DNMT3A is a de novo DNA cytosine-5 methyltransferase. It uses S-adenosyl-L-methionine to methylate CpG (and to a lesser extent non-CpG) cytosines in DNA, establishing methylation patterns that silence promoters, CpG islands, repeats and germline loci. It acts in the nucleus on chromatin as an oligomer, alone or together with the non-catalytic partner DNMT3L. An ADD domain reads unmethylated histone H3K4 and relieves autoinhibition of the catalytic domain, and a PWWP domain reads H3K36me2/3 and targets the enzyme to intergenic and pericentric heterochromatin. This horse entry, A0A9L0TK01, is an exon-skipping splice model of the horse gene (873 aa). It lacks the 36-residue exon that corresponds to human residues 339-374, inside the PWWP region, but it retains the full ADD domain and the full catalytic domain with all of its catalytic and CpG-recognition residues. Methyltransferase activity is therefore expected to be intact. PWWP-dependent heterochromatin targeting is probably lost.
Existing Annotations Review
GO Term
Evidence
Action
Reason
GO:0000122 negative regulation of transcription by RNA polymerase II
IEA GO_REF:0000107
MODIFY
Summary: DNMT3A represses transcription by methylating promoter CpGs. The methylation-dependent silencing term is more informative.
Reason: The Ensembl Compara transfer comes from human DNMT3A, which sits in the same PANTHER subfamily (PTHR23068:SF10). In human, DNMT3A recruitment leads to promoter hypermethylation and silencing. The work DNMT3A does is the CpG methylation, so the process term GO:0044027 (already a row in this review) states the mechanism. The horse model keeps every catalytic and CpG-recognition residue. This follows the refinement made in the human review.
The catalytic PC-motif cysteine (human C710 -> horse C671), ENV-motif E756 -> E717, the CpG-recognition TRD R836 -> R797 and the tetramer-interface R882 -> R843 are all conserved.
Summary: The gene has this localization, but in mouse it depends on the PWWP domain, and this product lacks a PWWP exon.
Reason: In mouse, Dnmt3a localizes to pericentric heterochromatin (major satellites) only when its PWWP domain is intact. PWWP disruption abolishes that localization. A0A9L0TK01 skips the exon that encodes human PWWP-region residues 339-374, so this product probably lacks the localization. Full-length horse DNMT3A, encoded by the exon-9-containing transcripts, would be expected to have it. The deletion has not been tested experimentally, so the row is left undecided rather than removed.
We show that the Dnmt3a and Dnmt3b proteins with their PWWP domains disrupted failed to accumulate in pericentric heterochromatin, resulting in their inability to methylate the major satellite repeats.
Summary: Heterochromatin localization is PWWP-dependent in mouse, and this product lacks a PWWP exon.
Reason: This is the same mouse Dnmt3a heterochromatin localization as the centromeric-region row. It depends on an intact PWWP domain, and that domain is truncated in this splice model. The localization is conserved for the gene, but this product's ability to localize there is not established.
We show that the Dnmt3a and Dnmt3b proteins with their PWWP domains disrupted failed to accumulate in pericentric heterochromatin, resulting in their inability to methylate the major satellite repeats.
The rest of the domain downstream is lost, so the PWWP fold is expected to be disrupted.
GO:0000978 RNA polymerase II cis-regulatory region sequence-specific DNA binding
IEA GO_REF:0000107
MODIFY
Summary: DNMT3A occupies promoters, but sequence specificity beyond CpG comes from partner transcription factors.
Reason: The human experiments behind this transfer are MYC- and WT1-associated promoter occupancy. In MYC's case, MYC recruits DNMT3A to its E-box targets, so the sequence recognition belongs to MYC. DNMT3A itself recognizes only the CpG dinucleotide. Chromatin binding records the observed occupancy without giving DNMT3A a transcription-factor-like activity. This follows the replacement made in the human review.
Summary: A meiotic XY-body localization from mouse. Which domain targets it there is unknown, and this product lacks part of the PWWP domain.
Reason: The XY body is a facultative heterochromatin compartment in spermatocytes. The domain that targets DNMT3A there has not been mapped. Heterochromatin targeting in mouse needs the PWWP domain, and that domain is truncated in this splice model, so this location cannot be transferred to this product with confidence. DNMT3A expression and localization have been seen in developing horse testis, but the antibody used does not distinguish splice forms.
