the best-supported activity of FTO can only be annotated with the substrate-generic GO:0035515, while the contested mRNA activity has a precise term of its own - an asymmetry that misrepresents the evidence.
Parent term: oxidative RNA demethylase activity
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005634 nucleus | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic assertion that FTO acts in the nucleus, matching direct human and mouse evidence. Reason: Nucleus is the dominant compartment for FTO in every study that has looked, and is independently supported on this gene by EXP, IDA and ISS rows. The IBA donor set includes mouse Fto and FTO itself, which is expected when the target carries its own experimental annotation for the term. Supporting Evidence: PMID:17991826 Consistent with a potential role in nucleic acid demethylation, Fto localizes to the nucleus in transfected cells. |
| GO:0006307 DNA alkylation repair | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Inherited AlkB-family alkylation-repair role, propagated from the FTO/Fto node. Reason: FTO does demethylate 3-methylthymine in ssDNA in vitro, which is the ancestral AlkB activity and a reasonable node-level inference. But FTO is excluded from double-stranded DNA by its active-site loop, and UniProt explicitly records that the ability to repair alkylated DNA and RNA is "however unsure in vivo". Retained as a genuine but peripheral capability, not as a core function. Supporting Evidence: PMID:18775698 We demonstrate here the oxidative demethylation of 3-methylthymine (3-meT) in single-stranded DNA (ssDNA) and 3-methyluracil (3-meU) in single-stranded RNA (ssRNA) by recombinant human FTO protein in vitro. |
| GO:0010883 regulation of lipid storage | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Phylogenetic transfer of an adiposity role, ultimately grounded in mouse Fto data. Reason: Mouse Fto loss- and gain-of-function affect fat mass, so a node-level adiposity role is defensible. But this is an organismal consequence of an RNA-modifying enzyme, several steps downstream of anything the protein does directly, so it is not a core function. See the IMP row for this term for a separate concern about its human evidence. |
| GO:0042245 RNA repair | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Inherited AlkB-family RNA alkylation-repair role. Reason: Same standing as the DNA alkylation repair IBA - the 3-methyluracil ssRNA activity is real in vitro and is the ancestral family function, but no in vivo repair role has been demonstrated for FTO. Kept, not core. Supporting Evidence: PMID:18775698 We demonstrate here the oxidative demethylation of 3-methylthymine (3-meT) in single-stranded DNA (ssDNA) and 3-methyluracil (3-meU) in single-stranded RNA (ssRNA) by recombinant human FTO protein in vitro. |
| GO:0035516 broad specificity oxidative DNA demethylase activity | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Phylogenetic assertion of the ancestral AlkB-type broad-specificity demethylase activity. Reason: Directionally right but a loose fit to the term definition, which specifies N1-methyladenine, N3-methylcytosine, N1-methylguanine and N3-methylthymine. FTO has no activity towards 1-methylguanine and none towards double-stranded DNA, and only low activity on ssDNA 1-methyladenine or 3-methylcytosine. Kept as a non-core residual family activity. |
| GO:1990931 mRNA N6-methyladenosine dioxygenase activity | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Phylogenetic assertion that the FTO clade acts on mRNA m6A. Reason: This is the contested substrate claim, and it is contested at the cellular rather than the biochemical level: isolated FTO clearly oxidizes m6A in single-stranded RNA, so the node placement is not unreasonable. What is disputed is whether internal mRNA m6A is a physiological substrate. Quantitative direct-RNA nanopore sequencing reports no change in mRNA m6A stoichiometry on FTO knockout, knockdown or inhibition, in AML and non-AML cells, while the same cells show the expected large rise in snRNA m6Am - an internal positive control that FTO activity really was lost. That evidence is currently a bioRxiv preprint (PMID:41279954), so it is treated here as sufficient to demote the claim from core, not to remove it. GOA already hedges the substrate question by carrying both this term and GO:1990984 tRNA demethylase activity; this review makes that hedge explicit. Supporting Evidence: PMID:41279954 We find that the stoichiometry of m6A sites throughout the transcriptome and especially at MYC-specific sites are unaffected despite depletion of FTO activity by knockout, knockdown, or pharmacologic inhibition. |
