Non-catalytic, substrate-binding subunit of the nuclear tRNA m1A58 methyltransferase. TRMT6 is a catalytically dead paralogue of the methyltransferase TRMT61A that has been repurposed as the substrate-binding half of the enzyme. Two copies of TRMT6 and two of TRMT61A assemble into a heterotetramer, a dimer of heterodimers, in which each tRNA binds across the dimer interface so that TRMT6 from the opposing heterodimer delivers the target base into the TRMT61A active site. In doing so TRMT6 splays the T- and D-loops of the tRNA apart, making adenosine 58, normally buried within the folded T-loop, accessible for methylation. TRMT6 therefore supplies substrate recognition, tRNA remodelling and site specificity for the complex, but has no catalytic activity of its own, and neither subunit is active without the other. The resulting m1A58 modification stabilises initiator methionyl-tRNA and, in tRNA3(Lys), removes Watson-Crick pairing at the base that terminates the HIV-1 primer-binding sequence. The complex has also been reported to deposit m1A on a small number of cytosolic mRNAs that fold into tRNA T-loop-like elements, at low stoichiometry; this mRNA substrate class is disputed, since biochemical validation of transcriptome-wide maps indicates that most mapped internal mRNA m1A sites are instead inosine produced by A-to-I editing.
Definition: An enzyme-substrate adaptor activity in which a protein binds a tRNA, remodels its tertiary structure, and presents a specific nucleotide of that tRNA to the active site of an associated tRNA-modifying enzyme, thereby conferring substrate selection and site specificity on the modification reaction.
Justification: GO can already say that TRMT6 is part of the tRNA (m1A) methyltransferase complex (GO:0031515) and, via contributes_to, that it enables the complex-level activity GO:0160107. What no existing term captures is what TRMT6 itself does, which is the mechanism resolved by the 2015 crystal structure - bind the tRNA, splay the T- and D-loops apart, and deliver a nucleotide that is buried in the folded substrate into a partner subunit's active site in trans. GO:0140767 enzyme-substrate adaptor activity is the closest existing term and is used for this gene's core function here, but it was framed for protein-substrate adaptors (for example in ubiquitin ligase complexes) and does not convey the substrate remodelling that makes this case distinctive. A tRNA-specific child would also serve the several other non-catalytic partners of tRNA-modifying enzymes that face the same annotation problem and currently attract only bare RNA binding or protein binding terms. Proposed only after checking that neither a suitable existing term nor a plausible alternative parent (GO:0008047 enzyme activator activity, rejected because TRMT6 is an obligate structural half of the enzyme rather than a regulator of an otherwise functional one) fits.
Parent term: enzyme-substrate adaptor activity
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005634 nucleus | IBA GO_REF:0000033 | ACCEPT | Summary: The tRNA m1A58 methyltransferase acts in the nucleus, where tRNA maturation occurs. Reason: Correct compartment, and unusually well corroborated for an IBA. TRMT6 itself has a direct immunofluorescence localisation to the nucleoplasm (HPA, GO_REF:0000052), and the same location is carried by the partner subunit TRMT61A. |
| GO:0031515 tRNA (m1A) methyltransferase complex | IBA GO_REF:0000033 | ACCEPT | Summary: TRMT6 is an obligate subunit of the TRMT6-TRMT61A heterotetramer; the phylogenetic assertion matches the human structural data directly. Reason: Core, and the single most informative annotation this gene carries. Complex membership is conserved from budding yeast (SGD S000005006, Trm6/Gcd10) and is resolved structurally in human. TRMT6 appearing in its own WITH/FROM list is expected and correct, since its own experimental complex annotation is one of the descendant evidences supporting the ancestral node, not a circular inflation of support. Supporting Evidence: PMID:26470919 a homologous but noncatalytic chain, Trm6, repurposed as a tRNA-binding subunit that acts in trans |
| GO:0005634 nucleus | IEA GO_REF:0000044 | ACCEPT | Summary: Electronic transfer of the UniProt subcellular location. Reason: Correct location, independently supported by IBA, ISS, NAS and direct immunofluorescence. |
| GO:0030488 tRNA methylation | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro signature IPR017423 (TRM6/GCD10 family) mapped to tRNA methylation. Reason: Correct and appropriately specific. This is the core biological process for TRMT6 even though it contributes to it non-catalytically, and it is the only informative BP term the gene carries. The domain-to-process mapping is sound because the TRM6/GCD10 family exists only as the non-catalytic half of this enzyme. |
| GO:0031515 tRNA (m1A) methyltransferase complex | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro-based assignment of complex membership. Reason: Correct, and confirmed independently by the crystal structure and by ComplexPortal. |
