GPC3 (glypican-3) is a cell-surface heparan sulfate proteoglycan attached to the outer leaflet of the plasma membrane by a GPI anchor. Its core protein is cleaved by a furin-like convertase into alpha and beta subunits that remain disulfide-linked, and it can be shed from the cell surface. GPC3 promotes canonical Wnt signaling by binding Wnt ligands and, through its heparan sulfate chains, the Frizzled receptors, and it inhibits hedgehog signaling by binding hedgehog ligands in competition with Patched and driving their endocytosis and degradation. Through these activities it restrains growth during development; loss of GPC3 causes the overgrowth disorder Simpson-Golabi-Behmel syndrome in humans. This donkey entry (A0A9L0J2B1, 565 aa) is an alternative-first-exon gene model of GPC3. Its first 43 residues are not homologous to the N-terminus of human GPC3 and do not include the human signal peptide. Downstream of that segment it is 96% identical to human GPC3 and keeps the convertase cleavage site, heparan sulfate attachment serines and GPI-anchor site. The canonical donkey GPC3 product, with a human-like signal peptide, is A0A9L0JGR2 (580 aa).
Summary: TreeGrafter transfer of a PAINT IBD placed at the eumetazoan glypican node PTN002247069 (seed mouse Gpc1). GPC3 can be released from its GPI anchor and shed, so a soluble form exists in the extracellular region.
Reason: Donkey GPC3 sits in the GPC3 subfamily (PTHR10822:SF4) inside the clade below PTN002247069, and there is no evidence of loss. Shedding of GPC3 is documented for the human protein, but its functional location is the cell surface, so the extracellular region is a secondary location. One caveat applies to this entry only. The A0A9L0J2B1 model lacks the human signal peptide (RESULTS-rereview.md), so the annotation holds for the GPC3 gene product with the canonical N-terminus and is not established for this transcript model.
A second proteolytic event may result in cleavage of the protein on the cell surface, separating it from the GPI-anchor and leading to its shedding from the cell surface
Summary: GPC3 is a GPI-anchored protein of the outer leaflet of the plasma membrane. PAINT placed plasma membrane at the euteleostome node PTN002585559 (seeds mouse Gpc3, rat Gpc3, human GPC3), and the donkey GPI-anchor site is conserved.
Reason: Shown experimentally for human GPC3 (PMID:14610063) and for its rodent orthologs, which seeded the PAINT node from which donkey GPC3 descends; the GPI-anchor asparagine is conserved in the donkey sequence. As for all secretory-pathway terms here, the A0A9L0J2B1 model itself lacks the human signal peptide, so the annotation describes the canonical GPC3 product of this gene.
Summary: GPC3 regulates Wnt and hedgehog signaling; the general term is true but uninformative.
Reason: This InterPro2GO family-level term is entailed by the more specific positive regulation of canonical Wnt signaling pathway row accepted in this review and by the proposed negative regulation of smoothened signaling pathway, which name the pathways actually regulated.
Summary: GPC3 is a cell surface proteoglycan; PAINT placed cell surface at the eumetazoan glypican node PTN002247069 (seeds Drosophila dally and dally-like, human GPC4, C. elegans lon-2).
Reason: The GPI-anchored, extracellular-facing location is where GPC3 engages Wnt, Frizzled and hedgehog. Supported by human data, by the family-level IBD and by the conserved anchor and glycosaminoglycan sites. This describes the canonical GPC3 product; the A0A9L0J2B1 model lacks the signal peptide.
Summary: TreeGrafter transfer of a PAINT IBD placed at the eumetazoan glypican node PTN002247069 (seeds Drosophila dally and dally-like, human GPC6, C. elegans and zebrafish glypicans). The only GPC3-specific evidence is a zebrafish morpholino study of gastrulation movements.
Reason: The PAINT placement covers the GPC3 subfamily and there is no evidence of loss, so the row is kept. However glypicans influence cell movement by modulating (mainly non-canonical) Wnt signaling, not by executing migration, and the GPC3-specific support (epiboly and convergence/extension defects in zebrafish morphants, rescued by mRNA) is a downstream organismal phenotype. It is therefore not a core function, and that morphant evidence is not used here as proof of participation.
