DNMT3B is a nuclear DNA cytosine-5 methyltransferase that helps establish DNA methylation patterns and regulate gene expression. Catalytically competent forms prefer CpG sites but also methylate non-CpG cytosines under tested conditions. Association with DNMT3L, other DNA methyltransferases and chromatin regulators influences activity and targeting. A distinct, catalytically inactive splice product, DNMT3B3, binds the nucleosome acidic patch and positions DNMT3A-containing complexes on chromatin. DNMT3B-dependent methylation contributes to repeat stability and transcriptional silencing; pathogenic variants cause a subset of immunodeficiency, centromeric instability and facial anomalies syndrome. Context-specific positive effects on gene expression are also reported. DNMT3B also helps maintain epigenetic repression of D4Z4 repeats. Hypomorphic variants can contribute to somatic DUX4 derepression and digenic facioscapulohumeral muscular dystrophy type 4 when repeat length and a permissive chromosome 4 haplotype provide the required genetic context.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000122 negative regulation of transcription by RNA polymerase II | IBA GO_REF:0000033 | ACCEPT | Summary: DNMT3B participates in transcriptional repression. Reason: The literal GO:0000122 IBD is on PANTHER:PTN002747452, an ancestor of the human DNMT3B leaf in PTHR23068. Mouse Dnmt3b evidence includes methylation-associated repression of Ment in a MYC-driven lymphoma model; the human target also has its own experimental repression annotation. These are positive grounds for the inherited regulatory function. The target appearing among the descendant sources is legitimate experimental grounding, not circularity. This process assertion does not require concluding that DNMT3B meets every mechanistic requirement of the distinct corepressor molecular-function term. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN002747452 SUPPORTS TRANSFER Exact IBD and target descent recovered; mouse experimental repression and human target evidence support the placement. |
| GO:0000122 negative regulation of transcription by RNA polymerase II | IDA PMID:17303076 Zinc-fingers and homeoboxes 1 (ZHX1) binds DNA methyltransfe... | ACCEPT | Summary: Tethered DNMT3B represses transcription in the ZHX1 study. Reason: The complete abstract reports enhanced repression when DNMT3B is directly tethered to DNA and associates with ZHX1. This supports the positive repression process at gene level. The original full Methods were not recovered, so the precise reporter and construct conditions remain curator-supported rather than independently reconstructed. This positive process evidence is distinct from the unresolved direction of recruitment required to assign a corepressor molecular function. Supporting Evidence: PMID:17303076 ZHX1 enhanced the transcriptional repression mediated by DNMT3B when DNMT3B is directly targeted to DNA. |
| GO:0000122 negative regulation of transcription by RNA polymerase II | IEA GO_REF:0000117 | ACCEPT | Summary: The matched ARBA repression prediction is supported by target evidence. Reason: DNMT3B satisfies the literal FunFam conditions in branch 41 of ARBA00029133. The predicted negative regulation is independently supported by the DNMT3B-tethered repression study and by conserved methylation-associated silencing. No erroneous paralog transfer or target-specific loss was identified. Propagation Review Root cause: NO FAILURE CORE Sources checked: ARBA:ARBA00029133 SUPPORTS TRANSFER Branch 41 of 95 matches the official target FunFam identifiers; human repression evidence supports the predicted process. |
| GO:0003677 DNA binding | IBA GO_REF:0000033 | MODIFY | Summary: Direct DNA-binding evidence supports double-stranded DNA binding. Reason: The DNA-binding IBD at PTN002747452 legitimately contains human DNMT3B and experimentally characterized mouse Dnmt3b. The source mouse record includes PMID:21378119, whose original Results also test human DNMT3B1 and DNMT3B3 against a double-stranded SAT2 DNA probe. The positive ancestral claim is therefore refined to the measured substrate class, without asserting exclusive binding to this repeat or equivalent affinity for every natural product. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: GRANULARITY MISMATCH Sources checked: PANTHER:PTN002747452 SUPPORTS TRANSFER Exact ancestral DNA-binding assertion and target descent recovered; mouse and human SAT2 duplex assays justify greater substrate precision. Proposed replacements: double-stranded DNA binding |