| GO:0005515 protein binding | IPI PMID:16043508 The bipartite structure of the tRNA m1A58 methyltransferase ... | REMOVE | Summary: IPI with TRMT61A (UniProtKB Q96FX7), the catalytic partner subunit. Reason: Bare protein binding carries no functional information. This is the most important interaction TRMT6 makes, and precisely for that reason it should be expressed as GO:0031515 tRNA (m1A) methyltransferase complex, which the gene already carries with IBA, IEA and IPI support, rather than as an uninformative binding term. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005515 protein binding | IPI PMID:28514442 Architecture of the human interactome defines protein commun... | REMOVE | Summary: High-throughput interactome detection of the TRMT61A (UniProtKB Q96FX7) interaction. Reason: Uninformative term; the TRMT61A interaction is captured properly by the complex annotation. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005515 protein binding | IPI PMID:32814053 Interactome Mapping Provides a Network of Neurodegenerative ... | REMOVE | Summary: Interactome-mapping screen contributing five partners for this gene (KAT5, LMO3, PPIA, SETDB1, YWHAG), reported as a single GOA reference. Reason: Uninformative term, and none of these partners has an established functional relationship to tRNA methylation. They are large-scale screen hits without follow-up, and this row adds five of the six protein binding annotations on the gene without adding any functional information. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | REMOVE | Summary: Proteome-scale AP-MS detection of the TRMT61A (UniProtKB Q96FX7) interaction. Reason: Uninformative term; the TRMT61A interaction is captured properly by the complex annotation. No more informative molecular function is supportable from this evidence, so the row is removed rather than rewritten. Removal does not imply the reported interaction is false. |
| GO:0005654 nucleoplasm | IDA GO_REF:0000052 | ACCEPT | Summary: Direct immunofluorescence localisation of TRMT6 to the nucleoplasm (Human Protein Atlas). Reason: The strongest and most specific localisation evidence on this gene, and the basis on which the weaker nucleus annotations can be trusted. Consistent with Reactome placing the methylation reaction in the nucleoplasm. |
| GO:0005634 nucleus | NAS PMID:16043508 The bipartite structure of the tRNA m1A58 methyltransferase ... | ACCEPT | Summary: ComplexPortal author statement placing the complex in the nucleus. Reason: Weak evidence code in itself, but the location is directly demonstrated by the HPA immunofluorescence annotation on this same gene. |
| GO:0031515 tRNA (m1A) methyltransferase complex | IPI PMID:26470919 Crystal Structure of the Human tRNA m(1)A58 Methyltransferas... | ACCEPT | Summary: ComplexPortal annotation based on the crystal structure of the human heterotetramer bound to tRNA3(Lys). Reason: The strongest evidence available for complex membership. The structure resolves the stoichiometry (two TRMT6, two TRMT61A) and shows that TRMT6 acts in trans, delivering A58 into the active site of the TRMT61A chain of the opposing heterodimer. Supporting Evidence: PMID:26470919 tRNAs bind across the dimer interface such that Trm6 from the opposing heterodimer brings A58 into the active site of Trm61. PMID:26470919 T-loop and D-loop are splayed apart showing how A58, normally buried in tRNA, becomes accessible for modification. |
| GO:0005634 nucleus | ISS GO_REF:0000024 | ACCEPT | Summary: Sequence-similarity transfer of nuclear localisation from the S. cerevisiae orthologue Gcd10/Trm6 (UniProtKB P41814). Reason: The orthology is unambiguous and tRNA A58 methylation is a nuclear step of tRNA maturation in both organisms. |
| GO:0006397 mRNA processing | IDA PMID:29072297 The m1A landscape on cytosolic and mitochondrial mRNA at sin... | KEEP AS NON CORE | Summary: Base-resolution mapping assigned a small set of low-stoichiometry cytosolic mRNA m1A sites, lying in tRNA T-loop-like structures, to the TRMT6/TRMT61A complex. Reason: Retained but demoted, and kept deliberately consistent with the parallel decision on TRMT61A. Even on its own account this paper describes a rare, sub-stoichiometric modification that cells largely avoid, and the 2026 biochemical validation of transcriptome-wide m1A maps narrows the genuine mRNA m1A set to a handful of sites. Not removed, because this is an experimental annotation whose full text is not in the publication cache and the curator read it. Note that GOA gives TRMT6 no mRNA methyltransferase MF term, only this process term, which is the right asymmetry, since the catalytic activity belongs to TRMT61A and TRMT6 would contribute to any mRNA methylation through the same adaptor role it plays on tRNA. Supporting Evidence: PMID:29072297 Within the cytosol, m1A is present in a low number of mRNAs, typically at low stoichiometries, and almost invariably in tRNA T-loop-like structures, where it is introduced by the TRMT6/TRMT61A complex. PMID:42337368 Using SCARPET, we show that newly mapped internal m1A sites are not m1A, but instead contain inosine from A-to-I editing. |