| GO:0005634 nucleus | IEA GO_REF:0000120 | ACCEPT | Summary: Electronic assignment of nuclear localization. Reason: Agrees with the direct human experimental evidence on this gene. |
| GO:0005737 cytoplasm | IEA GO_REF:0000044 | ACCEPT | Summary: Electronic assignment of a cytoplasmic pool from the UniProt subcellular location vocabulary. Reason: A real, cell-type-dependent cytoplasmic pool is documented directly, and it is where FTO reaches cytoplasmic tRNA. Supporting Evidence: PMID:30197295 As shown in Figure 2A, the relative proportion of FTO within the nucleus and cytoplasm varies across cell lines. |
| GO:0006307 DNA alkylation repair | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: Electronic transfer of the alkylation-repair role from mouse Fto and the InterPro FTO signature. Reason: Same judgement as the IBA and IDA rows for this term - an in vitro capability with no demonstrated in vivo repair role for FTO. |
| GO:0016607 nuclear speck | IEA GO_REF:0000044 | ACCEPT | Summary: Electronic assignment of nuclear speckle localization. Reason: Directly observed for human FTO by immunofluorescence. Supporting Evidence: PMID:22002720 We further show the partial colocalization of FTO with nuclear speckles, which supports the notion that m(6)A in nuclear RNA is a major physiological substrate of FTO. |
| GO:0035515 oxidative RNA demethylase activity | IEA GO_REF:0000120 | ACCEPT | Summary: Electronic assignment of the core RNA demethylase activity. Reason: Consistent with the multiple IDA rows for this term on this gene, and substrate-generic, so it is not affected by the dispute over which RNA class FTO acts on in cells. |
| GO:0035516 broad specificity oxidative DNA demethylase activity | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: Electronic assignment of the ancestral AlkB-type activity. Reason: Same judgement as the IBA and IDA rows for this term. |
| GO:0040014 regulation of multicellular organism growth | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: Electronic transfer of an organismal growth role from mouse Fto. Reason: Well grounded organismally - Fto-null mice are growth retarded and humans homozygous for the catalytically inactivating R316Q allele show postnatal growth retardation - but this is a whole-organism phenotype, not something the protein carries out. Supporting Evidence: PMID:19559399 Here we show that a R316Q mutation, inactivating FTO enzymatic activity, is responsible for an autosomal-recessive lethal syndrome. |
| GO:0042245 RNA repair | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: Electronic transfer of the RNA alkylation-repair role. Reason: Same judgement as the IBA and IDA rows for this term. |
| GO:1990931 mRNA N6-methyladenosine dioxygenase activity | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: Electronic assignment from the RHEA reaction and EC 1.14.11.53. Reason: Worth noting that this row is a mapping from a chemical reaction rather than from a cellular observation, and the reaction it maps from is itself now questioned: a 2025 NAR study finds that isolated FTO stops at the hemiaminal N6-hydroxymethyladenosine and that release of formaldehyde and adenosine happens non-enzymatically, whereas the GO term definition specifies that the enzyme "releases oxidized methyl group on N6-methyladenosine as formaldehyde". Kept non-core for consistency with the other rows of this term and because the net transformation is still observed. Supporting Evidence: PMID:40874592 The nascent hemiaminal product undergoes relatively slow non-enzyme catalysed fragmentation giving adenosine/formaldehyde. |
| GO:0001659 temperature homeostasis | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Orthology-based transfer of a thermogenesis phenotype from mouse Fto. Reason: A distal organismal phenotype of Fto-deficient mice. Retained because the mouse phenotype is real, but it is many steps removed from the enzyme's molecular activity and should not be read as a direct role. |
| GO:0010883 regulation of lipid storage | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Orthology-based transfer of an adiposity role from mouse Fto. Reason: Same judgement as the IBA row for this term - real at the level of mouse whole-animal physiology, but downstream of the molecular function. |