| GO:0003714 transcription corepressor activity | IDA PMID:17303076 Zinc-fingers and homeoboxes 1 (ZHX1) binds DNA methyltransfe... | UNDECIDED | Summary: Repression is established, but the current corepressor mechanism remains unresolved. Reason: The current GO:0003714 definition requires repression via binding to a DNA-binding transcription factor at a specific genomic locus. The accessible abstract establishes ZHX1 binding and enhanced repression of DNA-tethered DNMT3B, but does not resolve which partner recruits which or establish that defined transcription-factor bridge. The positive interaction and repression observations are retained; unavailable full Methods prevent adjudicating the narrower molecular-function assertion. This is not evidence that the complete experiments lack such a mechanism. Supporting Evidence: |
| GO:0003714 transcription corepressor activity | IEA GO_REF:0000117 | UNDECIDED | Summary: The corepressor ARBA rule matches, but its mechanistic interpretation is unresolved. Reason: The target matches branch 8 of ARBA00089152. Direct target papers support interaction-dependent repression, but the accessible original scope does not yet establish the DNA-binding-transcription-factor recruitment mechanism specified by GO:0003714. A correct rule match alone does not settle that term-level uncertainty. The prediction is neither rejected as a paralog error nor treated as disproven by inaccessible Methods. Propagation Review Root cause: UNRESOLVED Sources checked: ARBA:ARBA00089152 UNRESOLVED Branch 8 of 15 matches the target. Positive repression evidence exists, but the exact GO corepressor recruitment mechanism remains unresolved. |
| GO:0003714 transcription corepressor activity | IMP PMID:18567530 DNA methyltransferase 3B acts as a co-repressor of the human... | UNDECIDED | Summary: DNMT3B promotes hPc2-dependent repression, with the strict corepressor assignment unresolved. Reason: The complete abstract reports DNMT3B–hPc2/CBX4 interaction, increased FGFR3 expression after DNMT3B depletion, and methyltransferase-independent enhancement of hPc2 repression. These are positive functional results. They do not, within the accessible scope, resolve recruitment through a DNA-binding transcription factor as required by the current GO definition. Full original Methods and locus-level recruitment details remain inaccessible, so uncertainty is retained rather than denying the reported repression. Supporting Evidence: |
| GO:0003886 DNA (cytosine-5-)-methyltransferase activity | IDA PMID:16543361 Mutations in DNA methyltransferase DNMT3B in ICF syndrome af... | ACCEPT | Summary: Human DNMT3B performs DNA cytosine methylation. Reason: The recovered original includes recombinant methyltransferase assays and repeats relevant experiments with human DNMT3 proteins. The assay uses DNA substrate and radiolabeled S-adenosylmethionine; cell-based episomal methylation is a separate readout. This supports cytosine-5 methyltransferase activity. This source does not separately quantify non-CpG substrates, and no activity is transferred to all splice products. Supporting Evidence: |
| GO:0003886 DNA (cytosine-5-)-methyltransferase activity | IEA GO_REF:0000120 | ACCEPT | Summary: The enzyme and reaction mappings identify the correct cytosine-5 methylation chemistry. Reason: The matched ARBA00088999 branch, RHEA:13681 and EC:2.1.1.37 agree with direct human DNMT3B methyltransferase assays. The broad cytosine-5 activity is appropriate because the catalytic-domain study measures a preference for CpG together with lower CpA and CpT activity; CpG specificity is not exclusivity. Catalytically inactive DNMT3B3 remains a distinct product-specific case. Propagation Review Root cause: NO FAILURE CORE Sources checked: ARBA:ARBA00088999 SUPPORTS TRANSFER Branch 1 of 4 matches; its chemistry agrees with the independent EC/RHEA mappings and measured human activity. |
| GO:0003886 DNA (cytosine-5-)-methyltransferase activity | TAS PMID:10433969 Cloning, expression and chromosome locations of the human DN... | ACCEPT | Summary: The historical human DNMT3B catalytic assertion agrees with direct later measurements. Reason: PMID:10433969 establishes human DNMT3B cloning and its relationship to murine de novo methyltransferases; only its complete abstract was accessible. The curated catalytic statement is retained with independently inspected human activity evidence from PMID:16543361 and PMID:32083663. Those later assays corroborate the gene-level activity without being misrepresented as recovered experiments from the 1999 paper. |