| GO:0006397 mRNA processing | IDA PMID:29107537 Base-Resolution Mapping Reveals Distinct m(1)A Methylome in ... | KEEP AS NON CORE | Summary: An independent base-resolution study likewise assigned a small, evenly distributed subset of m1A sites carrying a GUUCRA tRNA-like motif to TRMT6/61A. Reason: Same position as the companion mRNA processing annotation. Two independent groups converged in 2017 on the same narrow, structure-driven substrate class, which is mechanistically coherent because the complex recognises a T-loop fold rather than a sequence, so an mRNA that adopts that fold is an adventitious substrate. Against treating it as core, PMID:42337368 shows by site-specific biochemistry that newly mapped internal mRNA m1A sites are in fact inosine, and concludes that genuine mRNA m1A is largely restricted to PRUNE1, ND5 and MALAT1. That paper is a discriminating assay rather than a blanket debunk, since the same method confirmed the equally contested internal m7G at high stoichiometry, but it does not mention TRMT6 or TRMT61A and so does not directly adjudicate this complex's role. Supporting Evidence: PMID:29107537 A different, small subset of m1A exhibit a GUUCRA tRNA-like motif, are evenly distributed in the transcriptome, and are dependent on the methyltransferase TRMT6/61A. PMID:42337368 These results strongly support the conclusion that m1A is not widespread in mammalian mRNAs but is instead largely restricted to the previously validated sites in PRUNE1, ND5 and MALAT1 RNA. |
| GO:0005654 nucleoplasm | TAS Reactome:R-HSA-6783492 | ACCEPT | Summary: Reactome places the TRMT6:TRMT61A methylation reaction in the nucleoplasm. Reason: Correct and specific, and independently confirmed by the HPA immunofluorescence annotation on this gene. |
| GO:0003723 RNA binding | HDA PMID:22658674 Insights into RNA biology from an atlas of mammalian mRNA-bi... | MODIFY | Summary: TRMT6 was recovered as an RNA-binding protein in a proteome-wide mRNA interactome capture experiment. True, but this is the only molecular-function term on the gene that says anything at all, and it says almost nothing. Reason: Correct in kind but far too general for a protein whose RNA-binding role is known at atomic resolution. GO:0000049 tRNA binding is a direct descendant of GO:0003723 (verified against the QuickGO ancestors endpoint) and is what the structural work actually establishes. Stated plainly, the replacement term is supported by the crystal structure rather than by the poly(A)-selected capture experiment cited here, which by design detects mRNA rather than tRNA binding; recovery of a nuclear tRNA-binding protein in that screen is nonetheless unsurprising given the reported mRNA T-loop-like substrates. If a curator prefers to keep the evidence and the term strictly matched, the alternative is to retain GO:0003723 for this row and add GO:0000049 with IDA support from PMID:26470919. Proposed replacements: tRNA binding Supporting Evidence: PMID:22658674 We identify 860 proteins that qualify as RBPs by biochemical and statistical criteria PMID:26470919 a homologous but noncatalytic chain, Trm6, repurposed as a tRNA-binding subunit that acts in trans |
| GO:0140767 enzyme-substrate adaptor activity | IDA PMID:26470919 Crystal Structure of the Human tRNA m(1)A58 Methyltransferas... | NEW | Summary: Proposed new annotation, not currently in GOA. TRMT6 recruits the substrate tRNA to the catalytic subunit and delivers the target base into its active site, which is what the GO definition of this term describes - an adaptor that brings together an enzyme and its substrate, contributing to substrate selection and specificity. Reason: TRMT6 currently has no molecular-function annotation that says what it does. This term is the closest existing fit for the mechanism resolved by the crystal structure, and is used as the molecular function of the core function for this gene. It is proposed with IDA evidence from the structural work rather than asserted as existing curation. A more precise tRNA-specific child of this term is suggested in proposed_new_terms. GO:0008047 enzyme activator activity was considered and rejected, because TRMT6 is an obligate structural half of the enzyme rather than a regulator of an otherwise functional one. Supporting Evidence: PMID:26470919 tRNAs bind across the dimer interface such that Trm6 from the opposing heterodimer brings A58 into the active site of Trm61. PMID:26470919 T-loop and D-loop are splayed apart showing how A58, normally buried in tRNA, becomes accessible for modification. |
Loading supporting content…
Download this section (compressed HTML)Q: Does TRMT6 have any function outside the TRMT6-TRMT61A complex? Every annotation it carries is either the complex, the complex's location, or a binding term, and no free TRMT6 pool or TRMT61A-independent activity has been reported. A negative answer is as useful as a positive one and would justify treating the complex as the sole unit of function.