| GO:0016706 2-oxoglutarate-dependent dioxygenase activity | IEA GO_REF:0000120 | ACCEPT | Summary: Electronic assignment of the Fe(II)/2-oxoglutarate oxygenase chemistry. Reason: This is the one molecular-function assertion on FTO that neither of the two live disputes touches. Whether FTO is best described as a demethylase or a hydroxylase, and whatever its physiological RNA substrate turns out to be, it is uncontested that it couples substrate oxidation to 2-oxoglutarate decarboxylation using an Fe(II) centre. Demonstrated for the orthologue at discovery and confirmed structurally. Supporting Evidence: PMID:17991826 We find that recombinant murine Fto catalyzes the Fe(II)- and 2OG-dependent demethylation of 3-methylthymine in single-stranded DNA, with concomitant production of succinate, formaldehyde, and carbon dioxide. |
| GO:0044065 regulation of respiratory system process | IEA GO_REF:0000107 | MARK AS OVER ANNOTATED | Summary: Orthology-based transfer of a breathing phenotype observed in Fto-deficient mice. Reason: An organ-system-level physiological readout in a knockout animal, with no proposed mechanistic connection to RNA demethylation and no supporting observation in human cells. Of the several organismal IEA terms on this gene this is the most remote from anything the protein does, and it makes the annotation set look as though FTO directly regulates respiration. |
| GO:0060612 adipose tissue development | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Orthology-based transfer of an adipose development role from mouse Fto. Reason: Consistent with the mouse adiposity phenotypes. Retained as a downstream organismal role, not a core function. |
| GO:0070350 regulation of white fat cell proliferation | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Orthology-based transfer of an adipocyte proliferation role from mouse Fto. Reason: Same standing as the other mouse-derived adiposity terms. Kept, non-core. |
| GO:0005829 cytosol | IDA GO_REF:0000052 | ACCEPT | Summary: Immunofluorescence localization to the cytosol (HPA). Reason: Consistent with the documented cytoplasmic pool, which is genuine but cell-line dependent; the nucleus remains the dominant compartment. |
| GO:0005634 nucleus | EXP PMID:22002720 N6-methyladenosine in nuclear RNA is a major substrate of th... | ACCEPT | Summary: Experimental nuclear localization of human FTO. Reason: Correct and uncontested. Supporting Evidence: PMID:22002720 We further show the partial colocalization of FTO with nuclear speckles, which supports the notion that m(6)A in nuclear RNA is a major physiological substrate of FTO. |
| GO:0005737 cytoplasm | EXP PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | ACCEPT | Summary: Experimental demonstration of a cytoplasmic FTO pool. Reason: Directly supported; the cytoplasmic fraction is where FTO reaches cytoplasmic tRNA. Supporting Evidence: PMID:30197295 As shown in Figure 2A, the relative proportion of FTO within the nucleus and cytoplasm varies across cell lines. |
| GO:0016180 snRNA processing | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | MODIFY | Summary: Captures the snRNA role but names the wrong process - the experiment shows covalent demodification of snRNA, not maturation of a primary snRNA transcript. Reason: GO:0016180 is defined as conversion of a primary snRNA transcript into a mature snRNA molecule. What was actually shown is demethylation of m6A in U6 snRNA and of internal and cap m6Am in snRNAs - covalent alteration of an already-transcribed snRNA. GO:0040031 snRNA modification is a child of GO:0016180 and is the accurate term; GO:0035513 oxidative RNA demethylation records the chemistry. This is the substrate class for which the evidence is strongest and least disputed, so getting the term right matters here. Proposed replacements: snRNA modification oxidative RNA demethylation Supporting Evidence: PMID:30197295 We further identified additional RNA substrates of FTO, including m1A in tRNA, m6A in U6 RNA, internal and cap m6Am in snRNAs. |
| GO:0005634 nucleus | IDA PMID:28002401 Reversible methylation of m(6)A(m) in the 5' cap controls mR... | ACCEPT | Summary: Direct nuclear localization of FTO. Reason: Correct and uncontested. The cached record for this reference is abstract-only and the abstract does not describe the localization experiment, so no verbatim quote is given; the assignment rests on the curator's reading of the full text. |
| GO:0005634 nucleus | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | ACCEPT | Summary: Direct nuclear localization of FTO. Reason: Correct and uncontested; the nucleus is where FTO acts on U6 snRNA and nuclear tRNA. Supporting Evidence: PMID:30197295 In HeLa, Mel624, and HepG2 cells, FTO predominantly accumulates in the cell nucleus. |
| GO:0005737 cytoplasm | IDA PMID:28002401 Reversible methylation of m(6)A(m) in the 5' cap controls mR... | ACCEPT | Summary: Direct cytoplasmic localization of FTO. Reason: Correct; consistent with the cell-line-dependent cytoplasmic pool documented in detail by PMID:30197295. The cached record for this reference is abstract-only, so no verbatim quote is given for the localization itself. |