Immunofluorescence of DNMT1, DNMT3a and DNMT3b on testis samples confirmed the differential expression and localization of these three DNA methylation transferases during horse development.
We show that the Dnmt3a and Dnmt3b proteins with their PWWP domains disrupted failed to accumulate in pericentric heterochromatin, resulting in their inability to methylate the major satellite repeats.
Summary: DNA engagement by the catalytic domain is CpG-directed.
Reason: In human DNMT3A structures, the target recognition domain makes specific CpG contacts (Arg836), and the enzyme flips its target cytosine out of the DNA. In horse, Arg836 is retained as R797, and the catalytic domain aligns completely. Unmethylated CpG binding is the specific descendant term and matches the human review's refinement.
The catalytic PC-motif cysteine (human C710 -> horse C671), ENV-motif E756 -> E717, the CpG-recognition TRD R836 -> R797 and the tetramer-interface R882 -> R843 are all conserved.
GO:0003886 DNA (cytosine-5-)-methyltransferase activity
IBA GO_REF:0000033
ACCEPT
Summary: This is the core catalytic activity. It is inherited within PANTHER subfamily SF10, and all catalytic residues are retained.
Reason: A0A9L0TK01 is in PTHR23068:SF10 (DNMT3A), the subfamily of human Q9Y6K1 and mouse O88508, which are the experimentally grounded donors for PTN000552637. Its catalytic domain aligns completely to human, and the PC-motif nucleophile (C671), ENV glutamate and CpG-recognition arginine are conserved. In mouse, PWWP-disrupted Dnmt3a still methylates non-satellite DNA efficiently, so the missing PWWP exon is not expected to remove catalysis.
Propagation Review
Root cause:NO FAILURE CORE
Sources checked:
PANTHER:PTN000552637SUPPORTS TRANSFER
The target is in PTHR23068:SF10 with human Q9Y6K1 and mouse O88508, the experimentally grounded descendants of this IBD node. Its catalytic residues are conserved, and nothing suggests the function was lost in this lineage.
The observation that these mutant proteins efficiently methylate other genomic sequences indicates that the PWWP domains are not required for enzymatic activities but are involved in functional specialization.
The catalytic PC-motif cysteine (human C710 -> horse C671), ENV-motif E756 -> E717, the CpG-recognition TRD R836 -> R797 and the tetramer-interface R882 -> R843 are all conserved.
GO:0003886 DNA (cytosine-5-)-methyltransferase activity
IEA GO_REF:0000120
ACCEPT
Summary: This is the core catalytic activity, consistent with the IBA row.
Reason: The combined electronic methods agree with the SF10 phylogenetic inference and the conserved catalytic residues. The PROSITE C5-MTase active-site match places the catalytic cysteine at horse residue 671.
The catalytic PC-motif cysteine (human C710 -> horse C671), ENV-motif E756 -> E717, the CpG-recognition TRD R836 -> R797 and the tetramer-interface R882 -> R843 are all conserved.
Summary: The protein is nuclear and acts on chromatin. The nucleoplasm and chromatin terms are more specific.
Reason: The nuclear IBD node PTN001497036 is deep in PTHR23068 and has experimental grounding in many members. Mouse mapping showed that the Dnmt3a nuclear-localization signals are basic stretches flanking the PWWP domain, and those stretches lie outside the skipped exon. DNMT3A binds nucleosomes and methylates linker DNA, so chromatin and nucleoplasm are the more specific supported locations. This follows the human review's refinement.
Propagation Review
Root cause:TERM SCOPING PROBLEM
Failure modes:GRANULARITY MISMATCH
Sources checked:
PANTHER:PTN001497036SUPPORTS TRANSFER
Nuclear localization is a deep, well-grounded inheritance in the DNMT3/DRM family, and the target's NLS regions are retained. The refinement to chromatin and nucleoplasm is about granularity and does not challenge the node.
These results suggested that the two basic sequences are functional NLSs and that they are both required for association with the major satellite repeats.
Summary: The protein is nuclear and acts on chromatin. This is refined in the same way as the IBA nucleus row.
Reason: This row makes the same claim as the IBA nucleus row and gets the same chromatin and nucleoplasm refinement. The NLS-bearing basic regions are retained in this model.
These results suggested that the two basic sequences are functional NLSs and that they are both required for association with the major satellite repeats.
