SKP2 (human, UniProt Q13309) — curation notes

Identity and architecture

SKP2 (FBXL1, p45^SKP2, cyclin-A/CDK2-associated protein p45) is a ~45 kDa F-box/leucine-rich-repeat
(FBXL-family) protein. UniProt Q13309 features an F-box domain at 95–137 and ten LRRs (138–385).
The F-box binds SKP1; the LRR solenoid plus accessory factors form the substrate-recognition surface.
The deep research report confirms identity and family placement and stresses that SKP2 is not a kinase
and not the catalytic ubiquitin-transfer subunit
[file:human/SKP2/SKP2-deep-research-falcon.md "SKP2 is the substrate receptor of SCF^SKP2 rather than an autonomous catalytic enzyme."].

Core molecular function: SCF(SKP2) substrate receptor

UniProt: "Substrate recognition component of a SCF (SKP1-CUL1-F-box protein) E3 ubiquitin-protein
ligase complex which mediates the ubiquitination and subsequent proteasomal degradation of target
proteins involved in cell cycle progression, signal transduction and transcription".

The Cul1-Rbx1-Skp1-F box^Skp2 crystal structure establishes the architecture: CUL1 is a rigid scaffold
whose globular domain binds RBX1 (E2 recruitment) and whose stalk tip
PMID:11961546 substrate-recognition module, holding the two

100 Å apart. SKP2 thus provides specificity, not catalysis
PMID:23911321
and its SKP1 contact is through the F-box
PMID:23911321.

Canonical substrate: Thr187-phosphorylated p27/CDKN1B, with CKS1B

p27 proteolysis is triggered by CDK-mediated Thr187 phosphorylation, and — uniquely among SCF
substrates — needs the accessory subunit CKS1
PMID:16209941. The
Skp1–Skp2–Cks1/phospho-p27 structure shows a composite receptor
PMID:16209941. Mutational work localizes the
CKS1 contact to Skp2 Asp-331
PMID:12813041 and shows
the requirement is for p27 turnover
PMID:12813041.
Independent work restates the model
PMID:16880511 Thr187-phosphorylated p27.

Functional consequence: loss of p45^SKP2 function blocks S-phase entry
PMID:7553852, and p27 removal
releases cyclin E/A–CDK2, giving a positive-feedback G1/S switch
[file:human/SKP2/SKP2-deep-research-falcon.md "p27 ubiquitination and proteasomal degradation promote G1/S progression."].

Additional degradative substrates (graded)

Non-degradative K63 branch: NBN/NBS1 in DSB repair

PMID:22464731 ubiquitination of NBS1 after DSBs,
which is required for NBS1–ATM interaction and ATM recruitment to foci;
PMID:22464731 recombination repair.
The production GO-CAM gocams/65a1f4f800001029/ models exactly this: SKP2 (Q13309) with
GO:1990756 ubiquitin-like ligase-substrate adaptor activity in GO:0070534 protein K63-linked
ubiquitination, upstream of NBN and ATM at the site of a double-strand break. This is a second,
non-proteolytic core activity, and the repository module modules/g1_s_transition.yaml independently
cites SKP2 as the metazoan CKI F-box adaptor.

Regulation (context for non-core calls)

Curation decisions taken in this review

  1. Bare GO:0005515 protein binding (50 IPI rows). Per repository policy, these were either
    MODIFYed to an informative MF where the cited study establishes one, or REMOVEd as uninformative
    (never asserting the interaction is false):
  2. MODIFY → GO:1990756 ubiquitin-like ligase-substrate adaptor activity for rows whose partner is a
    recognised SKP2 substrate (CDKN1B, CDKN1A, DUSP1, MYC, ORC1, CDT1, CDH1, HCV NS5A) or a component
    of the recognition bridge itself (SKP1, CKS1B) in a focused mechanistic/structural study.
  3. MODIFY → GO:0097602 cullin family protein binding for the CUL1 rows from focused SCF-assembly
    studies (11961546, 12504026, 12609982, 18239684, 27542266).
  4. REMOVE for proteome-scale screens (BioPlex 2.0/3.0, LuTHy, in-situ PLA network, ISG interactome,
    Polycomb complexome, neoPPI, AI-pipeline, multimodal cell maps, CRL AQUA proteomics), for partners
    that act on SKP2 rather than being acted on by it (FZR1/Cdh1, EP300, SIRT3, TRIM21, ASB2, HBx,
    SGT1/Hsp90), and for rows whose paper attributes the mechanism to a different F-box protein
    (SMAD4 via β-TrCP1 in PMID:16865698; RRM2 via cyclin F in PMID:22632967). The physical
    interactions remain recorded in IntAct; only the uninformative GO annotation is removed.
  5. Note on PMID:18239684: the supporting entity UniProtKB:A5D8W4 (cadherin-1) sits in a paper about
    APC/C Cdh1 (FZR1). Whatever the intended partner, the row is a bare protein-binding record and
    is removed as uninformative; no claim of curator error is made.
  6. GO:0042802 identical protein binding (IPI, PMID:22770219) is supported by direct dimerization
    data but is an acetylation-gated regulatory property, so KEEP_AS_NON_CORE.
  7. IBA rows (GO:0019005, GO:0031146, GO:1990756, GO:1905168) are accepted. For GO:1990756 and
    GO:1905168 the WITH/FROM includes UniProtKB:Q13309 itself — expected, since SKP2's own
    experimental annotations are among the descendant evidences the PAINT curator used to place the IBD;
    this is not circularity.
  8. Localisation. Nucleus/nucleoplasm = core (canonical p27 turnover is nuclear
    [file:human/SKP2/SKP2-deep-research-falcon.md "Canonical activity is predominantly nuclear and oscillates with the cell cycle."]).
    Cytoplasm/cytosol (real, acetylation- and AKT-gated; Reactome places CRL neddylation/CAND1 handling
    in cytosol) and nucleolus (HPA immunofluorescence only) are kept as non-core.
  9. Phenotype-adjacent process terms. GO:0042981 regulation of apoptotic process (IDA,
    PMID:23277542) is a downstream consequence of Myc-TD ubiquitination and SKP2 overexpression, so
    non-core. GO:0045087 innate immune response and GO:0051607 defense response to virus (IMP,
    PMID:27194766) rest on a single HCV/ISG12a study and are kept non-core.
  10. GO:0070936 protein K48-linked ubiquitination (IMP, PMID:27194766): the cached record is
    abstract-only and does not state the linkage, but K48 chains are the canonical degradative output of
    SCF(SKP2) and the curator read the full text — ACCEPT rather than second-guess.
  11. core_functions holds two entries, both with GO:1990756 as the molecular function: the
    proteolytic SCF(SKP2) branch (GO:0031146 + GO:0000082, nucleus/nucleoplasm) and the K63/NBN branch
    (GO:0070534 + GO:1905168, nucleus), each in_complex: GO:0019005. A third, cytoplasmic
    E-cadherin-degradation entry was drafted and then dropped: listing cytoplasm/cytosol as core
    locations would contradict the KEEP_AS_NON_CORE calls on those very annotations (validation flagged
    exactly this). The cytoplasmic branch is therefore described in description and raised in
    suggested_experiments instead.
  12. No proposed_new_terms. Every distinct activity SKP2 performs (adaptor activity, SCF
    membership, SCF-dependent proteasomal catabolism, K63 ubiquitination, positive regulation of HR
    repair, G1/S transition) already has an annotation; substrate relationships belong on has input of
    the ligase activity in GO-CAM, not as new terms on SKP2.