NCOA1/SRC1 (Q15788) is a broadly used p160 coactivator, not an AR-exclusive or universally required AR subunit. PMID:7481822(https://pubmed.ncbi.nlm.nih.gov/7481822/) establishes receptor interaction and enhanced steroid-receptor transcription. PMID:9223431(https://pubmed.ncbi.nlm.nih.gov/9223431/) distinguishes SRC1 splice isoforms and their thyroid-receptor coactivation. PMID:10567563(https://pubmed.ncbi.nlm.nih.gov/10567563/) directly supports SRC1 interaction with AR AF1 through residues 1053–1123: LXXLL-disrupted SRC1 retained AR potentiation in the tested context. This does not rule out LBD/AF2 recruitment in other experiments. The normal cache is abstract-only; indexed primary Figure 7 text was also checked as a locator, without claiming a full new article read.
The core is transcription coactivator activity. Nuclear receptor binding and transcriptional activation are supported independently across sources, including papers whose headline concerns another cofactor. PMID:15831516(https://pubmed.ncbi.nlm.nih.gov/15831516/) explicitly reports SRC1 ChIP recruitment to the RARA intronic module; this supports chromatin localization without making SRC1 the sequence-specific DNA-binding factor. PMID:17363140(https://pubmed.ncbi.nlm.nih.gov/17363140/) reports SRC1-dependent cooperation of HAT and SWI/SNF machinery in retinoid transcription. PMID:9556555(https://pubmed.ncbi.nlm.nih.gov/9556555/) extends coactivation to NF-kappaB p50, showing why an AR-only summary would be inadequate.
PMID:9296499(https://pubmed.ncbi.nlm.nih.gov/9296499/) reports intrinsic H3/H4 histone acetyltransferase activity of SRC1. The broad existing lysine acetyltransferase assertion is refined to histone acetyltransferase activity. The curated UniProt account explicitly cautions that the relevance of this activity is unclear; the review therefore does not inflate it into the obligatory mechanism of all SRC1 coactivation. Recruited CBP/p300/PCAF chemistry is distinct from SRC1's own reported activity.
Source-specific binding decisions use direct target statements from the cited studies, including receptor domains, PRMT2, UBCH7, TDG and AhR. The shared PMID16373399 PPAR/insulin/adipogenesis assertions remain uncertain: its accessible abstract emphasizes PGC1alpha and drug effects, but no misattribution is claimed without the missing full comparator assays. Likewise, the human developmental/behavioral transfers and direct-DNA-binding claims require their specific ortholog experiments; a domain name, brain expression or a knockout phenotype on another participant is not substituted for that evidence.
Normal fetch-gene seeded the review and immutable UniProt/GOA sources. Default Falcon research with Perplexity-lite fallback was attempted concurrently with publication caching; both providers failed (authentication/quota errors in .cache/module-audit-logs/NCOA1-research.log). No manual text was written under a provider-generated filename. Primary PubMed/PMC caches were fetched through the normal tools; source files were not manually edited. The normal cache frequently contains only abstracts. Positive target-specific abstract statements can support a claim; missing target detail is recorded as UNDECIDED, never as a wrong-gene conclusion based solely on the title. Full-text inspection was targeted to the passages specified below, rather than a claim to have read every source in full.
All original annotation IDs, labels, evidence codes, qualifiers, supporting entities and isoform fields are preserved. Only review judgments and author-supplied synthesis are changed. No NEW annotations are proposed. Current GO term definitions/obsolescence were checked through QuickGO (2026-10-26; diagnostic JSON in .cache/module-audit-logs/coregulator-terms.json). Obsolete original source IDs remain intact; author-proposed replacements use current terms. The PAINT assertions are not treated as pairwise transfers or rejected based on donor count. No independent reconstruction of their whole phylogeny is claimed.
Bare protein binding is refined when the cited evidence resolves a useful partner class. Uninformative but verified cofactor associations are removed under the current GO:0005515 policy, without denying the interaction. Unresolved pair tables, constructs or electronic source claims remain UNDECIDED. Chromatin occupancy does not by itself establish direct or sequence-specific DNA recognition. The original bile-transport pathway mappings are flagged as overextended: GO:0015721 explicitly describes directed movement by a transporter/pore or equivalent, whereas SRC proteins supply transcriptional regulation.
The review by Sun and Xu (PMID:32449767(https://pubmed.ncbi.nlm.nih.gov/32449767/)) supplies broader context: animal models establish many endocrine, neural and behavioral roles, and coregulator functions overlap without being identical. Its expression maps are not direct evidence that every human ortholog executes every brain-development process. The review's target-specific mechanisms and animal-model scope were inspected; no cancer association is promoted to an intrinsic core molecular function.
124 source assertions reviewed; 42 ACCEPT, 3 KEEP_AS_NON_CORE, 2 MARK_AS_OVER_ANNOTATED, 28 MODIFY, 5 REMOVE, 44 UNDECIDED. The status is COMPLETE with explicit unresolved claims, not a statement that all evidence gaps have been closed.