Summary: Family-level InterPro2GO mapping. GPC3 is a GPI-anchored cell-surface proteoglycan, not an extracellular matrix constituent.
Reason: PAINT placed extracellular matrix (seed mouse Gpc1) at node PTN000088172, and the GPC3 lineage did not inherit it. Neither the donkey TreeGrafter rows nor human GPC3's IBA rows include it, so the family-wide InterPro2GO term goes further than the phylogenetic curation. Glypicans are tethered to the plasma membrane by their GPI anchor or shed as soluble forms. Matrix residence has not been shown for GPC3, whose characterized functions all take place at the cell surface.
GO:0090263 positive regulation of canonical Wnt signaling pathway
IEA GO_REF:0000118
ACCEPT
Summary: GPC3 stimulates canonical Wnt signaling by binding Wnt and, through its heparan sulfate chains, Frizzled, promoting Wnt-Frizzled signaling complexes. PAINT placed this term at the bilaterian node PTN000850517 (seeds Drosophila dally and human GPC3).
Reason: Human GPC3 IDA (PMID:24496449) is one of the seeds of the PAINT node, and donkey GPC3 descends from it within the GPC3 subfamily. The donkey protein conserves the heparan sulfate attachment sites needed for Frizzled binding and the convertase site needed for Wnt modulation. The direction of the effect depends on context; in transfected CHO cells GPC3 repressed Wnt1 reporter activity (PMID:14610063). The stimulatory role is the one supported in HCC and by the IBD.
Positively regulates the canonical Wnt signaling pathway by binding to the Wnt receptor Frizzled and stimulating the binding of the Frizzled receptor to Wnt ligands
file:EQUAS/GPC3/GPC3-bioinformatics/RESULTS.md
the two heparan sulfate attachment serines (human S495, S509)
GO:1905475 regulation of protein localization to membrane
IEA GO_REF:0000118
KEEP AS NON CORE
Summary: TreeGrafter transfer of a PAINT IBD at the eumetazoan glypican node PTN002247069, seeded by rat Gpc4 and Gpc6. For GPC3, binding of Wnt or hedgehog at the membrane triggers endocytosis of ligand complexes, which changes how long they stay at the membrane.
Reason: The family-level PAINT placement includes the GPC3 subfamily, and GPC3-driven internalization of hedgehog and of Wnt-Frizzled-GPC3 complexes fits the term. This is a mechanistic aspect of its signaling roles, not a core function, and the seed evidence comes from GPC4/GPC6, not GPC3.
GO:0045879 negative regulation of smoothened signaling pathway
ISS PMID:18477453 Glypican-3 inhibits Hedgehog signaling during development by...
NEW
Summary: GPC3 inhibits hedgehog signaling by binding hedgehog with high affinity, competing with Patched for it, and inducing its endocytosis and degradation. This is shown for mouse Gpc3, which is in the same PANTHER subfamily (PTHR10822:SF4, glypican-3) as donkey and human GPC3.
Reason: GPC3 itself performs the regulatory step, binding and sequestering the ligand, so this is participation and not only necessity. Gpc3-null embryos showing hedgehog hyperactivity add genetic support. PAINT has no hedgehog-related IBD in the glypican family, and this is consistent with the function being specific to the GPC3 subfamily rather than family-wide. Only the mouse donor's acts-upstream annotations carry it, and PAINT did not propagate them at any family node. The transfer here is therefore restricted to the direct GPC3 ortholog (same subfamily, 96% identical over the shared core), not inferred from glypicans generally. The convertase cleavage site required for this function is conserved in donkey. Human GPC3 carries the same term by ISS from mouse.
ISS PMID:24496449 Glypican-3 binds to Frizzled and plays a direct role in the ...
NEW
Summary: Human GPC3 binds Frizzled receptors directly through its glycosaminoglycan chains, stimulating Wnt-Frizzled complex formation.
Reason: Direct GPC3-Frizzled binding (human GPC3, PMID:24496449) is the mechanism behind its positive regulation of canonical Wnt signaling, which PAINT propagates to this subfamily. The heparan sulfate attachment serines are conserved in donkey GPC3. The evidence code is ISS because the experiment was done on the human ortholog, which is in the same subfamily. Human GOA records only the process from this paper; the molecular-function term states the mechanism the abstract reports directly. The binding depends on heparan sulfate chains, so it is a property of the canonical, secretory-pathway product.
