A0A2I4G8T1 (gene LOC109005703, chromosome 3) is a full-length, 488-residue family-1 UDP-glycosyltransferase (GT1, UGT) from English walnut (Juglans regia). It has the two-Rossmann-fold GT-B architecture of plant UGTs: an N-terminal acceptor-binding glycosyltransferase domain (Pfam PF26168, residues 10-246) and a C-terminal nucleotide-sugar-binding domain (Pfam PF00201) carrying an intact 44-residue PSPG box (residues 348-391) with the conserved HCGWNS core and a terminal glutamine typical of UDP-glucose-using plant UGTs. PANTHER places it in subfamily PTHR48047:SF229 (UDP-glycosyltransferase 73C3-related) with Arabidopsis UGT73C3 and UGT73C4, and it is roughly equidistant (55-59% identity) from all members of the Arabidopsis UGT73C tandem cluster, which suggests it is a co-ortholog of that cluster rather than of any single member. Through descent from a PANTHER ancestral node whose characterized members transfer glucose from UDP-glucose to small lipophilic acceptors, it is inferred to be a UDP-glucosyltransferase that glucosylates small molecules. Characterized UGT73C relatives differ in acceptor (brassinosteroids, cytokinins, flavonols, trichothecene toxins, saponin aglycones), so the acceptor used by the walnut enzyme cannot be predicted. Its subcellular location has not been determined. No experimental characterization of this protein has been reported.
Summary: PAINT IBA from ancestral node PANTHER:PTN008730368, whose IBD for UDP-glucosyltransferase activity was seeded by six Arabidopsis enzymes: UGT73C1 (AT2G36750), UGT73C6 (AT2G36790), UGT73C5/DOGT1 (AT2G36800), UGT73C7 (AT3G53160), UGT89B1 (AT1G73880) and UGT89A2 (AT5G03490). The node therefore sits at least as deep as the UGT73C/UGT89 split, and the walnut protein (SF229, UGT73C group) descends from it. None of the seeds is in SF229 itself, but the term asserted is the donor-level activity shared across the node, not any seed's acceptor specificity. Target-level evidence agrees: the protein is full length, retains both GT-B domains and the PSPG box, and the PSPG box ends in the glutamine associated with UDP-glucose use. The earlier version of this review listed AT1G73880 as UGT73C3; it is UGT89B1 (UGT73C3 is AT2G36780, an uncharacterized SF229 member).
Reason: The walnut protein lies inside the clade to which the PAINT curator assigned UDP-glucosyltransferase activity, and its sequence shows no loss of the nucleotide-sugar-binding signature that would argue for divergence. Donor identity is not experimentally tested for this protein, but the phylogenetic inference is sound and is the most specific molecular function supportable.
Summary: InterPro2GO mapping from IPR002213 (UDP-glucuronosyl/UDP-glucosyltransferase). The domain is present and the claim is true, but GO:0008194 is the parent of GO:0035251 UDP-glucosyltransferase activity, which this gene already carries by IBA.
Reason: Correct but less specific than the evidence supports. Refine to the is_a descendant GO:0035251 UDP-glucosyltransferase activity already asserted for this gene by the PAINT IBA.
Summary: ARBA-derived assertion of glycosyltransferase activity. The claim is true (the protein is a family-1 UDP-glycosyltransferase), but it is a distant ancestor of GO:0035251 UDP-glucosyltransferase activity, which the gene already carries. The previous review marked this row as over-annotated; that action implies the claim is doubtful, which it is not.
Reason: True but very general. Refine to GO:0035251 UDP-glucosyltransferase activity, the specific descendant already supported for this gene by phylogenetic inference and by the conserved PSPG box.
PSPG box (44 aa, residues 348-391): WAPQLLILSHPSVGGFLTHCGWNSCIEGISAGVPMLTWPLFGDQ; last residue: Q
Core Functions
UDP-glucose-dependent glucosyltransferase that transfers glucose to small lipophilic acceptor molecules, inferred from descent from a PANTHER ancestral node of characterized UGT73C/UGT89 glucosyltransferases and from the conserved PSPG nucleotide-sugar-binding box. The physiological acceptor in walnut is unknown and cannot be inferred from the varied acceptors of Arabidopsis UGT73C relatives.
Q: Is the PANTHER placement of the walnut protein in SF229 (with UGT73C3/C4) rather than SF153 (UGT73C5/C6) or SF88 (UGT73C1) phylogenetically meaningful, given that the Arabidopsis UGT73C cluster appears to have expanded after the Brassicales-Fagales split?
Q: What are the in vivo acceptor substrates of this UGT73C-group glucosyltransferase in walnut, which accumulates naphthoquinones (juglone), ellagitannins and flavonoids?
Suggested Experiments
Experiment: Express recombinant A0A2I4G8T1 and screen UDP-glucose, UDP-galactose, UDP-xylose and UDP-glucuronic acid as donors against a panel of acceptors covering the substrates of Arabidopsis UGT73C relatives (brassinosteroids, trans-zeatin, flavonols, deoxynivalenol, triterpenoid aglycones) and walnut-specific phenolics such as juglone and its reduced form hydrojuglone.
Hypothesis: The enzyme prefers UDP-glucose as donor (as predicted from its PSPG box) and has an acceptor range that cannot be predicted from any single UGT73C ortholog.
Experiment: Build a maximum-likelihood tree of UGT73C-group sequences from Brassicales and Fagales genomes to test whether the walnut protein is a co-ortholog of the whole Arabidopsis UGT73C1-C6 cluster.
Hypothesis: The Arabidopsis cluster expanded within Brassicales, so no one-to-one ortholog of the walnut protein exists in Arabidopsis.
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.
ProtNLM2 made no predictions (GO or subcellular location) for this walnut protein and returned only the protein name "UDP-glycosyltransferase 73C1". The gene review agrees with the family placement but not with the single-member name: it places the protein in PANTHER subfamily PTHR48047:SF229 (UGT73C3-related) as a roughly equidistant (55-59% identity) co-ortholog of the whole Arabidopsis UGT73C tandem cluster, and supports UDP-glucosyltransferase activity by PAINT descent, with the acceptor substrate unknown. With no GO or location output there is nothing to score; the absence of predictions is recorded as assessed.