Suggested experts: Robert M Stroud, Janet Finer-Moore, James T Anderson
Q: Are the two nuclear-encoded mRNA m1A sites that survive SCARPET validation, PRUNE1 A58 and MALAT1 A7945, actually deposited by the TRMT6-TRMT61A complex? PMID:42337368 confirms them biochemically but never mentions either subunit, so the attribution currently rests on inference from the 2017 mapping studies rather than on a stated result.
Suggested experts: Samie R Jaffrey, Schraga Schwartz, Chengqi Yi
Q: Is TRMT6, rather than TRMT61A, the subunit that determines whether an mRNA is accepted as a substrate? Since TRMT6 performs the substrate binding and loop splaying, the reported preference for T-loop-like mRNA elements should be a property of TRMT6, and separating the two subunits' contributions would be the cleanest test of whether mRNA methylation is structural spillover.
Suggested experts: Robert M Stroud, Schraga Schwartz
Q: Do the four annotated TRMT6 isoforms (Q9UJA5-1 to Q9UJA5-4) differ in their ability to assemble with TRMT61A or to bind tRNA? The structural and biochemical work was done on the canonical form only, and none of the existing annotations is isoform-resolved.
Experiment: Measure the affinity of recombinant TRMT6 alone, TRMT61A alone, and the reconstituted heterotetramer for an authentic tRNA substrate and for an A58U mutant, by filter binding or biolayer interferometry. Comparing the three would show whether substrate recognition resides in TRMT6 and would supply IDA evidence for a GO:0000049 tRNA binding annotation to replace the current high-throughput RNA binding row.
Hypothesis: TRMT6 binds tRNA directly and independently of TRMT61A, and its tRNA binding, not TRMT61A's, sets the substrate specificity of the complex.
Type: Quantitative RNA-binding assay on separated subunits
Experiment: Mutate the TRMT6 residues that contact the T- and D-loops in the tRNA-bound structure, then test the reconstituted complex for assembly (by size-exclusion chromatography), for tRNA binding, and for methyltransferase activity. A variant that assembles and binds tRNA but does not support methylation would separate the adaptor function from the structural one.
Hypothesis: The loop-splaying contacts made by TRMT6 are required for catalysis by the partner subunit, so TRMT6 is a functional adaptor rather than a passive scaffold.
Type: Structure-guided mutagenesis with separation-of-function readout
Experiment: Reconstitute chimeric complexes pairing human TRMT61A with TRMT6 orthologues from species whose transcriptomes lack the reported mRNA m1A sites, and assay each chimera on an authentic tRNA and on a T-loop-like mRNA fragment. A shift in the tRNA-to-mRNA activity ratio that tracks the TRMT6 chain would localise substrate choice to TRMT6.
Hypothesis: TRMT6 determines whether an mRNA T-loop-like element is accepted as a substrate.
Type: Chimeric complex reconstitution and comparative substrate assay
What is not known — curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: TRMT6's molecular function is understood structurally but is not represented in its annotations. Its entire GOA molecular-function coverage is six rows of GO:0005515 protein binding plus one high-throughput GO:0003723 RNA binding, none of which states that it binds tRNA, remodels it, or presents A58 to the catalytic subunit.
OPEN CURATIONONTOLOGY MF_DARK
What is known: The mechanism itself is not in doubt. Three crystal structures of the human complex bound to tRNA3(Lys) resolve the heterotetramer, the trans delivery of A58 across the dimer interface, and the splaying of the T- and D-loops that exposes the target base. The gap is between what is known and what is annotatable and annotated.
Significance: TRMT6 is a textbook example of the protein binding failure mode - a protein whose function has been solved at atomic resolution for a decade yet which looks functionally uncharacterised in GO, because the only available molecular-function terms are uninformative ones. Any downstream analysis that scores molecular-function annotation depth would misclassify it as poorly characterised.
What would resolve it: Curation of GO:0000049 tRNA binding with IDA support from PMID:26470919 in place of the generic GO:0003723 row, together with a contributes_to GO:0160107 annotation, would close most of the gap immediately. Fully expressing the substrate-presentation mechanism additionally requires the new term proposed above.
Provenance (the field's own admissions):
Proposed term (ontology gap):
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)