ISS PMID:18477453 Glypican-3 inhibits Hedgehog signaling during development by...
NEW
Summary: GPC3 binds hedgehog proteins through its core protein, without needing its heparan sulfate chains, and sequesters them from Patched. Shown for mouse Gpc3 (same subfamily, PTHR10822:SF4).
Reason: Hedgehog binding is the molecular activity behind the inhibitory effect of GPC3 on hedgehog signaling. The binding is to the core protein, which is 96% identical between donkey and human GPC3 over the shared region, and the transfer is restricted to the GPC3 subfamily.
Positively regulates the canonical Wnt signaling pathway by binding to the Wnt receptor Frizzled and stimulating the binding of the Frizzled receptor to Wnt ligands
Negatively regulates the hedgehog signaling pathway when attached via the GPI-anchor to the cell surface by competing with the hedgehog receptor PTC1 for binding to hedgehog proteins
Q: Does GPC3 restrain growth in equids as in humans and mice, and is its role in hepatoblast and renal development conserved in donkey?
Q: Is the alternative-first-exon transcript that encodes A0A9L0J2B1 (ENSEAST00005057340) expressed in donkey tissues? If it is, can its product, which lacks the canonical signal peptide, enter the secretory pathway and be GPI-anchored, or should GO annotation be attached only to the canonical product A0A9L0JGR2?
Suggested Experiments
Experiment: RNA-seq or 5' RACE across donkey tissues to test whether the alternative first exon of ENSEAST00005057340 is used, with epitope-tagged expression of the A0A9L0J2B1 product to test signal-peptide-independent membrane targeting.
External Prediction Reviews
These computational predictions are reviewed separately from the GOA annotation set used for this review. The assessments below are from this project and do not constitute official GO annotations or endorsement by GO/UniProt. They are not included in the existing annotation review above.
Assessment of the single ProtNLM2 GO prediction for donkey GPC3 entry A0A9L0J2B1, an alternative-first-exon gene model of the GPI-anchored heparan sulfate proteoglycan GPC3 that lacks the signal peptide of the canonical product (A0A9L0JGR2). The prediction regulation of signal transduction (GO:0009966) is true of the gene. A rule-based hydropathy and n/h/c-region analysis finds no classical secretory signal in the alternative N-terminus, so the known cell-surface mechanism is unlikely to be available to this product, but the analysis is not a trained predictor and other contributions are neither shown nor excluded; the prediction is scored UNC. ProtNLM2 emitted no function text or subcellular location for this accession.
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-10-14
Review rationale: GPC3 regulates Wnt and hedgehog signaling as a GPI-anchored heparan sulfate proteoglycan at the cell surface; human GPC3 stimulates canonical Wnt signaling by binding Frizzled through its heparan sulfate chains. That mechanism needs entry into the secretory pathway, where the GPI anchor and heparan sulfate chains are added. A0A9L0J2B1 is an alternative-first-exon gene model whose first 43 residues are not homologous to the human N-terminus and lack the human signal peptide (P51654 SIGNAL 1..24). A Kyte-Doolittle and von Heijne-style n/h/c-region scan (signal_peptide_hydropathy.py) recovers the signal peptide and its cleavage after residue 24 in both human P51654 and the canonical donkey product A0A9L0JGR2, but finds no candidate h-region in A0A9L0J2B1; its most hydrophobic 7-residue window has a mean of 1.54, against 3.36 in the controls, and its longest apolar run is 5 residues. This is evidence against a classical secretory signal, so the known cell-surface mechanism is probably unavailable to this product. It is not positive refutation of the process term, however. The scan is a rule-based check rather than a trained predictor, the transcript itself is unconfirmed, and a contribution to signal regulation by the retained core protein through some other route is neither shown nor excluded. The call is therefore uncertain. The term is true of the GPC3 gene through its canonical product A0A9L0JGR2; donkey GOA already carries it and the more specific positive regulation of canonical Wnt signaling pathway (GO:0090263), so the prediction is not novel at the gene level.