CREBBP/CBP combines protein lysine acetylation with transcriptional coactivation. These are separate molecular claims, although they cooperate in many promoter contexts. The human-CBP preparation in the AML1 study acetylated histones and AML1; the MOZ–CBP fusion’s inhibitory behavior must not be attributed to normal CREBBP. The source states, “Human MOZ and CBP cDNAs were isolated” (PMID:11742995(https://pubmed.ncbi.nlm.nih.gov/11742995/), construct Methods). Together with its actual activity assays, this supports protein-lysine acetyltransferase activity rather than only one histone residue. Mouse CBP/p300 double deletion supports H3K18/H3K27 dependence, but is not a purified human CBP-only assay (PMID:21131905(https://pubmed.ncbi.nlm.nih.gov/21131905/)). The ASF1 work provides additional CBP-specific histone-acetylation evidence; its main article was read, with the supplementary access limit retained (PMID:24616510(https://pubmed.ncbi.nlm.nih.gov/24616510/)).
Promoter recruitment is independently supported. The rat SPT promoter experiments in human HepG2 cells show increased Sp1/PKA transcription with CBP; the paper reports enhanced activity “with coexpression of CBP” (PMID:12169688(https://pubmed.ncbi.nlm.nih.gov/12169688/), abstract). This does not mean CBP itself recognizes the Sp1 DNA sequence. P/CAF recruitment and native CBP–P/CAF association support cofactor complexes, while the acetylase activity assayed in the original P/CAF paper belongs to P/CAF/hGCN5 (PMID:8684459(https://pubmed.ncbi.nlm.nih.gov/8684459/)). The p21/NGF study assays p300 in rat PC12 cells; CBP’s background coactivator/HAT assertions remain biologically valid without relabeling those assays as CBP experiments (PMID:10362258(https://pubmed.ncbi.nlm.nih.gov/10362258/)).
Coactivation is not the only regulatory direction. The KLF4/HBG study contains endogenous human erythroid interaction and chromatin observations, together with a murine CREBBP expression construct that represses the assayed reporter. This supports contextual corepression and preserves the broader coregulator annotation (PMID:21539536(https://pubmed.ncbi.nlm.nih.gov/21539536/)). CBP’s acetylation of PAF53/POLR1E and fibrillarin provides other performed negative regulatory mechanisms in Pol I transcription (PMID:24207024(https://pubmed.ncbi.nlm.nih.gov/24207024/), PMID:30540930(https://pubmed.ncbi.nlm.nih.gov/30540930/)). None makes CBP a universal repressor.
The two official proteins reconstruct exactly as 2,442 and 2,404 residues. Q92793-2 replaces canonical residues 406–444 with one alanine, overlapping the TAZ1 recruitment region. The KIX, bromodomain and catalytic regions remain, but shared sequence does not establish equal recruitment or activity. The E4 study’s N-terminal functional interval also overlaps the altered region; that study does not authenticate isoform 2. The C-terminal ten-residue peptide used in the PDZ screen is shared by the two official sequences, which establishes peptide identity rather than full-length functional equivalence.
MRE11 lactylation represents an additional performed catalytic activity. The study reports, “MRE11 was specifically lactylated at its second DBD, K673.” (PMID:38128537(https://pubmed.ncbi.nlm.nih.gov/38128537/), Discussion). Actual Methods use purified GST-MRE11 and HEK293T-derived HA-CBP with lactyl-CoA; target mutation and depletion experiments connect modification to end resection and homologous recombination. This supports the existing lactyltransferase and repair-regulation assertions and a narrowly described catalytic unit. It does not establish that all cellular lactylation or every repair pathway is executed by CBP.
The separate PCNA study establishes direct fragment-level CBP–PCNA binding, damaged-chromatin association, PCNA acetylation and repair-associated turnover: “PCNA is a substrate of both CBP and p300” (PMID:24939902(https://pubmed.ncbi.nlm.nih.gov/24939902/), Results summary). Those positive observations do not yet settle the narrower damaged-DNA binding assertion. Protein-mediated recruitment to damaged chromatin and direct binding to lesion-containing DNA remain distinguishable mechanisms; both original lesion-binding assertions are therefore left unresolved without rejecting the positive repair contribution.
