Arrb1 (rat) review notes

Session 1 (initial review)

• fetch-gene resolved P29066 (ARRB1_RAT, Swiss-Prot, RGD:2156, GeneID 25387). 136 GOA rows
  in the stub (138 in the tsv; the two duplicated SynGO rows for postsynaptic density and
  glutamatergic synapse were merged by the stub generator).
• No deep-research provider available; literature synthesis is in Arrb1-deep-research-manual.md.
• Papers read in full text: PMID:10725339, 19171933, 19254952, 19339617, 20650893, 20705669,
  23604254, 37209686, 38175886. All others abstract-only.

Framing

Beta-arrestin-1 is a GPCR adaptor: it binds GRK-phosphorylated, agonist-activated receptors
(desensitization), links them to clathrin/AP-2 (internalization), and scaffolds signaling
(ERK, Src, AKT) and E3 ligases (AIP4/ITCH, MDM2) PMID:20650893. Nuclear/transcriptional roles
(p300 recruitment, CREB, LEF) are real but secondary.

Decisions worth flagging

• Phototransduction ISO (from mouse IMP PMID:11853756): the paper uses rod arrestin (Sag)
  knockouts PMID:11853756. QuickGO shows the
  mouse IMP is on Arrb1 (with MGI:2388373) and not on Sag for this paper. Looks like a gene
  mis-attribution at MGI; rat ISO row REMOVEd and a source defect recorded.
• Cysteine-type endopeptidase inhibitor activity ISO (from mouse IMP PMID:19915011): the
  mechanism shown is ERK1/2-p90RSK phosphorylation of Bad, not caspase inhibition. REMOVEd.
• GTPase activator activity ISO (from human IMP PMID:15611106, RhoA activation with Gq):
  RhoA activation is GTP loading, not GAP activity. MARK_AS_OVER_ANNOTATED.
• Negative regulation of ERK1/2 IMP (PMID:19254952): beta-arrestin-1 siRNA raised ERK only at
  the 2-min point PMID:19254952;
  consistent with the reciprocal regulation in UniProt. Kept non-core.
• PMID:19171933 (beta-arrestin-2 anti-apoptotic): beta-arrestin-1 knockdown reduced caspase-3
  cleavage in rat VSMCs PMID:19171933, opposite sign to the
  astrocyte IMP (PMID:24378649). Apoptosis effects of beta-arrestin-1 are cell-context
  dependent; all apoptosis rows kept non-core, no NEW apoptosis terms.
• No NEW terms proposed for processes; the only NEW row is the MF "ubiquitin-like
  ligase-substrate adaptor activity" (GO:1990756) for the AIP4-TRPV4 adaptor function, which
  is direct participation (the paper shows beta-arrestin-1 bridges the ligase to its substrate).

2026-10-28 review-fix pass

• Phototransduction ISO row: softened 'Beta-arrestin-1 is not studied' to what the abstract of PMID:11853756 (abstract only) supports, namely that it describes rod-arrestin-deleted and transducin-deleted mice and does not mention beta-arrestin-1.
• source_defects corrected_value is now MGI:MGI:98227 (mouse Sag, S-arrestin; MGI cross-reference of UniProt P20443), with the name in corrected_label.
• Removed supported_by self-citations of the gene's own -deep-research-manual.md where a primary citation in the same block already supports the row (the reviewer's own conclusion is not independent evidence). The validator now warns that no annotation cites the deep-research file; that warning is accepted deliberately, which sets the status to DRAFT.