Summary: There is a nucleoplasmic and euchromatic DNMT3A pool, and it does not depend on PWWP-mediated heterochromatin targeting.
Reason: In mouse, PWWP-disrupted Dnmt3a was not lost from the nucleus. It stayed nuclear and lost only the large heterochromatic foci. This is the location that fits this splice model best.
However, deletion of a large portion of the PWWP domain abolished the ability of Dnmt3a and Dnmt3b to accumulate in large nuclear foci, and all transfected cells showed pattern B
Summary: A minor cytoplasmic pool is reported for mammalian DNMT3A. This is not where the enzyme does its work.
Reason: The IBD node PTN002747444 rests on mouse and human cytoplasmic observations of Dnmt3a/Dnmt3b, and the target falls inside that clade. DNA methylation itself happens in the nucleus. The cytoplasmic pool is kept as non-core, matching the human review.
Propagation Review
Root cause:NO FAILURE NON CORE
Sources checked:
PANTHER:PTN002747444SUPPORTS TRANSFER
The target falls inside the DNMT3A/DNMT3B clade below this IBD node. The cytoplasmic pool is real but is not where catalysis happens.
Unlike Dnmt3a, which was concentrated on heterochromatin, Dnmt3a2 displayed a localization pattern suggestive of euchromatin association.
GO:0006346 DNA methylation-dependent constitutive heterochromatin formation
IEA GO_REF:0000107
UNDECIDED
Summary: Pericentric major-satellite methylation by DNMT3A needs an intact PWWP domain, and this product lacks a PWWP exon.
Reason: For the gene, this process is real. DNMT3A methylates major-satellite repeats at constitutive heterochromatin, and the human review keeps it as non-core. In mouse, deleting or mutating the Dnmt3a PWWP domain severely reduced major-satellite methylation while other targets were still methylated. This splice model lacks the exon encoding human 339-374 in the PWWP region, so it would probably not take part in this process. The prediction is untested, so the row is undecided rather than removed.
Strikingly, with the deletion of the PWWP domains (Dnmt3aΞPWWP and Dnmt3b1ΞPWWP), the ability of Dnmt3a and Dnmt3b1 to methylate the major satellite repeats was severely compromised, whereas their activities for all the other sequences examined were not significantly affected.
Summary: The methyl acceptor is cytosine C5 in DNA.
Reason: IPR001525 is the C5 cytosine methyltransferase family. Its mapping to methyltransferase activity is correct but general. DNA (cytosine-5-)-methyltransferase activity is the supported descendant, and the gene already carries it.
Summary: Gene regulation by DNMT3A works through CpG methylation. The specific descendant term is supported.
Reason: The InterPro ADD_DNMT3A mapping is correct but general. GO:0044027 is a descendant via epigenetic regulation of gene expression. It is already a row in this review and states the methylation mechanism.
Summary: The human source for nuclear-matrix association cannot be checked here, and the human review also leaves it undecided.
Reason: The human source (PMID:12138111) is abstract-only in the cache, and the abstract does not describe a nuclear-matrix fractionation. Ortholog transfer cannot resolve a claim that is unresolved at the source.
Summary: The human source is a recruitment event that depends on a pseudogene lncRNA, and that RNA has not been shown to exist in horse.
Reason: In the human experiments, DNMT3A and EZH2 are recruited to the PTEN promoter by the PTENpg1 antisense RNA, which comes from a pseudogene. The transfer to horse requires a horse counterpart of that RNA, which has not been established. The generic DNMT3A methylation activity does not establish this specific RNA-guided process.
These observations intrigued us to investigate whether EZH2 and DNMT3a may directly interact within an protein:RNA complex together with PTENpg1 asRNA Ξ±.
Summary: DNMT3A self-associates into tetramers through its catalytic-domain interfaces, which are retained here.
Reason: Human DNMT3A homotetramerizes, and R882 at the RD interface is needed for an active tetramer. In horse, R882 is retained as R843, and the whole catalytic domain aligns. Self-association is an assembly property that supports catalysis, not a function in its own right, so it is kept as non-core, matching the human review.
The catalytic PC-motif cysteine (human C710 -> horse C671), ENV-motif E756 -> E717, the CpG-recognition TRD R836 -> R797 and the tetramer-interface R882 -> R843 are all conserved.
