CCND1 (cyclin D1, P24385) — curation working notes

Identity

Human CCND1 at 11q13.3 encodes cyclin D1 (295 aa), historically BCL-1 and PRAD1; one of
three mammalian D-type cyclins (CCND1/2/3). UniProt P24385 CHAIN 1..295 with a
Cyclin N-terminal domain at 28..152, a disordered C-terminal region 262..295, and
MOD_RES 286 phosphothreonine. Not to be confused with CCND2/CCND3 or the cyclin D1b
splice variant, whose altered C terminus removes the Thr286 degron
(CCND1-deep-research-falcon.md, "Alternative splicing produces cyclin D1b, whose altered
C terminus removes important localization and degradation controls").

Core molecular function: non-catalytic CDK4/6 activator

Cyclin D1 is a regulatory subunit with no intrinsic catalytic activity; the kinase
chemistry is done by CDK4/CDK6 within the holoenzyme.

Consequence for review: the two orthology-transferred catalytic annotations
(GO:0004672 protein kinase activity, GO:0016301 kinase activity, both IEA from mouse
Ccnd1 P25322) are wrong in kind and were removed. The activator/regulator terms
(GO:0061575, GO:0016538, GO:0043539) and the complex terms (GO:0000307, GO:0097128,
GO:0097131) carry the real biology. GO:0043539 (generic protein Ser/Thr kinase activator
activity) was modified to the CDK-specific child GO:0061575.

GO-CAM agreement: gocams/index.tsv types CCND1 in both the AMBRA1 model
(61e0e55600000624) and the FBXO32 model (680ad14200006016) as
GO:0061575 cyclin-dependent protein serine/threonine kinase activator activity acting in
GO:0000082 G1/S transition of mitotic cell cycle — the same MF/BP pair adopted in
core_functions.

Assembly, localization and turnover

Consequence for review: in all of these, cyclin D1 is the substrate of the ligase, not
a participant in the ubiquitination process, so no ubiquitination/catabolism process term
was proposed (proposed_new_terms: []). The machine-readable substrate relationship
belongs on the ligases as has input; this is raised in suggested_questions.

CDK-independent transcriptional corepressor role

Genuine second function, explicitly curated by UniProt: "Exhibits transcriptional
corepressor activity with INSM1 on the NEUROD1 and INS promoters in a cell
cycle-independent manner (PubMed:16569215, PubMed:18417529)".

So GO:0000122, GO:0003714 and GO:0017053 were accepted, and the corepressor role appears
as a second core_functions entry.

Reviewing the bare protein binding (GO:0005515) block — 76 rows

Repository practice: resolve to an informative MF where the cited evidence supports one,
otherwise REMOVE as uninformative (which does not assert the interaction is false).
Applied by partner:

WITH/FROM partner action
CDK4 (P11802), CDK6 (Q00534), CDK2 (P24941), mouse Cdk4 (P30285) MODIFY → GO:0019901 protein kinase binding
FBXO31 (Q5XUX0), UHRF2 (Q96PU4), CDC27/APC3 (P30260) MODIFY → GO:0031625 ubiquitin protein ligase binding
HDAC3 (O15379) MODIFY → GO:0042826 histone deacetylase binding
AR (P10275) MODIFY → GO:0035257 nuclear receptor binding
INSM1 (Q01101), NRF1 (Q16656) MODIFY → GO:0008134 transcription factor binding
p21/CDKN1A (P38936), p27/CDKN1B (P46527, Q96TE0), BRCA1, SARS-CoV N, Ralbp1, SMARCA4, SMAD4, FBXW7 REMOVE (uninformative)

Notes on two of these: GO:0050681 "nuclear androgen receptor binding" is obsolete
(QuickGO), hence GO:0035257 nuclear receptor binding for the AR row. The three rows from
PMID:35512704 come from a mutation-directed neo-interaction screen, so they are not
evidence for a wild-type molecular function and were removed rather than resolved.
GO:0019899 enzyme binding (WITH USP2, O75604) was modified to the informative child
GO:1990381 ubiquitin-specific protease binding (UniProt: "Interacts with USP2").

Substrate-of-a-process vs participant: the DNA-damage rows

PMID:19412162 supports IDA rows for GO:0006974 DNA damage response and GO:0031571
mitotic G1 DNA damage checkpoint signaling. The mechanism is that damage stabilizes
FBXO31, which degrades cyclin D1, producing G1 arrest — i.e. cyclin D1 is the terminal
target
whose removal implements the arrest, not a transducer of checkpoint signalling.
Per the CLAUDE.md participation rule (being required for, consumed by, or acted on by a
process is not participation) both were marked over-annotated rather than accepted;
neither was removed, since the observation itself is sound and the papers' full text was
not consulted beyond the abstract. The CDK-independent HR/RAD51 role reported in the deep
research ("Reducing cyclin D1, but not merely inhibiting CDK4/6, decreased RAD51
recruitment after DNA damage") is a different claim, not evidenced by this reference,
and was not used to accept these rows.

Pleiotropic / propagated rows

Not proposed as new terms

Comparator check run before declining a ubiquitination/catabolism term for the substrate
role: GO's convention places the process on the enzyme. Cyclin D1's own GO-CAM
appearances (gocams/61e0e55600000624, gocams/680ad14200006016) already model it in the
role GO intends — as the CDK activator and as the ligases' input molecule — so there is no
gap to fill. proposed_new_terms: [].