Nutrient-responsive OTUD3 acetylation and enhanced OTUD3 nuclear localization are supported by actual modifying-enzyme and substrate-mutant experiments. Human OTUD3 is explicit, whereas assay host alone does not establish the CBP insert’s species. The obesity-associated G288D phenotype is not explained simply by altered acetylation or localization (PMID:35675826(https://pubmed.ncbi.nlm.nih.gov/35675826/), modifying-enzyme Results and constructs Methods). Ku70 K542 acetylation and cytoplasmic CBP accumulation are another distinct substrate context (PMID:15023334(https://pubmed.ncbi.nlm.nih.gov/15023334/)). These observations make nuclear localization nonexclusive.
The p53 study separates chain extension from necessity for degradation. Its two-step assay primes p53 with MDM2, removes that enzyme and adds CBP with fresh ubiquitination components. The source confirms, “MDM2 had been effectively removed before the addition of CBP polypeptides.” (PMID:19805293(https://pmc.ncbi.nlm.nih.gov/articles/PMC2752525/), externally recovered original, two-step-assay Results). Bacterial CBP fragments and endogenous human cytoplasmic fractions provide additional support. A contextual GO:0061630 ubiquitin protein ligase annotation therefore captures actual catalytic work. No additional degradation-process annotation is proposed.
The narrower GO:0034450 definition currently specifies a product thioester, which does not describe the measured amide-linked chain extension. This is an ontology question rather than a reason to erase the positive result. The existing 47 term ancestries and five target GO-CAM activities contain no covering ubiquitin MF. The cited construct source names human U47741; however, its pcDNA3.1 vector name differs from pRSV in the E4 article (PMID:10500189(https://pubmed.ncbi.nlm.nih.gov/10500189/), constructs Methods). Main Methods and the figure caption also differ in quantity units. Neither discrepancy is silently resolved, and no unaccessed supplementary or image evidence is claimed.
Partner classes are assigned from the actual partner identity and evidence, not from co-localization requirements. TP53, DNA-binding transcription factors, non-DNA-binding coregulators, beta-catenin, RNA helicases, histones and enzyme partners are distinguished. Positive generic interactions remain when a finer supported class is unavailable. Binding a catalytic partner does not give CREBBP that partner’s activity.
TRERF1 binds promoter DNA, so the original non-DNA-binding coactivator-binding term is too narrow. DAXX and CITED2 require the broader coregulator-binding class because their source-specific regulatory roles are not uniformly activating. NAP1L1’s histone-chaperone role does not make it a protein-folding chaperone; its histone-dependent binding and native co-IP evidence remain positive. Binding POLR1E alone does not establish binding the assembled Pol I core.
The ASC2/NCOA6 fractionation paper reports partial CBP comigration but a negative co-IP in its ASCOM fractions. That observation is preserved and the paper is flagged as miscited for a positive association at that specific assay scope. Independent CBP–NCOA6 evidence supports the retained refined binding claim (PMID:12482968(https://pubmed.ncbi.nlm.nih.gov/12482968/), ASCOM Results/Figure 2B; PMID:10866662(https://pubmed.ncbi.nlm.nih.gov/10866662/)). It does not follow that every CBP–NCOA6 association is absent.
The PDZ screen measures an isolated CREBBP ten-residue C-terminal peptide against 25 PDZ-domain constructs. These are positive domain-binding observations, not native full-length encounter tests (PMID:36115835(https://pubmed.ncbi.nlm.nih.gov/36115835/)). The authentic correction changes specified figure labeling and formatting and is preserved separately (PMID:36477203(https://pubmed.ncbi.nlm.nih.gov/36477203/)). The importin study similarly compares WT and K13E CREBBP residues 5–19 against engineered KPNA4 pockets. Actual Table 1 shows WT binding and weaker or abolished K13E binding; literal mutation-feature wording is not evidence that WT lacks binding. This supports the binding class without proving native full-length nuclear import (PMID:39009827(https://pubmed.ncbi.nlm.nih.gov/39009827/), Table 1 and adjacent Results).