| GO:0005515 protein binding | IPI PMID:11735126 Dnmt3b, de novo DNA methyltransferase, interacts with SUMO-1... | KEEP AS NON CORE | Summary: DNMT3B associates with SUMO1 and is a SUMOylation substrate. Reason: The source-specific IntAct records and complete abstract support N-terminal interaction assays with SUMO1 and modification of DNMT3B by SUMO1. These positive observations are retained under the project policy for generic binding. They do not establish DNMT3B as a SUMO-conjugating enzyme or, without resolving covalent versus noncovalent recognition, justify a dedicated SUMO-binding function. The later noncovalent result for the S270P disease mutant is not automatically assigned to wild-type protein. |
| GO:0005515 protein binding | IPI PMID:11735126 Dnmt3b, de novo DNA methyltransferase, interacts with SUMO-1... | MODIFY | Summary: Targeted interaction with Ubc9 supports ubiquitin-like conjugating enzyme binding. Reason: The literal partner P63279 is UBE2I/Ubc9, the SUMO E2 enzyme. The original abstract and matched Y2H, pull-down and co-immunoprecipitation records identify the DNMT3B N-terminal interaction in a SUMOylation study. This targeted interaction and partner enzyme class support the binding refinement; DNMT3B itself is the modified substrate, not the conjugating catalyst. Original full Methods remain unavailable. Proposed replacements: ubiquitin-like protein conjugating enzyme binding Supporting Evidence: PMID:11735126 Deletion analysis and colocalization experiment demonstrated that Dnmt3b interacts with SUMO-1 and Ubc9 at its N-terminal region. |
| GO:0005515 protein binding | IPI PMID:16357870 The Polycomb group protein EZH2 directly controls DNA methyl... | KEEP AS NON CORE | Summary: EED association places DNMT3B in a Polycomb-associated interaction context. Reason: Matched EED–DNMT3B records include HeLa co-immunoprecipitation and pull-down evidence. The inspected publisher supplement supports the experimental context. EED is a noncatalytic PRC2 component, so its association is not evidence of binding a histone methyltransferase enzyme or of DNMT3B catalyzing histone methylation. Retain the positive generic interaction under project policy. The linked figure correction is disclosed in the reference assessment. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:16357870 The Polycomb group protein EZH2 directly controls DNA methyl... | MODIFY | Summary: Targeted EZH2 interaction supports histone methyltransferase binding. Reason: EZH2 is the histone methyltransferase partner. Source-specific pull-down records and the original supplementary Methods/corrected supplementary panel description support interaction of DNMT3B constructs with GST–EZH2 fragments, alongside endogenous HeLa association. This is stronger partner-recognition evidence than membership in an untargeted complex alone and supports GO:1990226. The corrected lane presentation is disclosed; no affinity, full-length natural-product mapping, or DNMT3B histone-catalytic activity is inferred. Proposed replacements: histone methyltransferase binding Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:16983096 Recruitment of the de novo DNA methyltransferase Dnmt3a by K... | KEEP AS NON CORE | Summary: LANA physically associates with DNMT3B, although downstream recruitment preferentially involves DNMT3A. Reason: The actual original tests DNMT3B despite its DNMT3A-focused title. GST–LANA fragments bind in-vitro-translated DNMT3B, and tagged proteins co-immunoprecipitate in HEK293T cells. These are positive target-specific interaction results, not a purified all-component affinity measurement. The strong DNMT3A relocalization and promoter-recruitment results must not be transferred to DNMT3B. Retain this viral interaction as non-core; failure to retrieve the pair from a separate interaction-database query is not negative evidence. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:17303076 Zinc-fingers and homeoboxes 1 (ZHX1) binds DNA methyltransfe... | KEEP AS NON CORE | Summary: ZHX1 binds DNMT3B and enhances its repression in the reported tethering assay. Reason: The complete abstract maps an interaction between the DNMT3B PWWP region and a ZHX1 homeobox region and reports nuclear colocalization. The positive interaction is retained. A homeobox annotation and repression result alone do not settle the direction and locus specificity of a transcription-factor recruitment mechanism, so no more specific mechanistic binding term is imposed from the accessible abstract. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:18953337 De novo DNA methylation promoted by G9a prevents reprogrammi... | MODIFY | Summary: Targeted G9a interaction supports histone methyltransferase binding. Reason: The original reports G9a/EHMT2 ankyrin-dependent Dnmt3a/b recruitment, with recombinant/in-vitro-translated pull-down, cellular co-immunoprecipitation and endogenous mouse embryonic-stem-cell evidence. G9a is a histone methyltransferase, so these targeted interaction experiments support the partner-class binding term. The study includes mouse and human constructs; the species of every cited Dnmt3b plasmid was not independently resolved, and a human host is not proof that all expressed proteins are human. The imported human partner assertion is preserved with that limit. Proposed replacements: histone methyltransferase binding Supporting Evidence: PMID:18953337 G9a recruits Dnmt3a and Dnmt3b through its ANK domain |
| GO:0005515 protein binding | IPI PMID:19798101 Np95 interacts with de novo DNA methyltransferases, Dnmt3a a... | KEEP AS NON CORE | Summary: UHRF1 associates with DNMT3B in a chromatin-regulatory context. Reason: The original primarily studies mouse Np95/Dnmt3 proteins, but explicitly reports endogenous human DNMT3B–UHRF1 co-immunoprecipitation in HEK293T cells in its supplementary result. Thus this is not a mouse-only or wrong-gene citation. The association is retained without assigning UHRF1 ubiquitin-ligase chemistry to DNMT3B or inferring a direct binary interface from co-immunoprecipitation. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:22094255 Oxidative damage targets complexes containing DNA methyltran... | KEEP AS NON CORE | Summary: EED and DNMT3B occur in oxidative-stress-associated complexes. Reason: The matched NCCIT records report density-gradient colocalization and spoke-expanded co-immunoprecipitation in the oxidative-damage study. The accessible abstract independently describes complexes containing DNMTs and Polycomb proteins. These support association, not direct EED recognition or a distinct DNMT3B enzymatic activity. The nominal full-text cache lacks the operative Methods/Results, and the original was not recovered; the positive curated association is retained with that explicit access limit. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:22094255 Oxidative damage targets complexes containing DNA methyltran... | KEEP AS NON CORE | Summary: EZH2 associates with DNMT3B in the reported oxidative-stress complex. Reason: These source records are gradient colocalization and spoke-expanded co-immunoprecipitation in NCCIT cells. They support a positive complex association without independently resolving direct EZH2 recognition in this experiment. The separate targeted GST–EZH2 study supports the more specific binding refinement on its own row; it does not convert this source into a binary assay. Retain the source-specific generic association under project policy and disclose the unavailable operative original text. |
| GO:0005515 protein binding | IPI PMID:24074865 Dnmt3L antagonizes DNA methylation at bivalent promoters and... | KEEP AS NON CORE | Summary: DNMT3B–EZH2 association changes in the presence of Dnmt3L. Reason: The recovered original Figure 6D and Methods report endogenous DNMT3B/EZH2 co-immunoprecipitation in HEK293 cells with inducible FLAG-tagged murine Dnmt3L. This supports the association and its experimental modulation. The purified interaction experiment elsewhere concerns Ezh2–Dnmt3L, not purified DNMT3B–EZH2. The cellular complex result is retained without inventing a DNMT3B interface or treating the murine regulator as human. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:27728808 Enhancing the Cytotoxic Effects of PARP Inhibitors with DNA ... | KEEP AS NON CORE | Summary: DNMT3B is recovered with EZH2-containing complexes in the damage-response study. Reason: Matched NCCIT records include fractionation-associated co-immunoprecipitation with and without hydrogen peroxide, several represented by spoke expansion. The original main text explicitly describes DNMT1/3B-containing Polycomb/histone complexes. These positive results support association; the detailed supplementary target assays were not independently recovered, and neither their absence from the inspected main text nor the predominance of DNMT1 in the narrative negates DNMT3B. No direct histone-enzyme recognition is asserted for this source. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32968275 Structure of nucleosome-bound DNA methyltransferases DNMT3A ... | KEEP AS NON CORE | Summary: DNMT3B3 forms a functional chromatin-targeting complex with DNMT3A. Reason: The original structure and reconstitution establish a human DNMT3A–DNMT3B3 complex on nucleosomes. The deposited full DNMT3B3 sequence exactly matches Q9UBC3-3, which is catalytically inactive but binds the nucleosome acidic patch and positions active DNMT3A toward linker DNA. Generic protein binding is retained as a positive source assertion; the informative nucleosome-binding role is synthesized in the product-qualified core rather than inferred as autonomous DNMT3B methylation from this association. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | KEEP AS NON CORE | Summary: BioPlex detects a DNMT3A–DNMT3B association in HEK293T cells. Reason: The exact source record identifies DNMT3A as bait and DNMT3B as prey in a HEK293T affinity-purification/mass-spectrometry experiment, with both sequence/taxon identifiers authenticated. This corroborates the pair at gene level. AP-MS does not identify the natural DNMT3B product or establish a binary interface by itself; the independently studied DNMT3B3 nucleosome complex is kept distinct. Retain the positive generic association under project policy. |
| GO:0005634 nucleus | IBA GO_REF:0000033 | MODIFY | Summary: The conserved nuclear localization can be refined to supported nucleoplasmic and chromatin pools. Reason: The exact nuclear IBD is at PTN001497036 and the human target descends from it. Its broad cross-lineage grounding is valid, including target experimental evidence. Human HPA immunofluorescence supports nucleoplasm, and the sequence-authenticated DNMT3B3 reconstitution establishes a chromatin-bound targeting role. These provide a more informative location set without rejecting the ancestral nucleus assertion. Nucleoplasm and chromatin describe supported, nonexclusive pools at gene level; this does not assign both pools to every natural product or claim that the cited experiment resolved both. Current chromatin ancestry does not contain a typed nucleus edge, but that graph limit does not negate the experimentally supported nuclear chromatin pool. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: GRANULARITY MISMATCH Sources checked: PANTHER:PTN001497036 SUPPORTS TRANSFER Literal nuclear IBD and target descent are present. Independent human evidence refines the broad location; no lineage loss is alleged. Proposed replacements: nucleoplasm chromatin |
| GO:0005634 nucleus | IDA PMID:17303076 Zinc-fingers and homeoboxes 1 (ZHX1) binds DNA methyltransfe... | MODIFY | Summary: Nuclear ZHX1 colocalization is compatible with more precise independently supported DNMT3B pools. Reason: The source abstract reports nuclear DNMT3B/ZHX1 colocalization but does not resolve subnuclear anatomy. The refinement uses independently inspected HPA nucleoplasm evidence and DNMT3B3 nucleosome-targeting experiments, while retaining the original study at its actual nuclear resolution. Nucleoplasm and chromatin describe supported, nonexclusive pools at gene level; this does not assign both pools to every natural product or claim that the cited experiment resolved both. Current chromatin ancestry does not contain a typed nucleus edge, but that graph limit does not negate the experimentally supported nuclear chromatin pool. Proposed replacements: nucleoplasm chromatin |
| GO:0005634 nucleus | IEA GO_REF:0000120 | MODIFY | Summary: The matched nuclear-location prediction is refined using independent human localization evidence. Reason: ARBA00026330 branch 539 matches the target FunFam/taxon conditions, and UniProt SL-0191 supplies the nuclear-location mapping. Both are sound broad predictions. HPA supports nucleoplasm and the DNMT3B3 study supports chromatin association. Nucleoplasm and chromatin describe supported, nonexclusive pools at gene level; this does not assign both pools to every natural product or claim that the cited experiment resolved both. Current chromatin ancestry does not contain a typed nucleus edge, but that graph limit does not negate the experimentally supported nuclear chromatin pool. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: GRANULARITY MISMATCH Sources checked: ARBA:ARBA00026330 SUPPORTS TRANSFER Branch 539 of 1293 is an exact match; SL-0191 is compatible. Refinement is based on independent human target evidence. Proposed replacements: nucleoplasm chromatin |