| GO:0035515 oxidative RNA demethylase activity | IDA PMID:28002401 Reversible methylation of m(6)A(m) in the 5' cap controls mR... | ACCEPT | Summary: Direct demonstration that FTO removes the N6 methyl group from the m6Am cap nucleotide. Reason: Core molecular function. This is the substrate-generic RNA demethylase term and it is supported here by the cap m6Am result, which is one of the two best-established FTO activities and is not challenged by either current dispute. Supporting Evidence: PMID:28002401 Moreover, we find that m6Am is selectively demethylated by fat mass and obesity-associated protein (FTO). FTO preferentially demethylates m6Am rather than N6-methyladenosine (m6A), and reduces the stability of m6Am mRNAs. |
| GO:0035515 oxidative RNA demethylase activity | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | ACCEPT | Summary: Direct demonstration of demethylation across several RNA classes. Reason: Core molecular function, and the single best-supported statement about FTO - the same study establishes activity on snRNA, U6 m6A, tRNA m1A and mRNA, so the generic term holds regardless of how the substrate dispute resolves. Supporting Evidence: PMID:30197295 We find that FTO binds multiple RNA species, including mRNA, snRNA, and tRNA, and can demethylate internal m6A and cap m6Am in mRNA, internal m6A in U6 RNA, internal and cap m6Am in snRNAs, and N1-methyladenosine (m1A) in tRNA. |
| GO:0061157 mRNA destabilization | IDA PMID:28002401 Reversible methylation of m(6)A(m) in the 5' cap controls mR... | KEEP AS NON CORE | Summary: Cap m6Am demethylation by FTO renders transcripts susceptible to decapping and shortens their half-life. Reason: A real, directly measured downstream consequence, but it is an effect of the enzymatic activity rather than a process FTO executes, and the size of the mRNA-level effect is part of what is currently under re-examination. Retained, non-core. Supporting Evidence: PMID:28002401 FTO preferentially demethylates m6Am rather than N6-methyladenosine (m6A), and reduces the stability of m6Am mRNAs. |
| GO:0061157 mRNA destabilization | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | KEEP AS NON CORE | Summary: FTO-mediated demethylation lowers the abundance of transcripts carrying internal m6A. Reason: Same judgement as the other row for this term. Note this particular result is tied to internal mRNA m6A, the contested substrate, so it inherits that uncertainty. Supporting Evidence: PMID:30197295 FTO-mediated demethylation has a greater effect on the transcript levels of mRNAs possessing internal m6A than the ones with cap m6Am in the tested cells. |
| GO:1990931 mRNA N6-methyladenosine dioxygenase activity | IDA PMID:25452335 Meclofenamic acid selectively inhibits FTO demethylation of ... | KEEP AS NON CORE | Summary: In vitro m6A demethylation by FTO, assayed while characterising meclofenamic acid as a selective FTO inhibitor. Reason: The in vitro activity is not in doubt and this paper is a careful biochemical and structural study of it. What is demoted is the inference from in vitro m6A turnover to internal mRNA m6A being a physiological FTO substrate, which is what the term asserts. Kept on the gene, non-core, with the dispute recorded on the IBA row for this term. Supporting Evidence: PMID:25452335 Here, we have identified meclofenamic acid (MA) as a highly selective inhibitor of FTO. |
| GO:1990931 mRNA N6-methyladenosine dioxygenase activity | IDA PMID:26457839 Fluorescein Derivatives as Bifunctional Molecules for the Si... | KEEP AS NON CORE | Summary: In vitro FTO demethylation assayed during development of fluorescein-based inhibitors and co-crystal structures. Reason: As for the other IDA rows of this term - solid in vitro enzymology, demoted only because the physiological relevance of internal mRNA m6A as the substrate is contested. Supporting Evidence: PMID:26457839 Here, we identified that fluorescein derivatives can selectively inhibit FTO demethylation, and the mechanisms behind these activities were elucidated after we determined the X-ray crystal structures of FTO/fluorescein and FTO/5-aminofluorescein. |
| GO:1990931 mRNA N6-methyladenosine dioxygenase activity | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | KEEP AS NON CORE | Summary: Direct demonstration that FTO demethylates internal m6A in mRNA. Reason: This is the strongest single piece of evidence for the term, and the same study is candid that the effect is small relative to other substrates: it reports that tRNA m1A is dramatically affected by FTO loss while m6A and cap m6Am in polyadenylated RNA are not noticeably changed in Fto-knockout mouse cells. That magnitude contrast, from a paper that supports the mRNA activity, is itself a reason to keep the term but not to treat it as the core function. Supporting Evidence: PMID:30197295 Interestingly, the m1A level in tRNA is dramatically affected by FTO-mediated demethylation in comparison to m6A and cap m6Am in polyadenylated RNAs in Fto knockout mouse cells and tissues. |