Several positive curated assertions have limited source resolution. The SNAIL paper’s abstract directly reports SNAIL binding and acetylation, whereas the GOA pair names TP53; that exact pair is retained with curator deference, not described as abstract-verified (PMID:25314079(https://pubmed.ncbi.nlm.nih.gov/25314079/)). The NFE4 and IRF2 abstracts foreground PCAF or p300; their CBP assertions remain qualified rather than rejected from incomplete text. The Pokemon–SMAD4 paper reports HDAC1 recruitment and reduced SMAD4–p300/CBP interaction, with SMAD2/3 activation unchanged; it does not report CBP acetylation of SMAD2/3 (PMID:25514493(https://pubmed.ncbi.nlm.nih.gov/25514493/)). The positive contextual TGF-beta annotation is retained at its curated cofactor scope.
The viral N studies are also kept distinct. One original uses a human CBP catalytic fragment and reports faint acetylation despite its PCAF-focused abstract (PMID:33894414(https://pubmed.ncbi.nlm.nih.gov/33894414/)). The other’s actual publisher main and Extended Data Figure 4 support CBP gain/loss effects on N K375 acetylation in human cells, without establishing CBP insert species from the host (PMID:34239064(https://pubmed.ncbi.nlm.nih.gov/34239064/)). Its 2026 correction replaces specified panels/source data and explains statistical formatting; no retraction or correction of that particular CBP experiment is asserted (PMID:42557414(https://pubmed.ncbi.nlm.nih.gov/42557414/)). Corrected raw measurements were not independently reanalyzed.
The source has 281 complete GOA records grouped into 278 deterministic assertions. Three extra literal rows share their grouping keys; all are preserved, including 116 raw IPI records represented by 113 grouped IPI assertions. No source field or ordinary qualifier is changed.
The actual PTHR13808 tree contains 435 nodes; its human target has an exact 2,442-residue MSA sequence. All six ancestral assertions are on the target path at PTN000351795. The 36 donor joins retain 259 literal occurrences from complete current donor records. Self-source records are legitimate grounding, not circularity. Mouse and rat p300 chromatin ChIP provide donor evidence at chromatin-associated DNA resolution, not purified human CBP DNA-affinity measurements (PMID:18458156(https://pubmed.ncbi.nlm.nih.gov/18458156/), PMID:20702579(https://pubmed.ncbi.nlm.nih.gov/20702579/)). All 16 ARBA rules have current exact target matches; current matches are not reconstructed historical rule premises.
The 65 Reactome roots have positive typed paths to one of two literal Q92793 entities, nuclear or cytosolic. Event-specific compartment assertions remain distinct from their partners’ catalytic steps. Five complete target activities in four GO-CAM models were inspected, retaining their molecular inputs and causal edges; this is not a claim to have read every activity in those models.
The source evidence index retains full source identities, raw record projections, sequence checks, actual access boundaries and public paper locators. Cached publication bytes remain unchanged. An externally recovered original can have been read even when the protected cache is abstract-only; those two access statements are kept separate. The automatic research request timed out and its fallback returned an authorization error, so these are manual evidence notes rather than a fabricated provider report.
The E1A studies report binding and competitive recruitment rather than a shared E1A acetylation mechanism. The TRAM study maps CBP interactions with E1A and p53, including competition with MDM2 (PMID:10196247(https://pubmed.ncbi.nlm.nih.gov/10196247/)); the Ad12 comparison separates CBP binding from a p300-specific repression result (PMID:10049825(https://pubmed.ncbi.nlm.nih.gov/10049825/)). The SRCAP study links CBP association to increased transcription and reports E1A disruption of that interaction (PMID:10347196(https://pubmed.ncbi.nlm.nih.gov/10347196/)).
The chetomin paper is a primary HIF study whose detailed abstract experiment concerns p300 CH1; the retained CBP TAS annotation draws on its coactivator context without transferring the p300 assay (PMID:15261140(https://pubmed.ncbi.nlm.nih.gov/15261140/)). The ZMYND8 paper is a bromodomain-protein damage-localization screen, with the exact CREBBP nuclear panel left at curator-grounded scope because it was not personally recovered (PMID:25593309(https://pubmed.ncbi.nlm.nih.gov/25593309/)).
Publication-cache selection preserves the exact existing main files. Some contain only an abstract or abstract plus Discussion despite a broader availability flag. Richer original passages used above remain distinct read evidence with public original locators and source hashes in the evidence index; a cache-selection change does not turn those passages into text present in the protected cache.
