Tp53 (rat, P10361) curation notes

Re-review 2026-10-10

What GOA changed

• GOA refresh seeded 135 new rows (now 351 total). Most are qualifier or donor splits of
  existing ISO terms: MGI now uses acts_upstream_of_or_within for mouse Trp53 (MGI:98834)
  phenotype-based annotations, and human TP53 rows are split by isoform (P04637-2, -4, -7 etc.).
  About 60 carried terms the review had not seen, mostly mouse phenotype terms
  (development, lymphocyte, cardiac, metabolic).
• New IBA rows for chromatin and nucleus with is_active_in, an IEA protein tetramerization
  row with involved_in, and two acts_upstream_of IMP rows for PMID:30514107.
• One row is retired: GO:0005515 IPI PMID:15872011 (nucleophosmin).

Pending rows (135): 29 ACCEPT, 77 KEEP_AS_NON_CORE, 24 MARK_AS_OVER_ANNOTATED, 1 MODIFY, 4 UNDECIDED

• Mouse developmental phenotypes (in utero/embryo development, somitogenesis, gastrulation,
  CNS/cerebellum/heart development, septum morphogenesis, organismal growth, lymphocyte
  lineage/thymic differentiation) were marked over-annotated. The donor studies are
  compound mutants (Gcn5, Jmjd5, Rps6, Mdm2/Mdm4, PACT, ASPP2) in which p53 activation
  causes the defect PMID:26334721
  PMID:17000767.
• Mitochondrial matrix rows (ISO/ISS) kept as non-core:
  PMID:22726440
• Endoplasmic reticulum rows over-annotated (p53 is a substrate of the ER ligase synoviolin)
  PMID:17170702.
• Necroptotic process (GO:0070266) MODIFY to programmed necrotic cell death (GO:0097300): the
  p53 mechanism is cyclophilin D/permeability-transition necrosis, not death-receptor necroptosis.
• Protein import into nucleus, protein stabilization, chromosome organization, histone
  deacetylase regulator activity, gene expression, regulation of tissue remodeling marked
  over-annotated (p53 is cargo/upstream trigger, or the term is generic)
  PMID:15657445.
• UNDECIDED (4):
• GO:0035861 site of double-strand break (mouse IDA PMID:24550317): the cached paper only
  mentions 53BP1 ["tumor protein P53 binding protein foci formation"]; possible Trp53bp1 confusion at MGI.
• GO:1903799 negative regulation of miRNA processing (PMID:21522133): the paper shows
  p53-dependent miR-29 transcription, not processing.
• GO:0045861 negative regulation of proteolysis (PMID:23277542): the paper studies
  "p53-independent c-Myc-induced apoptosis".
• NOT GO:0001701 in utero embryonic development (PMID:9425892, a correspondence with no
  abstract) conflicts with the positive row from other compound mutants.

Changed actions on existing rows (7)

• GO:0005515 protein binding (IPI PMID:11278372, TAFII31): MARK_AS_OVER_ANNOTATED -> MODIFY to
  GO:0001223 transcription coactivator binding PMID:11278372.
• GO:0005515 (IPI PMID:8389468, WT1): MARK_AS_OVER_ANNOTATED -> MODIFY to GO:0061629
  PMID:8389468.
• GO:0005515 (IPI PMID:12464630, nucleostemin): MARK_AS_OVER_ANNOTATED -> REMOVE (generic,
  no more specific MF supported; interaction not disputed).
• GO:0005515 (IPI PMID:15872011, nucleophosmin, retired): MARK_AS_OVER_ANNOTATED -> REMOVE.
• GO:0005759 mitochondrial matrix (IEA SL-0170): MODIFY (-> outer membrane) -> KEEP_AS_NON_CORE;
  the earlier "not in the matrix" premise is contradicted by PMID:22726440 and PMID:24101517.
• GO:0031000 response to caffeine (IEP PMID:16544096): KEEP_AS_NON_CORE -> MARK_AS_OVER_ANNOTATED;
  caffeine was used as an ATM inhibitor ["caffeine (ATM kinase inhibitor) inhibited Ser-15 phosphorylation"].
• GO:0010038 response to metal ion (IEP PMID:17466256): KEEP_AS_NON_CORE -> MARK_AS_OVER_ANNOTATED;
  the platinum drug acts via DNA damage ["attack on DNA, leading to fast activation of p53 and caspase-3"].

Other edits

• Added supported_by (verbatim UniProt CC, deep-research or cited-paper quotes) to 185
  existing rows that had none; added the cited PMIDs and UniProtKB:P10361 to references.
• Description: removed curation commentary ("captured here", "GOA set"); added the
  transcription-independent necrosis role.
• core_functions unchanged (TF activity, DNA-damage response, p53-mediated intrinsic
  apoptosis, G1 checkpoint).

Open questions

• See suggested_questions for the three donor-side checks behind the UNDECIDED rows.
• Whether the many cell-type-specific apoptosis/proliferation ISO terms from mouse should
  transfer to rat at all remains a judgment call; they were kept as non-core.

Review follow-up 2026-10-17: in response to PR #4629 review, two rows that had inherited the
shared "compound-mutant (or sensitized) genotype" rationale were rewritten against their own
donor papers. GO:0021549 cerebellum development: the donor PMID:24013501 (Hamard et al. 2013,
mouse IDA) is a single-locus Trp53(deltaCTD/deltaCTD) knock-in, not a compound mutant; reason
and source comment now argue from that paper (the deleted C-terminal domain mis-regulates p53
output, cerebellar p53 levels unchanged), the unrelated Jmjd5 quote (PMID:26334721) was replaced
by verbatim quotes from PMID:24013501, and the reference was added. GO:0009792 embryo development
ending in birth or egg hatching: the donor PMID:16702401 (ASPP2) shows lethality enhanced on a
p53+/- background (synthetic lethality, loss-of-function direction), so the "inappropriate p53
activation" mechanism was removed; MARK_AS_OVER_ANNOTATED is kept on necessity-vs-participation
grounds (and the lethality is largely postnatal). The other nine rows with the shared comment
have IGI donors that fit it and were left unchanged.

Review follow-up 2026-10-17 (second round, PR #4629): GO:0001756 somitogenesis still carried
the shared "compound-mutant" mechanism and the borrowed Jmjd5 quote (PMID:26334721). The mouse
IGI's recorded reference, PMID:17320900, is a Dlk1/Notch adipogenesis cell-culture paper
(abstract only; no p53, no somites), so no mechanism is taken from it; MGI records the IGI
partner as Kat2a (MGI:1343101). The nearest readable Kat2a/p53 somite evidence is PMID:17325035
["We show here that elimination of the p53 tumor suppressor allows Gcn5 −/− embryos to survive
longer and to develop somites"], which is quoted as such, not as the confirmed source. The row
keeps MARK_AS_OVER_ANNOTATED on the participation argument (p53 acts upstream of presomitic
mesoderm survival under either reading), the source entity is marked UNRESOLVED, PMID:17320900
is listed in references with a MISCITED reference_review, and the reference mismatch is raised
as a suggested question for MGI. GO:0007417 central nervous system development now quotes its
own donor, PMID:16199867 (Npm1 knockout: p53 activation, apoptosis, p53-null rescue), instead
of the generic Jmjd5 sentence; MGI records the IGI partner as Lig4 (MGI:1335098). The
GO:0001701 summary no longer says all three donors show lethality from p53 activation: the
Msh2 donor (PMID:9288110) reports lethality in Msh2/p53 double-null mice, the loss-of-function
direction. Validation: 0 errors; the pre-existing GO:0001701 consistency warning remains.