| GO:1990984 tRNA demethylase activity | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | ACCEPT | Summary: FTO removes N1-methyladenosine from tRNA, with a measurable effect on translation. Reason: Core molecular function. Of all the reported FTO substrates this is the one with the largest measured cellular effect size, it is seen in both nuclear and cytoplasmic tRNA pools, it holds in Fto-knockout mouse cells as well as in human knockdowns, and it comes with a functional readout. Nothing in either of the current disputes touches it. Supporting Evidence: PMID:30197295 We also show that FTO can directly repress translation by catalyzing m1A tRNA demethylation. PMID:30197295 Interestingly, the m1A level in tRNA is dramatically affected by FTO-mediated demethylation in comparison to m6A and cap m6Am in polyadenylated RNAs in Fto knockout mouse cells and tissues. |
| GO:0005654 nucleoplasm | TAS Reactome:R-HSA-8857692 | ACCEPT | Summary: Reactome placement of FTO in the nucleoplasm. Reason: Consistent with the direct nuclear localization data. |
| GO:1990931 mRNA N6-methyladenosine dioxygenase activity | IMP PMID:26458103 Dynamic m(6)A mRNA methylation directs translational control... | KEEP AS NON CORE | Summary: FTO perturbation changes 5'UTR m6A on heat-shock transcripts, inferred as FTO acting as the m6A eraser on those mRNAs. Reason: This is the mutant-phenotype leg of the contested claim, and it is the reason the term cannot simply be demoted without argument - GO:1990931 on this gene carries IBA, IDA, IEA and IMP. The experiment is genuine and the curator read the full text, so it is retained. The caveat is methodological rather than about this paper: the m6A changes here, as throughout the FTO mRNA literature, are read out by antibody-based mapping, the specificity of which for FTO-dependent changes is what the direct-sequencing reassessment challenges. Non-core, with the dispute recorded rather than resolved. Supporting Evidence: PMID:26458103 Upon heat shock stress, the nuclear YTHDF2 preserves 5'UTR methylation of stress-induced transcripts by limiting the m(6)A 'eraser' FTO from demethylation. |
| GO:0005634 nucleus | IDA PMID:26458103 Dynamic m(6)A mRNA methylation directs translational control... | ACCEPT | Summary: Direct nuclear localization of FTO, by cell fractionation alongside the other m6A writers and readers. Reason: Correct and uncontested, and consistent with every other localization row on this gene. Supporting Evidence: PMID:26458103 Consistent with this notion, both the m6A "writers" (METTL3, METTL14, WTAP) and the "eraser" FTO were predominantly present in the nucleus, whereas the majority of the "reader" YTHDF2 resided in the cytosol |
| GO:0010883 regulation of lipid storage | IMP PMID:26287746 FTO Obesity Variant Circuitry and Adipocyte Browning in Huma... | KEEP AS NON CORE | Summary: Lipid-storage phenotype attributed to FTO on the basis of a study of the FTO locus, whose own conclusion assigns the mechanism to IRX3 and IRX5. Reason: The source paper perturbs an enhancer inside FTO intron 1, not FTO protein: the causal variant rs1421085 disrupts an ARID5B motif and derepresses a preadipocyte enhancer acting at 1.2 Mb on IRX3 and IRX5, and the authors name those genes, not FTO, in their conclusion. Attributing the lipid-storage phenotype to the FTO gene product therefore conflates the locus with the protein - a confusion UniProt flags explicitly in its own OBESITY comment. The term is nonetheless retained as non-core because independent mouse Fto protein data support an adiposity role, which is what the IBA and orthology IEA rows for this term rest on. Per project policy an experimental annotation is not removed here; the cleanest curation fix would be to re-home this particular reference to IRX3 and IRX5. Supporting Evidence: PMID:26287746 The rs1421085 T-to-C single-nucleotide variant disrupts a conserved motif for the ARID5B repressor, which leads to derepression of a potent preadipocyte enhancer and a doubling of IRX3 and IRX5 expression during early adipocyte differentiation. |