GO:0044027 negative regulation of gene expression via chromosomal CpG island methylation
IEA GO_REF:0000107
ACCEPT
Summary: This is the core process. DNMT3A itself catalyzes the methylation step that silences CpG-island promoters.
Reason: DNMT3A performs the key step of this process, cytosine methylation at CpG islands. The catalytic and CpG-recognition machinery is conserved in this splice model. In mouse, promoter and other non-satellite targets were still methylated by PWWP-disrupted Dnmt3a.
The observation that these mutant proteins efficiently methylate other genomic sequences indicates that the PWWP domains are not required for enzymatic activities but are involved in functional specialization.
GO:0045892 negative regulation of DNA-templated transcription
IBA GO_REF:0000033
MODIFY
Summary: The inherited repressive role is better stated as silencing by CpG-island methylation.
Reason: The target descends from IBD node PTN002747444. Human DNMT3A appears among the donors because it has its own experimental evidence, which is the expected pattern and not circular. The mechanism of the repression is DNA methylation. GO:0044027 states that mechanism, as in the human review.
Propagation Review
Root cause:TERM SCOPING PROBLEM
Failure modes:GRANULARITY MISMATCH
Sources checked:
PANTHER:PTN002747444SUPPORTS TRANSFER
The target descends from this IBD node, and human DNMT3A grounds it with its own evidence. The refinement to CpG-island methylation-dependent silencing states the mechanism and does not challenge the node.
Summary: Human DNMT3A is recruited to MYC target promoters through a MYC interaction. This is a context-specific way the enzyme is recruited.
Reason: Binding to sequence-specific transcription factors is one way DNMT3A is targeted to particular loci. The human MYC interaction is the source. It has not been mapped to the PWWP exon missing from this model, and the rest of the protein is near-identical (97.3%). The interaction is kept as non-core, matching the human review.
Summary: The lncRNA partners in human and mouse are specific RNAs, and neither has been shown to have a horse counterpart.
Reason: In human, the RNA partner is the PTENpg1 antisense RNA from a pseudogene. In mouse, it is the Dum lncRNA. The RNA-binding region of DNMT3A has not been mapped, and conservation of these RNAs in horse has not been shown. General lncRNA-binding capacity may well be conserved, but the evidence available here does not establish it for horse.
These observations intrigued us to investigate whether EZH2 and DNMT3a may directly interact within an protein:RNA complex together with PTENpg1 asRNA Ξ±.
Summary: Methylation of its own catalytic cysteine is a side reaction that DNA substrate suppresses.
Reason: Mouse Dnmt3a slowly methylates its own active-site cysteine in DNA-free assays. The reaction is not seen when CpG DNA is present. The catalytic cysteine is conserved in horse (C671), so the capacity should transfer. A physiological protein-methylation role has not been established, so this is kept as non-core.
In the presence of CpG containing dsDNA, which is the natural substrate for Dnmt3a, the transfer of the methyl group from AdoMet to the flipped target base was preferred and auto-methylation was not detected.
GO:0141196 transposable element silencing by piRNA-mediated DNA methylation
IEA GO_REF:0000107
KEEP AS NON CORE
Summary: In the male germline, DNMT3A catalyzes de novo methylation of transposons targeted by piRNAs. This is a developmental use of its core activity.
Reason: In this process, DNMT3A writes the methylation that piRNA pathways direct, so it participates rather than merely being required. The ortholog data come from mouse male germ cells, and the catalytic domain is fully conserved in horse. Horse has no rodent-specific DNMT3C, which does not argue against a DNMT3A contribution. Whether this splice form is made in horse germ cells is unknown. The row is kept as non-core, matching the human review.
Immunofluorescence of DNMT1, DNMT3a and DNMT3b on testis samples confirmed the differential expression and localization of these three DNA methylation transferases during horse development.
The catalytic PC-motif cysteine (human C710 -> horse C671), ENV-motif E756 -> E717, the CpG-recognition TRD R836 -> R797 and the tetramer-interface R882 -> R843 are all conserved.
Summary: DNMT3A forms catalytic heterotetramers with DNMT3L or DNMT3B3. The complex term is true but broad.