| GO:0005634 nucleus | NAS PMID:32083663 Structural insights into CpG-specific DNA methylation by hum... | MODIFY | Summary: The broad nuclear statement is refined without treating a catalytic fragment as a localization assay. Reason: PMID:32083663 establishes human catalytic-domain DNA methylation and structure, not a whole-cell localization experiment for every product. The nuclear statement is consistent with independent HPA nucleoplasm evidence and the DNMT3B3 nucleosome-bound complex. Nucleoplasm and chromatin describe supported, nonexclusive pools at gene level; this does not assign both pools to every natural product or claim that the cited experiment resolved both. Current chromatin ancestry does not contain a typed nucleus edge, but that graph limit does not negate the experimentally supported nuclear chromatin pool. Proposed replacements: nucleoplasm chromatin |
| GO:0005634 nucleus | TAS PMID:10433969 Cloning, expression and chromosome locations of the human DN... | MODIFY | Summary: The historical nuclear assertion is refined with later direct human evidence. Reason: The complete PMID:10433969 abstract concerns human cloning and expression; the original full localization methods were not recovered. The broad curated nuclear assertion is not rejected. Independently inspected HPA and DNMT3B3 experiments justify the gene-level nucleoplasm/chromatin refinement without pretending to recover the 1999 subnuclear assay. Nucleoplasm and chromatin describe supported, nonexclusive pools at gene level; this does not assign both pools to every natural product or claim that the cited experiment resolved both. Current chromatin ancestry does not contain a typed nucleus edge, but that graph limit does not negate the experimentally supported nuclear chromatin pool. Proposed replacements: nucleoplasm chromatin |
| GO:0005654 nucleoplasm | IDA GO_REF:0000052 | ACCEPT | Summary: Human immunofluorescence supports nucleoplasmic DNMT3B. Reason: The current HPA target record is authenticated by ENSG00000088305 and Q9UBC3. Its ICC/IF summary and location element explicitly report supported nucleoplasm. This is a read of the assay record, not an independent interpretation of microscopy pixels, and it does not resolve natural splice products or exclude chromatin-associated protein. Supporting Evidence: GO_REF:0000052 |
| GO:0005654 nucleoplasm | TAS Reactome:R-HSA-212222 | ACCEPT | Summary: Reactome places the DNMT3B participant recruited by PRC2 in nucleoplasm. Reason: The saved event path is R-HSA-212222 input 212315, whose member 212316 is human DNMT3B at nucleoplasm. The participant is annotated as residues 1–853 of Q9UBC3. This supports the location assertion in the modeled recruitment event; it is not an assay of every isoform. Supporting Evidence: Reactome:R-HSA-212222 |
| GO:0005654 nucleoplasm | TAS Reactome:R-HSA-212269 | ACCEPT | Summary: DNMT3B is a nucleoplasmic participant in the modeled DNA/PRC2 methylation complex. Reason: The catalyst path 212269 → 212303 → 212288 → 212315 → 212316 reaches human DNMT3B in nucleoplasm. DNMT3B supplies DNA cytosine methylation, whereas the PRC2/EZH2 component supplies histone methylation. The location is accepted without transferring the complete complex reaction to DNMT3B alone. Supporting Evidence: Reactome:R-HSA-212269 |
| GO:0005654 nucleoplasm | TAS Reactome:R-HSA-4655374 | ACCEPT | Summary: The SUMOylation event places its DNMT3B substrate in nucleoplasm. Reason: R-HSA-4655374 explicitly uses participant 212316, human DNMT3B in nucleoplasm, as an input to SUMO1 modification. The location assertion is supported. Being the substrate of SUMOylation does not give DNMT3B SUMO-transferase activity or justify a new process-participation annotation. Supporting Evidence: Reactome:R-HSA-4655374 |
| GO:0005654 nucleoplasm | TAS Reactome:R-HSA-9710490 | ACCEPT | Summary: The modeled GSDME-promoter methylation event contains nucleoplasmic DNMT3B. Reason: The saved catalyst/complex/member path from R-HSA-9710490 reaches participant 212316. This establishes the source model’s DNMT3B location. It does not independently measure the GSDME reaction or establish equivalent promoter targeting by every DNMT3B product. Supporting Evidence: Reactome:R-HSA-9710490 |
| GO:0008168 methyltransferase activity | IEA GO_REF:0000002 | MODIFY | Summary: The broad methyltransferase mapping can be specified as DNA cytosine-5 methylation. Reason: The literal InterPro2GO mapping from IPR001525 correctly identifies methyltransferase activity. Human DNMT3B methylation assays establish the DNA cytosine-5 chemistry, so the general term is refined to GO:0003886. This is a specificity improvement rather than a failed family transfer. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: GRANULARITY MISMATCH Sources checked: InterPro:IPR001525 SUPPORTS TRANSFER Exact InterPro2GO mapping recovered; direct target chemistry supports a more specific term. Proposed replacements: DNA (cytosine-5-)-methyltransferase activity |