| GO:0090335 regulation of brown fat cell differentiation | IMP PMID:26287746 FTO Obesity Variant Circuitry and Adipocyte Browning in Huma... | MARK AS OVER ANNOTATED | Summary: Brown/beige adipocyte differentiation phenotype attributed to FTO, but demonstrated by the authors to be mediated by IRX3 and IRX5. Reason: Unlike regulation of lipid storage, this term rests on this single reference, and this reference is a locus study whose experiments - haplotype-specific enhancer assays, CRISPR repair of rs1421085, IRX3/IRX5 knockdown and overexpression - never manipulate FTO protein. The beige-to-white shift is explicitly attributed to IRX3/IRX5 derepression. Marked as over-annotated rather than removed, in line with project policy on experimental annotations and because mouse Fto does have reported effects on adipocyte differentiation; but as it stands this annotation transfers a neighbouring genes phenotype onto FTO. Supporting Evidence: PMID:26287746 Our results point to a pathway for adipocyte thermogenesis regulation involving ARID5B, rs1421085, IRX3, and IRX5, which, when manipulated, had pronounced pro-obesity and anti-obesity effects. PMID:26287746 This results in a cell-autonomous developmental shift from energy-dissipating beige (brite) adipocytes to energy-storing white adipocytes, with a reduction in mitochondrial thermogenesis by a factor of 5, as well as an increase in lipid storage. |
| GO:0016607 nuclear speck | IDA PMID:22002720 N6-methyladenosine in nuclear RNA is a major substrate of th... | ACCEPT | Summary: Partial colocalization of FTO with nuclear speckles. Reason: Directly observed and biologically sensible for an enzyme acting on nuclear RNA. Supporting Evidence: PMID:22002720 We further show the partial colocalization of FTO with nuclear speckles, which supports the notion that m(6)A in nuclear RNA is a major physiological substrate of FTO. |
| GO:0035515 oxidative RNA demethylase activity | IDA PMID:22002720 N6-methyladenosine in nuclear RNA is a major substrate of th... | ACCEPT | Summary: The founding demonstration that FTO oxidatively demethylates m6A in RNA, in vitro and with a matching direction of effect in cells. Reason: Core molecular function. Two caveats are worth recording without changing the action. First, chemistry: a 2025 study reports that isolated FTO predominantly forms the hemiaminal N6-hydroxymethyladenosine and that the demethylated product arises by slow non-enzymatic fragmentation, unlike ALKBH5 and AlkB - so "demethylase" may describe the net outcome rather than the catalysed step. Second, substrate: the cellular m6A changes reported here were measured on bulk mRNA. Neither caveat disturbs the substrate-generic claim that FTO oxidatively demethylates RNA, which is independently supported by the snRNA and tRNA work. Supporting Evidence: PMID:22002720 We report here that fat mass and obesity-associated protein (FTO) has efficient oxidative demethylation activity targeting the abundant N6-methyladenosine (m(6)A) residues in RNA in vitro. PMID:40874592 The results imply that, at least in isolated form, FTO preferentially acts as a hydroxylase, producing a hemiaminal product, rather than a demethylase, distinguishing it from ALKBH5. |
| GO:0035516 broad specificity oxidative DNA demethylase activity | IDA PMID:20376003 Crystal structure of the FTO protein reveals basis for its s... | KEEP AS NON CORE | Summary: Crystal structure of FTO bound to 3-methylthymine, with biochemistry defining its single-stranded substrate preference. Reason: A real in vitro activity with a structural explanation, but narrower than the term implies - FTO strongly prefers 3-methylthymine and 3-methyluracil, is inactive on 1-methylguanine, and is excluded from double-stranded DNA by an active-site loop. Retained as an ancestral, non-core capability of the AlkB-like domain. Supporting Evidence: PMID:20376003 Structural comparison shows that this loop selectively competes with the unmethylated strand of the DNA duplex for binding to FTO, suggesting that it has an important role in FTO selection against double-stranded nucleic acids. |
| GO:0035515 oxidative RNA demethylase activity | IDA PMID:18775698 Oxidative demethylation of 3-methylthymine and 3-methyluraci... | ACCEPT | Summary: Oxidative demethylation of 3-methyluracil in single-stranded RNA by recombinant human FTO. Reason: Correct, and the earliest direct demonstration that RNA is a better FTO substrate than DNA. Supports the substrate-generic RNA demethylase term independently of the m6A literature. Supporting Evidence: PMID:18775698 In addition, these two proteins can catalyze the demethylation of 3-meU in ssRNA with a slightly higher efficiency over that of 3-meT in ssDNA, suggesting that methylated RNAs are the preferred substrates for FTO. |