Reason: The structures show DNMT3A-DNMT3L and DNMT3A2-DNMT3B3 assemblies. No more specific DNMT3A-complex term is in use, so the broad term is kept as non-core, matching the human review.
the DNMT3A-DNMT3L-DNA complex in which two DNMT3A monomers simultaneously attack two cytosine-phosphate-guanine (CpG) dinucleotides
Core Functions
De novo DNA cytosine-5 methyltransferase. It methylates CpG sites in chromatin and silences CpG-island promoters and other targets. In this exon-skipped horse model, the catalytic domain, ADD domain and tetramerization interface are retained, but PWWP-guided heterochromatin targeting is probably compromised.
The observation that these mutant proteins efficiently methylate other genomic sequences indicates that the PWWP domains are not required for enzymatic activities but are involved in functional specialization.
The catalytic PC-motif cysteine (human C710 -> horse C671), ENV-motif E756 -> E717, the CpG-recognition TRD R836 -> R797 and the tetramer-interface R882 -> R843 are all conserved.
Q: Is the exon-9-skipping DNMT3A transcript (ENSECAT00000098563) expressed in horse tissues? If so, does the resulting PWWP-truncated product lose pericentric-heterochromatin targeting while staying catalytically active?
Suggested Experiments
Experiment: Use RT-PCR across exons 8-10 of horse DNMT3A in testis and somatic tissues to measure exon-9 inclusion. Then compare the localization of GFP-tagged full-length and exon-9-skipped horse DNMT3A, and their major-satellite methylation, in Dnmt3a/3b-deficient ES cells.
External Prediction Reviews
These computational predictions are reviewed separately from the GOA annotation set used for this review. The assessments below are from this project and do not constitute official GO annotations or endorsement by GO/UniProt. They are not included in the existing annotation review above.
Both ProtNLM2 predictions are correct but less specific than supported annotations the gene already carries. Gene-expression regulation is refined by GO:0044027 (CpG-island methylation), and nuclear localization by nucleoplasm. The PWWP-region exon skip limits heterochromatin-targeting claims but leaves the NLS regions, ADD domain and catalytic domain intact.
LSP β Less precise than existing annotation Review score: 2/2
Prediction method: ProtNLM2 Β· Version: UniProt API snapshot 2026-09-08
Review rationale: Gene-expression regulation is correct for DNMT3A, which silences genes by de novo CpG methylation, and the horse model retains the ADD and catalytic domains. The gene already carries the accepted, more specific descendant GO:0044027 negative regulation of gene expression via chromosomal CpG island methylation, which states the mechanism; the predicted term is its ancestor (via epigenetic regulation of gene expression). The PWWP-region exon skip changes heterochromatin targeting but not methylation of other genomic sequences.
Supporting Evidence:
PMID:29414941: "the DNMT3A-DNMT3L-DNA complex in which two DNMT3A monomers simultaneously attack two cytosine-phosphate-guanine (CpG) dinucleotides"
PMID:23042785: "DNMT3A recruitment is increased, with hypermethylation being associated with silencing of gene expression."
PMID:30478443: "Substitutions in the PWWP domain abrogate binding to the histone modifications H3K36me2 and H3K36me3, and alter DNA methylation in patient cells."
PMID:15456878: "The observation that these mutant proteins efficiently methylate other genomic sequences indicates that the PWWP domains are not required for enzymatic activities but are involved in functional specialization."
LSP β Less precise than existing annotation Review score: 2/2
Prediction method: ProtNLM2 Β· Version: UniProt API snapshot 2026-09-08
Review rationale: DNMT3A is nuclear, and mouse deletion mapping places its two NLS basic stretches around the PWWP domain, outside the exon skipped in this horse model, which also retains the chromatin-associated ADD and catalytic domains. The gene already carries the accepted, more specific descendant nucleoplasm (GO:0005654). Horse testis immunofluorescence is relevant but does not identify this accession or resolve compartments in the cached abstract.
Supporting Evidence:
PMID:15456878: "These results suggested that the two basic sequences are functional NLSs and that they are both required for association with the major satellite repeats."
PMID:23042785: "co-expression of WT1 and DNMT3A proteins in nuclei of blastemal cells in human fetal kidney and Wilms' tumours."
PMID:12138111: "Unlike Dnmt3a, which was concentrated on heterochromatin, Dnmt3a2 displayed a localization pattern suggestive of euchromatin association."
PMID:38705424: "Immunofluorescence of DNMT1, DNMT3a and DNMT3b on testis samples confirmed the differential expression and localization of these three DNA methylation transferases during horse development."