| GO:0009008 DNA-methyltransferase activity | IEA GO_REF:0000002 | MODIFY | Summary: The DNA-methyltransferase prediction is refined to its measured cytosine-5 chemistry. Reason: IPR030488 maps to DNA-methyltransferase activity and identifies the DNMT3B ADD domain context. Direct human biochemical evidence supports the more specific cytosine-5 term. The ADD-domain mapping is not itself proof that every shortened natural product retains catalytic activity. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: GRANULARITY MISMATCH Sources checked: InterPro:IPR030488 SUPPORTS TRANSFER The literal mapping is correct but broad; independent target assays determine substrate chemistry. Proposed replacements: DNA (cytosine-5-)-methyltransferase activity |
| GO:0009008 DNA-methyltransferase activity | NAS PMID:10325416 The human DNA methyltransferases (DNMTs) 1, 3a and 3b: coord... | MODIFY | Summary: The historical DNA-methyltransferase statement is refined with direct human catalytic evidence. Reason: The accessible original abstract reports human DNMT expression and alternative transcripts, while discussing mouse catalytic precedents. It is not independently reconstructed as a purified human enzyme assay. The accepted broad enzyme statement can nevertheless be refined to DNA cytosine-5 methyltransferase activity using the later direct human assays, without inferring equal activity of the alternatively spliced transcripts. Proposed replacements: DNA (cytosine-5-)-methyltransferase activity |
| GO:0009008 DNA-methyltransferase activity | NAS PMID:10555141 DNA methyltransferases Dnmt3a and Dnmt3b are essential for d... | MODIFY | Summary: The conserved de novo methyltransferase function has a defined cytosine-5 chemistry. Reason: The complete source abstract reports mouse gene-targeting and methylation evidence, together with human ICF mutations. Those organism-specific observations are retained. Independent human biochemical assays justify refining the broad DNA-methyltransferase term to cytosine-5 activity; no purified human experiment is attributed to the 1999 mouse study. Proposed replacements: DNA (cytosine-5-)-methyltransferase activity |
| GO:0009008 DNA-methyltransferase activity | NAS PMID:11102980 Genetic variation in ICF syndrome: evidence for genetic hete... | MODIFY | Summary: Human genetic evidence is consistent with the independently measured cytosine-5 methyltransferase activity. Reason: The source abstract reports DNMT3B mutations in a subset of patients with ICF syndrome and discusses repeat hypomethylation. It does not establish equal enzymatic effects for all variants. The gene-level DNA-methyltransferase assertion is refined using independent direct human methylation assays, rather than treating variant association as a new catalytic measurement. Proposed replacements: DNA (cytosine-5-)-methyltransferase activity |
| GO:0010468 regulation of gene expression | IEA GO_REF:0000002 | MODIFY | Summary: Gene-expression regulation includes supported negative and context-specific positive effects. Reason: The IPR030488 mapping is correct but broad. The supported refinements preserve three nonexclusive contexts: RNA polymerase II repression, methylation-based epigenetic silencing, and the distinct positive BAG1 expression effect. Mouse inducible Dnmt3b-loss experiments link promoter demethylation to Ment derepression; human lymphoma data are correlative. Human DNMT3B variant studies independently connect D4Z4 hypomethylation with conditional DUX4 derepression. Together with established methyl-transfer chemistry, these support a performed epigenetic silencing role without asserting a new human CpG-island-specific or constitutive-heterochromatin formation assay. The positive BAG1 context is not generalized to the methylation mechanism. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: GRANULARITY MISMATCH Sources checked: InterPro:IPR030488 SUPPORTS TRANSFER The exact broad regulatory mapping is supported; target and ortholog evidence distinguish transcription repression, methylation-based epigenetic silencing and a positive BAG1 context. |