| GO:0035516 broad specificity oxidative DNA demethylase activity | IDA PMID:18775698 Oxidative demethylation of 3-methylthymine and 3-methyluraci... | KEEP AS NON CORE | Summary: Oxidative demethylation of 3-methylthymine in single-stranded DNA by recombinant human and mouse FTO. Reason: Same judgement as the other rows of this term - genuine in vitro ssDNA activity, no demonstrated in vivo repair role, and strictly single-stranded. Supporting Evidence: PMID:18775698 They showed negligible activities against 3-meT in double-stranded DNA (dsDNA). |
| GO:0008198 ferrous iron binding | IDA PMID:20376003 Crystal structure of the FTO protein reveals basis for its s... | KEEP AS NON CORE | Summary: Binding of the catalytic Fe(II) centre, from the FTO crystal structure. Reason: Correct and undisputed, but it describes the cofactor rather than what the enzyme does, so it is not a core function. No supporting_text is quoted because the cached record for this reference is abstract-only and the abstract does not state iron binding explicitly; the assignment rests on the curator's reading of the structure. |
| GO:0005634 nucleus | ISS PMID:17991826 The obesity-associated FTO gene encodes a 2-oxoglutarate-dep... | ACCEPT | Summary: Nuclear localization inferred from the mouse orthologue. Reason: Superseded in practice by the direct human evidence on this gene, but correct. Supporting Evidence: PMID:17991826 Consistent with a potential role in nucleic acid demethylation, Fto localizes to the nucleus in transfected cells. |
| GO:0006307 DNA alkylation repair | IDA PMID:18775698 Oxidative demethylation of 3-methylthymine and 3-methyluraci... | KEEP AS NON CORE | Summary: Repair-type demethylation of 3-methylthymine in ssDNA, shown in vitro. Reason: The biochemistry is solid but the biological process assignment is an inference from it. No cellular or organismal alkylation-repair phenotype has been shown for FTO, and UniProt records the in vivo relevance as unsure. Kept as non-core. Supporting Evidence: PMID:18775698 We demonstrate here the oxidative demethylation of 3-methylthymine (3-meT) in single-stranded DNA (ssDNA) and 3-methyluracil (3-meU) in single-stranded RNA (ssRNA) by recombinant human FTO protein in vitro. |
| GO:0042245 RNA repair | IDA PMID:18775698 Oxidative demethylation of 3-methylthymine and 3-methyluraci... | KEEP AS NON CORE | Summary: Repair-type demethylation of 3-methyluracil in ssRNA, shown in vitro. Reason: Same standing as the DNA alkylation repair row. Retained as a non-core in vitro capability, with no demonstrated in vivo repair function. Supporting Evidence: PMID:18775698 In addition, these two proteins can catalyze the demethylation of 3-meU in ssRNA with a slightly higher efficiency over that of 3-meT in ssDNA, suggesting that methylated RNAs are the preferred substrates for FTO. |
| GO:0006400 tRNA modification | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | NEW | Summary: Proposed new annotation - FTO alters the modification state of tRNA by removing N1-methyladenosine, in both nuclear and cytoplasmic tRNA pools. Reason: GOA carries the molecular function (GO:1990984 tRNA demethylase activity) but no matching biological process. tRNA m1A levels rise on FTO knockdown in three human and mouse cell lines and in Fto-knockout MEFs, so the process assignment is directly supported by the same experiments. Added because this is one of the two substrate classes for which the evidence is not disputed. Supporting Evidence: PMID:30197295 Both nuclear tRNA m1A and cytoplasmic tRNA m1A increased upon FTO knockdown, with cytoplasmic tRNA m1A showing a larger increase (Figure 5C). |
| GO:0017148 negative regulation of translation | IDA PMID:30197295 Differential m(6)A, m(6)A(m), and m(1)A Demethylation Mediat... | NEW | Summary: Proposed new annotation - FTO represses translation as a direct consequence of demethylating m1A in tRNA. Reason: The authors state the causal link explicitly and demonstrate it, and it is the clearest cellular consequence yet attributed to any FTO activity. Adding it makes the functional output of the uncontested tRNA activity visible in the annotation set, where currently only the mRNA-stability consequences of the contested mRNA activity are represented. Supporting Evidence: PMID:30197295 We also show that FTO can directly repress translation by catalyzing m1A tRNA demethylation. |
Loading supporting content…
Download this section (compressed HTML)Q: Is N6-hydroxymethyladenosine (hm6A) or its 2'-O-methyl counterpart detectable in cellular RNA, and if so is it the true physiological product of FTO rather than a transient intermediate? If the hemiaminal is the enzymatic product and demethylation completes non-enzymatically, should FTO be classified as a hydroxylase rather than a demethylase in GO?
