SCP2 (human) — curation notes

UniProt: P22307 (SCP2_HUMAN). HGNC:10606. Gene SCP2 on chromosome 1.

Bifunctional gene / two protein products

The SCP2 gene encodes two protein products from alternative promoter usage /
processing, sharing a common C-terminal SCP2 (sterol-carrier) domain:

The displayed sequence in P22307 is the 547-aa SCPx form; the SCP2 domain is
residues ~433–543 (FT DOMAIN 433..543).

Thiolase / branched-chain and bile-acid beta-oxidation (SCPx)

Disease (SCPx / SCP2 deficiency = LKDMN, MIM 613724)

Lipid-transfer / sterol-carrier function (SCP2 domain)

Localization / peroxisomal import

Notes on annotation actions

2026-11-03 ORGANELLES batch 4 (mitochondrion rows)

Rows: GO:0005739 IEA (GO_REF:0000044, SL-0173), ISS (GO_REF:0000024, WITH UniProtKB:P32020)
and EXP (PMID:17157249), all located_in. No mitochondrial sub-compartment rows.

Checks:
- UniProt P22307: [Isoform SCP2] Mitochondrion tagged ECO:0000269|PubMed:17157249 and
ECO:0000250|UniProtKB:P32020; [Isoform SCPx] peroxisome only.
- PMID:17157249 (Stanley et al. 2006, full text cached incl. Experimental Procedures): the
string "mitochondri" never occurs. Its only cell experiment is GFP fused to mature SCP2
PMID:17157249,
and the leader is described as cleaved in peroxisomes
PMID:17157249.
The earlier summary that this paper describes a mitochondrial transit peptide was wrong.
Recorded as a source defect (WRONG_REFERENCE, UNIPROT, PROBABLE: supplement unseen).
- The ISS donor P32020 is mouse (not rat as the old summary said). UniProt REST (2026-11-03):
P32020 [Isoform SCP2] Mitochondrion {ECO:0000269|PubMed:11003606, PubMed:26901662}.
PMID:26901662 (mouse MA-10 cells, full text cached):
PMID:26901662
PMID:26901662
but the crude 11,000 g fraction was checked for catalase only, not an ER marker, and
PMID:26901662.
QuickGO also lists mouse HDA rows (PMID:14651853, PMID:18614015).
- MitoCarta3.0: SCP2 listed; evidence "Rickettsial homolog, mito protein domain+";
SubMitoLocalization "MIM"; MSMS "50-75crude", 9 tissues; HPA "Peroxisomes (Supported)".

Outcome: all three rows MARK_AS_OVER_ANNOTATED -> KEEP_AS_NON_CORE. The earlier reason
("not a core site") argued non-core, not over-annotation; the minor mitochondrial pool of
isoform SCP2 has mouse fractionation/imaging, mouse proteomes and MitoCarta3.0. The EXP row
keeps the same action because the claim stands on that other evidence, and carries the
source defect. Added PMID:26901662 to references; revised the PMID:17157249 review note.
Stay at the root: MitoCarta's "MIM" is the only sub-mitochondrial line and the presequence
was shown non-functional.

2026-11-05 ORGANELLES ER batch 3 (endoplasmic reticulum rows)

Rows: GO:0005783 IEA (GO_REF:0000044, SL-0095) and ISS (GO_REF:0000024, WITH UniProtKB:P32020),
both located_in, both KEEP_AS_NON_CORE. No ER sub-compartment rows. Mitochondrion rows from
batch 4 untouched.

Checks:
- UniProt P22307: "[Isoform SCP2]: ... Endoplasmic reticulum {ECO:0000250|UniProtKB:P32020}";
"[Isoform SCPx]: Peroxisome" only. So both ER rows rest on mouse P32020.
- UniProt REST P32020 (2026-11-05): [Isoform SCP2] "Endoplasmic reticulum {ECO:0000269|PubMed:11003606}"
(sole source; the 2016 paper PMID:26901662 supports Mitochondrion/Peroxisome/Cytoplasm only,
and only cites earlier ER pools).
- PMID:11003606 (abstract-only, mouse L-cell fibroblasts stably overexpressing SCP-2):
PMID:11003606
plus SCP-2 stimulation of microsomal phosphatidic acid synthesis in vitro.
- Soluble protein, PTS1 -AKL, no signal peptide/TM: any ER pool is on the cytosolic face.

Outcome: actions unchanged (KEEP_AS_NON_CORE), close call against MARK_AS_OVER_ANNOTATED
(overexpression-only localization) stated in the ISS reason. ISS summary corrected (donor is
mouse, not rat). Isoform restriction recorded as a source defect (Gene_Product_Form_ID ->
P22307-2) on the ISS row. Stays at the root. Added PMID:11003606 to references.
core_functions[1] still lists GO:0005783 for the lipid-transfer function; left as is (the
microsomal transfer data are the functional link), but worth a second look.
- Correction (2026-11-06): settled in review of PR #6028. GO:0005783 was removed from the
lipid-transfer core function's locations (cytosol only), and its description now names
donor and acceptor membranes rather than an ER location, since no assay places the transfer
activity at the ER.