The current IBD for transcription regulator complex (GO:0005667, original row 206) includes an additional FlyBase donor, FB:FBgn0261617, compared with the original GOA snapshot. The other five IBA donor sets match. Original WITH/FROM values are preserved; current donor membership is not represented as historical equality or as a reason to challenge the conserved node placement.
The recovered NRC paper provides the original construct and binding Methods and the CBP-specific Results. Full-length GST-tagged human NRC associates with endogenous CBP in 293T extracts, whereas GST alone and the more highly expressed NRC 1–852 fragment do not. This supports the existing coactivator-binding refinement through a cellular association assay, without establishing a purified direct pair or a minimal interaction interface. The authors distinguish possible C-terminal contact from effects on NRC folding. The result remains separate from the later ASCOM-fraction negative co-IP (PMID:10866662(https://pmc.ncbi.nlm.nih.gov/articles/PMC85955/), construct/binding Methods and CBP-association Results; PMID:12482968(https://pubmed.ncbi.nlm.nih.gov/12482968/)). No CREBBP isoform-specific result or figure-pixel read is inferred.
The project’s Curation instructions explicitly retain supported positive generic protein-binding annotations as KEEP_AS_NON_CORE when no evidence-backed narrower class has been established. This standing instruction applies to the 17 retained generic associations. It does not turn an uninspected source panel into verified evidence; curator deference and assay limits remain explicit.
Reference identity, access and support for a particular assertion are separate questions. Twelve reference assessments now say UNVERIFIED where the actual read boundary leaves the precise CREBBP source claim unresolved. Their existing positive annotation decisions are preserved with curator deference. Direct target statements in an abstract can still verify a bounded claim; an abstract-only flag alone is not a reason to downgrade every source. Short literal anchors have been added for catalytic, coactivation, partner-binding and cytoplasmic claims, with source-level quotation totals checked across this journal and the review. The index records their hashes and review locators rather than repeating the quotations.
The earlier journal statement that independent NRC evidence refines the PMID:12482968 source row is superseded here. Rereading the complete available main Methods, Results and Discussion found a negative CBP co-IP in ASCOM fractions, with partial comigration, and no positive CBP–ASC2 assay in that text. That original IPI row is now UNDECIDED; its source fields remain unchanged. The separate positive NRC row remains refined on its own evidence (PMID:12482968(https://pmc.ncbi.nlm.nih.gov/articles/PMC140670/), ASCOM fractionation Results, Figure 2B experiment; PMID:10866662(https://pmc.ncbi.nlm.nih.gov/articles/PMC85955/), CBP-association Results, Figure 14A experiment). Complete figure captions and pixels were not newly read. The limited negative result does not imply that all CBP–ASC2 interactions are absent.
The coactivator core now also cites the Cart1 study’s explicit CBP transcriptional stimulation, while preserving its p300-specific localization and acetylation distinctions (PMID:12929931(https://pubmed.ncbi.nlm.nih.gov/12929931/), structured abstract). Histone and nonhistone lysine acetylation remain one catalytic unit with explicitly stated substrate breadth. The contextual p53 chain-extension activity is described at its demonstrated cytoplasmic scope and the authored new molecular-function assertion explicitly uses enables; original GOA qualifiers are unchanged.
Current GO:0070577 is obsolete. Its replacement guidance points to histone-reader activity, which carries an additional chromatin-adaptor meaning; the existing histone-binding refinement already represents the measured acetylated-H4 interaction without manufacturing a redundant new assertion (PMID:24361270(https://pmc.ncbi.nlm.nih.gov/articles/PMC3923519/), structure and peptide-binding Results). The damaged-DNA-binding rows remain unresolved: actual PCNA interaction and recruitment to UV-damaged chromatin do not distinguish lesion-selective DNA contact from protein-mediated recruitment, and absence of a resolving assay is not a demonstrated contradiction (PMID:24939902(https://pmc.ncbi.nlm.nih.gov/articles/PMC4117764/), chromatin-association Methods and Results).
For IFNAR2, the current term definition does not encode agonism or extracellular binding. The abstract explicitly reports intracellular receptor contact, but exact-term annotation usage is predominantly interferon ligands; the checked intracellular comparators do not establish a precedent for this exact term. The refinement therefore retains its explicit contact-only scope and records the convention question (PMID:17923090(https://pubmed.ncbi.nlm.nih.gov/17923090/)). Full-text access is not claimed.