Suggested experts: Christopher J Schofield, Wei Shen Aik
Q: Does any quantitative, antibody-independent method - direct RNA sequencing, or chemistry-based single-nucleotide mapping - detect FTO-dependent changes in internal mRNA m6A stoichiometry in any cell type or physiological condition? A positive result in some context would reconcile the two literatures; a consistent negative across contexts would mean the mRNA m6A eraser model should be retired.
Suggested experts: Samie R Jaffrey, Chuan He
Q: If FTO inhibitors such as FB23-2 remain cytotoxic in FTO-deficient AML cells, what target actually mediates their anti-leukemic effect, and do FTO degraders - which do not rely on active-site occupancy - act through mRNA m6A, through snRNA m6Am and splicing, or through tRNA m1A and translation?
Suggested experts: Jianjun Chen, Chuan He
Q: Does FTO-dependent snRNA m6Am affect pre-mRNA splicing, and is this the route by which FTO loss changes mRNA abundance without changing mRNA m6A stoichiometry?
Suggested experts: Samie R Jaffrey
Q: Which FTO activity accounts for the lethal human R316Q polymalformation syndrome and for the growth phenotypes of Fto-null mice - tRNA m1A demethylation and translational control, snRNA m6Am and splicing, or mRNA m6A?
Experiment: Quantify hm6A and hm6Am in RNA isolated from wild-type and FTO-knockout cells using workflows explicitly designed to preserve labile hemiaminal adducts (rapid quench, low temperature, avoidance of the acidic or aqueous steps shown to destroy these adducts), with isotopically labelled hm6A standards. Compare against the same samples processed by conventional LC-MS/MS workflows to measure how much signal the standard protocol destroys.
Hypothesis: FTO's enzymatic product on cellular RNA is the hemiaminal N6-hydroxymethyladenosine, and reported demethylation reflects non-enzymatic decay during sample handling.
Type: quantitative LC-MS/MS of modified nucleosides
Experiment: Run matched quantitative m6A measurements on the same RNA samples from FTO-knockout, knockdown, inhibitor-treated and control cells by three orthogonal methods - direct RNA nanopore sequencing, an antibody-free chemical mapping method such as GLORI, and MeRIP-seq - across several cell types including primary cells and a tissue with high FTO expression. Include snRNA m6Am by CROWN-seq in every sample as the positive control for loss of FTO activity. Pre-register the analysis so that discordance between methods is interpretable.
Hypothesis: Internal mRNA m6A is not a physiological FTO substrate; the reported FTO-dependent m6A changes are artefacts of antibody-based enrichment.
Type: comparative transcriptome-wide RNA modification mapping
Experiment: Test each published FTO inhibitor and degrader for cytotoxicity in isogenic FTO-knockout versus wild-type AML lines, and rescue with catalytically dead FTO (R316Q) versus wild-type FTO. Couple this to thermal proteome profiling or chemoproteomics in the knockout background to identify the off-target protein responsible for any FTO-independent killing.
Hypothesis: The anti-leukemic activity of FTO inhibitors is largely FTO-independent.
Type: isogenic chemical-genetic target deconvolution
Experiment: In FTO-knockout and control cells, measure tRNA m1A stoichiometry per isoacceptor by misincorporation-based tRNA sequencing alongside ribosome profiling and matched RNA-seq. Determine whether translational-efficiency changes track the isoacceptors whose m1A changes, and whether they are rescued by wild-type but not R316Q FTO.
Hypothesis: The dominant cellular consequence of FTO loss is altered translation via tRNA m1A, not altered mRNA stability via m6A.
Type: ribosome profiling with tRNA modification sequencing
Experiment: Combine CROWN-seq quantification of snRNA m6Am with deep long-read transcriptome sequencing in FTO-knockout versus wild-type cells, and test whether U1 snRNA carrying a non-methylatable transcription-start nucleotide phenocopies FTO loss for the affected splicing events.
Hypothesis: FTO regulates splicing through snRNA m6Am rather than through mRNA m6A.
Type: snRNA modification mapping with long-read splicing analysis
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)