| GO:0010628 positive regulation of gene expression | IMP PMID:18413740 DNA methyltransferase 1 and 3B activate BAG-1 expression via... | KEEP AS NON CORE | Summary: DNMT3B positively affects BAG1 expression in the reported cellular context. Reason: The complete abstract reports lower BAG1 expression after DNMT1 or DNMT3B knockout and increased expression after overexpression, with CTCFL/BORIS recruitment and promoter chromatin changes. This positive source assertion is retained as a contextual regulatory role. The cache’s full-text flag does not supply the missing operative Methods/Results, so those details are not independently reconstructed and DNMT3B is not assigned histone methyltransferase chemistry. Supporting Evidence: |
| GO:0051718 DNA (cytosine-5-)-methyltransferase activity, acting on CpG substrates | IBA GO_REF:0000033 | ACCEPT | Summary: The inherited CpG methyltransferase function agrees with the measured substrate preference. Reason: The literal CpG-methyltransferase IBD at PTN002747452 is on the target lineage, grounded by mouse Dnmt3b and Dnmt3c experimental sources. The lineage judgment is not weakened by donor count or by the rodent-specific biology of Dnmt3c. Direct human DNMT3B assays independently establish efficient CpG methylation. Lower CpA/CpT activity does not contradict the ability to methylate CpG; it only prevents describing CpG as the exclusive substrate. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN002747452 SUPPORTS TRANSFER Exact CpG IBD and target descent recovered. Both mouse experimental annotations and direct human chemistry support the inherited capacity. |
| GO:1902494 catalytic complex | IPI PMID:32083663 Structural insights into CpG-specific DNA methylation by hum... | MODIFY | Summary: The catalytic complex is specifically a DNA methyltransferase complex. Reason: The human DNMT3B catalytic-domain/DNMT3L structure and reconstitution identify a DNA methyltransferase complex, a typed descendant of catalytic complex. The deposited DNMT3B sequence is the canonical 571–853 fragment with an N-terminal GHM remnant; it is not a full-length natural isoform. The separate nucleosome-bound DNMT3A–DNMT3B3 complex provides another supported context for the same complex class, without assigning catalytic activity to inactive DNMT3B3. Proposed replacements: DNA methyltransferase complex |
| GO:0031491 nucleosome binding | IDA Q9UBC3-3 PMID:32968275 Structure of nucleosome-bound DNA methyltransferases DNMT3A ... | NEW | Summary: Sequence-authenticated DNMT3B3 anchors a DNMT3A-containing complex to the nucleosome acidic patch. Reason: The full deposited DNMT3B3 sequence in PDB 6PA7 entity 8 exactly matches the current 770-residue Q9UBC3-3 product. Reconstitution used human DNMT proteins expressed in Sf9 and Xenopus core histones. Cryo-EM identifies DNMT3B3 acidic-patch contacts; AlphaScreen/native-gel assays and charge-reversal mutants support nucleosome binding. R740E/R743E reduce binding and chromatin association in human DKO8 cells while the DNMT3A-containing complexes retain naked-DNA methylation. This is an accessory targeting role, not intrinsic methyltransferase activity of DNMT3B3. Current typed ancestry shows no redundancy with the existing DNA/protein-binding assertions. Supporting Evidence: |
Loading supporting content…
Download this section (compressed HTML)Q: In the ZHX1 and hPc2 repression systems, which DNA-binding transcription factor recruits DNMT3B at the endogenous target locus, and does the recruitment direction meet the current GO corepressor definition?
Q: Which of the remaining sequence-defined natural DNMT3B products have catalytic activity or chromatin-targeting roles comparable to the authenticated DNMT3B3 accessory product?
Q: How do native DNMT3B/DNMT3L and DNMT3A/DNMT3B3 assemblies distribute across nucleoplasmic and chromatin-bound pools in relevant human cell types?
Experiment: Use endogenous tagging, reciprocal domain-selective perturbation and locus-resolved occupancy assays to distinguish DNMT3B recruitment by a DNA-binding factor from repression after artificial DNA tethering. Measure methylation and transcription separately.
Hypothesis: DNMT3B repression in the ZHX1/CBX4 systems depends on a defined locus-specific recruitment interface.
Type: Endogenous recruitment and separation-of-function analysis
Experiment: Express sequence-verified natural products, measure methylation on matched naked DNA and nucleosomes, and test complex formation with DNMT3A and DNMT3L under controlled stoichiometry. Include inactive DNMT3B3 as an accessory-function comparator.
Hypothesis: Sequence-defined natural splice products have separable catalytic and nucleosome-targeting capacities.
Type: Product-resolved biochemical reconstitution
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)